<%BANNER%>

Photocatalytic Oxidation of Hazardous Air Pollutants Using Silica-Titania Composites in a Packed-Bed Reactor

Permanent Link: http://ufdc.ufl.edu/UFE0022107/00001

Material Information

Title: Photocatalytic Oxidation of Hazardous Air Pollutants Using Silica-Titania Composites in a Packed-Bed Reactor
Physical Description: 1 online resource (127 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: mercury, methanol, photocatalysis, silica, titania
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This work centered on the optimization and scale-up of photocatalytic reactors employing silica-titania composites (STC) for two applications involving the abatement of hazardous air pollutants (HAPs) from industrial facilities: (1) degradation of HAPs, particularly methanol, emitted from pulp and paper mills; and (2) recovery of mercury emitted from chlor-alkali plants. STC were synthesized with varying pore sizes (50 ?, 120 ?, and 260 ?) and titania loadings (0-60%) and were tested for the removal of methanol from a humid air stream. The efficiency of methanol oxidation was dependent on the surface area of the STC and the space time of the gas in the reactor. For 120 ? 12% and 260 ? 12% STC irradiated with UVA light, a lag time of 1.0 s and 1.2 s, respectively, was observed before mineralization began. After this lag time, which was zero for the 50 ? 12% STC, the data followed pseudo-first order reaction kinetics and the rate constant, k, was 0.40 s-1 for all pore sizes. Using the 50 ? STC, the efficiency was further improved by using a 4% titania loading and UVC lamp, which generated a higher photon flux compared to a UVA lamp. The presence of hydrogen sulfide in the gas stream decreased methanol removal efficiency and resulted in sulfur dioxide and sulfate oxidation byproducts. When compared to other catalyst supports, the STC was more efficient in a low-humidity gas stream with a relative humidity (RH) of less than 0.22% at 23?C. In a high humidity gas stream (RH = 95% at 23?C), the efficiency of the STC was inhibited by water vapor due to its surface chemistry and performed similarly to titania-coated activated carbon. When compared to titania-glass spheres, the use of an adsorbent catalyst support resulted in higher degradation efficiencies. Based on the promising bench-scale results, a 40 ACFM pilot reactor was fabricated employing a packed bed of STC and a 4.3 s space time through the packed bed. The pilot reactor achieved methanol removal rates up to 66 ? 7% with less than 1 ppmv formaldehyde production at steady state. A pilot-scale photocatalytic reactor packed with STC was tested at a chlor-alkali facility over a three-month period. This pilot reactor treated up to 10 ACFM of end-box exhaust and achieved 95% mercury removal. The pilot reactor was able to maintain excellent removal efficiency even with large fluctuations in influent mercury concentration (400-1600 ug/ft3). The STC pellets were regenerated ex-situ with hydrochloric acid and performed similarly to virgin STC pellets when returned to service. Based on these promising results, two full-scale reactors with in-situ regeneration capabilities were installed and operated. After optimization, these reactors performed similarly to the pilot reactor. A cost analysis was performed comparing the treatment costs (i.e., cost per pound of mercury removed) for sulfur-impregnated activated carbon and the STC system. The STC proved to be both technologically and economically feasible for this installation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022107:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022107/00001

Material Information

Title: Photocatalytic Oxidation of Hazardous Air Pollutants Using Silica-Titania Composites in a Packed-Bed Reactor
Physical Description: 1 online resource (127 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: mercury, methanol, photocatalysis, silica, titania
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This work centered on the optimization and scale-up of photocatalytic reactors employing silica-titania composites (STC) for two applications involving the abatement of hazardous air pollutants (HAPs) from industrial facilities: (1) degradation of HAPs, particularly methanol, emitted from pulp and paper mills; and (2) recovery of mercury emitted from chlor-alkali plants. STC were synthesized with varying pore sizes (50 ?, 120 ?, and 260 ?) and titania loadings (0-60%) and were tested for the removal of methanol from a humid air stream. The efficiency of methanol oxidation was dependent on the surface area of the STC and the space time of the gas in the reactor. For 120 ? 12% and 260 ? 12% STC irradiated with UVA light, a lag time of 1.0 s and 1.2 s, respectively, was observed before mineralization began. After this lag time, which was zero for the 50 ? 12% STC, the data followed pseudo-first order reaction kinetics and the rate constant, k, was 0.40 s-1 for all pore sizes. Using the 50 ? STC, the efficiency was further improved by using a 4% titania loading and UVC lamp, which generated a higher photon flux compared to a UVA lamp. The presence of hydrogen sulfide in the gas stream decreased methanol removal efficiency and resulted in sulfur dioxide and sulfate oxidation byproducts. When compared to other catalyst supports, the STC was more efficient in a low-humidity gas stream with a relative humidity (RH) of less than 0.22% at 23?C. In a high humidity gas stream (RH = 95% at 23?C), the efficiency of the STC was inhibited by water vapor due to its surface chemistry and performed similarly to titania-coated activated carbon. When compared to titania-glass spheres, the use of an adsorbent catalyst support resulted in higher degradation efficiencies. Based on the promising bench-scale results, a 40 ACFM pilot reactor was fabricated employing a packed bed of STC and a 4.3 s space time through the packed bed. The pilot reactor achieved methanol removal rates up to 66 ? 7% with less than 1 ppmv formaldehyde production at steady state. A pilot-scale photocatalytic reactor packed with STC was tested at a chlor-alkali facility over a three-month period. This pilot reactor treated up to 10 ACFM of end-box exhaust and achieved 95% mercury removal. The pilot reactor was able to maintain excellent removal efficiency even with large fluctuations in influent mercury concentration (400-1600 ug/ft3). The STC pellets were regenerated ex-situ with hydrochloric acid and performed similarly to virgin STC pellets when returned to service. Based on these promising results, two full-scale reactors with in-situ regeneration capabilities were installed and operated. After optimization, these reactors performed similarly to the pilot reactor. A cost analysis was performed comparing the treatment costs (i.e., cost per pound of mercury removed) for sulfur-impregnated activated carbon and the STC system. The STC proved to be both technologically and economically feasible for this installation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022107:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101109_AAAABF INGEST_TIME 2010-11-09T15:54:42Z PACKAGE UFE0022107_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 21101 DFID F20101109_AAAXIA ORIGIN DEPOSITOR PATH stokke_j_Page_112.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
a43d3c9a627cc9e35d4e9dc120429edc
SHA-1
6d8d1e9879d7eb8db414e596b6aab61d98e59c41
92 F20101109_AAAXHM stokke_j_Page_002.txt
eaa947b00051b6dd949746352a9d91b5
bf27929cd7abe905dfae83472893002b316f143d
25271604 F20101109_AAAXGY stokke_j_Page_048.tif
cefa8df9c26eb2ed68c315493ecbf5c0
c7a40b42f6a981db1931b49a252d49a3468708fe
7752 F20101109_AAAXIB stokke_j_Page_100thm.jpg
79c2d41eef6929f8ed5d4007d2a0c892
d9447542ef098e65ea1f9c0e7ef074e0ce1ffd6e
5365 F20101109_AAAXHN stokke_j_Page_006thm.jpg
0e3293432160a77bed875c3bff39ed81
0354b006c3b764db496c4ca15297162e18a71c49
714 F20101109_AAAXGZ stokke_j_Page_075.txt
173cdebd7089ea54f5ff1bfe6dc19697
b6a707f7feeb26067f478c121d359ce54c37a288
8423998 F20101109_AAAXIC stokke_j_Page_115.tif
176a1fdd6aa4070e052710376d94b929
f91b9010b1967abcd87424c9af3ed3f44d78b2a3
15929 F20101109_AAAXHO stokke_j_Page_026.QC.jpg
8ee5c8b638f28902091d9f0707e0a038
ce30bef634a6eeb4884abc913c0c83f29700fffc
55562 F20101109_AAAXID stokke_j_Page_014.pro
bd223bbee9b5a2296b84b9ea789a8f13
27ea26d3c763d54964429004b253c362d5f58b4b
53379 F20101109_AAAXHP stokke_j_Page_078.pro
3bd203596dfe6e77ffa078aeaeeea1e4
8d86a4573671d16901613253e371b4674ed2ecaa
50652 F20101109_AAAXIE stokke_j_Page_060.pro
cb1ee493efe0e01582d3f141352d279b
55dba095abc2249e2a9b5f792bfb15b8f5e99ac7
F20101109_AAAXHQ stokke_j_Page_032.tif
19b792ee1879c38e867dfd7dc12c31b5
1345d605901b934eaa2fb0af7e060c7eb903f5e9
1051940 F20101109_AAAXIF stokke_j_Page_085.jp2
e54ce494c3e925d166c0d801e259a09e
6f91e43ffb2b38f19b47b057a36ebeed91bf4cf5
F20101109_AAAXHR stokke_j_Page_097.tif
5f156d3b3dea4e9e375b870152df0df0
1b84b9af4f8adacc90b4403b9f38f8de2bac682a
25039 F20101109_AAAXIG stokke_j_Page_031.QC.jpg
f5c755f21a3c1c569b9eccf3f4e0a2f3
65ce1faff639653b30857f39480a346f7446b964
40329 F20101109_AAAXHS stokke_j_Page_087.jpg
d2fab1d545713a8809bd127e001a6ff1
18bd0585a552783e49ed138a4ffb76cc1aac4588
87719 F20101109_AAAXIH stokke_j_Page_078.jpg
bd91cd20d742481e95d074325a58ff22
d158ada25483e5e0e4e9fbf24fb243fb1a5900ee
7391 F20101109_AAAXHT stokke_j_Page_008thm.jpg
4bc8050d3d44fc3595cb1bc2648eb83e
2df2728b86633a54cb783bce44d9adbaf6a9bd2f
55535 F20101109_AAAXII stokke_j_Page_053.pro
3a3281214a77c014bf5f3a2844602384
fa1bba56d5a7aff2c61311cba295217f12069ada
F20101109_AAAXHU stokke_j_Page_001.tif
8b2483eb28f043897d60bec0d84850d5
5eacf0769d8f8d712e7a25963049398799649d93
89389 F20101109_AAAXIJ stokke_j_Page_084.jpg
81ec104af117a7c226f2116d7cbd4552
7e5b67a17a8eb67c44733e447fa06cd34f8d4f1a
52466 F20101109_AAAXHV stokke_j_Page_106.pro
ebd1b1ab1f4a06d951a4ec0c75ea032b
c96c03fae5918b1bd1a7221944dd0e5f3b3838f5
26277 F20101109_AAAXIK stokke_j_Page_119.QC.jpg
5b2eccecafe92f21da925dc748c0371b
6e4a18949cf6fc80c1188c8544b4e016620695cc
F20101109_AAAXHW stokke_j_Page_124.tif
b22c3aeb542a7803dbeccbb4a75b3a0b
efe20b6f8222f4363b0cc14aeb6c03c201df3547
F20101109_AAAXIL stokke_j_Page_126.tif
10cc01fee61a9847ef4e520057a2255f
38033e1bb6b5fdd243b2a30e96a623223aaf8e9c
54600 F20101109_AAAXHX stokke_j_Page_025.pro
d856d02c8c35a5443ff83cd78e359c65
243bfff6d72fe624a6c38defe0e6ead23ee40700
190622 F20101109_AAAXIM UFE0022107_00001.xml FULL
27bb9ca502f37180c824cdab1c5cb88b
e98edbcc69c1987504a6835281d69cb7010d65d4
3888 F20101109_AAAXHY stokke_j_Page_096thm.jpg
490b72735abde2e68da015e005dae780
bda362db80572eab704f6488f0e7e31609d27176
81266 F20101109_AAAXJA stokke_j_Page_016.jpg
be91ac83480dbb79c348ad42ddac808b
1a2bd05088276111ec2cbf1f76a689b8335319ae
2044 F20101109_AAAXHZ stokke_j_Page_120.txt
aa47aec190041d275b31f56d50c3dfd9
6c558a189ba4e9808fbdd131fbcf1be512b527c4
81623 F20101109_AAAXJB stokke_j_Page_017.jpg
6dfd49ae6904d1da16f8b56bb00e5441
3b0150398510889e8133d337f398883fe7006bc2
63110 F20101109_AAAXJC stokke_j_Page_019.jpg
e86304aa408dbb54f13961097e444992
94262b43067ff10f62a89012e498cc8b221e84eb
28842 F20101109_AAAXIP stokke_j_Page_001.jpg
a8383d4b90372f3989f51c3af18eedd3
1d01034a67f22216d2c505bf8635408dd5372d24
56601 F20101109_AAAXJD stokke_j_Page_020.jpg
81e83eda67538009c30d3da36b173f9f
e4c52d429ed4bc232753192c5451679cac5ee3e6
10129 F20101109_AAAXIQ stokke_j_Page_002.jpg
91719c6a7925321242d05c87e672a664
cb59179d00d0815c7d1a068ec66c2c11870bed3f
83538 F20101109_AAAXJE stokke_j_Page_021.jpg
7dfa32b5b900d39b093a6ea4acdbd225
2e1547d84a9bc723d4cf725f76aa65c44fa18aa4
10381 F20101109_AAAXIR stokke_j_Page_003.jpg
ff5db8ac2fd8e933a2a233ab40c6ec1b
3b51115f7e615866882b444ec24fdc26503b481b
85327 F20101109_AAAXJF stokke_j_Page_022.jpg
47f4aa7c10250fc086599dc3e8e0dddb
c8096cf82b130fb0781d36cf3636fcf203116d20
64945 F20101109_AAAXIS stokke_j_Page_004.jpg
5c44748e7ae1a7ad5866b501d5a1e74d
06c94bcbbe356eca4a7e983550cdd1e22fff6f1d
84448 F20101109_AAAXJG stokke_j_Page_023.jpg
deaf68923dccb72eb8bfe2e4ab808a84
41f247a8a8c2c2cafd257158e0ba52944733449f
83114 F20101109_AAAXIT stokke_j_Page_005.jpg
d9f450ace59d93932a232f3385d330b9
64c42f8f2b925da4ee7ba421ca514298f759f345
81250 F20101109_AAAXJH stokke_j_Page_024.jpg
d754ed672545ac9075393ed929018e08
c6d83ad21126f879e68f43fad4c7b4a8218a7ae9
86863 F20101109_AAAXIU stokke_j_Page_006.jpg
18897481e10faeb369cc028f40bb1a94
30ea99bb25443fde1686b12baefc33ad28ec7080
87627 F20101109_AAAXJI stokke_j_Page_025.jpg
8ada79da2872cfb1ba5592ff209173d1
ac2f33a6b5339ede5fe59ed44613ae956fbad284
46117 F20101109_AAAXJJ stokke_j_Page_026.jpg
4f49c08db72a5989f5148feea91e4f0f
7573e8b8caa6d85e36c84ec6b5669297159d8a87
44147 F20101109_AAAXIV stokke_j_Page_007.jpg
643d95c93361bdba1153e1af376c0fb0
dabe9a5a8c6248ae049e2cbc26e63245db841270
86823 F20101109_AAAXJK stokke_j_Page_027.jpg
1177c3a9a7aade92b14268242e1147ec
95e87cc2e228d9fc9112b7e2a4ccc62e107c8fc6
102566 F20101109_AAAXIW stokke_j_Page_008.jpg
126abb8d046761f3d096ef6bf0d5aeaf
14e04dfb0ce43d557e34cce38229a484cf213730
87364 F20101109_AAAXJL stokke_j_Page_028.jpg
88895a30afbb84d609ed1ba1283d5727
5089c08d2fa73223e3ac5a2430567faca564b99e
100316 F20101109_AAAXIX stokke_j_Page_009.jpg
079d85e2839e46926a468ffbc03e12e5
1b15e90852b919554e44e0c6dc9395c78000fff2
86946 F20101109_AAAXKA stokke_j_Page_052.jpg
1f8b0403e811fa0db60a88ed24320ca8
3f9be249e023c945dfe8d2dd5a9fb1e004d3c550
90249 F20101109_AAAXJM stokke_j_Page_029.jpg
dbf1e2254a3ed52eeed8a3db44ec1982
cc35f8f2d33c7801530bc6979dc3b26cafbb84f1
86236 F20101109_AAAXIY stokke_j_Page_013.jpg
8ab8c169c301bf699e5c1e57a2ccc0f0
3fcc571b818f05c700c29dd6d5606c98f32e629a
84155 F20101109_AAAXKB stokke_j_Page_055.jpg
c03c469d96397f0eafcaf012f0194560
c191434cbfdebc7a6c3ce41e00608b2a666525dd
90202 F20101109_AAAXJN stokke_j_Page_030.jpg
905ac57f9ca9964f495fcf1437ceb333
e1b7c9c99684d34a8b93848ffe1e9eae8782a3ad
90296 F20101109_AAAXIZ stokke_j_Page_014.jpg
424038013441eb10fac5ff598945fd8d
29654c80ce773b25ec33662399b3278d5f715220
87207 F20101109_AAAXKC stokke_j_Page_056.jpg
9051282feed46dcf1ed431c58e982bf6
6041ef59583c503aad0f0de16ebe04e93ca7ec86
81431 F20101109_AAAXJO stokke_j_Page_031.jpg
7e0f05155ed2af3b92d653268f058dce
8b2d83a7e337aa8686646235419c81b4c678b79f
86567 F20101109_AAAXKD stokke_j_Page_057.jpg
8a454762e933886c44d09545423128d8
6fb2ee2c0f3e5faf010f51ec3785c8b61c4f9cff
88938 F20101109_AAAXJP stokke_j_Page_032.jpg
89825347c7aa5e65b36f5d35b5574ec6
08a2bbf15fe66c9c89437d5af07cf8ee461cc9b0
84881 F20101109_AAAXKE stokke_j_Page_061.jpg
e747e69231f0652f100a81161baef362
7bf0f5cc89cfac7e405d005de804e10e6d4e24c2
80234 F20101109_AAAXJQ stokke_j_Page_036.jpg
a13c35960751629b844cffb3e2504e6a
43f25b86e9be4ea3eee3c9c0295e3cb5c7971b51
82975 F20101109_AAAXKF stokke_j_Page_062.jpg
595cdd6756e73e0840aba280dd23c1d4
cc9a83fa5f467a19d336807bfefcd4346105409c
77955 F20101109_AAAXJR stokke_j_Page_038.jpg
bea02b98155d66cb5c810dee18e7f281
fe06bfbc32e23ebe04315bc6f3b3da702413c52b
85081 F20101109_AAAXKG stokke_j_Page_063.jpg
abab13e44a2ac23b6d66baf25de1aff9
544a0d641f8c6fc72f80e3e897d26f46a7e6c630
89029 F20101109_AAAXJS stokke_j_Page_039.jpg
3d6330e4137f7bb8b00f167e814946e0
73fcf86777cbfb4248db57f5bd184645c2ecedd9
77781 F20101109_AAAXKH stokke_j_Page_064.jpg
4a6ab7fde0cc342263bd71448c26fb40
1baf06bfd7fe939257a97ed397165a68b0d0e03c
80351 F20101109_AAAXJT stokke_j_Page_041.jpg
7d6ec7b42c53f2f42910c625eebc5e6f
d927faf1ff12ab81e0c5f52812b8261866fe309d
90152 F20101109_AAAXKI stokke_j_Page_066.jpg
8a70d306a5b8358830c70ab97d78f34d
e37e4ae57ec8d0b5176e5ee02af029b71dac23a9
76693 F20101109_AAAXJU stokke_j_Page_042.jpg
5bd16fdf940e98370b4e979cc9adfd03
812155e89575a8bb6bc2e75a64353b84da093f2d
87987 F20101109_AAAXKJ stokke_j_Page_067.jpg
48608510830c8ccbb7ecb67537952817
3908375c2acd7d8119334e6d3eee0d08789f592b
56205 F20101109_AAAXJV stokke_j_Page_043.jpg
e2ff3a1bbd91f4e13ab23ff88d0ce83a
a649726275b97cf5bfbcecf9a9f7813005d57ebc
48759 F20101109_AAAXKK stokke_j_Page_069.jpg
e23b6bfb5a5fabb1dae7158e0044bab6
2447ac1df4b0f8a1aec4686c2d8c18394ad609b7
88957 F20101109_AAAXJW stokke_j_Page_047.jpg
24d77fa791d248118f085c5d31d2e67b
54b9feedf2a917692f9182427adfe5bbb87f12d1
41310 F20101109_AAAXKL stokke_j_Page_070.jpg
aa7abf023631d7eba84d33cc0f1ecc29
1fe4079dbda475ef6b29cb087904abbe23453e4b
87977 F20101109_AAAXJX stokke_j_Page_048.jpg
11693aa06891b25be123a61e34dfe22b
473034917a37140b33e64e21128c48bf7a0a58ef
24888 F20101109_AAAXLA stokke_j_Page_099.jpg
330faee0e35c3960f7425008c856687a
2dbe64cfbdd4f8a0aefd96047361fbccd8142e8e
32354 F20101109_AAAXKM stokke_j_Page_073.jpg
b4621af68f9b233cbb9ac9ff96292956
1a3f51ea6844820110ade0d489213803a8cf74cb
44740 F20101109_AAAXJY stokke_j_Page_050.jpg
cf29b5d83e4920ddcac36d471360fede
6d370853e1c83f1a111e3c39d2f2d42293e51ff2
34000 F20101109_AAAXKN stokke_j_Page_075.jpg
833b2d5bf3a8d0e55924e656684ca79c
b3a9f70c31c8bd909f5199bd57007844c7588431
41075 F20101109_AAAXJZ stokke_j_Page_051.jpg
453bf9861ccf31455dea3937716f2930
6110c9f935f35414df7021108860390f5e2c105b
89335 F20101109_AAAXLB stokke_j_Page_101.jpg
bfc35e732d2e7a6ede6a9a64983bceb8
48607b7be19e2a305c754b6caa164404a6fcb07c
43386 F20101109_AAAXKO stokke_j_Page_077.jpg
444c1c48a32069976ebbc41c4727fcf5
ec822a0c33a47ea4d4273c55a6bb29455b0ef0ae
86344 F20101109_AAAXLC stokke_j_Page_102.jpg
bb678fc71ffd577e614731eeba5afdf7
c802456ed91ee81b4c2583ee5315ed97c0b8d6a8
88528 F20101109_AAAXKP stokke_j_Page_079.jpg
ddfd06a1cd11c0ff5aa4ad9a473f3488
1f7e94904116f3964c80b6d1fc85ef95bbd051d1
83206 F20101109_AAAXLD stokke_j_Page_104.jpg
4bc9bfd02d09c9a49a1d024f950e7ca0
7839e0ed985378c42471514338f5f0dc655c9b97
99916 F20101109_AAAXKQ stokke_j_Page_080.jpg
985c8e106713b162cbe1a5ded492faa1
8684bff623970672adc9d0d6d93fd4d4f74371c9
83618 F20101109_AAAXLE stokke_j_Page_105.jpg
860af3e53a22c394bc1648bd3d3c9da2
fca73e2f343ff1577df22c98b13a979dbe2d7758
81684 F20101109_AAAXKR stokke_j_Page_085.jpg
85cff48aa33da6eded411213232d4536
94f8d7936b0c4b263a625a7dd0ad7d24b4e0a3ff
83258 F20101109_AAAXLF stokke_j_Page_106.jpg
7ed12e202d4c3bad28873cd34bb356a1
65aea74240212b303abb17f33b19b625964ea45a
38546 F20101109_AAAXKS stokke_j_Page_088.jpg
dea19f4aae312f3fb47bf4bfdf02a38e
2942805d9a3b747d84ba1290e76fbbd9495d282a
79976 F20101109_AAAXLG stokke_j_Page_107.jpg
8423b6df46872fd4fbd75768e0a3d143
00626a659c563dcfcb2a53ab693ee43783a50f6d
41559 F20101109_AAAXKT stokke_j_Page_089.jpg
f43ce8584704a558a00a51d4998fbb7e
8e277dc4d6a75bec59325b8b272b7ee99c9319ca
85587 F20101109_AAAXLH stokke_j_Page_108.jpg
feafc699c0612e2a3fd5683032aa2c47
25614ae62f094140984b337e73e4449e5e947e72
86793 F20101109_AAAXKU stokke_j_Page_090.jpg
51521df4bd71be13c92093524f1aeeb7
fd4e1b1a908d4f75806d1a7d7d056c8909e3d731
86202 F20101109_AAAXLI stokke_j_Page_109.jpg
7dcb5509c4f5a247f9adae4939c7711f
9716d7ba0735a89acd974da33ae06eb662e50e51
85888 F20101109_AAAXKV stokke_j_Page_091.jpg
d75650623c4501385e3862d341106619
7f73d66498c7d9fe13140c4a5e8d33c701df8649
48288 F20101109_AAAXLJ stokke_j_Page_111.jpg
1bcb038d81e5dbb58783978124cd849e
63291893949206b521b9a3c6773067121ca54a61
83873 F20101109_AAAXKW stokke_j_Page_092.jpg
5a84e0d1064aa3e3b92469268955f3dd
44ff90954014a51282905652ea1d815c5b0d1bab
24261 F20101109_AAAXLK stokke_j_Page_113.jpg
ac0ebf58fd3200ac4e3d7121f7192ca3
108f91ecf3d664be787f7df4108df53d684dc986
89056 F20101109_AAAXKX stokke_j_Page_093.jpg
3efb42c43d6704826f96296443669e75
bb5ab84e999e889f55d53a8c12def4c02ea5aedd
1051984 F20101109_AAAXMA stokke_j_Page_011.jp2
87fc9765cf1947ba5e88846ea13fb1c1
4693a52d40df7c58574295e8da5bfd6ac9b59f7f
25488 F20101109_AAAXLL stokke_j_Page_115.jpg
00430e4a0ea1f8677d64ac734ce1d9dc
c7cb0ff4f6f551433164e85f06fe5dd68433ea97
88639 F20101109_AAAXKY stokke_j_Page_094.jpg
6349dc3e14638300587168689ae87211
2a0229e450fb1a5396e98c7dd6db8ad9141d3ba4
885251 F20101109_AAAXMB stokke_j_Page_012.jp2
c661c68e87c8053d23482fdc014e70e4
277497bf2920e8af94fd348d86c804dbc3028fde
33055 F20101109_AAAXLM stokke_j_Page_116.jpg
c5a50587843be42f98775fd9550ea9b1
14b7a6ec9efea12705ccd55e545c5b5702a98131
63350 F20101109_AAAXKZ stokke_j_Page_095.jpg
b15d40f3de4f14e7f8dd8e33c539d569
18ec39f7534eb67780261db8057085e0f9f2ed0b
83729 F20101109_AAAXLN stokke_j_Page_117.jpg
8ba65dbdd32a1d49c87c62eeb16fac8e
a8404dbd41496963adbca27dc65b678865cd9a41
1051969 F20101109_AAAXMC stokke_j_Page_013.jp2
f5da3116bf8845f153e35e08952ba430
43eec198db052a090533bd714462142f2c6dc57b
76788 F20101109_AAAXLO stokke_j_Page_120.jpg
73042c482e8d059d07258df78feeaacd
142630f3f7b742c2b7a2863926ee9215faee6aa9
1051970 F20101109_AAAXMD stokke_j_Page_015.jp2
d829c0b8e21bcceed0a738d75a833f9e
f852241132086429e38c8c78088f097a3ee34864
95543 F20101109_AAAXLP stokke_j_Page_121.jpg
a59c98d15fdd50423effc62f0611eec4
e742fe47d111e9c8ba75985ba2b5e1c1597045d4
1051908 F20101109_AAAXME stokke_j_Page_017.jp2
d4e32e355f5b6e04def57b5a5aceb158
b911f85df096bbcd2753d0ca6a9ff867b04a6fdc
99989 F20101109_AAAXLQ stokke_j_Page_122.jpg
78cd91a1d9c7faf07960acd218ff7c93
095ed02421825211282beef58cfdf3b867204b9d
907012 F20101109_AAAXMF stokke_j_Page_019.jp2
d870f105a8855a3a969eb5ea41c74aa6
92441829740445a442ebd2b558c60fcaedc7d7b4
103112 F20101109_AAAXLR stokke_j_Page_123.jpg
c1aa528f776af6cb073eb2adba68bb27
4c0ecaee226ad2449d0a4a01c7090fc8af49b6f2
81243 F20101109_AAAXMG stokke_j_Page_020.jp2
dc025b5a425418aa8d42a5b8a55a56ec
47fe2051440e5c833fc9d9a93274be52cfc7daf6
92329 F20101109_AAAXLS stokke_j_Page_126.jpg
d71752f4731c036a3aa6d76ab03a9391
ee0b95e87bc1af12227cae0cf5da26c818fd3a0b
1051985 F20101109_AAAXMH stokke_j_Page_021.jp2
8dfd95a8d6a7f7943b37632603ad0a8e
732739e3e82a7d1ba5dac2c70cb795e59db0b0ec
32373 F20101109_AAAXLT stokke_j_Page_127.jpg
b451db85c251eaddabd5bc7a641846d8
cb60b7fad1410076d372a15eb2b7dc294e1f58e8
1051924 F20101109_AAAXMI stokke_j_Page_027.jp2
8333f6bca8812fa1a35734d3ba8ede03
c84c78e447fd79c144100408b8dace85391c4013
275933 F20101109_AAAXLU stokke_j_Page_001.jp2
2e6a8431cfff5fda75b58afc6305d971
b2ebf01e4740c21333720c8ed417c364ba9fa759
F20101109_AAAXMJ stokke_j_Page_029.jp2
c1b8787c356b627e27515bc96aa0f307
2f00e6672c5aa5d411a26d217cfc686ade06a5e1
30157 F20101109_AAAXLV stokke_j_Page_003.jp2
5b7787331af335247200e1311e7afc96
d68c35befbd866949294af3656015178114bf044
1051982 F20101109_AAAXMK stokke_j_Page_030.jp2
b47af68f0e65f8502f295d02e55e33e2
580fe5f5517de2fae470141009771b6ba9229a13
1051960 F20101109_AAAXLW stokke_j_Page_006.jp2
bed1260ad2de2c38f276d0abba10331c
10fbbdcd50740ba3f6859258b6448271d053b8d0
1051894 F20101109_AAAXNA stokke_j_Page_057.jp2
92ed9269772ef2a06188a2a8d78fdc9d
aaf701cadd0e8535aa8eb124b73f6333181d297b
1051974 F20101109_AAAXML stokke_j_Page_031.jp2
3a7cce6feb3a439bbb13f2b498d5b6aa
5ee1a4df6ff12103ecf2abc199d8108370ebc9d5
1051983 F20101109_AAAXLX stokke_j_Page_007.jp2
d536c550af2ed16dd06df625d01756bd
52a4c0a51b3ec650b34a3aa1da688e9db0075b4e
1051986 F20101109_AAAXNB stokke_j_Page_058.jp2
87e6808aa79ce057f67d43767aa91c36
9369282810d4bbd809fce8cf44dbabeb86965be7
1051914 F20101109_AAAXMM stokke_j_Page_034.jp2
cd029bed1d01b6d941e04ebacdab41cf
0efd19c84c97dea7722682f1b99ece24e1464401
1051978 F20101109_AAAXLY stokke_j_Page_008.jp2
7ae17f882935b2169300df531df91f39
19e1e2013d280031b01d42d43db280b24c401609
1051968 F20101109_AAAXNC stokke_j_Page_061.jp2
f28c3ff034ad3255f90cd8fd9d57ba5d
2132d47aed26df2387c19b793dc196c210bbd168
1051980 F20101109_AAAXMN stokke_j_Page_036.jp2
2216c1137772b81854701cfa3c313fd3
42e8d4f91a79cc63999b2a5c8aec049b249016a9
1051975 F20101109_AAAXLZ stokke_j_Page_009.jp2
bd8d5981c7cf42a3a8f6cdba8e0d4f1f
c4223678c2696b95ec3e32a5bd1797f9ffdcd03e
1051951 F20101109_AAAXMO stokke_j_Page_038.jp2
c2311d35090aec4e9917afc21c22c44a
682acfb8bff16b31fe43aea73459ccdb6ad8753c
F20101109_AAAXND stokke_j_Page_063.jp2
8c04a3c8e27ad648060ae5ec5ff99966
d77b9c78a29f269d62f9369c033a3640204e473d
941981 F20101109_AAAXMP stokke_j_Page_040.jp2
e15b0d5223f90858a59fb7ee83a4b502
7d3a1471fbd9e02d592139d47c15cd5699648371
F20101109_AAAXNE stokke_j_Page_064.jp2
059252e8699522a86e2d9157ba271ae9
71a1cc5b86f73bcf9fa38c3fad6e589095f88ced
F20101109_AAAXMQ stokke_j_Page_041.jp2
022af62283d14eab0a0778333e4ff1dd
b004d756c721a641fe6d584eab833d142c7127c4
F20101109_AAAXNF stokke_j_Page_066.jp2
2968aab29fe2bbf79dc14ba31a6d1da8
f60518f662eaab453cc77b300adcf99bf4588aaa
1051967 F20101109_AAAXMR stokke_j_Page_045.jp2
14988936f5034021ab8f8d3a1963c6ec
3bc998c5cc0b7429ff594197dba3f4fccb19b118
F20101109_AAAXNG stokke_j_Page_067.jp2
c7197e0a4677d0b7c62c81e2e9b07ab4
f09da5bf8be3905582b3838328ae635e68ddcdf2
1051937 F20101109_AAAXMS stokke_j_Page_046.jp2
015ba89933a4d8c42023b25f31ac30c7
e21942aed30e84c06a1fc41162effe161ad89d1d
527159 F20101109_AAAXNH stokke_j_Page_068.jp2
ed241ac9dc4b4f59415b1078f6ecf3fb
eb6c56e01ba368ac17dec4929511eb7e07e47b30
F20101109_AAAXMT stokke_j_Page_047.jp2
77660e557a139f2ca96c9a1cee3718bf
3930ba4f6513c7806a74bdc2ae0320c421a030b6
56800 F20101109_AAAXNI stokke_j_Page_069.jp2
43de81161ba64950ccdb58977a41ea4a
0a461b2713b5d826c07f47a346412fe3c6802f16
1051926 F20101109_AAAXMU stokke_j_Page_048.jp2
56ef40f5bcf961a7d88b029b4c8bc7a3
2348d773ad6f73cc1da0aa8fb850d53219818ed4
44892 F20101109_AAAXNJ stokke_j_Page_070.jp2
9e7fd20260026679d16a62c09864df10
f650d502757893adf8716a8801658e4b41899c46
503635 F20101109_AAAXMV stokke_j_Page_050.jp2
cfa2233afb843edd0c96130e0bb33c0a
f35d3f289b0c66aadf5b8dc9c6e02cdec7921c2e
205529 F20101109_AAAXNK stokke_j_Page_072.jp2
0a4a50f8eeb06173f5e14fdd334fa8ba
e6e1fdaa491d0de5dcd0e608c7780f7541217b47
1051959 F20101109_AAAXMW stokke_j_Page_052.jp2
acaf7a3de2ada0397a104fda28e5d36d
bd854d3e73c6a1ee30b1ce6e329ff70d279b27ad
389376 F20101109_AAAXNL stokke_j_Page_073.jp2
2a452cd3f51f25a6ca488d8dd09ba350
ac0b7b599a14de3f8a7bff90d7c4a7fafe4c66b8
1051939 F20101109_AAAXMX stokke_j_Page_053.jp2
049fb4a20bf1924e6187750f78e76f9f
50fc3632b5ed427c9888d5676cc5a5c46e8dc97e
F20101109_AAAXOA stokke_j_Page_100.jp2
7bbfa12ebb98d38f89169ed896007b37
2c17a42a451014c6eeb38dae4723c5713a1d35f9
384357 F20101109_AAAXNM stokke_j_Page_074.jp2
3b81bf0e0a1e169f27263094bfce7186
aa51706d0749feaa016834061028d993e1ea116a
1051957 F20101109_AAAXMY stokke_j_Page_054.jp2
d917e17ba069567a964b4e57dd53333c
5905517caa3b05fc87d6ebaff1b296af0bdc3daf
1051966 F20101109_AAAXOB stokke_j_Page_102.jp2
a3107ad0a673b9fc48b2df7407e3cb36
6291713b5d0973edae201ff051a260b455a00c70
339638 F20101109_AAAXNN stokke_j_Page_075.jp2
9c82ed2421cb9bcf49f77cc617bc0dee
5f934a38fa6a261ee19ecf25fe70f2c7c92ec1f6
1051973 F20101109_AAAXMZ stokke_j_Page_055.jp2
c0ebfd06b8dc0175b330b90b801ab97d
54dfdcbcbd9b69d796f43608d7e67b8473f0d5f3
1051962 F20101109_AAAXOC stokke_j_Page_104.jp2
f93114223bd01c16a3ad9fd13a5d12a3
a946b16a82782c351839b1d6b3935cd1dff85e17
43926 F20101109_AAAXNO stokke_j_Page_077.jp2
890b2afba8963629c4d682bf55a2c230
003b50d1d7012bf7556cd37a765f3251da800f88
1051954 F20101109_AAAXOD stokke_j_Page_106.jp2
a270dc0ce5b6ecc2d2135dc4be3ed1d9
bda3e6eade7548e89b0de26b7307c26b63f5563c
F20101109_AAAXNP stokke_j_Page_080.jp2
84cd4f50f826be2cca3c15f284d0aec5
95f1a70826ba6a8f7959bf59d364d0ad357a917b
F20101109_AAAXNQ stokke_j_Page_083.jp2
27453f9379c868aa1166b4e3edcb3786
208cc415ec7ec03c62dd094afea4f190fe79ea7f
1051953 F20101109_AAAXOE stokke_j_Page_107.jp2
518c4e8e10e7a99a2552261e4aa16d4c
db8e51fd0408aa345924d7443f712a3d3db7d443
471393 F20101109_AAAXNR stokke_j_Page_086.jp2
e1fa5a0fe4df0482e64c2b77ed65c771
5e4bb45b21b961ad4f164e4dc75d5e84b66d51a0
1051945 F20101109_AAAXOF stokke_j_Page_108.jp2
072b36afe16313590d2d050e803fbe7a
15ae501f99534b300b4195a29bc2081a5dc400b1
857814 F20101109_AAAXOG stokke_j_Page_111.jp2
bd5d0a3d2fca3dd1e1da7843251a8257
483aca0a56ea856ef0693da3c24f9a69e2f1b913
496153 F20101109_AAAXNS stokke_j_Page_089.jp2
031e0fc2b0cd5fee5117917bb1f801fd
1b2e5565cf9e72bc899bac732ff1a577eadbca76
370816 F20101109_AAAXOH stokke_j_Page_112.jp2
ab3e394131c41fb56ada3bfc859421b0
09262219ce312a3da16ae52ec5b7e84d88c19183
1051956 F20101109_AAAXNT stokke_j_Page_090.jp2
801a5632f96af8568e8c4602ffea54bb
cd4ab3beeec1559013683896ebc4e81bfa5074b4
420924 F20101109_AAAXOI stokke_j_Page_114.jp2
06e3d75a25a733dc1c7bcd88591786a0
9faad15732d6ff5b51cd620c005429aba16e384b
1051979 F20101109_AAAXNU stokke_j_Page_091.jp2
c88e7d172c266221538dc205838703ed
64b49a1085630a433715a1bc9a2af0f00eb5ce40
220865 F20101109_AAAXOJ stokke_j_Page_115.jp2
a209be28da15acfb3d3d42f3d9d26db3
2060338efc07971b76c6e588a963a3b829b90dea
1051949 F20101109_AAAXNV stokke_j_Page_092.jp2
01d9865bcd150b0aa0d76cc28a97ab60
5d8aca0c4d07f618f2c5552c7ad33a55086e13f5
1051965 F20101109_AAAXOK stokke_j_Page_117.jp2
387b9d8642cb5b8425656a02ddd21f74
ebaa28b46522270daeea1c7ea217bf267bdb0fdf
1051981 F20101109_AAAXNW stokke_j_Page_094.jp2
27f23c9af9cde5cfb85f459c825d7d71
492573ab95f58937139ff2be3a2cb36789a122e9
F20101109_AAAXPA stokke_j_Page_014.tif
7a6ededd077e804bb91225dd2293a491
9908e5fd3d81a73a5809ef57e7c3ab8718e9c275
F20101109_AAAXOL stokke_j_Page_118.jp2
0a7f766ac4762af98dbb4ab67e386a04
e5111500384a115a6184a02dd12b2867f58560db
F20101109_AAAXNX stokke_j_Page_096.jp2
5a5c2daf81cc3949b68cf565e655a1f6
70ea301929f769f29b62bf8d7e267140f80295cc
F20101109_AAAXPB stokke_j_Page_015.tif
3e2a6397444084eef07fcfcad1654750
8bc25b10aebda87e639ca152c72c308e5a5f97f0
F20101109_AAAXOM stokke_j_Page_119.jp2
74759cd69e069ddf48770b97fa02bb7e
affa9260b14c2a595a5dbe41cc48b0c5e9f7e9ad
509311 F20101109_AAAXNY stokke_j_Page_098.jp2
6879fa9290f0fe6ab1567cd6aa4eaf78
c2d7b237e606bfe2db131a2ba3ea917d5a9fff33
F20101109_AAAXPC stokke_j_Page_016.tif
35e3155845c356b1c5f493b1927e4f64
d195313edc462ff2bddc91b94a22e2cdd7592557
F20101109_AAAXON stokke_j_Page_120.jp2
7425b54a45604a22ffc077101fd3d34e
944e4bee10c4fca880d780842497e2e6160bcda3
494068 F20101109_AAAXNZ stokke_j_Page_099.jp2
29566ba437391a4982afe865844d99a9
1d8f0b86a1289dfec6bf2413a6ba07b6edfe2ef9
F20101109_AAAXPD stokke_j_Page_017.tif
86c1117e2db25cf3610391e9a605b719
20275497b6e0f6e0a41b0ba392a14024fafd3a01
F20101109_AAAXOO stokke_j_Page_121.jp2
0a2812ae00eaf834c0a704d79af59933
b3be3eb383d89b79a70ac18248f95a4614518743
1053954 F20101109_AAAXPE stokke_j_Page_020.tif
a3a4d5c55b22691865b462ca715efe2d
3e87c8da8126377ded4f0c8565a0b9dd05b01801
1051977 F20101109_AAAXOP stokke_j_Page_124.jp2
492c037b504d0dd848cc04eeee02bbe3
2bd879a2355c6cd97a7fa9e7c174da1603520c84
F20101109_AAAXOQ stokke_j_Page_125.jp2
2a13515999f48d2b3d27cfd8fa2235f1
50892472174700066b6558c80581211720136f1f
F20101109_AAAXPF stokke_j_Page_021.tif
4112e7c3c2255583b52bce8a90069dcb
61d7b4bf879e45e0cba02c6e450dc966735c28e8
1051904 F20101109_AAAXOR stokke_j_Page_126.jp2
b3893aedf61b68159bbd7fc122cf1d34
b76f170add5caa753adfc4bb55a1087c58c28acf
F20101109_AAAXPG stokke_j_Page_023.tif
5c26b4d3e65666c69f5cfcbfbc7521c3
adccdd750c60fa3b0c7b49a4d5eba5af5c64a8c3
F20101109_AAAXOS stokke_j_Page_004.tif
1357c15e3452e9b33e171fc872e28ae8
a7daee508fbd7312b169ab8c34cede338d45bc21
F20101109_AAAXPH stokke_j_Page_025.tif
0e06f92da4ad18eff59ee4f952801b51
b93e5bb97fa9ad819f7d90be30b2628ecf8042bf
F20101109_AAAXOT stokke_j_Page_005.tif
3ceb5fc5d5f151c9f02d10ac76472d44
ba10f23e9a969b7e1a654d1706e61451e6e8222c
F20101109_AAAXPI stokke_j_Page_026.tif
75ab0a3d429bad09824d77fb0e866447
8a72170b96f5b17f133b83294b01cf2740ff9c34
F20101109_AAAXOU stokke_j_Page_006.tif
b4eeac8e16ad10ddf6fa57ccad64ebca
6c2973a5edfed9c00c0f30696e16612f97455316
F20101109_AAAXPJ stokke_j_Page_027.tif
5da7b3a2e3b17b5583cb8b5a07da1418
efd343979d151ad684c134dc23e830eaa96e2855
F20101109_AAAXOV stokke_j_Page_007.tif
60e11cf9186d9225e3a3de82b87d5266
e8456ae45fbd55a4f3067f6d2af21e35d567254d
F20101109_AAAXPK stokke_j_Page_028.tif
922abace576e583b472217c324db13c2
4748b07f78be48d131d4fc2a6d1a10ac352d6450
F20101109_AAAXOW stokke_j_Page_008.tif
d3b89b902fef0d10796865948eb2e48d
2b29656d4f9fadbe110085e2409511dd7cc6ad12
F20101109_AAAXPL stokke_j_Page_029.tif
3bd1f49159b87250dcafc76fc135b076
dd943e556ef71a23801813861d973ac6546adce0
F20101109_AAAXOX stokke_j_Page_010.tif
6f16c63edafe14f8f24dfb4b49645cc4
05ef732315affd9d8eaf0bb0fe7cb3f36b98baea
F20101109_AAAXQA stokke_j_Page_053.tif
32271aa5a1f19952a30c8f7dd8b53056
2203cc3ebeb1d93b7705422a9561184478f5188c
F20101109_AAAXPM stokke_j_Page_030.tif
54a37279353eb9060c321d2a3c986070
4fd29715423104891e3041bb22a290512c6f7a3a
F20101109_AAAXOY stokke_j_Page_012.tif
e00c2086071e9f7bc35c44168f850b0b
cdb22c793bf014e91b24a30ff442f5e76d874021
F20101109_AAAXQB stokke_j_Page_056.tif
4103895bcd9343dc9da2e6470c14bdf3
c401e0aecc8a0e0782f3d2dc0377cc86688f0770
F20101109_AAAXPN stokke_j_Page_031.tif
e19c492acdbb5aa4b84d956ff103cc77
b0286059a62ac2c678cd680f9def98ac69a35096
F20101109_AAAXOZ stokke_j_Page_013.tif
177f5e78216f0ead0157fdc8f0d7d67f
61ff55eabd9297cbaf89afb5cb5d56c763d38b8b
F20101109_AAAXQC stokke_j_Page_060.tif
a8ef5ad843c3784d080a9bddbdbb5d2e
06d0bb79066cd1531a7170006f9a9b21d23ad00e
F20101109_AAAXPO stokke_j_Page_034.tif
bf9d9654e63d1bf9e4e83bd4afac3729
e418da38770468836cd420dca7c805852b2a33d8
F20101109_AAAXQD stokke_j_Page_062.tif
fdd11739b2935a45da3da780c4faf076
6cb8636bbdae6788c58ddbc62eca0905706ad54d
F20101109_AAAXPP stokke_j_Page_036.tif
5039b08fa3044433b4cd729b8bf5c71d
b3f8d4f96998405b6c532d5c112cdb807a83fff6
F20101109_AAAXQE stokke_j_Page_064.tif
b39e8df0a3902bc3930ea2351670b182
d32aa3ca694e01a6e178e950e6f7c331e94637b0
F20101109_AAAXPQ stokke_j_Page_038.tif
defa4c357c0b7840b9073e5568500938
818eb3a72d4ed5242884a229519bb646170a7757
F20101109_AAAXQF stokke_j_Page_066.tif
517d1bf8dc30034e2fe7f28a9be719d6
4f670ba43a4b12ad4dc8ae0c69cfb80f5691b782
F20101109_AAAXPR stokke_j_Page_039.tif
9cbae02780a1aac769e1a7eb29c3c998
28546331d649d3677426ee4062fad7217064ba77
F20101109_AAAXPS stokke_j_Page_041.tif
4a4342a635820a77552448fc2a4c0e1e
e5044a8b35b4b099a2302d41dea056d2d23df3b7
F20101109_AAAXQG stokke_j_Page_067.tif
6a69c36ef5135424dc7d9ce14da62295
25dce8fc9591b17d6c15d10aa06fa7861c527660
F20101109_AAAXPT stokke_j_Page_042.tif
1987cd7fa74eb7f41c2067c94a6d429c
f143d6ecbaa2c12d5b039f654afa2244542e9c02
F20101109_AAAXQH stokke_j_Page_068.tif
369d6f17bfe0567944fe2d8ebbd587c0
6ab1e01844951f3b65bf22351eb67c57fed38972
F20101109_AAAXPU stokke_j_Page_044.tif
bbc58872d4312f96f2d40d8df2af524e
32b3e3a8b6dff217886693de33836e92c48fec6f
F20101109_AAAXQI stokke_j_Page_069.tif
a16fa372da7e515fcf5261f90e92a01b
ac81ee3c1ecd97baf1454cdda952a1a05de766f8
F20101109_AAAXPV stokke_j_Page_045.tif
34459c9d734910ca1a5f708180136624
66c2473d1d64db6f9711f2af489ee8a46255b55f
F20101109_AAAXQJ stokke_j_Page_070.tif
31d3624a25fb566a40085fcf0a6539f4
63a84eeacb03943ac031e4d78ddb061f91035ffd
F20101109_AAAXPW stokke_j_Page_049.tif
c76e1ae019718699541f4a6d5bdf42f5
5db889cf7dc0db44b05196f9a834a52f8c59322b
F20101109_AAAXQK stokke_j_Page_071.tif
ebe78c00999831d230680223b40c8928
aa82b41a97d4f8b460928a03a5feae0d6e15a93d
F20101109_AAAXPX stokke_j_Page_050.tif
80c0e21bd04ba6c76d05556e11f22fae
98945db09bb984a74b80386022dde428cf84f870
F20101109_AAAXRA stokke_j_Page_096.tif
18e448a6d823c64c1fcc5a640d80e82f
04cec5b2c3bab2132e3a2d23eb826adef32f5014
F20101109_AAAXQL stokke_j_Page_072.tif
220078218e83b753ca935d68cce30146
900636fa219345dda792cbe50a9344579922f480
F20101109_AAAXPY stokke_j_Page_051.tif
2b39a17645564b9012adb7bed2c6e8ee
de52d93319464acca65c2cf6e79c803cdce51e33
F20101109_AAAXRB stokke_j_Page_099.tif
be5f613ddf78f1787e5161effb29fc61
c21f0820283df46c6ade847c40e6e2b0837d18f2
F20101109_AAAXQM stokke_j_Page_073.tif
3729ea03e3843458c36907e6cc9bff58
1b76b807fddc886d35b8494c8356fd7a9531754b
F20101109_AAAXPZ stokke_j_Page_052.tif
891dd4948c1a4559840bc87e6fd49c6e
bce4b9c8619d0af2b8366ad6a2535b49827d4467
F20101109_AAAXRC stokke_j_Page_100.tif
2df845cc1c9578c4e4de525ae49400b5
fefdae2b5c3cd790dc70bef0abc52f3a88ca2c56
F20101109_AAAXQN stokke_j_Page_074.tif
c9ffe1736ecad35ef8781ff8b1a6b18a
862b0e0e7c2315f719b553bda2b6e8830716b304
F20101109_AAAXRD stokke_j_Page_101.tif
104d0e94d155e91936d4919fd8492bfd
491aaec4e221e3054be984183ab30d4036b78e8e
F20101109_AAAXQO stokke_j_Page_076.tif
704a1a9c6efd35a4d7689ea8f7e0945e
ec344a364c913ffdabf0aca5421bf0860786bc55
F20101109_AAAXRE stokke_j_Page_103.tif
8355e9b7cdc0786af349568371c75548
e2318baa33df50f6c72ff94b4ad861ce810215a3
F20101109_AAAXQP stokke_j_Page_077.tif
973b5930783209cb80bd4d65697b8585
448401293bde8b2854261fdb5da23c3518913049
F20101109_AAAXRF stokke_j_Page_104.tif
a3d60c1c724209c34684e7eec680b040
d681429a0ffc37be1c52a55474828d012c6a1e0a
F20101109_AAAXQQ stokke_j_Page_078.tif
3cbc58bc812134b9a0ca45ce4a080840
ca3f88de69e98c051a38c44e3ad8549d650090bd
F20101109_AAAXRG stokke_j_Page_105.tif
9b34e9c500a0ee9c9cf2b4e0aa393304
f0950f40cbca757b0b5681e9ab02fb2b9f7ddddd
F20101109_AAAXQR stokke_j_Page_079.tif
7b06814cbaa972552974d0b3e234dc01
8c5fe7451114d3de53c005b052dd85818b647713
F20101109_AAAXQS stokke_j_Page_080.tif
4616acb0e27b41cf99cc9b3094afa0b8
64a89682c8811f0ef8044f7a04ac1c4fbc9e4914
F20101109_AAAXRH stokke_j_Page_111.tif
801187d69190b502506868dbda696212
2ba475da6c7a85c55966480b15572d8dc4b0c084
F20101109_AAAXQT stokke_j_Page_086.tif
537a9e903a533e85f7470ffc2a4079d8
dda3ceae57e58d64a2ec37423de354e3bf62dda4
F20101109_AAAXRI stokke_j_Page_114.tif
b0b3d3f24fc1d299aa51a33b3c640cbc
58b8098fee1316aab8fbada8a931a28a23d4e719
F20101109_AAAXQU stokke_j_Page_087.tif
989f49a8cd860ffe42ad23b1733b9a79
091cd5c82aa630ea94661f6c472d372b8a9a8d3b
F20101109_AAAXRJ stokke_j_Page_116.tif
033804415e6290529c48b670dffb5dc1
84a101880d52d96a020a9bc096521dfa5ed2ff02
F20101109_AAAXQV stokke_j_Page_088.tif
2a39eda78f7e2e6b977183c6b3ab9731
6143ee88f77479e995ede8cce3ea3aec75a10053
F20101109_AAAXRK stokke_j_Page_117.tif
f62018c5a069133b5c685ec48ee30224
65d21b4dfd9fe9d90d1130c43226816bb81dfd6b
F20101109_AAAXQW stokke_j_Page_090.tif
59b3edd626b2f7f1cc33a6df8425e7ea
42c1a5c92b112a62def00b4705ad0a804b81ef7e
51697 F20101109_AAAXSA stokke_j_Page_017.pro
bdff8974735fbcf384f5f3971a65c2b5
a0b040b816468acb2966cb0a9a4af7952def6d0a
F20101109_AAAXRL stokke_j_Page_118.tif
312c7b9161743c07bd147bd189bade64
2d6217d70de623f3baef83b482026964b4a8896f
F20101109_AAAXQX stokke_j_Page_091.tif
fafb0873e783b2f292b17fb0e43f79d7
cb992971de03fc98ef20b01c315f9a97b6084890
41008 F20101109_AAAXSB stokke_j_Page_019.pro
808e20bcbda98d084b231797ada13646
cc28fde38f0088120892d6203ef1df484c20ab6d
F20101109_AAAXRM stokke_j_Page_120.tif
0a9132e4fbf125a969485a050a0e8de0
1ce841e09b2f58e03947a6c3fec719fb1080804e
F20101109_AAAXQY stokke_j_Page_092.tif
7b88e6539643f21d1453e4f651e0e258
415e82fcb9e0df0e23bad109714580c3e39a11ac
41806 F20101109_AAAXSC stokke_j_Page_020.pro
32dd039796ffb7689de5b7b36f9a0c6e
7252f2b1aeb655fb5627591a8c127c17b6096df3
F20101109_AAAXRN stokke_j_Page_121.tif
0de548a5a9ff0ddb55230b286d538058
4f4bf598aa94eeecf4290e3c9480b0b98485677a
F20101109_AAAXQZ stokke_j_Page_093.tif
2209356296ef3f37d9bcab4ac662851a
f0cc1f23db6fab55cbb4af490bc8a5ddbdaced98
54784 F20101109_AAAXSD stokke_j_Page_022.pro
7f7168e059fd0156f8938d6c7eb995f4
0a126354e1b5715d6f362df0765a49b0ab06db99
F20101109_AAAXRO stokke_j_Page_122.tif
d6b9f999599b3ffcd82dd1f2de447fd7
1aa484c8abcc2347017ac6c93a57d17cefa12bee
51254 F20101109_AAAXSE stokke_j_Page_024.pro
3476abab44f2dab606c58b60156b23aa
11bcf9621c0a6175bac45ea2909fa7af64281bb8
F20101109_AAAXRP stokke_j_Page_123.tif
7ad0710419fd7d6d1e4deee44b51aed1
c04e0dfd55f8df52f19f2c95a8319b87d24c9820
54799 F20101109_AAAXSF stokke_j_Page_027.pro
a0cb48ba5d5b782e09e9240d908f84cf
83ea2022fd9c8b8927207f07b1e0e45fc380a30a
F20101109_AAAXRQ stokke_j_Page_127.tif
1d90444443941a2304c6e935d44562d2
e8da209b35e1fa155540e74d4b0e982a5270f958
54505 F20101109_AAAXSG stokke_j_Page_028.pro
34f4a81cbf2ebe669b065aeb31fd3da9
80e125bbda53f5fde10725330f68a8c4ed709992
8907 F20101109_AAAXRR stokke_j_Page_001.pro
6e011edfeaa1392b74bd955d5fda1568
b56c46f6b3e6fa2b6de87d929adf718ff7d6db85
57567 F20101109_AAAXSH stokke_j_Page_029.pro
1ae3581eee65b51056c4100506d90298
ee2a5eec34c6bb35c48dda3d661bce211a85df8a
1066 F20101109_AAAXRS stokke_j_Page_003.pro
702895adafbe89987430aac03de3b1ee
dfc6faf5e00993f5bd5978f5da8e42a62702858e
37670 F20101109_AAAXRT stokke_j_Page_004.pro
2a6eab01ada9393c7de3ef4be81e2ac2
8d2a8c2b02fe49f8ad57df0fc08de92fa1d15d80
57144 F20101109_AAAXSI stokke_j_Page_030.pro
6c0535f13d372cbd808e852f95419ad1
389db26bca7dd30e3033dbf93f7fd717c9634b48
91798 F20101109_AAAXRU stokke_j_Page_005.pro
c1421c0335673842287ea97c6c0fe764
f702c52d38d6a514cd8b62627a16c294b2680f5d
50869 F20101109_AAAXSJ stokke_j_Page_031.pro
f0d1605364faf1ac6b1a96f16f4dd9f9
f6ffeb811c70ae893cc7b6000737142aa6e58fee
74737 F20101109_AAAXRV stokke_j_Page_008.pro
9ccf70882f3cc5131e6623253c069ccc
49de41b4230632708b8c3b8f1ecf74567a1f6ead
55924 F20101109_AAAXSK stokke_j_Page_032.pro
d1fbae79244db974c416666318294361
14c6d1f1cbc5b7f93f3a9c90c623887bdfbbe0d0
78577 F20101109_AAAXRW stokke_j_Page_009.pro
657d2638f5462948fdbbb3e8b9a6bba4
647f2e31ebc3957dadc5af68c621afab721dd00f
51209 F20101109_AAAXSL stokke_j_Page_033.pro
d5723ce70a612bd723547608eaae14b0
d1ec23b0cd04e3ec946abc5b50f88643069404fb
31080 F20101109_AAAXRX stokke_j_Page_010.pro
ec6ffdbb5328dcefaaf4360cfd80d10a
383c85771901407288a21fa8daeb7649c0853fe3
54561 F20101109_AAAXTA stokke_j_Page_057.pro
48316e40994b5a9c2a836b45bc99679a
b3f9e1284b91254660a8532d2a112e4ba3a08b7d
50291 F20101109_AAAXSM stokke_j_Page_034.pro
5020464197b11b0a959bc5bda51eba1c
2e5294b64416dc10f01b27f5f14eed42a1cc08d0
50126 F20101109_AAAXRY stokke_j_Page_011.pro
32e94b50c5e5b523c148a8b807674d25
732245f4262d3f345a32350ac4909c2e7dca2a48
51811 F20101109_AAAXTB stokke_j_Page_059.pro
b289c20a0aa60d52b0bd51a5beeb552f
2037a14f85ac958903710de69c284e9d770c15a7
21334 F20101109_AAAXSN stokke_j_Page_035.pro
1e35428b931f62166114e1fc364c0e51
3ced17a6aaeb710fe709f168656893504ee4ba62
39163 F20101109_AAAXRZ stokke_j_Page_012.pro
81e5aeb932ade003282341d9cca139a8
37c890836ddbfd175a59e6d0d23d4a46bdbcd52e
53577 F20101109_AAAXTC stokke_j_Page_061.pro
b9018e249a5d5bf09335c142ee4a93f9
de4b06f8f44f33c56938a6b6a943dab3f11496a4
54760 F20101109_AAAXSO stokke_j_Page_037.pro
dc0695db5705d8dd5b702abd9ffe3670
829d974b3f8112c0ea733a3fbbee8a4cd536e0e1
51719 F20101109_AAAXTD stokke_j_Page_062.pro
a82aa3001c1bf22095f025e2856afd28
05debca08d06bb91218efb7212115ae835297f68
53008 F20101109_AAAXTE stokke_j_Page_063.pro
5488b3fc1e1191523d6cf1fa6dba9d29
aa560d895bc619b5d1cea8e22027259b9ccfd0a9
48253 F20101109_AAAXSP stokke_j_Page_038.pro
0ee7aeb5bfe155bc8a053764e6c5af98
81cdf7f528d75a9784dc9ed871365cc2c57613be
53341 F20101109_AAAXTF stokke_j_Page_065.pro
b382200fa16bca112f35f2de95b61003
c7ac5e58d17d415b6997dce6942d6b319f1d6267
43703 F20101109_AAAXSQ stokke_j_Page_040.pro
98048ca77aeb949031cd6e47c7661236
b3fed6f82d82a7c1e70b88ef1334515e777cc949
55781 F20101109_AAAXTG stokke_j_Page_067.pro
d8ec1d96070ddbe6e2a4cd707c759c40
48dc2f3fc4514027550e84b85e73aa42e45fff87
46582 F20101109_AAAXSR stokke_j_Page_042.pro
c488fe9bc67dc766d7b9398d004537f8
3c2763e037ca9054eb38a1c477a4dffc24c76e5f
11004 F20101109_AAAXTH stokke_j_Page_071.pro
9feb9ea4feeeccdaeb9d77e176a1faa5
91e684abf4032fa422aeb0190d26781e01c13cdc
36357 F20101109_AAAXSS stokke_j_Page_043.pro
a3d8c7af8cc414cbcb086294031b9ce6
e524b737ad8fedb4028ddc48167e845c139d0a13
4720 F20101109_AAAXTI stokke_j_Page_072.pro
8bb900647d6ae2801ded54e41ea819ee
35860f2e04584a316ab370bf9ff7dcad24b3ae79
12115 F20101109_AAAXST stokke_j_Page_044.pro
97b2937937c83136ddc7c9a13d32ad56
054ff2738927a97715abb0af028d7f9ef8251373
54765 F20101109_AAAXSU stokke_j_Page_045.pro
1e5ae4bd9e1a1d860b3b3b8dbabacfe0
195343195325f3f905ca45e9f3031437e8a8e890
14245 F20101109_AAAXTJ stokke_j_Page_073.pro
e708e2d3f6963b50b548a2cbc796c0bd
8ac689753650f6c507dee2d339c6113d76076b9c
54493 F20101109_AAAXSV stokke_j_Page_046.pro
96cf4fd22d25552631ea2e8cb3dba728
b1015c56607f5221b9dee05593ef5ffbdd8d9b4c
12538 F20101109_AAAXTK stokke_j_Page_074.pro
44dafca01d9eb32d8a7fee657edec29d
ee31ae68ef7e2a36880a894bbf5a062121cb4058
55751 F20101109_AAAXSW stokke_j_Page_047.pro
59bc423d20405470b347d66a981c292e
4a25a0ee4420587c1fcb78facdb6c02bb42145ca
50507 F20101109_AAAXUA stokke_j_Page_107.pro
e75b8402ed254ece895d41340528db3b
d371528ce5d1159602a8c5af4a60f82b5c5561dd
64519 F20101109_AAAXTL stokke_j_Page_080.pro
3ceef3a1d98b9312d8c0952f55811606
ec138c5eddc9fadfd030d47c0f204e2b8e9e0f0d
22886 F20101109_AAAXSX stokke_j_Page_049.pro
8cde181b6b6c1252c0c18d103d7b8140
f33a593881c83e172ef3dff18e115767cd69f812
F20101109_AAAXUB stokke_j_Page_108.pro
86ebe259ecf7b7f6130f8ed1d0858410
58756a24d9c71f8f69c0bf6e7293687946831144
52661 F20101109_AAAXTM stokke_j_Page_082.pro
19152d9077f43c7f6d61e3173546af63
f96818109e4f2aa7e200a2b0c8cabd69150a84e8
55352 F20101109_AAAXSY stokke_j_Page_054.pro
33b26faa6a0a4b4c98d2be78fa50765e
3be46dd74626ca4b57e905ef8cd821bff760f886
54649 F20101109_AAAXUC stokke_j_Page_109.pro
5401e13946777a3f123f52fa2ab0e45f
6ba2c6c755a77228871180722e25cab0ccd66f0e
54196 F20101109_AAAXTN stokke_j_Page_083.pro
09c5d779fb86a5a56a76be9100421bcb
abc6f571718027367f163607b9e64bb06a1842a1
52810 F20101109_AAAXSZ stokke_j_Page_055.pro
930a3a52f65ddb663cf9e3dcc9d98286
0d82729119a9f328b55c00aa43ab19c9f6a733c9
18070 F20101109_AAAXUD stokke_j_Page_111.pro
68d28ea6485b368d8877e831c2115c7d
98d8f16c641b6c58f8284f68b7e3dc8aabe69d13
21906 F20101109_AAAXTO stokke_j_Page_086.pro
aad9dffff9ec0437cf320272f68a0323
c2b340190f1a63d874d630fb35872913f4022f95
1898 F20101109_AAAXUE stokke_j_Page_112.pro
763e087b28927b7aca773c34a05532b8
a7446d35ef3910aac2bbb46b7cbb4982a54a0c21
53107 F20101109_AAAXTP stokke_j_Page_090.pro
ca4c3b45a68b3559572a631549927bcf
a2fbe6a64782103a8d077c31b2264a5e9d643e95
10556 F20101109_AAAXUF stokke_j_Page_115.pro
0434a29a75a2e9a0c33dd6b47566123f
a1e953bcd055ff46cf5034dd5af559f9f9e9f1bd
53868 F20101109_AAAXTQ stokke_j_Page_091.pro
cd8a6b00ee3d042845760e541c904a0c
b2e9fdb600a8dadd4c3abdabe3982255ae7a1ea4
16814 F20101109_AAAXUG stokke_j_Page_116.pro
eb9d362c5b495472a041718bac50a5f6
569c4976db53177226f08da5d8f9873782d6d74e
55994 F20101109_AAAXTR stokke_j_Page_093.pro
7e52fe87092603924a8d777aa5b5e912
c43459e6c239a4e5db880a6f53a1482a35143e54
17562 F20101109_AAAYAA stokke_j_Page_043.QC.jpg
4800391ab8e88b8648e1cfe02dcd98e1
468018e9a8a72762291aa5ad9699bb2d8b13cf23
56325 F20101109_AAAXUH stokke_j_Page_118.pro
232bcff7fffd273a46ce7665e9e0474b
789059097b6591d961d060cd614706b4835a052f
56883 F20101109_AAAXTS stokke_j_Page_094.pro
a0ea5dc59c2fd8a3d7d77d97bb09c0c6
b17685b133d46a9139148d9352206d38700aaef3
14914 F20101109_AAAYAB stokke_j_Page_044.QC.jpg
26fcec7585f9a7916c17dedffb56792e
ca371ead2888211d8e7ad782e30859ecb0dd3539
52265 F20101109_AAAXUI stokke_j_Page_119.pro
5001a0b0ce09e49af735a9f5d2c611d5
eb35066170ff78d0a65eaf03b77243353eb0dc4d
39425 F20101109_AAAXTT stokke_j_Page_095.pro
a82794692e286fb46a6ab28171ced484
8ca629e7982dc68a6aba89f7440c7003d59d06bc
27679 F20101109_AAAYAC stokke_j_Page_045.QC.jpg
208827be928bae118962403e0818ddba
4d8d31098bfe4f9352d4cbd1ce2e5e547f8e8dbd
65003 F20101109_AAAXUJ stokke_j_Page_121.pro
7db26295c291f16022d39b87e9ffdb69
34fc96a608e247c1f72e43bd9240ab9d00e5a233
10204 F20101109_AAAXTU stokke_j_Page_098.pro
58c67be6a6cb693630e81e3c6d7927be
7cc5e18a5ca7a50932d94e143776e2dcb9d2ec27
7353 F20101109_AAAYAD stokke_j_Page_045thm.jpg
b78ceddaf28d7dc74edb4a414538b746
9c2819886790c62cd9a66924b63d12d8d90937c8
2994 F20101109_AAAXTV stokke_j_Page_099.pro
6bec48acdf0c4bae1f58962c96051a8f
ae0df90aaa554566b5e8043cfd1f1b7993d0d179
7368 F20101109_AAAYAE stokke_j_Page_046thm.jpg
3cefe2e867b69b0ed91c9e4f204ee9b0
9e3c361931777594abe8ca13e07dcd60208c7664
63653 F20101109_AAAXUK stokke_j_Page_122.pro
e0795b2e2a5597d67389c40391bb6535
f7b77f69950b406870b6b774d6b5951156e6f41b
55432 F20101109_AAAXTW stokke_j_Page_100.pro
932b880e2f2b909a6c42162ba549f59e
21829eb5b41f38f84913c121165e2ee1dded9999
7446 F20101109_AAAYAF stokke_j_Page_047thm.jpg
eafc514dc64c85fe08cd6c94019b1546
46a6bde74fe8ff89e94a43bb4c316f21a061c8df
64726 F20101109_AAAXUL stokke_j_Page_123.pro
0ad3699ab196a2f03cfc44c5a36e8a18
a1670741c752ea61d4233ec3ab31172a47c6534f
57431 F20101109_AAAXTX stokke_j_Page_101.pro
2b609d76595ba65a7cb6c2cdc766696a
4f93fcf6b6ef0d78a7c45f1d93a6a5c012a42a2a
27984 F20101109_AAAYAG stokke_j_Page_048.QC.jpg
f74d96d8cf210e92ee2884982b8a4a4d
8e8a9806ec8ef3ca79eb19bace0ae68b3f7495d5
2200 F20101109_AAAXVA stokke_j_Page_014.txt
3d7ece627ef7b5dbcd88bef30f4d5b17
11b1b167a2e824ec7c219df2f66cd60bc04bfa9f
7411 F20101109_AAAYAH stokke_j_Page_048thm.jpg
cdf6771b2210cebcc507ad818c0c6320
4c71db3d4e57f4e66cd34ab09caa2da6eb9f26eb
64034 F20101109_AAAXUM stokke_j_Page_124.pro
928602f4177b61e2760e03b36d935f6a
1a781c866e3525b50d9e095fdfe61ffb3b52b339
55660 F20101109_AAAXTY stokke_j_Page_102.pro
dd8a06f3eca36e6a3f3dfb3e62462be9
261c6c5b0bc62d2cdf7af9e536c5feb9d2b90e31
1968 F20101109_AAAXVB stokke_j_Page_015.txt
7db696b857472ef7a51f9e35c7b360f2
c30d2ddc33220d59ebd047b9339c1f0d64d7be06
14805 F20101109_AAAYAI stokke_j_Page_049.QC.jpg
0a2998a19365d82b14c8626f406ba3e1
91b0c3f298449c76bffc98fb5cf15c1ceaf69848
67070 F20101109_AAAXUN stokke_j_Page_125.pro
c133f60ba45ace054c512bca4780ea69
d3f7147b8b41833bdd63794e41b788d0d1336291
57237 F20101109_AAAXTZ stokke_j_Page_103.pro
7daafc173f5bfd1be02f2d8462dae423
148161c907e249ac31f0a4e77d22bf6c92b4cad5
2212 F20101109_AAAXVC stokke_j_Page_016.txt
7b6fb55e16108fd630c389bca48269f4
c087f3b3246a2c15d0bc7e8259932f90c31ad8b7
4165 F20101109_AAAYAJ stokke_j_Page_049thm.jpg
81350575314ac948a69f0a61b6fd4935
ccf0c199fbc3145e9f712153bfd0eb48c303d5ea
56810 F20101109_AAAXUO stokke_j_Page_126.pro
3d2b9d8772744cc01efa155d66f675dc
19a83b0c4b57bd9bc8a9f570f5366214560a58f7
2179 F20101109_AAAXVD stokke_j_Page_018.txt
ffa2302cc1273621960bae5b4e15a3e8
c7af4e73401b820953a9289215838b318bba5858
14692 F20101109_AAAYAK stokke_j_Page_050.QC.jpg
1e48a1f8b06e9ede6d9510191836a3bd
11bfb2a26664573da8104877244d3226f88b5bca
15887 F20101109_AAAXUP stokke_j_Page_127.pro
3f354145cb851133c650aaa1ea7c2e1a
4bbe1287c0a6d34bcb714fc25efa1f45896a6a94
1679 F20101109_AAAXVE stokke_j_Page_019.txt
815332f13378d86829ed3fc523d86960
4e0b61071845b90555c421e84e90deb3c4fb0440
4336 F20101109_AAAYAL stokke_j_Page_050thm.jpg
88cd3dd0c55feb1c66b063dc78eea926
f1fd2c04a9da2b06e26eca3c418a489e5141ceb7
512 F20101109_AAAXUQ stokke_j_Page_001.txt
0f205d9acddd7e285d14679e15702ddc
c34cc338f4747a9536cdfe2d09c75a8122d49aaf
2062 F20101109_AAAXVF stokke_j_Page_020.txt
a23840aa9196270e47a1e7588045943c
26f31c923b58aaa307b9c6f1379c9717bce646ac
26021 F20101109_AAAYBA stokke_j_Page_062.QC.jpg
77b815c84a008e55e001607aa7ce7484
c3a213b82432452f61614a7f735910cc33119e3d
4253 F20101109_AAAYAM stokke_j_Page_051thm.jpg
12a1686853016ce49c49a48e414e0dbb
8e503eab0bd6fb2dac941c7bbefbfb8f08dfd1d5
97 F20101109_AAAXUR stokke_j_Page_003.txt
0bd329a599a122578df3a331cba1cd8f
4cc47501efeaafbb77ad1fc5c15951075b8016ab
2076 F20101109_AAAXVG stokke_j_Page_021.txt
4fafbeccbb13b77000670b656f2941e0
d86a062aa3e999c28f0725dce5f4880dad65e153
7336 F20101109_AAAYBB stokke_j_Page_062thm.jpg
011527381b671bc3bd0b8dbd8b68c982
8b63889eb34eff3ea1e69f3bbb47533aaa393373
27538 F20101109_AAAYAN stokke_j_Page_052.QC.jpg
1aa08292dff0a50be18d32fa4bd5a276
d43edb5ce420ea1e8598a55c9b605c7aaf415d54
1542 F20101109_AAAXUS stokke_j_Page_004.txt
4d177548a1e5aee5966f0fc44b50b275
952072cbc852f60319234d503a6b8225d38ce051
2177 F20101109_AAAXVH stokke_j_Page_023.txt
eee1af9961bd1e95f985732ff239e94b
f23f3aeb3f33ba05fa77b5986c08c32186b370e5
27064 F20101109_AAAYBC stokke_j_Page_063.QC.jpg
291e586ba20191b2fc646b814a9909a0
b210349c6d42f3a1987df967d2e840e9438fcc1a
7434 F20101109_AAAYAO stokke_j_Page_052thm.jpg
2b3569735895bd649fcfe63c5590e913
d8e7e17362f57996fa72e12d90ed4fcac2ad7966
3260 F20101109_AAAXUT stokke_j_Page_006.txt
e306b7e45c219623ed853c36ef2d2653
dcaf8726575453c0a4328b7b9c90103f6db8b8d9
2020 F20101109_AAAXVI stokke_j_Page_024.txt
6d3cd4c0ea6e67788f78e6d850498d13
4f6761494e4478c2b700c899cf0e3dbbca4702d4
24627 F20101109_AAAYBD stokke_j_Page_064.QC.jpg
6ff20ef1973532b0605fafc7f7e35582
26af3760fb2ebe06f0dfa305577e5534c3785aef
27706 F20101109_AAAYAP stokke_j_Page_053.QC.jpg
3975b0ca85efb28a350a0adc3fb2fec5
87b0582897ffce009ea092356d8ef791218c2205
1359 F20101109_AAAXUU stokke_j_Page_007.txt
e93610ed47a97009e35ee1b88f1bc6d3
3d0dab1f7da1f2ffbc056c298c0f69716747308b
1797 F20101109_AAAXVJ stokke_j_Page_026.txt
c6a3b0fe7c270c77e9aace32415ff6dd
872a7465850164ebd018b4a3109c598433d56838
27145 F20101109_AAAYBE stokke_j_Page_065.QC.jpg
35fd55a85b8c43f206aa4babc5f8929e
0a9227ebc4295598e3ce7a7b560d8628b9705a2c
7510 F20101109_AAAYAQ stokke_j_Page_053thm.jpg
7a097a1846fddb930f990fff5b71f4e7
33332650195d691a94d8950679c1b83c20c04537
3048 F20101109_AAAXUV stokke_j_Page_008.txt
8ec14a9cf9f3d3877d59322c7ab0b5cd
22104bc8b6ad00e1b672d81efd04cd85a184d317
2156 F20101109_AAAXVK stokke_j_Page_027.txt
b3a750e83942ef65f26f023d2449658d
271401ce662a34f8fd5896ad678fb481699f68bd
28699 F20101109_AAAYBF stokke_j_Page_066.QC.jpg
1df85b98cd0030930577fd056f30d5cf
2371bb80a08affd6d1ba487cc1f37c7dcb8d93f4
27581 F20101109_AAAYAR stokke_j_Page_054.QC.jpg
90e75be3bf3ff2df73c3b43beb38d69c
f1a82bef205a6de66fdd83f4f338977a6acbd35e
3146 F20101109_AAAXUW stokke_j_Page_009.txt
bffccbcd264963b6ccc1f70ef41acc64
a874a8b6794e21e8b0491e5b301ad4a895e69ae8
7457 F20101109_AAAYBG stokke_j_Page_066thm.jpg
4dde189f1b253878a6e6f0d7a8b67955
e24a4cdb0433fa4dc33440c31092029827fd23ea
7605 F20101109_AAAYAS stokke_j_Page_054thm.jpg
d7381af0540e07dd6f45f9633ca8346c
b12c2c0794034dc989e698f61583cc9ca8cd6883
1263 F20101109_AAAXUX stokke_j_Page_010.txt
6970b51c2cdd1f7819aeb020118c6914
c785d6db9312365b09d50c248b2da6048ad7adc5
804 F20101109_AAAXWA stokke_j_Page_051.txt
993b950b78860c8c08ef76fb7885ce17
bff4621e8fd3030a5c3ac71080e478fbb99ca9f8
2252 F20101109_AAAXVL stokke_j_Page_029.txt
c15a6acdbac00b505356a6d5b2209b6d
737dafa6cff299306e3e1531d8c6804fa4011a3e
16921 F20101109_AAAYBH stokke_j_Page_069.QC.jpg
7420c9dfffc45e60bcfe4a7f854038fb
b0174e88417d07b7911f3032f759a87abe2756fc
26017 F20101109_AAAYAT stokke_j_Page_055.QC.jpg
f95173a3e3a7db6e85e199a60f1b25f7
d3128988014a50a63da89d95c8fb5fc6f9b525c9
1557 F20101109_AAAXUY stokke_j_Page_012.txt
c5d07cb1eb51b69b4babd9089681fb29
1f8efc79a74e6152cfd966160a14e8edab4caa67
F20101109_AAAXWB stokke_j_Page_052.txt
3c58d152ca166566007689e9029bc565
0b139db1a49c77fa56a84f9b846d0f0b9372923c
2011 F20101109_AAAXVM stokke_j_Page_031.txt
4525aa23a1bb667ee45c5f9d85c2caf1
91df1ed304fcc6c8d1cff3834469b686ddd1a03f
5295 F20101109_AAAYBI stokke_j_Page_069thm.jpg
831724d439d0d228ab1ce2eb80e44b3e
2687621343735a710d06fb3df9712eb5fd05fdb4
7560 F20101109_AAAYAU stokke_j_Page_056thm.jpg
339c559d0ce6332aa7501f602dc29b8d
9272d6be232724f6a1286b0f430eee63e9b7e0ad
2194 F20101109_AAAXUZ stokke_j_Page_013.txt
d73b750d0ac1ced53dc23dc70f00659d
c5860bc6cdaaeb5d20869008b4a64ed09ed8be14
2198 F20101109_AAAXWC stokke_j_Page_054.txt
d548d155c7e55655300a370c748caadc
e5640974634094c7704a0c30db702e2e55d0eba8
2095 F20101109_AAAXVN stokke_j_Page_033.txt
268664d47b04fb2e391f88954d1fc16f
982ea3857a3cbf970900473e9e364d248307cb71
14283 F20101109_AAAYBJ stokke_j_Page_070.QC.jpg
835d96df9e9071b2dcf3cabfc585743b
4c6929feb7e5d6f00e0275065a98befa1ddb5003
2079 F20101109_AAAXWD stokke_j_Page_055.txt
db67ccc00136c8e77a775a380564dd5c
402730b52a4bf03a84a321695d81eaa5b3e8f0a3
895 F20101109_AAAXVO stokke_j_Page_035.txt
c1ff84e1c0381b9dc8933f27241dacb4
26fa3d77e1b566e551db9f00df6ac40c000386bd
4466 F20101109_AAAYBK stokke_j_Page_070thm.jpg
c97fb94addfcd2879d07258ad1df39db
8d2d9b83cf2e27f33648a1e69149c5badddb5aba
27440 F20101109_AAAYAV stokke_j_Page_057.QC.jpg
9b9303e562b6ddde6b99f4090848401b
eac94d501c1b776ed89f076aaefc8eec5dfa619b
2225 F20101109_AAAXWE stokke_j_Page_056.txt
04e30dc481a45642d695ea2fecfa4084
bb222af369a10f6e55a0b0261d0d3c87c7cb78f8
1997 F20101109_AAAXVP stokke_j_Page_036.txt
22773002bffe1e027c873cbcc1f29797
34ee90f58aba3e96e5d3cd5d4de388e094790b90
3825 F20101109_AAAYBL stokke_j_Page_071thm.jpg
b895fabca9a2544b3a30cba0764da0e3
b6b683bbc6a332391a130db9948d2f68b04ab13e
7517 F20101109_AAAYAW stokke_j_Page_058thm.jpg
75720bf5d779f15ad96c7f448c84b177
84b77d4b3501817270bf2305a967b91a568207dd
2222 F20101109_AAAXWF stokke_j_Page_058.txt
eec720a0f79379ba66ef05dec6e6756a
ebbb54dc77a18709b7ef67e0c8b7688b7904042e
2146 F20101109_AAAXVQ stokke_j_Page_037.txt
f8c374fb117f07da595b9e1fd0829ecc
9cb8d78397f6d6be37d7894f4bc1cd0e47093b0c
7128 F20101109_AAAYCA stokke_j_Page_082thm.jpg
ec130fd4d866804e43c832081ac55dde
a30e41f966888ef20e8917f8f53baaf10ba687b7
7845 F20101109_AAAYBM stokke_j_Page_072.QC.jpg
07d76af8b1cee09e1c6215e820660693
7cdb01c5e3aa5bf5f611ec4fca84fcf557f46468
7028 F20101109_AAAYAX stokke_j_Page_059thm.jpg
3026fa6143ed262fd28dfd615e6bb729
289fb687942d4fc44d7ec3faeac029081a8de9cc
2127 F20101109_AAAXWG stokke_j_Page_059.txt
5be7049602769bec7bc9cb1af4ab52b5
8ff4c26d8202c59a6bcf612d99599cc124433804
1970 F20101109_AAAXVR stokke_j_Page_038.txt
a3352179dddb70efda72245563abe078
ea19f892b3b9132f51f34efec1962b364dde9d18
28090 F20101109_AAAYCB stokke_j_Page_083.QC.jpg
f909e6a3d1cb259be0a3045262e01349
b33d33d370eb442d5aca5709f1a997d859617a4a
2874 F20101109_AAAYBN stokke_j_Page_072thm.jpg
253c1cef02c619909f62d1af61d8c705
a6ae1cd19029ef0404ce22e27f5ac53d8da32723
26582 F20101109_AAAYAY stokke_j_Page_061.QC.jpg
1d844a4964edea6cee6d383861904c88
31cad9dae77a4de2452f2397c877215727fc977f
2034 F20101109_AAAXWH stokke_j_Page_060.txt
7587ca717525b86580227811ee08db04
d3e0ef133ece673d4ca6b6604da3e9fb641ec681
1854 F20101109_AAAXVS stokke_j_Page_040.txt
4b298a5cff135bb1cca0e1199004d0a9
88b35b50285c5de39d4c8e233f0eef61e753cc4b
28680 F20101109_AAAYCC stokke_j_Page_084.QC.jpg
4f9e6c68276e48147bbf1a3ae5de2d58
34a05603f0e636a84a555a500b0212fd76879ce8
10465 F20101109_AAAYBO stokke_j_Page_073.QC.jpg
a4248c6b577ef78eb78ef1c98ddeffca
71e4e0f3143a753ced1848595a3619f35155916a
7191 F20101109_AAAYAZ stokke_j_Page_061thm.jpg
b7170ed61f4aeabab8d1e73f2598a165
2eb1085684ad38b87cbb67d47cde83f1d4b14081
2126 F20101109_AAAXWI stokke_j_Page_061.txt
1e294c453869803b1eeac3637af6fd5a
a1d565c3f18d73406dd11690cefaa0343db22368
1881 F20101109_AAAXVT stokke_j_Page_042.txt
42b944964b79afab7e7f72982e127cbb
2c8bf2e9286316c7b1009af28e95e085ec343b48
7834 F20101109_AAAYCD stokke_j_Page_084thm.jpg
e97f66cebf32adbb792410d97cbf02a4
ab4432dce084699d9841189bc10dc36c2f695699
12315 F20101109_AAAYBP stokke_j_Page_074.QC.jpg
a4940b017b6383e7507bac4ec3e8cfcf
5fcba7da615a58f4e7ef45064b97dc1b3921d1a9
2037 F20101109_AAAXWJ stokke_j_Page_062.txt
abc5d11c428c2794d3756ea9f808ca86
ab65c3628ad9dcf735641c63ab42d5f05205d057
1629 F20101109_AAAXVU stokke_j_Page_043.txt
85684b79c5e0b63dc8caf1e1d165bf17
ac2ef38bc65eff6172c51dfe9ea3875d05409ec9
25542 F20101109_AAAYCE stokke_j_Page_085.QC.jpg
eb85460866bb01557a886f9708397caf
e58d5eddf032a6077dafdff15e060e62c3619afe
11934 F20101109_AAAYBQ stokke_j_Page_075.QC.jpg
5e6cc5dc116386196a7f2ae7a9c2e620
b5404ba3636eed301dc1b814923e8c93d419432b
1198 F20101109_AAAXWK stokke_j_Page_068.txt
f60e75ead55a40ab0690d080870d5d77
2b3b50acf584d630dc722f766cb364b96c265a77
588 F20101109_AAAXVV stokke_j_Page_044.txt
0298d52372075e8956a2e9597668de51
97b0777dbf4061821cec6dda26ab6d48771e8510
7040 F20101109_AAAYCF stokke_j_Page_085thm.jpg
8b39a2c9cf3225900069e4a29b275cae
a4264534fbe3646ed1fcae5567b5b33d9b0a57a5
3644 F20101109_AAAYBR stokke_j_Page_075thm.jpg
0599ae57d0247f62768b88ae5d470cd8
e0ef69288220c2e3208d59d65bd84115d2313b7d
1028 F20101109_AAAXWL stokke_j_Page_069.txt
71308fe4c3dcf93c607e01ea73d0b0d2
653ca37c2e8a7da87f9bfcbe8326a07fe4f23b71
2139 F20101109_AAAXVW stokke_j_Page_046.txt
fea303f2f8cd0dd5d7133bc0011ff5dc
36c8d6bce2e175a6430fd133369744274df2cd68
12855 F20101109_AAAYCG stokke_j_Page_087.QC.jpg
ab5b8589bdf93d8ac2bcfcecd236fb0f
61fad656ed4e5d5993de76f2e90d424eff911f96
12906 F20101109_AAAYBS stokke_j_Page_076.QC.jpg
57a79e64eda222c2b28b040a0eb68b25
3412b99ba816b3076f416359e4c7d5ce79ab3c8f
1829 F20101109_AAAXXA stokke_j_Page_095.txt
36673c9b81fdc77e9a8aca9288467d30
d6c64d292f69e996c52f82304b20ef3013eefc5f
2195 F20101109_AAAXVX stokke_j_Page_047.txt
9901cd90a140aa1e27e0c3827e8f68f3
ba9b74273f5737516ce1bac6fe83ccf96af7994b
11690 F20101109_AAAYCH stokke_j_Page_088.QC.jpg
e7e1c0c42aeff8338296159caafef94b
83fb9bb357e042bf78a659bd63ea42e19edbb550
4154 F20101109_AAAYBT stokke_j_Page_076thm.jpg
3647bb7cc3202e798e1f24ca90dc6f7b
9f21c3a20acb6f2c2bad1f7a43582a1dc4db797e
263 F20101109_AAAXXB stokke_j_Page_096.txt
81ca1f11043d0e420c6cfd8fb622b554
3d84eab33a6a2f30b163a249cf2692a476dd9860
F20101109_AAAXWM stokke_j_Page_078.txt
9f006f4525281d4e41b6851dfa73a755
666bc9434ca952e4733cb0e0a2dde11471963e92
1091 F20101109_AAAXVY stokke_j_Page_049.txt
303fd91dcf1401ee4941064eee93800d
a3a4c0e7eadcafcef76b58dfa4dd6909ab1e1c6f
4109 F20101109_AAAYCI stokke_j_Page_088thm.jpg
d687e34d9ca09319bf2aa70fe8d46d50
675943bd0c2b9fe75777ff9bd1565414a6a10502
14553 F20101109_AAAYBU stokke_j_Page_077.QC.jpg
bb45ef277b28364a491636baf7f579a8
2758f1dc72ed2d1f0211b82f39f3b760e16c0f3b
325 F20101109_AAAXXC stokke_j_Page_097.txt
ebdf64f0b0cc451e6a859f33d4e47148
89663a50dfa05f089ac9b004c06f9bad0c813822
2221 F20101109_AAAXWN stokke_j_Page_079.txt
88df746ddf5b2f2db6260598e1adfce7
36c4b2d446900daa10733a1abcb531f201cfba63
1086 F20101109_AAAXVZ stokke_j_Page_050.txt
ebb16387c24c80682edbb583910fa6fa
27923611885c088b84ffe0dcc5bf906e887508fe
13404 F20101109_AAAYCJ stokke_j_Page_089.QC.jpg
9dbc5d0c6eaae01e12f09e3b39ece357
10093c68a80d74e1fb19e4d8129d6fa2764dcf1b
4209 F20101109_AAAYBV stokke_j_Page_077thm.jpg
4d69021d37de19cd3eec9d7f7a3bc0b2
b57ae01e3a07d618de6524efec5c877c7f0614fa
2247 F20101109_AAAXXD stokke_j_Page_100.txt
863dee6addb5f9777b58c812742a5ef8
3a07a07599261743b2314b377a93509d92ff0aee
2580 F20101109_AAAXWO stokke_j_Page_080.txt
b226f11530ff095432b0297e41d88a5d
84bf0ebc096d1f8c156ba74c16a25d541c739ead
4012 F20101109_AAAYCK stokke_j_Page_089thm.jpg
399d79f7d25e64ef489040840803c0d4
9d2d5d82bdd656552b0f18eaeafabad2261290e1
2183 F20101109_AAAXXE stokke_j_Page_102.txt
b73b26cf8ea5fb9e7472fb8f555065d2
38291e835168cae099bc23a5d913b8ab25801aac
F20101109_AAAXWP stokke_j_Page_081.txt
82a6e90cbc5d286b345b15a6e2cd688a
45cbbd92a3dccbecfa84274a12b233cb7da0f89c
27446 F20101109_AAAYCL stokke_j_Page_091.QC.jpg
d39d4ffd4a96c4c9150fe6afbce2e7f0
173bcfb811e1127376c4e315a79ee2e0eedbc8fb
26737 F20101109_AAAYBW stokke_j_Page_078.QC.jpg
35f9c41dec27b963986caa2a3f037edf
b6175184ed2275bfc1f6e41b8805afac86aada58
2245 F20101109_AAAXXF stokke_j_Page_103.txt
2606f2d9509c68eeaad0e3c6dd3b64a1
d23b2ac851759a0d7d012247f7fdb88abf9a9bd9
2163 F20101109_AAAXWQ stokke_j_Page_083.txt
581f6f2b8bd2d43c1ae304e1960c8c3b
b30dc2f3ee216dbd3e535fb353726c49f9b9014d
28211 F20101109_AAAYDA stokke_j_Page_101.QC.jpg
bad80be399de18426a3e6a756f3abf74
c8e2841c677147ab7eed6cce05b5ec2fb0e311e1
7266 F20101109_AAAYCM stokke_j_Page_091thm.jpg
28b20378624b44164ac820f02c5e9a13
c7259a8c55c6c4fd236b03aba1089fb39182734a
30268 F20101109_AAAYBX stokke_j_Page_080.QC.jpg
e447175ac32a67983a2f55973b6688f8
78d691a155fbfd38a93de35bc1fbc5d2c1769a48
2101 F20101109_AAAXXG stokke_j_Page_104.txt
11b8807ed3dccb3a4806d7e9cb03dfeb
b40f78edd21e7edad79da15c9e323c25bfa1267d
2078 F20101109_AAAXWR stokke_j_Page_085.txt
f2986165788c52ee731f171714a827fd
a734c7866741d198c4f8384a178169ebe20681e4
27373 F20101109_AAAYDB stokke_j_Page_102.QC.jpg
67186f9623e55b33c76476a2f77eff36
725412a32f92f5edf03573f5bbc23f4bee754814
26945 F20101109_AAAYCN stokke_j_Page_092.QC.jpg
2ce479a370898d23179de06e7a0f5f12
cb0c6ea6ed703ee1b7d3f76b21f6889d56e45be4
7516 F20101109_AAAYBY stokke_j_Page_081thm.jpg
66308270c165efac852dfe22b12b740d
c01817fa994188e9ae53b6596378f9048b38a458
2148 F20101109_AAAXXH stokke_j_Page_105.txt
4c94140323c039d050c4ab374ce642ed
ce3817852715e8c74f88cb6152e8ab8aeb8f3ef1
1323 F20101109_AAAXWS stokke_j_Page_086.txt
c7ea139414976208ca875d9419026fbc
f2d8fe7e63f1f503833e3e09f85870cfe9defe7a
28373 F20101109_AAAYDC stokke_j_Page_103.QC.jpg
b0ec45de9405ef8bf832b92791b19903
89681d8fbc3757d72d3546a46228527ca84dc05a
7242 F20101109_AAAYCO stokke_j_Page_092thm.jpg
a2f157532f04709e7454c2bce56044e8
dc06bf6dfafd9178f468ebad890922d8259f1a22
26851 F20101109_AAAYBZ stokke_j_Page_082.QC.jpg
e24442ed9ae52c669a7f574556121a65
c0051a7455ea8dc5c47147eba49579eef9f90af3
2072 F20101109_AAAXXI stokke_j_Page_106.txt
4c08e032a845fef6fc4b93c016dbcad1
6db199a541066e6b6ae65d2b748625a919c8b313
1203 F20101109_AAAXWT stokke_j_Page_087.txt
2c5397fc75a0b7295a22cff047dd79d1
f02176751a802e76eb16a48d2683387e9eb7ba08
F20101109_AAAXAA stokke_j_Page_061.tif
464c504f2753ee60c17df67656b2e9a8
b64c9d7ea32543a8217b43fc39b7d349b2f56722
7603 F20101109_AAAYDD stokke_j_Page_103thm.jpg
87a5cbb57a9383a3cb44cf6154972c8d
d1cf4691d0363032116b0aa83c20f21f856da365
7430 F20101109_AAAYCP stokke_j_Page_093thm.jpg
978e8b03b6ea7461708f13d7ebd39285
a9d1e3d34cbd879d415e91151474222b8046f3c9
F20101109_AAAXXJ stokke_j_Page_107.txt
41ddb4d12f9542d19e55e905319b02a4
5b317180e9b51b56315068ca716c8a39b52a2163
747 F20101109_AAAXWU stokke_j_Page_088.txt
b5195aa0fea4b1accffdded05c0a3029
155a454f6cc7c5a200c720a12008f7ad250b5706
F20101109_AAAXAB stokke_j_Page_108.tif
8e76d9dd5e3354227ec1cb15440b8e96
06277bcb49f8acd6d240662ca46f666cab51a85f
7448 F20101109_AAAYDE stokke_j_Page_104thm.jpg
73b49f01f1445aa15078df2b5e644c8e
2baf5f0cd1c07d82992f2df67536010d44402462
28004 F20101109_AAAYCQ stokke_j_Page_094.QC.jpg
5d8ba68483c887f1f44d92fb1da1f401
ac56701ea77b0015ebcf5a5123c1935f197badae
F20101109_AAAXXK stokke_j_Page_108.txt
1143a53bb8894d5b1b8c9e7c11b7731e
ec0016478e72aedac49683ffbdd084b45fe493ba
709 F20101109_AAAXWV stokke_j_Page_089.txt
1de6124359476a7bbcd0061070e06667
d468c13b73dacce40eaf94e5abda1b54c910df63
2285 F20101109_AAAXAC stokke_j_Page_101.txt
15695dfc63366bd9519add2eab6d2b90
29ba2bf31870203403ca9946707dc3e3aea9a7d6
7366 F20101109_AAAYDF stokke_j_Page_105thm.jpg
2aadb6e74a977d5902b9f849afe7d2f2
8115458c09ce1d3f3bb717cacd7f53907eef920b
7561 F20101109_AAAYCR stokke_j_Page_094thm.jpg
45a4212b1b2498e7d0c1446bd29abe4c
9b1b970fe507f0edd75b0050ba3a0af5c9153c5c
989 F20101109_AAAXXL stokke_j_Page_111.txt
8679ded62f217f305e88b3a3adf52180
348dcab2bcab355725c09919f4f8e6a8630c0326
F20101109_AAAXWW stokke_j_Page_090.txt
870cd5aad75c2760f5e656ed9c369e25
33e20346bcd8d9ea71ad60ee8c832c0f7fe338ca
7059 F20101109_AAAXAD stokke_j_Page_119thm.jpg
76efec87a294cc6fb7a39dd4ee8105ce
38bdbf6c0296df2cb5ae93c8535b725fad3ebcfe
26002 F20101109_AAAYDG stokke_j_Page_106.QC.jpg
3c406158242b0812ea0b3eb7c71d8d18
f3b3546f4893c525178222f3f8b23b7568e33d63
19477 F20101109_AAAYCS stokke_j_Page_095.QC.jpg
83b25d50b4cb12a0e853770b8c1cda61
457c69759feaee4cddb7c5d23685e71f8fca25ab
135 F20101109_AAAXXM stokke_j_Page_112.txt
d5d7bd10ee1d494b486723686c86e382
ed26abfdd828f1f2eb07fb8035368b0a99e884cf
2115 F20101109_AAAXWX stokke_j_Page_091.txt
a430dbdb278bb36e8ec3b4d8ef0fbfe9
0e46a693498dcb6fd0ce4a72e791712c9fa2a10e
878722 F20101109_AAAXAE stokke_j_Page_004.jp2
3d47d1542121fcf739227d23bc0f65ba
17400f7efc61ee5abc2402ba8fdbc9150aa2f54c
19951 F20101109_AAAXYA stokke_j_Page_004.QC.jpg
ba1da60a320df92c6d5bea6d4a6aab81
31ba73eb7aa30f0fe24cdfe21407545a6402e0e4
7494 F20101109_AAAYDH stokke_j_Page_106thm.jpg
c1c429d0d62ea59a91c758e0ac277e25
7ab1dd82e173b9ba4e875d34f9564ae529eb539d
12273 F20101109_AAAYCT stokke_j_Page_096.QC.jpg
77ff6a68fa42a0f7747fe0b88cc777ad
684f73a4cb0557651df4ce44e1fb4fe5bcefe2ec
2152 F20101109_AAAXWY stokke_j_Page_092.txt
8fa98e8ae26675ffd7125777b76331c2
ef175c8a0c1989f5911df0153454e113857d1b49
5022 F20101109_AAAXAF stokke_j_Page_043thm.jpg
37e033c54c83645fecae054d500b8c13
b373af0d927052e5210a7602619db91ce7b86345
5946 F20101109_AAAXYB stokke_j_Page_005thm.jpg
98f4d64a7f5d4af6192d9f45cfacfd41
a27d3177a6f2197ab8c5d5f525655d1a5b2469a1
26022 F20101109_AAAYDI stokke_j_Page_107.QC.jpg
1a943893a68def01d8f3f4ca88c176b6
be5a4ea3692917843f62970336903177094b4df3
3317 F20101109_AAAYCU stokke_j_Page_097thm.jpg
b1fe51085f951a70f00c36b5139db2ab
ba10a18259c1afe46f6a4792e4503ff72ad41ed1
2199 F20101109_AAAXWZ stokke_j_Page_093.txt
87cde248aeac3780a761705690ed7c0d
80a25594175cb947189e11eec419bb488417cd5a
28407 F20101109_AAAXAG stokke_j_Page_047.QC.jpg
8b10f5dddfccc1aab3765f1ebe1220d1
a89c65e22fd6aba9597b2722576cc9a2852c091b
12486 F20101109_AAAXYC stokke_j_Page_007.QC.jpg
8d4957844d4e86f561e95ce26870ad3d
16558650183c46197ce7d16a839370ef71773549
707 F20101109_AAAXXN stokke_j_Page_113.txt
a1093caed934e74f47e84ab91f0e4dc5
4330e0c656d6e6d93f3577a53b27f6c9d8f0a4fb
26969 F20101109_AAAYDJ stokke_j_Page_108.QC.jpg
5c632e26ee49d56224c3f31be4f97a43
03e7baa02ee7c1d90e38dbd39ca13146ae3941ce
11793 F20101109_AAAYCV stokke_j_Page_098.QC.jpg
48ddc70e540bcaaaa8b8623946f1af81
ecdfbb40db14c33ed8c3d444fb9745a5ea60332e
F20101109_AAAXAH stokke_j_Page_123.jp2
8ed452b24df6e671ee1f51ecb99216bb
78bd2e006b02fc84334ab2b5d349f1c520bc7532
3575 F20101109_AAAXYD stokke_j_Page_007thm.jpg
3d3fbb28223b24a9b46bf82942a404cf
2514cf7ef8439bb786eef4fd6e8f7d4add25c6af
687 F20101109_AAAXXO stokke_j_Page_114.txt
a8fe924a8b7e98f13157519414147f01
5d4bb9db076594c5456bf2cbf06804dbc69cbe7e
26782 F20101109_AAAYDK stokke_j_Page_109.QC.jpg
102883711542250284d69cef02f1394d
eb5fcb405c30d8dacbc7d4c9175eff4227af9fc9
3729 F20101109_AAAYCW stokke_j_Page_098thm.jpg
d575bf52047dc419cee82a4b400bfcce
bf4d7755914d216c2099b35163d69068abd7e34a
496034 F20101109_AAAXAI stokke_j_Page_051.jp2
b0f61bd362e46acf0be91762b500931b
805bd90e894be0630057109f73ba75731f50ae3f
28698 F20101109_AAAXYE stokke_j_Page_008.QC.jpg
0c27539beb0e636d71afc5f33e77a281
739926f91f6714504d56f0fea43f01de9b807ba2
2085 F20101109_AAAXXP stokke_j_Page_117.txt
a13828bc1c5c18e4bb67052f57920d78
8e70a326c731368e1fffbce3a4f930a38befac99
18913 F20101109_AAAYDL stokke_j_Page_110.QC.jpg
fb5241110d818db459695ff87229bbf8
400cfe642a0c6d1327a1291f2018d66d11ea17a6
27738 F20101109_AAAXAJ stokke_j_Page_067.QC.jpg
28d2ad811c71ec83ddc60c38e63d1e71
02d31e3e1a206b73bee31ccda67fb5d08c05e8e6
7502 F20101109_AAAXYF stokke_j_Page_009thm.jpg
57fd8792f914b48a77fc739db073af4a
d1a73816dcc1070ed0d45101ff93e3006c4f8efa
2204 F20101109_AAAXXQ stokke_j_Page_118.txt
bb2bf66f60c681c4cae869d6aeb3a18b
44f8fb0b61d9360e8ddf1b0d6686918788ce96fa
28587 F20101109_AAAYEA stokke_j_Page_122.QC.jpg
89f5704286137f71f55498a790914d9e
6261502eb232aa2d08e007a2692f58f0cafb7e40
13429 F20101109_AAAYDM stokke_j_Page_111.QC.jpg
dac011117c53766f1172ea748d2d9d0c
4b669b8fd2ca67dd6fc58cef4b3aae374f19badd
8211 F20101109_AAAYCX stokke_j_Page_099.QC.jpg
bbe0c9aa02c4006491d1ee4ac4f66273
6398abe5695eae43cebdff0b35c5d7abe97dfc0d
979 F20101109_AAAXAK stokke_j_Page_077.txt
ff1b4a584959e7a5f35e7dcac9b14664
1218e6440efd7c86a0397fc3102ebefa3efa955f
12514 F20101109_AAAXYG stokke_j_Page_010.QC.jpg
80d50c04e018080407a40872085a0c4c
79b5f4bfd1a4ce86422644b356d35b7e5f69d82f
2611 F20101109_AAAXXR stokke_j_Page_121.txt
6c95870e47c1574f11aa7f04b2774c9f
7187f3888bc25eb52d76fd88305bac579e076414
7545 F20101109_AAAYEB stokke_j_Page_122thm.jpg
d788b4000f4556c7f9da38dfcb9a7268
8efba6116464ac2fed6766d8c9791c73db63c919
7014 F20101109_AAAYDN stokke_j_Page_112.QC.jpg
3ae6ced80769659cf054cf9d8316766c
48df76a2e76503bdfed29af008c9bb86394f9f3a
2820 F20101109_AAAYCY stokke_j_Page_099thm.jpg
fb73462a3d5f3a1ab0afaae04207858d
6298049ec87be8c7a5002eaee7d5d8903432e4c0
586 F20101109_AAAXAL stokke_j_Page_098.txt
3d62b7235b9739a44c6ff5e3cd92484b
9510c5ea466de6f5af3ac06904dfdb3a5c9635b2
7193 F20101109_AAAXYH stokke_j_Page_011thm.jpg
60d3e3e4c43f417a9911ce7098220e3a
28acd481493f92a99ddbd3f35204de1d0d00dda9
2584 F20101109_AAAXXS stokke_j_Page_124.txt
5fc3d34315e9c1b222006a604d08a8e6
8929421a6ed9c7cacce69791ec7dc3cd6164a851
7566 F20101109_AAAYEC stokke_j_Page_123thm.jpg
856cae3079dda7458544852648bf7187
4ce4da997f8d71cc326367b84d712d606ee65d49
2810 F20101109_AAAYDO stokke_j_Page_113thm.jpg
d7b4b8acc14500628554a5d4d6471240
99edc2b667796bf0aacba81bb96cc5f6f16a8945
28210 F20101109_AAAYCZ stokke_j_Page_100.QC.jpg
87eb7fe933c0e75dbe4ff3a71301fba7
acab553ade0b50ffe59098a9e8afda694fdaf75e
1051972 F20101109_AAAXAM stokke_j_Page_005.jp2
a29e6ace6c764ecf549c5ddc02f32ae1
2268568ce50b9351aff8b594edbc34f8b9238148
20748 F20101109_AAAXYI stokke_j_Page_012.QC.jpg
17997fb936e7c9bae34bc22be8ea7255
ee72010b556c46c3b8d5cc6453b9c51409476973
2692 F20101109_AAAXXT stokke_j_Page_125.txt
3618df3239dfc999c72aeb378ccf542f
1192bb9add9ac0ba97b74b5b723f81fbcc2fda93
3973 F20101109_AAAXBA stokke_j_Page_087thm.jpg
23eea4dc23baea923a11cd9c5bf2a84f
4ed48b09b28f101de9324e963a32dbe79009902c
7374 F20101109_AAAYED stokke_j_Page_124thm.jpg
9bb37400511601973c984d4e9dd24c08
df48d3293a08a5043a6e427815b9f4731ec94809
9194 F20101109_AAAYDP stokke_j_Page_115.QC.jpg
83f03f1240c99a8aab8ad0f3d51cc6ca
56b6e97fed46757ff89fee4af6243672002a7f0b
365474 F20101109_AAAXAN stokke_j_Page_116.jp2
97e7bda4cd676996e18270782b5ed50b
3826a3ec4ce18cf19a81639ebe355d2f77f1c993
27029 F20101109_AAAXYJ stokke_j_Page_013.QC.jpg
f8df798ad709f87782818012bce1f09d
fe7aee0ff51704c4b02552926bc763d9971b1ff8
2310 F20101109_AAAXXU stokke_j_Page_126.txt
dee8a988a461fd3eaf52a9d530d1edd9
a975d3e05eb0d3f0d1b84083c6fdd91a04ef9fb5
4575 F20101109_AAAXBB stokke_j_Page_044thm.jpg
7aa86651579089a457fc47eb38ff3cf3
abaab204c823e6bb6095bad6f9be6850cf0d0233
29200 F20101109_AAAYEE stokke_j_Page_125.QC.jpg
528780d4308cf64bec19b2c24049ce5f
f158952da981646202a89ba1d2259ec78dc289bf
2524 F20101109_AAAYDQ stokke_j_Page_115thm.jpg
69e608926b309baac352311f9954ceea
ea40c1a46a44825fc36d9e9ccfdc2733a879c1d0
90735 F20101109_AAAXAO stokke_j_Page_100.jpg
c3e71dcfa049efe04fedd3c9aec1f1ea
06d9273367255bbb363df431e94708639ead4f58
7212 F20101109_AAAXYK stokke_j_Page_013thm.jpg
0a98958188e68fc1ae15eb19b03e5b71
9e4c65675f173fa8a19f20a580cbf046e6850976
669 F20101109_AAAXXV stokke_j_Page_127.txt
11134fadc5afe3c7e284fa81c92cf247
0adda89e6d89d225bcbd0e56f8316c345a0dec1a
7558 F20101109_AAAXBC stokke_j_Page_101thm.jpg
78f2be0c98f6ca12f19bf869af6e52d2
8c63868ccc86d4ac9c76ad63b8e9823aeab6becf
7622 F20101109_AAAYEF stokke_j_Page_125thm.jpg
968481967500c288a9aab6c3711b41fe
f8e0be0cf3c54d9b9dc81184cfcdf094d6d2ca47
10754 F20101109_AAAYDR stokke_j_Page_116.QC.jpg
43adb0ad11b409e62fdc66de43136942
7b97c8d8fc294ea012ac36a5130c2d08fce72226
80015 F20101109_AAAXAP stokke_j_Page_059.jpg
41a1ed65e60ae50faa7d8ce7443783c3
42e4152b59d832fcbc1ed8753a0111af314ef949
28463 F20101109_AAAXYL stokke_j_Page_014.QC.jpg
977774da1d6305ff751605d56dad96be
c780e6f29fe0ceb1a0885a1c4a3eb9e90bcd937f
2457 F20101109_AAAXXW stokke_j_Page_001thm.jpg
3e4b494898b7dddfabc2ce632029dd3c
2e06d7bb641e7ebb49bb32619c7539cd9de6dab3
7290 F20101109_AAAXBD stokke_j_Page_028thm.jpg
8247d1a80647ac7409b05687eecd5b66
3ce21c232c5f56d8ef04c528f7fec91de3b47b9c
26050 F20101109_AAAYEG stokke_j_Page_126.QC.jpg
b7ef3382253d3ccdc75d57d1ee8834cc
cbdc46a60893b6a0c3d0e58f8199c9bb9c104ceb
3203 F20101109_AAAYDS stokke_j_Page_116thm.jpg
fa983d8ef14c174b2580c527dbb3bbe0
9b0fd552050fe4b94e1da1eed298e57251e7b754
2159 F20101109_AAAXAQ stokke_j_Page_022.txt
9d64ea90ecdc80b12d0b350df7251083
6249d80e67e0f1f55971411d1662d1fc1220f937
7708 F20101109_AAAXZA stokke_j_Page_025thm.jpg
0e4f3ad48d2a5361a968341ae09991b4
4d3a9ba293c6271af20290cbcfa4e1e18af417b8
7173 F20101109_AAAXYM stokke_j_Page_015thm.jpg
b24e7442aafe673cb3b227b3e3bc4d0d
602bfbb8072f6f6ef6a05a54d9cbe57d26a8c99a
810027 F20101109_AAAXXX stokke_j.pdf
eb4c71445f46f9ef73c032a4cfb08bde
9c02dbd468e7dd27bb6bcac88679902425677a6e
2039 F20101109_AAAXBE stokke_j_Page_017.txt
96a483b8f5506a52e8ae66f3c4767a41
a435634794a23a8fffb5ef98600d0a246d0314aa
6685 F20101109_AAAYEH stokke_j_Page_126thm.jpg
c21df2dc79d45d281ab5136b7e558cfa
62c752bf530a35808491e2900ab3dff32dd1e732
26025 F20101109_AAAYDT stokke_j_Page_117.QC.jpg
fa3465e22345830eb4ca82adc9dd1e4f
655d30a3bb1da8606830ef40cb8d5f25fceb9df0
52440 F20101109_AAAXAR stokke_j_Page_104.pro
aadb09fe0bc8c6f0040ef3ab625f1345
625c7807efa1e3fa1c74642d52308980d9aba354
4960 F20101109_AAAXZB stokke_j_Page_026thm.jpg
ab7b5c41f65b7a53a8adf2c745673f93
616b51353d116d5f868195ce8e61dfac665d5ae3
25093 F20101109_AAAXYN stokke_j_Page_016.QC.jpg
7502f1d7b5241cc88a3f3098a404031d
e247d6696eaba43b8a83533e89dd4f613fb36023
3110 F20101109_AAAXXY stokke_j_Page_002.QC.jpg
56f802b2d5003222ba16ee7f6b2c757f
679823797e365fcb4055afbe7114b124de0e6b70
24513 F20101109_AAAXBF stokke_j_Page_060.QC.jpg
57b70bf3c54fc3e252a887d64432feb7
eea7d77bfbc8fa892d7228e09a0219fe219bf1b8
9999 F20101109_AAAYEI stokke_j_Page_127.QC.jpg
a5abd81065b1e52188c64e57381212a3
177f2a0ed419e7d9663789df65e4b7c288fb998b
7117 F20101109_AAAYDU stokke_j_Page_117thm.jpg
24d9c9303a4960c151c0d5377a3bfedd
00a611dbd49bd40ec47da8eeeff0fd3c58823882
27819 F20101109_AAAXAS stokke_j_Page_025.QC.jpg
aba763e64246e2780f363363b3b3dcac
5bb40aa9482ff38d2235d32653269c3bf5e9b86e
7413 F20101109_AAAXZC stokke_j_Page_027thm.jpg
936bdcbc0349fa53f39ebdd31a1dd4c1
55308611b42dba8d3cab52b1c35d05ace7d1bf76
1345 F20101109_AAAXXZ stokke_j_Page_002thm.jpg
4d48777dcec4e84f6e29afcd6acfddff
cc4c5de1a5f1c094bc323dec973e7385a30144b2
F20101109_AAAXBG stokke_j_Page_105.jp2
6a38f05ca5703fcd03c0b81c2f604c2f
bff009b5d0fd15c063c1cdeac817672778a8a8f1
3169 F20101109_AAAYEJ stokke_j_Page_127thm.jpg
40132d1e2d9e6074177bc3126ab30e87
a29016af3802771c64d36ef0cde99d6ec1a7c05a
29012 F20101109_AAAYDV stokke_j_Page_118.QC.jpg
b5caa152b11cea9fcfad172f0e8af616
23f91c3d853030533327ecc876508bd4c172127b
2237 F20101109_AAAWWB stokke_j_Page_053.txt
0d511f092dc02ee24b95cdc2c1808466
45b4774602cef4ed2fb9337a1c7094890376ee4b
F20101109_AAAXAT stokke_j_Page_028.txt
2263c324effef180f83dfaca6f86e7a5
27bf7ff4b75838a6d6b247bee5bfe58e1c263bb0
27999 F20101109_AAAXZD stokke_j_Page_028.QC.jpg
f5df5c8535fbdde099b23001c2a33d61
b6f681cc0fed765ebed7c319fbe61d3d3da5bb78
6983 F20101109_AAAXYO stokke_j_Page_016thm.jpg
a50faa587207ab49a45b1d2920da465c
3c97e6c7253b61b668253c9e8d09fbe09167de62
1051911 F20101109_AAAXBH stokke_j_Page_042.jp2
fe39bd67bf636ea2124459368a48b007
5411282fc110412bcb9dbdd03d5ec9eaa98c49f4
147300 F20101109_AAAYEK UFE0022107_00001.mets
912d027c664c6ed7e82720888ca662cd
622970814f9bb546e7a527d4ce4aa23b9cc910ce
7626 F20101109_AAAYDW stokke_j_Page_118thm.jpg
60a3e0e5c912eab63f02cce4ac0136d7
5697225846bc91f9bdc7416d8f87f25d846c0065
7580 F20101109_AAAWWC stokke_j_Page_014thm.jpg
fbb4bc4edd1dc284d7473399496eb139
4d247b318486710aefc9be58ab982629d7d82311
F20101109_AAAXAU stokke_j_Page_022.tif
6ebf5306f0ea19de8868c7df5122f431
da9236ccb2df18a7678a09c1ead978e9682fda75
28839 F20101109_AAAXZE stokke_j_Page_030.QC.jpg
a0e844ccfb9c3c5945d25f93a8e3c4df
8ca0df98ef6f31a018f70eb53ef4f15ee3fad7b3
25780 F20101109_AAAXYP stokke_j_Page_017.QC.jpg
6b408321827469801c45c1d1732b4bb4
4db0e292c8f9ece9f9d6c14a7a62264da925b326
11825 F20101109_AAAXBI stokke_j_Page_076.pro
b5bf9ff579043119c46f43200d6ef0d4
6f19772eec128f2d08c4fe27200b387ecc244253
21751 F20101109_AAAYDX stokke_j_Page_120.QC.jpg
cd704edf74d8be3e5867f388a925c43e
f8d5ba2ffbb46773b33b27b140ca18a3faef0b38
27305 F20101109_AAAWWD stokke_j_Page_002.jp2
b2c118ecd2e70df0b4b767eb618558af
fef881716e3575e507bd99767706c0099a4c0b3e
3402 F20101109_AAAXAV stokke_j_Page_073thm.jpg
010e7579a3c0ab60f8472e8a3a12dfd3
8bb4c1d90f88680e913c46dbce11492c61b46325
7786 F20101109_AAAXZF stokke_j_Page_030thm.jpg
574623e286c24ffe8b4bebf691345666
33dbff22ef9cfa2279a9cc5c9b75678785711788
7112 F20101109_AAAXYQ stokke_j_Page_017thm.jpg
5ba822d5ba26cfd172eaf35cba676acd
c23b600a23fe1e9ee0fe9808db55b7e2855bb852
64731 F20101109_AAAXBJ stokke_j_Page_012.jpg
7b354f871be5fe36a13e5c559a6053cb
535f0075924636deb99eccd639ff1bbfb6ef6bde
F20101109_AAAWWE stokke_j_Page_024.tif
3ab97da221af382c5a2089b264e9daf4
8edf4952246ddf1b3aec1ff3388cdd9d85fec891
7146 F20101109_AAAXZG stokke_j_Page_031thm.jpg
da7cb2f5b1168598ea3c039e8fe02c40
4f65f3d4e1cde2100980893df429e6e562d71470
27593 F20101109_AAAXYR stokke_j_Page_018.QC.jpg
ff870361927054748902893a98ab830e
0d91b7d2329b4907a9e0b6846882b720a3935ce8
688 F20101109_AAAXBK stokke_j_Page_071.txt
e54dfbc59f3c11dcd0a1d7c7d625838c
6d7ec0e9951729bbddba3da33b8eb957fe25c2ff
5876 F20101109_AAAYDY stokke_j_Page_120thm.jpg
ef2793d44e252c5ad3ecb69a1f12b73b
cdaf9ebbb113cdff4840daca7509bf8481714a14
7323 F20101109_AAAWWF stokke_j_Page_021thm.jpg
e263b12cd2138e1efb820d20e3690507
bfad1c692519316916c22bb1a161361dad99ce61
F20101109_AAAXAW stokke_j_Page_106.tif
185bcc0ab042c0a97e9d505e9987da8c
7e6ebea66f5446bb34ccf805801f3e9ef94aa37d
28752 F20101109_AAAXZH stokke_j_Page_032.QC.jpg
672e7fa49e24eb300e2f3dbd8a4f3f83
0524ac0972020acae33a4f336e4e02f14cb4baea
7306 F20101109_AAAXYS stokke_j_Page_018thm.jpg
a3ce6e683da8f7e9b9f4626a470de916
19dc9e606c716de1b024cf86130ac0ac7d763ff9
7420 F20101109_AAAXBL stokke_j_Page_065thm.jpg
9d6effad6ab0bc8dfe0b7b1324cb37b8
82eee3478ed068ece5cd1fb3d206c0ba393fb3c1
7371 F20101109_AAAYDZ stokke_j_Page_121thm.jpg
9c628eb9ea1008dbcfe3d39af63f8346
6ef9e0ee6d39667ab449d017eaaf47c273c16ba1
56909 F20101109_AAAWWG stokke_j_Page_058.pro
87ee16d5bfab735e5240884c76f1f493
a2630f04414492b42bba825c91b0a270b966c0be
53979 F20101109_AAAXAX stokke_j_Page_013.pro
c3a7c28dc01d7aa25fd535498b1c2d91
953d996141639b6e00192562f2ccacc77291f9c4
7475 F20101109_AAAXZI stokke_j_Page_032thm.jpg
9264fef287a5d1c8f385dfbbb3bba4f6
1ed76ce99cad790cd5a4599e853c8d7834c2a5b6
19996 F20101109_AAAXYT stokke_j_Page_019.QC.jpg
70469be8d127745de17fb1547c550f40
e7cc6a01c585937e7d3a3e4c2f5a7ac8e25ab5b2
49622 F20101109_AAAXCA stokke_j_Page_044.jpg
d3395e8da0def0aa5ebb4fc1c9bcf2ab
6021eeb45fe5418c0e42921447c967c1bf82b82c
3738 F20101109_AAAXBM stokke_j_Page_005.txt
952f3631c797b367fbcc908335f194c9
e9a8092eebdb85335eb764c6fff2ab670f4fb853
1051950 F20101109_AAAWWH stokke_j_Page_101.jp2
8227c6a8593e8e03c535bbdf8e162c34
306f0427eb8de538d82d4cd32faeb4312e7f0c55
F20101109_AAAXAY stokke_j_Page_003.tif
89af5f22977cfdf6d5ca828a40ac4513
1496e9546ac85781c953c83acc1877f7ac4bed99
23961 F20101109_AAAXZJ stokke_j_Page_033.QC.jpg
140398194585758930b54087505837f2
d0b98ced33970a98cbb3014163cf46d64df1b081
5578 F20101109_AAAXYU stokke_j_Page_019thm.jpg
529427a61110b8965457c040350c4aac
ef27fe122db49c11f43f836d7c0847b83c000fe3
1051907 F20101109_AAAXCB stokke_j_Page_023.jp2
7486b2f2d7306b71e39215a46e099a6a
83dc9bab5ebefa92db4376bde5e033a4645fca8c
938 F20101109_AAAXBN stokke_j_Page_002.pro
72e9da635bdc9ef7af97eba444b106d5
5dc0575a86c653e7d7f09287cc3dc77f4d67a306
2089 F20101109_AAAWWI stokke_j_Page_082.txt
5d5148d106eb258c2187e0150bc8e7f0
f5170e9976442b9d27465fae8acd7b629a3725f0
F20101109_AAAXAZ stokke_j_Page_103.jp2
7e3566e082d305b0d76134217d621693
7d045a1ebc59849c5ee744ad314397cf391e33e6
6301 F20101109_AAAXZK stokke_j_Page_033thm.jpg
cef37c3a1c8305c5c1d9eb96c4a72f43
a5eb9dda32c99188b81a96b231128c1f528cfe77
26269 F20101109_AAAXYV stokke_j_Page_021.QC.jpg
875a8fe67ccf97f239ae0d2f43f9e8dd
c3efea43d71d0e0ec754497b74442ba3eac2f40f
3818 F20101109_AAAXCC stokke_j_Page_114thm.jpg
5db5559afe327cf54986b5927091268c
850d099f4961437295466d59def57054994caa40
56512 F20101109_AAAXBO stokke_j_Page_066.pro
c1ca5336517687a92d87d82b64b6f780
04c8cef47a97571a4bbe69ca21dce984e0835405
80952 F20101109_AAAWWJ stokke_j_Page_110.jp2
4597b5fa6364e2c5c1ad7bc0227ed830
e3e00755be061670d597b717f3d49ce22e454ef0
25480 F20101109_AAAXZL stokke_j_Page_034.QC.jpg
98ebf5fb935f93c26b38f5dacee5d31a
6eddb35bed21f57267b64d35e7c047b6137d7a9a
26985 F20101109_AAAXYW stokke_j_Page_022.QC.jpg
aaf34c68ad27447506828ef8c9177b5d
5a4b3f9ce234dfbef133f208b8cf6f0d5c42f9d1
86664 F20101109_AAAXCD stokke_j_Page_054.jpg
79658c521100269c54ebc9d9501fcad9
a9a391ecf7dd49bbb9b34f3c04d72b3f3fd6446a
1051927 F20101109_AAAXBP stokke_j_Page_062.jp2
b71ce324e591f60cc9023560b05be17a
cc7b01b39566ebfea4f4cde265cbb9749e66851d
980 F20101109_AAAWWK stokke_j_Page_116.txt
ec85d025eb292d829fdd1670fa9494a9
4fc006363e50388945363bfc215ff6a5ed37a1c2
11797 F20101109_AAAXZM stokke_j_Page_035.QC.jpg
19574ce38fafa89843d81f15e8a1e76b
366de679dcd6284e5a1e9b2142b72d7a883a6c79
26727 F20101109_AAAXYX stokke_j_Page_023.QC.jpg
413367ad14c60fc0a1c2d32cfb15aaac
a113d9a3cc44d7d7fc8310335ccb60009ab1b3d3
2268 F20101109_AAAXCE stokke_j_Page_030.txt
116aa3d15006e94236bf7c5b9f4ecabd
1804855fa77d081d5e080e4b916c88eeea624705
1051936 F20101109_AAAXBQ stokke_j_Page_039.jp2
6ea88718eb0efeadca9cd871215a3722
5d8dd9c319d83ca132ad6ecc42373d501f534c65
F20101109_AAAWWL stokke_j_Page_094.tif
d361072c87873103b0259c7067fdc46d
67c081f03c92ec9aefa9687bacc0387895eaa2d9
3606 F20101109_AAAXZN stokke_j_Page_035thm.jpg
5b6eca68e43a72ad37da1c9d57e528ac
2ba87ec3165e740de5ea9134fd85207a239de9ad
25979 F20101109_AAAXYY stokke_j_Page_024.QC.jpg
53029568bcc123472fde14d43eea65cb
3da8b15fc3108dc99d74554b43502b749c0a02e8
F20101109_AAAXCF stokke_j_Page_003.QC.jpg
71e93e4f529893131ac47e44c39d1238
31f2164cf546c10e299c7f081e0b80b8ecbcfbc4
F20101109_AAAXBR stokke_j_Page_109.txt
28839aea4e827c0ed9274a4dad382588
c61ab94fd8f6c343a65a7dab8c26cd8b67754cc2
F20101109_AAAWWM stokke_j_Page_047.tif
6433490451c7df7ea7de53133ae83741
ebdb37302feea0c284800a21a2807487ee35242d
25161 F20101109_AAAXZO stokke_j_Page_036.QC.jpg
7f41712ead12129fc4912e8eb0b3090d
2c979363e3302525a85af1d5a6f215baa448f15d
7087 F20101109_AAAXYZ stokke_j_Page_024thm.jpg
7116e9ad359a9bee86014c6d14015a90
037be3f78630fd5601fe89528fb1106212c3bb75
14277 F20101109_AAAXCG stokke_j_Page_070.pro
5f60b5bae1e172cc58db3767fb231abf
7ece9aad8930012591e7b8a473ad9ae39702c3d0
567424 F20101109_AAAWXA stokke_j_Page_049.jp2
3e93e80e603d3bd0461438ac3feb73bc
0deb9727e87dda95dd03c5a465ec956a8f97695c
22620 F20101109_AAAXBS stokke_j_Page_005.QC.jpg
2931e07b7df3487c247a505f3cec9241
a43efbe42b0f76037067ad9effc8b425a5c4dcc5
233891 F20101109_AAAXCH stokke_j_Page_113.jp2
63ef13f35f166c70db745e066576b10e
5a411ca931f57fbf2de7cfba15aed2debc59dce4
7447 F20101109_AAAWXB stokke_j_Page_090thm.jpg
573e0d964ab5ec350ca71e2578de3321
1d2ed9a817ef2e051725a1e871d39c28ae04d3ce
1791 F20101109_AAAXBT stokke_j_Page_110.txt
d90a1899964af3f3609076f56d02d1b6
eac637ec19541b8747122712fbe2ae54440de405
F20101109_AAAWWN stokke_j_Page_054.tif
e67f34d1f218750f63a3d3242da0bdb9
4ec58cc3ee211d24589dc3d0e5fb9df3590d7bf1
6803 F20101109_AAAXZP stokke_j_Page_036thm.jpg
bb57cb482f30d03f690b68fa0a93ff79
c091f1184626e932071e892ea78563e773636df4
5932 F20101109_AAAXCI stokke_j_Page_096.pro
e7f2b7a177f1f9b0be1b91677830568a
7d2ab83b4bff46c5a3c1f569f9a3e05a9a21099b
82143 F20101109_AAAWXC stokke_j_Page_119.jpg
2f4c4c2497e969552cae5f6961960029
936b94f4629e939223a1e51a9901b465de11d94e
49923 F20101109_AAAXBU stokke_j_Page_015.pro
5778c420a2abba6d4ad07c8660abf50e
833dbbcbb1b20874fa64edc58c2ba1280d8f24c6
1051891 F20101109_AAAWWO stokke_j_Page_109.jp2
3b95ecd53dc7f465deee215d43690c4e
fc59e37d5959d976ed19caa36a07f3b88176f1ed
28198 F20101109_AAAXZQ stokke_j_Page_037.QC.jpg
17538d54a3e3a591fb6814b341135b05
e48ef5ef5c1a24f54afdd68576d564290a9e8c29
F20101109_AAAXCJ stokke_j_Page_095.tif
9cdaa2dfd85741acbe53b4ba32909a16
f0d344f7c79aed0995ff10d10e53a000bb1d52ef
1040 F20101109_AAAWXD stokke_j_Page_073.txt
5328c2e0cd994bc739ca8bc9f4b85c03
80a177b8ac0a25491dd7290eaa126e2273f4d0af
7268 F20101109_AAAXBV stokke_j_Page_083thm.jpg
6d811ce211bd865977bd4e6af888e86e
3b8becf5923a3a69ef11aaa69258fd764785ec16
7487 F20101109_AAAWWP stokke_j_Page_109thm.jpg
526ec1205a26b0a604cfef43a2a63878
4e0cbd6f8c15bc0c0b6796f0b4118443774bd345
7614 F20101109_AAAXZR stokke_j_Page_037thm.jpg
1713fa4f8cb00196529f3b6f6adcabe6
ce235de4de28b6f0fa8e7427d15fc4c1efb3967f
90476 F20101109_AAAXCK stokke_j_Page_118.jpg
2903a34bc08661097fd69cbebaab7601
89e6637ba1cac624cadaf10b3258a9974efca7ee
2154 F20101109_AAAWXE stokke_j_Page_011.txt
ef2d5bb9133d3fcb5ccc296369a50e11
029c8d70b5a90e609733d21cd61817c43eec0425
77867 F20101109_AAAXBW stokke_j_Page_060.jpg
8d1586a4477ddf58622f33553762d28c
440b813e0b9a17e42d0b581b7f020ae25871c636
F20101109_AAAWWQ stokke_j_Page_019.tif
b4b55192bbae4ebff1ef087b46b15b27
38e6acaf2c8f7f4cb7dfd8da6fccfa1d520f7399
6913 F20101109_AAAXZS stokke_j_Page_038thm.jpg
6ebd00085276a64eedfdbd3bc68b3b60
4f88fb3f9fb9d73ab4a566e046dd7c9a8368c7ce
7328 F20101109_AAAXCL stokke_j_Page_078thm.jpg
2f6ea73330a2cddd2704f4393c7c3ff0
eb8de43f6958dec7a576ab937a5f0b9f7ca839eb
85298 F20101109_AAAWXF stokke_j_Page_082.jpg
32e05b9db739a3ab1ec038fa79cab4f3
7bad7ea9376961e9e54008fbd95ef88e387b4e3f
1051941 F20101109_AAAWWR stokke_j_Page_060.jp2
9f6240775fbeb924517d08351e1ab70a
7cbfd5cce0fa7d2b7cdd4b5728b809882679f013
28150 F20101109_AAAXZT stokke_j_Page_039.QC.jpg
5ac19c5e91b2d6186fec7ca74da323ab
318295bd18f97f14bca01c9c3d305e91dd9712ec
88774 F20101109_AAAXDA stokke_j_Page_058.jpg
8f06b6f4af6404cc894119ee5b335bcc
04a4c7e322f743d9087bdd3474f659a32dc91e9f
2122 F20101109_AAAXCM stokke_j_Page_065.txt
f3d38b4512918359c8da4ebdc40c7e69
544f0b01bbde3474d90343fad7ff8e5c0b912029
19061 F20101109_AAAWXG stokke_j_Page_020.QC.jpg
8cbaac3c3e5b06b1ea435dd16fc33b94
904f538bdaf47bd5f9539400d5c437629ef2df17
13997 F20101109_AAAXBX stokke_j_Page_068.QC.jpg
c118eee7488efa9d40c3af9058d4b986
240b031af6f86d5ae50ba5f0b216f77bef4a1cd3
F20101109_AAAWWS stokke_j_Page_084.tif
3d6a0420bc87c390feaaa2df63d4ceb8
2ceff701a912cd7453934429134aab5866c5d663
21478 F20101109_AAAXZU stokke_j_Page_040.QC.jpg
0428d10f159120ad869cb6637906198e
ef0519bf463c62559ded56e6651449e56de0837a
48044 F20101109_AAAXDB stokke_j_Page_036.pro
3405d61b307e02697c2b7d47ee52f213
24ba83a758ae44a96763ca03d2d61852fd6e6dc6
8636 F20101109_AAAXCN stokke_j_Page_113.QC.jpg
2020fbaddec53e3fbc44c29874383995
496f15e3eec23fbf3a74035f41375272cde96315
F20101109_AAAWXH stokke_j_Page_046.tif
8745c8c21b998821a5381a2d46b3d987
f6e2588582f606f0168e712ad62c04f69c092fcd
F20101109_AAAXBY stokke_j_Page_043.tif
c34372fefd6e02f78cc6990842fa45f9
0c7598b0a1a2f17e6a56492884bb35d7ca472aa3
54168 F20101109_AAAWWT stokke_j_Page_016.pro
146b4072771225e50a7454de69e58121
1dcc3860681615f1bac9781cbf25ac46f83b011a
6387 F20101109_AAAXZV stokke_j_Page_040thm.jpg
946b6bd6e2244a860eb49df4e75dd7a7
3114548ac25525fdf8b9a39297572b4c3fd21b33
F20101109_AAAXDC stokke_j_Page_011.tif
dd75497d8e9738ee9e6255b092b16271
a488caa0ca4a678c99e05ae1e27b48f937f3af1e
18571 F20101109_AAAXCO stokke_j_Page_087.pro
17fa9f0b09f33bd41f414f27593a0178
7db41af160d2c1b663c41337d816b0516827d0b6
F20101109_AAAWXI stokke_j_Page_059.tif
a17f26a8383fbef3f1002a0e7527ab52
878a1f7b9744a5ac5ced43b07e305be1b50ca2ea
F20101109_AAAXBZ stokke_j_Page_058.tif
1ea5531243fc0d5c9722c85027a58c36
90ff3ba5ffb77ca3b3efd721c472615fc3895f31
F20101109_AAAWWU stokke_j_Page_074.txt
5cbecd16cb696235299677ea7abeaea1
fd3c3bfc12029600f72d56014d309e5b690c5ac6
26039 F20101109_AAAXZW stokke_j_Page_041.QC.jpg
0804788a172bde6357f30d8899aee6a9
88ff088307d090ec5a8df9c81a881b80d0f7861a
87994 F20101109_AAAXDD stokke_j_Page_045.jpg
14def8adf6bd1283d5f86c0dbdc195a1
8e40a018f1d572f42ea9640fd3253b2343cf01b2
51527 F20101109_AAAXCP stokke_j_Page_085.pro
f35c658a988bb95b71f3843db2929b69
06e5a0c61c0ec29e0794458e2ecf0b23b41ebd73
28257 F20101109_AAAWXJ stokke_j_Page_079.QC.jpg
75b366b569baaef8ee888276e6015540
6cc4135c67915c34b099a77f50dfce7c11828626
14487 F20101109_AAAWWV stokke_j_Page_077.pro
2605734359bd180470ec5c5730c18b06
7073552bfad8b698897a843b9664a81b2ca8841a
7187 F20101109_AAAXZX stokke_j_Page_041thm.jpg
c0e98b8481e09dc6899fce0ec32fca01
f465a5e1e7e98a9ae59897e34a6a71135a1b780b
36273 F20101109_AAAXDE stokke_j_Page_114.jpg
ada5481a6e4818552731494add6d297d
03201ddb864d46328ab78664d3fbe889d5ab5ba0
56218 F20101109_AAAXCQ stokke_j_Page_056.pro
074122a12078ce1e8cc8c8cf573cfe5c
1a74a43e3acc3b863d500af1c738561390e0180b
29456 F20101109_AAAWXK stokke_j_Page_123.QC.jpg
a94cfd316e03349a6e0c3da94f4e046d
be22d80ff2579d703aa6d65a90cf836367c81e89
86928 F20101109_AAAWWW stokke_j_Page_018.jpg
bf24196021e0841e785dd2aa64781a9e
c7d39c971eb94f2bca7e2d96b3fc80fc8804b637
24711 F20101109_AAAXZY stokke_j_Page_042.QC.jpg
3745f066afba58c926e164887bff828a
7a39546e7f6724144caa04cb5e40b97d7b9e1e37
56775 F20101109_AAAXDF stokke_j_Page_081.pro
bae061e83da7d54ea7cb1b23639b8fe9
ad4f5158d9cf6edcefd700d8444d38d5d1d2d4cd
49002 F20101109_AAAXCR stokke_j_Page_035.jp2
d9bab801618519154e4896ad7daa6a19
f896da9b5c5121344178b9b8127056994c61fd22
37946 F20101109_AAAWXL stokke_j_Page_086.jpg
dd4875d02f55662f9b55d62e3db1abd7
817261af40388ce9531bb6853a805e9169fbe6e0
27941 F20101109_AAAWWX stokke_j_Page_058.QC.jpg
69eaa4bd6e6313e0c2ba6aaa6cc6853b
b091ef9f9bdba7aeb790aae68ac32a54025e26f6
6751 F20101109_AAAXZZ stokke_j_Page_042thm.jpg
a0e1cdf878971ec2bd454e014042380c
1cf92aa647a886dcd20b69edec88ab16f3a52b5b
7489 F20101109_AAAXDG stokke_j_Page_108thm.jpg
1d5aad48aa69243086ecb570bd60f944
778d86239f27018094d5077954dba0fd93644259
2224 F20101109_AAAWYA stokke_j_Page_039.txt
38157d013b3e1891d79fbaeb61008aa2
e42efa88436a063002aa3b9fe8a7d6d3630db10c
F20101109_AAAXCS stokke_j_Page_081.jp2
363d743adc3964f7bd4ed316e4518c19
418d2f85ed65bb684a7e34faf5b6a8fd89aa2e68
F20101109_AAAWXM stokke_j_Page_098.tif
c5f62238ac1a4d1400da167ed9d1646f
652e0e99994ec49a993c465878810a3f6fa8c564
5558 F20101109_AAAWWY stokke_j_Page_110thm.jpg
f523a0822784ca2df98adba23211470d
22977fc06c3c972bc16cd757a8d04b41b12c3aaf
49098 F20101109_AAAXDH stokke_j_Page_120.pro
347511fdf159701f066e6454dbcf1a42
34a385d8bc23a25a30b9e132920a85d46101bc08
85060 F20101109_AAAWYB stokke_j_Page_065.jpg
e016cafd81bc6277b92008adf7a469fe
7367cbd9b5e5a0806a69af4a472394e2a3f63cf5
F20101109_AAAXCT stokke_j_Page_067thm.jpg
eafd8db637c26bda0a0a174a840fa5e7
07dde61b6ab31bc7f79ee755d35e2c7ba56210f4
24615 F20101109_AAAWXN stokke_j_Page_068.pro
692196a6279ea3bb24a8d0e74898fb40
7f77e0a4c76453fe3966c8d956350b47173cc678
68483 F20101109_AAAWWZ stokke_j_Page_040.jpg
ddbd1df4bd8e881accadec74feca4a4d
a1a7ce6c45d045fd3e9059960798b0d266650b32
2218 F20101109_AAAXDI stokke_j_Page_045.txt
9820e8fb09c1b4551adaab05f6a79f0b
ae446c820e3bdc54c8907c2d06e210feb46c70ec
782 F20101109_AAAWYC stokke_j_Page_115.txt
55b4ef1b8b31424cec6ec92c670d47af
b56e43e2b3768be3a56f0118311e1c5349f733b6
7555 F20101109_AAAXCU stokke_j_Page_029thm.jpg
a2a3a29ee967e4f3f96ae8f5c2c73670
3de4e32774169caa3200509e3b50970af0b4e218
F20101109_AAAXDJ stokke_j_Page_112.tif
6bdc5f17ae7e6374ac7e2dd603b39527
3d207a45bd13cf85d07134e30140ab9eb0afee1a
F20101109_AAAWYD stokke_j_Page_033.tif
e6f26a6a6ff56fa17f3c8d05f1a75dc7
c35c931d5eda8517012660595afafa2d0141e8d2
2581 F20101109_AAAXCV stokke_j_Page_122.txt
3df649bacaadb9b3bfc18bca45595b72
199f8a9b7b983760c330b7060b52334974a1711c
28675 F20101109_AAAWXO stokke_j_Page_029.QC.jpg
d48acfe424cee209abdab94f7d9c331d
ac5880f93a1c69811dd0288883dedbc15c0042bc
1360 F20101109_AAAXDK stokke_j_Page_003thm.jpg
26a2a54567ff5fe43ee37add9aa803f9
5e5974b501c0647552efa87e7c86edd870df1d50
439057 F20101109_AAAWYE stokke_j_Page_076.jp2
0ec5b9b602a21ea1ceaceb2e497edf33
d9332f59290cbbdf8cb304bb0163a8bfd29701a6
5600 F20101109_AAAXCW stokke_j_Page_004thm.jpg
ce208f25fbf4b89f25cca4d561125afb
ada46c6e6ed0512798c6900416f0f58273e72602
148 F20101109_AAAWXP stokke_j_Page_099.txt
8795f57d5d363939a8a4bc365d3d5d1d
e645bd32cbf99710c10d3a2afc389c14b0b8eb60
57349 F20101109_AAAXDL stokke_j_Page_084.pro
f7a829ad4cf99f8633a4c96c146ed856
9c7a67e77c7de250616e6ab9ab2e1a623d25585a
379070 F20101109_AAAWYF stokke_j_Page_127.jp2
8c3830de85679722e3764997f4da9350
4b9c51dbf6222e8c4a6f1c8acf171d64ea5b907d
25973 F20101109_AAAXCX stokke_j_Page_038.QC.jpg
4260684f40f3f2b8d9e21aac88ebcff0
31cff6bab5af8719e8006a85f16fc97bd6bf6d5e
1961 F20101109_AAAWXQ stokke_j_Page_064.txt
6702ebbe5ba9463e370fcba444e65ba7
1d24631e027be918d3530ce3cdefb5382d651ee9
F20101109_AAAXDM stokke_j_Page_037.tif
25f0881f463b3721e1829416fee10788
dcfad27a72c6730f3afcdc292351481a7b30919b
63109 F20101109_AAAWYG stokke_j_Page_026.jp2
2aa68b301ea8a56d7c03ea67caa471ef
5ed749f74dc8dbef4f607e6bd5cf4d415de83999
F20101109_AAAWXR stokke_j_Page_119.tif
91d75df13dbadfe1302b466e28ea2a5f
4ec6a5898cba31abfb0bfd7735c0f970eed9f518
81149 F20101109_AAAXEA stokke_j_Page_015.jpg
5088a161d79aea30fd5454f4d623556f
e8afa523b5c92238d099d598a2acf99b2eec7a6b
27650 F20101109_AAAXDN stokke_j_Page_124.QC.jpg
8ad895641b2adcf270edd6467ede5fe8
df81a5500aa4b23fed950f264763e0037ea697bc
45004 F20101109_AAAWYH stokke_j_Page_068.jpg
63128f64e3c845fb6cb32bed34c0cfa9
8ad6a2c8813696ee125408b86b7fd2ad80863207
26663 F20101109_AAAXCY stokke_j_Page_105.QC.jpg
66eb4e33ca8b807e29c936b4ffb1c80f
33dd340304c40bdb2ca8f7c03435e288bd54b18e
54907 F20101109_AAAWXS stokke_j_Page_023.pro
b9ba26c15b99e018347647dff471465f
61785f3acacd91473e32781a7e43e315bcfe19ac
F20101109_AAAXEB stokke_j_Page_009.tif
1d205dde24ef9097ea0620604fe8dd39
d4bdd060adf4a9664d125f67d8a6d490bd99381b
54129 F20101109_AAAXDO stokke_j_Page_052.pro
e9ff4504e07befbda86dced0d7d1013d
b4c8031bb529aa0c19e1f78e24a268379a06c187
3937 F20101109_AAAWYI stokke_j_Page_111thm.jpg
7d96c2b48d5bdba7708864ca07e26892
d20497209d2842326fdb7d29699ae80a3b258096
55374 F20101109_AAAXCZ stokke_j_Page_018.pro
31fb01416714990d2d1091171fb0e3d9
938d2db6d2b78f438ad25b7663168e5935d79c1b
F20101109_AAAWXT stokke_j_Page_125.tif
3112bdb781ab7bd6902319a9332ab3a7
d5f494895d7324fdacf10789c84a1a9b8f3ee3c1
2609 F20101109_AAAXEC stokke_j_Page_123.txt
a4c15de499ca34ac3c98426699b437e7
a0cf2cce0c34ee0debdf91e8faf17e701833b41e
2116 F20101109_AAAXDP stokke_j_Page_048.txt
28dc468958070a545f805e09b90e4c34
37ef3d106b4bce219ed34a9d27b4aa74219ef635
12862 F20101109_AAAWYJ stokke_j_Page_050.pro
d88289da2607ded236b5756737c4783c
b898796e766e48ef1aa94873ff09e4dca99f8ee9
1051971 F20101109_AAAWXU stokke_j_Page_032.jp2
b4631c6af8d59dc96d04974ef55a2401
d5ad167db65ad217888835c247b95d3ce6edeec4
82086 F20101109_AAAXED stokke_j_Page_034.jpg
4aa85939b94fef54a54cf9b31905aa48
b7f245d587872b731755516eb1d8f96d274e85c1
F20101109_AAAXDQ stokke_j_Page_016.jp2
150ac2f841461a72c8b1ea201229ed79
8ec19dbfd85be77d9b85065e09e665159be55acb
F20101109_AAAWYK stokke_j_Page_059.jp2
40e5ea85c5933901e0c81929fc0f289f
6e398bdd44d08dab9a2a337c24e6eae7c9e7dae4
51128 F20101109_AAAWXV stokke_j_Page_117.pro
70856aa31b3d078d3e66ee9254455bab
5193cf9d5296cca3bc18bf6827751c5b0790d18a
13027 F20101109_AAAXEE stokke_j_Page_114.pro
9131ac400f263b18cb6b1e6ec62ebe52
258616cc80d8f446a12666573a3eb17d97d5521c
52670 F20101109_AAAXDR stokke_j_Page_021.pro
a24eef94e56fbd5cf0d6ea01cc88d98f
366d33ba42b6ac2dcac49d59571de321c6054632
57385 F20101109_AAAWYL stokke_j_Page_110.jpg
afc48a77f383dc6bb0a7d9639288d65f
68cf057b3e27a03e3cbd77e3adb35501ffd202ab
2178 F20101109_AAAWXW stokke_j_Page_025.txt
b2c63b6d5ce709aaaf8d6f35d414e790
dc06a885b38ccfc3db9ba988aee63da09cc6bb23
3602 F20101109_AAAXEF stokke_j_Page_010thm.jpg
11b7a54d1c9bf09557f698f5e286dfe5
bae6d48e5cebadce592dd22b9cb0380ec38451f4
7235 F20101109_AAAXDS stokke_j_Page_055thm.jpg
bb650659f98f8fd597f1ba1473bb0bdd
96d1839656305988f966bf047d9f25a0e649d422
F20101109_AAAWYM stokke_j_Page_085.tif
981addbd4070fb9c1ee2ca7b809df008
0dd43162195c36240e8dc28e7b458b2042576af4
F20101109_AAAWXX stokke_j_Page_074thm.jpg
165cf868b83b1826b0bdfea8174281e7
5e709055cf025a21e4839a5c2d3439c61906eda6
53893 F20101109_AAAXEG stokke_j_Page_048.pro
681a302da93393f71dde3e78a04c5c1c
9dbeaf6319b594cd82d1bfd39c4ab1a32f63f8fd
632098 F20101109_AAAWZA stokke_j_Page_044.jp2
78998d89a51f6d2287cf427f92ea5b5e
c9054a81dcbe257c31d82e0b3d50bece232add40
44236 F20101109_AAAXDT stokke_j_Page_010.jpg
9a08093c01b4dccbc3098c852b400550
86f84e5de97c3e5a9a2a6e03818631a65e2bf245
37175 F20101109_AAAWYN stokke_j_Page_074.jpg
8c68184ea44b73e1b950e793f934d564
0dcd98990620d6c602fda7ba8992f7c08603e091
22217 F20101109_AAAWXY stokke_j_Page_072.jpg
29e6b6b66e566b190879c2cd7e750fdf
644607ec30e1e8ba53da74372719a4c102b36cfe
F20101109_AAAXEH stokke_j_Page_089.tif
e02181f3fc84fdf1bb931bb73d858911
4d8dd1d5d5d1641db9f6cf075747b169adf3b8c1
1051920 F20101109_AAAWZB stokke_j_Page_084.jp2
4b122337570927eae0b4197d336b20be
d7d429560cb388ca8ad0709e71a87e26e24f6108
7058 F20101109_AAAXDU stokke_j_Page_034thm.jpg
6f631895ed2d1ee844ec9ea8897711a6
ea14a05a580af627784cac25194078556d4635ed
7375 F20101109_AAAWYO stokke_j_Page_057thm.jpg
f19427c57b50b28966ebaf025a18701b
6e0ed8a993b724e5c70d36a5714836a9d511ac4d
28495 F20101109_AAAWXZ stokke_j_Page_093.QC.jpg
0ce1378e51d10112fd7d3a18a94080ae
d291f7937ea6edef66bce037c2caec9b491b54b8
499016 F20101109_AAAXEI stokke_j_Page_087.jp2
8c32263c7e9a4b553529d0c8bd03c8ef
5c11f935e9d7f348d5008d529c34e4527fd62a87
F20101109_AAAWZC stokke_j_Page_079.jp2
b4c70691155f39e072688d48c552b651
98e0257116461fd88bc59915a0941493d0149f92
F20101109_AAAXDV stokke_j_Page_083.tif
621eec9c11e750f2adb6f33763557ad0
a89353f1e8add0fcb00b6dfc27c7d82096f1a572
7624 F20101109_AAAXEJ stokke_j_Page_039thm.jpg
43a5ba7cf656d33bd8d6f71e64741e9e
f033dbb73a0187fb97026693dc95c5b60154508f
7410 F20101109_AAAWZD stokke_j_Page_102thm.jpg
18b73c2e801fec65062d4031bbe23c40
7769a6c7b7456567515a61d74155c69a54f9f0b8
85152 F20101109_AAAXDW stokke_j_Page_011.jpg
19883b66a9c19ef203b4aa988e2ae450
e86b9a239c41c3f914a10d29f1511095c5708eea
1051934 F20101109_AAAWYP stokke_j_Page_022.jp2
d8dc0ce2e3f7f74de97f55e5e990efe1
b71be3dbd3ed2df21efaacd9445fc23fa79e5dbe
7120 F20101109_AAAXEK stokke_j_Page_107thm.jpg
1f764bb4afdc7e665dbc178fb101e45c
d2abe9641bd53969dd52dd847b678d4e9da31cb0
7132 F20101109_AAAWZE stokke_j_Page_023thm.jpg
a1b53c7982ce990505784ad8e2c39daf
01ba25aaa4722095703b98d5af4ce0ea674b23e4
2082 F20101109_AAAXDX stokke_j_Page_063.txt
a9040e72c9d7703b71c1eef2d1a981b5
b8cdb81a87344783cd170aee29e1dfafd09d8735
297 F20101109_AAAWYQ stokke_j_Page_072.txt
e834a4f5d65bf983e5385f64811c8f65
0618fd1c2f4c10ab380b1c4a8f2ad2ca45c89716
7041 F20101109_AAAXEL stokke_j_Page_060thm.jpg
1cc0b91aa87775997a9b54d5a67b95ac
cbb48fe01d809c20c51d026a1566b3b62c7fa26b
F20101109_AAAWZF stokke_j_Page_056.jp2
1c6a227c3c423293934ed492514e9751
982a9f21ba103e70d8a702349de617d964aec46a
F20101109_AAAWYR stokke_j_Page_057.tif
258baa53a58a1e4b5c1fbc0115788f96
525da755c99741e76dd1682de9e05b17d4ee0438
F20101109_AAAXFA stokke_j_Page_014.jp2
abda434ae82c61f91b83433fa97317fa
54ef0e87998951a5237c4294e80be3f7a0d8d49b
814744 F20101109_AAAXEM stokke_j_Page_095.jp2
ad08dac70fe14d07cf920f5b42b65818
6cb8f660cae42a72b6aa53d18c2d01fb4bc80119
463332 F20101109_AAAWZG stokke_j_Page_088.jp2
c4d5e51ab810778a18ee9766566f69d3
58a4d3458f3e285c055abedc0acf7018ff1e5b15
8219 F20101109_AAAXDY stokke_j_Page_001.QC.jpg
842c154ac0ba9a8fb261d9360a4a109a
ca5dfdff7157953ffbcfdf4e7965ec9bc797e128
39580 F20101109_AAAWYS stokke_j_Page_076.jpg
ea814b300df305aa09d80e8da940cd6f
0e4288b239e3aa036fb3f66175df70d91f03cf5c
1974 F20101109_AAAXFB stokke_j_Page_041.txt
a68e7b91166cc62c9a827d9691f340dc
f971d36abf2817c696afd22a01142b5a3c9b0061
2211 F20101109_AAAXEN stokke_j_Page_066.txt
ccb522b8610375f457b716cced0237d1
ecf2e2a410d30060a5907b8868b8f0fe7272b78d
14082 F20101109_AAAWZH stokke_j_Page_088.pro
4a4a532e9047fd6fb4f4c845e889ae84
0cea327767cf661899a353ac9d9493ff7eecb4af
740218 F20101109_AAAWYT stokke_j_Page_043.jp2
54f213edf217829171ffd7e73d5bb003
7735198a592b822735890e0f3797aa52ab5105bc
102673 F20101109_AAAXFC stokke_j_Page_125.jpg
e98d0b48b3f5f4c690343ad44f44b9e2
8dfe9c87f1b14234c5d7a75023d8012ff0c56c08
F20101109_AAAXEO stokke_j_Page_107.tif
74f9c9c5593b12b2f1cf7efa1d5dc9dc
2b6afe8e01692c3f4a25eac13e7e4210d2cb746e
37871 F20101109_AAAWZI stokke_j_Page_110.pro
e5192170dfe4c3bcd0a11d2eab500529
45e4e2d922aea59456fb6c52b051d250e6bd42a6
12949 F20101109_AAAXDZ stokke_j_Page_051.QC.jpg
c188298ff64e6544f8e99e99026cef96
68662404307d13f3ba99abbdab45896a4a6bfac7
25397 F20101109_AAAWYU stokke_j_Page_011.QC.jpg
56c6543c29f6f0e63ba90779b1882761
e1559da999e8702bdf795385452c5e59deab5d56
87911 F20101109_AAAXFD stokke_j_Page_053.jpg
f1d57bcf3d4f79dcd707d9ef9491dc3e
3a8d422b7925a562ccd21ea0fe86e8b3d9aceb22
F20101109_AAAXEP stokke_j_Page_037.jp2
0e848d6b415691e51f150b5c812f0d23
058ab51a3b4d4c55833e5eaed4410c9bb59789b2
2073 F20101109_AAAWZJ stokke_j_Page_119.txt
bfed5c87f6f64d0a99d891c873240fa7
45297b0dbcb242908820f34dbbe2844f041e6736
36041 F20101109_AAAWYV stokke_j_Page_035.jpg
dcbfa129fa3b9b71ae336b2f094f95f2
7a6d2f7d3977c86f3ceb87fd802f0ca41f7b02b8
2188 F20101109_AAAXFE stokke_j_Page_067.txt
6034ec9e3db3ba7055ead3ba41a68a7c
933c622e4ff8e98f87268991e042efd1d3549f7f
87829 F20101109_AAAXEQ stokke_j_Page_083.jpg
4b33f7f9e0bb88a84eb03d59eb1db7d0
6605a5a4aa442c709c18db99d6d274d16a1dafdb
5729 F20101109_AAAWZK stokke_j_Page_012thm.jpg
d2fd6c87e48aff1224e29175211747da
b26d753c3da0e4b325f2ebd2b7db92c93c13872a
F20101109_AAAWYW stokke_j_Page_040.tif
9d23167bbeccf2bf7f18f40c813867fd
908bc9fa1b95165794f8a1f27ff1655394d6065a
27380 F20101109_AAAXFF stokke_j_Page_121.QC.jpg
95227deeed59686d07258c82cd5c3637
dcda2efc31afcd69809251660667ad38f6a85009
26932 F20101109_AAAXER stokke_j_Page_046.QC.jpg
5f7d61ab0f634dceb8301442c756fb2d
6cb41da6671b38a283493293eef69823992ab545
93462 F20101109_AAAWZL stokke_j_Page_124.jpg
1befabfbe18c8f4f9e83add6b8d1b14a
03e8b66322445d517b1f2298659ed73e8404856c
45405 F20101109_AAAWYX stokke_j_Page_049.jpg
5b941db70f67b0b990d07140a8379d55
569a04aaa1f30a8a523223661b771c1afa539072
28711 F20101109_AAAXFG stokke_j_Page_009.QC.jpg
3dc7b1954aeee6c86e3864c0e5afdf27
058d98b9f20d6067e2eddf72d2274a74262e9d81
108291 F20101109_AAAXES stokke_j_Page_033.jp2
26a3339b994f946b59cb6eb5a2ccefaa
30cd817a65a2186afaec5400850d3a292a83dc0d
34466 F20101109_AAAWZM stokke_j_Page_071.jpg
09458ab6e9bdeea25a790fe2c2469fee
4dc241f1f1ffccffa6400fa9dc86b784472c118e
53148 F20101109_AAAWYY stokke_j_Page_092.pro
494adfbfbdad8d149c9e7eab7e5c9daf
f52f1ed62ee71e52ffdf2fe2212eb220513296bb
9412 F20101109_AAAXFH stokke_j_Page_113.pro
3003c854bf8c779dcbbc62859cd4581f
d867676261dc19c8000075482a5f0914b9110b71
F20101109_AAAXET stokke_j_Page_102.tif
0cd52020b9c62a69e64b0120afe53107
0ba52ae3e628ee187c96f417985e34b818fe4d31
6949 F20101109_AAAWZN stokke_j_Page_064thm.jpg
a599a0fa68bb495a83e664d1eed46e64
73670c346c548418b1978b8a1f7101e50e449cf3
F20101109_AAAWYZ stokke_j_Page_110.tif
24405ca38f3352ff11d412d1b37ddf76
c680c52fbddd939edf8538a2dd876a43ed046794
F20101109_AAAXFI stokke_j_Page_109.tif
fef9856738897304a988d28059757a19
004cdc38e30a274920b6ca6be80794d89e0fab6f
7535 F20101109_AAAXEU stokke_j_Page_079thm.jpg
af52a6beac46b8eea44fed41335c55b5
c422533b0cda9ff773aa2ba705489c1eacc2d5b9
11652 F20101109_AAAWZO stokke_j_Page_071.QC.jpg
a5458e37c9676e60ec2ada768f84dee1
5552c622d545c70ba155a9228c4ffe2e4bc9ce96
F20101109_AAAXFJ stokke_j_Page_082.jp2
b805c5edaea224499d69efacf7c1646d
d10f20c03d9ec50f5ca845525daf01ab71329e5a
F20101109_AAAXEV stokke_j_Page_081.tif
15fb2265831292b90b13f73793889f88
2f711ced6ee5eac4fda08ce53e2360d6311cbe2d
27773 F20101109_AAAWZP stokke_j_Page_027.QC.jpg
0eeb76ddf900d5ea40c0c745da2f8c9f
9c7a16f4daeee18ecd56512ed5584b322dd5f7e9
F20101109_AAAXFK stokke_j_Page_055.tif
78d8739d995542887f994463153ae534
6cc5b105f58e2482f0aac9f842727dbd4bb736b4
2230 F20101109_AAAXEW stokke_j_Page_094.txt
9a46a28c112750bfb1cc4b24fcc6ab4f
bc5bed11d183b29bd30ad68280996afb072cffa3
F20101109_AAAXFL stokke_j_Page_028.jp2
803e092bb80e9afabdaa39347679df35
94aaee5ef0d8ff8b9c435e6186312bb14773b6dd
7053 F20101109_AAAXEX stokke_j_Page_063thm.jpg
8ede8a7a3a7f637d18a6f316a43a589f
588b5e03afb91f8323f38673995b2dcc2e24103c
F20101109_AAAWZQ stokke_j_Page_078.jp2
544b8b2ef49dc2f87554db2778f9648a
4896dce0f4d29693a481b0a14089b94b734c9b83
753 F20101109_AAAXFM stokke_j_Page_076.txt
8535cb896edee39bb4d0d922aa30a9fe
5f629356e82ffc90c99c2602816518efa42c4306
28466 F20101109_AAAXEY stokke_j_Page_081.QC.jpg
9000d548fd22fcafc0d18963a418ba29
45f47b1a81cbb859eb570897e991a800723c9b66
F20101109_AAAWZR stokke_j_Page_113.tif
09a43a44b4c6aa975af0c2a8b72ff7cb
7550635832df6346314b234f249d4a0efe550bd7
22561 F20101109_AAAXGA stokke_j_Page_006.QC.jpg
7be5d2a8c637f838cba04a26fabd7d88
ba4d2b4c95ce1e59caf1b27f9a2d29d9abf7462a
36459 F20101109_AAAXFN stokke_j_Page_026.pro
889f8151763af17ed1773d2ac2532e8d
c6734ae8a42110529d384a5f383957ea445afea9
5335 F20101109_AAAXEZ stokke_j_Page_095thm.jpg
4e883d06fd08611ea29df10d6eedd692
32775c67a425266a1979245aed137bfbd28bed9a
F20101109_AAAWZS stokke_j_Page_075.tif
004c30211da7dafbaeb83766762a8a7a
a9a3926bb6b7f1569034d4016d843081eea6545c
F20101109_AAAXGB stokke_j_Page_025.jp2
cf5b65cf1f577fa68ecf32e072ebc557
0d88de405e4449adf3a83a0da89df6016ad65ae9
18419 F20101109_AAAXFO stokke_j_Page_069.pro
e66d0a83844f1228da004c4c22b39262
9b87627b4e7e13697e1f2c9baadefd9a9c8d0aa1
85367 F20101109_AAAWZT stokke_j_Page_046.jpg
e8ed57bcfc2b13849a4aaaf51054ef9e
327eb2fbac20729c3da4de43e5e9f39432657e67
10453 F20101109_AAAXGC stokke_j_Page_075.pro
0045bf7f023ff94b4fd080591ff5466b
1706a11c8bcbd5ed5d94dfb52b80dd7fd9b25eb3
2274 F20101109_AAAXFP stokke_j_Page_084.txt
949fd5fa826b915060e165db6138193e
30bccbebfaaf29dae10e8a317e40ea03dd03fc56
F20101109_AAAWZU stokke_j_Page_024.jp2
f538769efe2b48cea289bd00e12b56b5
e67727c264fcbacaedb238dd8e79d040b4a9675a
680 F20101109_AAAXGD stokke_j_Page_070.txt
3abaae975cf24862eb9ae03b1bd98c84
95869554ebe839f2b6b10d64c352ea6df55899fe
27058 F20101109_AAAXFQ stokke_j_Page_090.QC.jpg
68b5084df68397ca1a7ea02862eb7abb
42e5e2e087c0308bf09b65181364826a668c4296
12101 F20101109_AAAWZV stokke_j_Page_086.QC.jpg
9a20ec399cda5ffe9b374df2b4d53c5a
200921d37730dba7ff46c0330eff768f95206b4f
F20101109_AAAXGE stokke_j_Page_010.jp2
fb200217ead0ea893bd226bc5416ad2f
1292ce74cb333ea08c6b57b5ac5f7fa8987da826
F20101109_AAAXFR stokke_j_Page_082.tif
ddb23b90d8d375a5a684ca07ee7d6be1
2eeb1e685137025aba408584d1673ab89cf85b41
F20101109_AAAWZW stokke_j_Page_063.tif
318817f71edf39b3fd8a2f98297e8652
1013c6540c78683896cdf061b97c23d9ac8c956f
F20101109_AAAXGF stokke_j_Page_065.tif
d3358ccc20d525f285b2e7b396afa062
ee4c753d2bf137dd5cbd50880a8fe301ce060ba4
27886 F20101109_AAAXFS stokke_j_Page_056.QC.jpg
6b65f0ba736891078fd8870c0c642db9
d6820f55de1c3aff57b675638b4d6c1bccdb85cc
39666 F20101109_AAAWZX stokke_j_Page_096.jpg
ca3b3ea20b11ea5d38c72735859d46a1
c4b75cfbcd7de0695b7450b514994db6175c24a1
2220 F20101109_AAAXGG stokke_j_Page_032.txt
bb5271d7d5ed7325008374dc673f21ab
8a8c1509561311a7e8525970302fd3fc63f94cc4
F20101109_AAAXFT stokke_j_Page_035.tif
4ed6998fcea98927d26e127e490637a0
45d0bfa369f4f60deca1346ef6a0f0560a44589e
2142 F20101109_AAAWZY stokke_j_Page_057.txt
b95351800c4b70d8bad02d11b52a4918
cd1fc3f78ee57542e4a0bb6c28c0fb05f0671c9e
47997 F20101109_AAAXGH stokke_j_Page_064.pro
0dc2e98718ba370ee44666be720e082c
d38f00eae18541acd84994f82cfacf144a08af87
54486 F20101109_AAAXFU stokke_j_Page_105.pro
7ff1da4ed10f18c834f9d58fcc160836
a49f6de7900ecdbe52c8ad38862fc816180be6d6
14144 F20101109_AAAWZZ stokke_j_Page_051.pro
d1ee4b06f995fc53265d3d87675dc000
6aaa750ef9f98bda33ab2f67390897ecfdd638d7
72780 F20101109_AAAXGI stokke_j_Page_033.jpg
1be83977676fcecab33d42ed556ba229
bf2e77f1df759ec2d0a212f91d1cb0f60f97ec1c
48688 F20101109_AAAXFV stokke_j_Page_041.pro
f820526f648585ce8bf19bfa1bf586cf
9840c129caedf994d57b76aed858c8c02c90e463
80488 F20101109_AAAXGJ stokke_j_Page_006.pro
bea87a68764c71df3759aba200a8acde
1f3e243fe48ab46e4c66468500f424e4d2f77048
7369 F20101109_AAAXFW stokke_j_Page_022thm.jpg
0136d8692bdae4aebdeeba6d2859b28f
0bae47e1ba854881ab382e3dd188baf64362fb1f
F20101109_AAAXGK stokke_j_Page_097.QC.jpg
e468f30dfb6ec6308f0c7ef19696b556
f932fc9a787f43008a44dc2f4a191cee95070714
87634 F20101109_AAAXFX stokke_j_Page_037.jpg
a981fe54e7c2788986b1837c72887bba
4820591cf7dfa2c379b77bf1285913c6aaa7d590
4196 F20101109_AAAXGL stokke_j_Page_068thm.jpg
76030ca4ec7fd89fceba6dd1e8899b3a
5ad9e52dde61e8a613128e4cf942431ebbac4bba
30072 F20101109_AAAXFY stokke_j_Page_097.jpg
add464f3c83fad289b9c0cb33aec62ad
5b05b50fa9b40aedbde2793515419b93a66af351
88535 F20101109_AAAXHA stokke_j_Page_103.jpg
0e39f4c2af84d195cde26799e0b207c4
6b79a9f494acdda3fdc155375bbb6e55e2f3cfd6
4452 F20101109_AAAXGM stokke_j_Page_097.pro
61611779c82bc2d8ec8734adf07f1d15
ac40a18bfeb22cce826355a71465d0d22665d033
F20101109_AAAXFZ stokke_j_Page_018.tif
f7948632714797975d200b44d34980ea
df654e6e7b6fb9027db4e5bc9823e80fdb3da91b
2607 F20101109_AAAXHB stokke_j_Page_112thm.jpg
f858db78261dc02f7743b09c39134556
87b40940d8030f499882d48c9211af48e9ed42f0
F20101109_AAAXGN stokke_j_Page_122.jp2
56dc8ebe2daa5e8ec49246f6798c30d7
451b2b9610924718f0aca6b3fafa93151e7b576b
14934 F20101109_AAAXHC stokke_j_Page_089.pro
fa464f054c8264b2511efdeaa3fbd89f
20e03b45459d7b78cc2f7e1c3a605f729e6bb1a1
89659 F20101109_AAAXGO stokke_j_Page_081.jpg
f08d96aee6c4dae660a6a29478037f04
f7b44e649ed8fed7c698702a5cce2266668fff37
F20101109_AAAXHD stokke_j_Page_065.jp2
a43813b9fceafb43d7dec1d640c315fb
c752cfe95b1c0da97b9cda194dadbeaa571061e2
56110 F20101109_AAAXGP stokke_j_Page_039.pro
6f1cc712668a08b05bcc90c99e57dd66
31057aa9aca9a18df1ea5c1d1901272048097f99
7817 F20101109_AAAXHE stokke_j_Page_080thm.jpg
1e9c3f70dfded523abf91900583e17bc
74cbed9ef0804d545e07fead663ad77daac18289
374639 F20101109_AAAXGQ stokke_j_Page_071.jp2
cee2f7ce05d409eb44fd733ab9116234
825bd46704f441ad7dad693facc9fdf3c9cb97b1
F20101109_AAAXHF stokke_j_Page_002.tif
e7d97c33bf8e6dd050b06f6a0f774178
a005f99081c9d5d656372439d9a0e1fc200eb365
56484 F20101109_AAAXGR stokke_j_Page_079.pro
659001471719dde44a19b33d11f3ce6c
20ac774234bb10fc5e6917cb8bafa62ef1f158ed
12039 F20101109_AAAXHG stokke_j_Page_114.QC.jpg
f8a7256f8bbed814b454491b07cd66d2
7c9b60ce750177848a0b00df95385aea5aea984b
24348 F20101109_AAAXGS stokke_j_Page_059.QC.jpg
94008d827002b11013b7c8551153344c
0a7ebd77f44eddf5c8bbb34fb0969cca1b392bb4
5655 F20101109_AAAXHH stokke_j_Page_020thm.jpg
714168d57eddb85d6a8a13c72e9bab72
357d5e1517329ccf41821a3a8cb4171b13f92445
38961 F20101109_AAAXGT stokke_j_Page_098.jpg
e614adc33072e947578a08c05ce90621
4e4e07a0d7f35a80219316e5797b5b6ada0df6de
F20101109_AAAXHI stokke_j_Page_018.jp2
fc62f60a2d2a96efae22abcade973003
a4f5a34f3c618e14712c202b477a2ae90ae0a345
350346 F20101109_AAAXGU stokke_j_Page_097.jp2
9b52255434cc5bd58962051929c550d4
74fdfb34fc4fff5edcb544901698b921534731e0
3380 F20101109_AAAXHJ stokke_j_Page_086thm.jpg
90e01a8493da983bb662da9d585a8057
d703b517acc10573da8b2c4a0efcec4febbf0b11
1981 F20101109_AAAXGV stokke_j_Page_034.txt
4852321c77c0559c665b02a3c9d364fc
3feae1b7ebd8226b92b3199f387921501d184c3d
26184 F20101109_AAAXHK stokke_j_Page_104.QC.jpg
6abfcc03c1b8606ad8b0adfd26184191
81da138c698d6e9df186c0a08178c56665287ad1
25240 F20101109_AAAXGW stokke_j_Page_015.QC.jpg
5baaa4b390e9bf3f88f1275ecbe9dd7f
068911f4d92ce268b41b7256b4da3b1af3f4599f
33172 F20101109_AAAXHL stokke_j_Page_007.pro
4a6507287cd0a0416df66b210e497f24
417060619882b84158de4f3f72ae3f80f00b91de
1051943 F20101109_AAAXGX stokke_j_Page_093.jp2
4852f657a869e635e928cd232540c8b0
fc31458dafb02fde7e1e5f9374f87f2890be5001







PHOTOCATALYTIC OXIDATION OF HAZARDOUS AIR POLLUTANTS USING SILICA-
TITANIA COMPOSITES IN A PACKED-BED REACTOR




















By

JENNIFER MORGAN STOKKE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008

































2008 Jennifer M. Stokke


































To my parents, Wayne and Susan









ACKNOWLEDGMENTS

I would first like to thank my parents and family for the love, guidance, and support that

they have provided through the years. I would also like to thank my advisor, Dr. David Mazyck,

for the guidance and opportunities that he has provided during my graduate education. I will

certainly take away many lifelong lessons from this experience. I am grateful for the guidance

and suggestions provided by my advisory committee, Dr. Paul Chadik, Dr. Chang-Yu Wu, and

Dr. Hassan El-Shall, as well as those of Dr. Angela Lindner.

I would like to thank Rick Sheahan and Joe Sines of MicroEnergy Systems, Inc (Oakland,

MD) for services related to pilot reactor fabrication and testing, Jim Stainfield from NCASI

(Gainesville, FL) for providing analytical support, and Christina Akly, Heather Byrne, Teri

Lierman, Paloma Rohrbaugh, Brendon Blum, Miguel Morales, Vanessa Pineda, Gustavo Avila,

and Aly Byrne for their assistance in conducting laboratory studies. I am appreciative of the

Department of Energy and American Forest Product Association for sponsoring a portion of this

research via grant number DE-FC36-03ID14437.

Finally, I would like to thank past and present members of my research group for their

support: Ameena Khan, Heather Byrne, Jennifer McElroy, Matthew Tennant, Thomas

Chestnutt, Morgana Bach, Jack Drwiega, Miguel Morales, Vivek Shyamasundar, William

Kostedt, IV, Christina Ludwig, and Alec Gruss.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

LIST O F TA B LE S ............................................................................................ .............

LIST OF FIGURES ................................. .. ..... ..... ................. .8

A B S T R A C T ............ ................... ............................................................ 1 1

CHAPTER

1 IN T R O D U C T IO N ....................................................................................... .................... 13

Degradation of HAPs Emitted from Pulp and Paper Mills..................................................14
Recovery of Mercury Vapor Emitted in End-Box Exhaust at Chlor-alkali Facilities............16

2 L ITE R A TU R E R E V IE W ...................... .. .. ......... .. .............................. ..........................20

P hotocatalysis ................................................................20
U se of A dsorbents as Catalyst Supports........................................................................ .. ...23
Photocatalytic D egradation of M ethanol ..................................................... ...... ......... 25
Photocatalytic O xidation of H 2S ............................................................................. ............30
Enhanced Mercury Recovery using Photocatalysis.................... ........................ 32

3 EXPERIMENTAL ............... ................................................. ........ 36

Synthesis of Photocatalytic M aterials..................................................................................36
Silica-Titania Com posite (STC) Pellets ........................................ ....... ............... 36
TiO2-Coated A activated Carbon (A C) ........................................ ......................... 37
T iO 2-C oated G lass Spheres ..................................................................... ..................38
Characterization of Photocatalytic M aterials...................................... ................................ 38
Bench-scale Reactor for Methanol and H2S Removal Studies..............................................39
Pilot R eactor for M ethanol R em oval ................................................................. ............... 40
Analysis of Methanol and Oxidation Byproducts ............................................................41
Analysis of H2S and Oxidation Byproducts ................................ ...............41
M ercu ry A n aly sis ....................................................... ................. 42
C o lb u rn j-fa cto r ............. ..... ............ ................. ....................................................4 3

4 CH A R A CTER IZA TION O F STC .............................................................. .....................45

5 OPTIMIZATION OF METHANOL DEGRADATION USING STC PELLETS IN A
BENCH-SCALE REACTOR ........................................................... .. ............... 52

A d so rp tio n ........................................ ...................................................................................5 3
Simultaneous Adsorption and Oxidation of Methanol .........................................................53









Form ation of Photocatalytic Byproducts...................................................... ..... .......... 54
Effect of Space Time on M ethanol Degradation ....................................... ............... 56
M ass Transfer ................................................................ .... ...... ........ 58
K in etics.................... ............ .... ..................................................... 60
Effect of TiO2 Loading on Methanol Degradation............................................................61
Effect of UV Wavelength on Methanol Degradation..........................................................64
Effect of H 2S on M ethanol D egradation...................................................... .......................65

6 EFFECT OF CATALYST SUPPORT ON THE PHOTOCATALYTIC
DEGRADATION OF METHANOL IN A PACKED-BED REACTOR ..............................78

Methanol Adsorption and Oxidation in a Low Humidity Gas Stream...................................79
Methanol Adsorption and Oxidation in a High Humidity Gas Stream ................................82
W after V apor A dsorption ............................................................................... ....................83
Reactor Scale-up using TiO2-doped M materials ............................................ ............... 84

7 PILOT STUDIES FOR METHANOL DEGRADATION............................................. 90

ST C Synthesis for Pilot-Scale Studies......................................................... .....................90
UV Light Distribution in a Packed Bed of STC .......................................... ............... 92
P ilo t S tu d ie s ................... ......................................................... ................ 9 2

8 DEVELOPMENT OF A REGENERABLE SYSTEM EMPLOYING STC PELLETS
FOR MERCURY REMOVAL FROM END-BOX EXHAUST AT A CHLOR-ALKALI
F A C IL IT Y ................... ......................................................... ................ 1 0 0

Pilot-Scale Packed Bed Reactor .................................................. ............................. 100
P ilot Stu dy R esu lts................................. ..................................................... ............... 10 1
F ull-scale R actor ................................................................................ 104
Economy ic A analysis ..................................................................... .............. .. 107
C a p ita l C o sts ............................................................................................................ 1 0 7
O & M C osts.................................................. 108
Economic Feasibility ............................. .................. ......... 109

9 CONCLUSIONS .............. ........ ..... ..................... ..117

L IST O F R EFER EN CE S ......... ......................................................... ..................... 121

BIOGRAPHICAL SKETCH ............. ...........1....................127









LIST OF TABLES


Table page

4-1 BET surface area, total pore volume, and calculated pore size for the STC
synthesized with varying concentrations of HF and TiO2. ...........................................49

5-1 Sum m ary of experim ental conditions. ......................................................................... 68

5-2 Weisz modulus values for variable space time experiments. .........................................68

6-1 BET surface area and average pore volume of TiO2-doped sorbents.............................85

6-2 Surface area and TiO2 loading per reactor volume. .................... .......................... 86

7-1 BET surface area, pore volume, and pore size analysis for pilot STC. .............................95

7-2 UV intensity measurements through packed beds of STC of varying depths ...................95

7-3 Results of pilot studies with variable potentiometer settings ...........................................95

8-1 Sum m ary of pilot experim ents ......... ................. ................... ..................... ............... 110

8-2 Full-scale performance data for three operation cycles. ............................................... 110









LIST OF FIGURES


Figure page

3-1 Bench-scale reactor set-up used for methanol/H2S adsorption and photocatalytic
oxidation studies. ....................................................... .................. 44

3-2 Reactor drawings. A) 8 mm annulus reactor. B) 25 mm annulus reactor.......................44

4-1 Measured and expected surface area data for STC ................... ............... ............... 49

4-2 Nitrogen adsorption/desorption isotherms. A) STC with varying pore sizes and
constant TiO2 loading (12%). B) 50 A STC with varying TiO2 loadings (0-60%). .........50

4-3 Pore size distributions. A) STC with varying pore sizes and constant TiO2 loading
(12%). B) 50 A STC with varying TiO2 loadings (0-60%). .....................................51

5-1 Adsorption breakthrough curves for STC pellets of varying pore sizes (50 A, 120 A,
and 260 A) and constant TiO2 loading (12%) ........................ ............................... 69

5-2 Methanol removal using STC pellets. A) Methanol removal using STC of varying
pore sizes illuminated with UVA light. B) Extended study for 50 A 12% STC pellets....69

5-3 Effluent formaldehyde concentrations from STC pellets of varying pore sizes (50 A,
120 A, and 260 A) when illuminated with UVA light ................................. ............... 70

5-4 Normalized effluent methanol concentration for 50 A 12% STC illuminated with
U VA light at various space tim es. ........................................... ............................ 70

5-5 Effluent formaldehyde concentration for 50 A 12% STC illuminated with UVA light
at various space tim es. ......................... ...................... ............. ...... .. ...... 71

5-6 Effluent methanol and formaldehyde concentrations at steady state at varying space
times. A) 50 A. B) 120 A. C) 260 A STC............................ ............................... 71

5-7 Effluent methanol and formaldehyde concentrations at steady state for variable face
velocities and constant space time (4.3 s) for 50 A 12% STC........................................ 73

5-8 Linear regression of L-H model using mineralization rates achieved at various space
times. A) 50 A 12% STC. B) 120 A 12% STC. C) 260 A 12% STC.......................... 73

5-9 Effect of TiO2 loading in 50 A STC on methanol removal when illuminated with
U V A light .................. ..... .... ........ .......................................74

5-10 Effect of TiO2 loading on formaldehyde production at steady state using 50 A STC
illum inated w ith U V A light. .................................................................... ...................75

5-11 Percent transmittance of UVA light through 50 A STC with various TiO2 loadings........75









5-12 Effluent formaldehyde concentrations using 50A 12% STC irradiated with UVA and
U V C lig h t ................................ ................... ....................... ................ 7 6

5-13 Predicted C/Co versus absorbed light flux for 50 A 12% STC, 4.3 s residence time,
Co = 50 ppmv and 95% relative hum idity..................................... ......................... 76

5-14 Effluent concentrations of H2S, methanol and oxidation byproducts from 50 A 4%
STC illuminated with UVC light (Co methanol = 50 ppmv, Co H2S = 50 ppmv). ............77

5-15 Effluent concentrations of H2S and SO2 from 50 A 4% STC illuminated with UVC
light (C o H 2S = 50 ppm v).......................................................... ....77

6-1 Normalized effluent methanol concentration for titania-doped materials used in the
dark (adsorption only) and with UV light (adsorption and oxidation). ..........................86

6-2 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials ................87

6-3 Normalized effluent methanol concentration for TiO2-doped materials used in the
dark and with UV light in a high humidity gas stream (RH = 95%). ..............................87

6-4 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a
high-hum idity gas stream (RH = 95% ) ........................................ .................................... 88

6-5 Water vapor adsorption breakthrough profile in a high humidity gas stream (RH =
95% )............................................................................................ ..... ................88

6-6 Normalized effluent methanol concentration for TiO2-doped materials used in a large
annulus reactor (25 mm) and high humidity gas stream (RHi = 95%) ..............................89

6-7 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a
25 mm annular reactor and high humidity gas stream............................................... 89

7-1 Mixing assembly for blending the raw ingredients for pilot-scale STC synthesis. ...........96

7-2 Pilot-scale m olds for STC synthesis. ............................................................................ 96

7-3 Specialty heat chamber for pilot-scale STC synthesis................................................ 96

7-4 Alzak box test system for measuring UV light penetration through packed beds of
v ariou s depth s. .......................................................... ................. 97

7-5 Process flow diagram for methanol degradation pilot studies........................................98

7-6 General arrangement drawing of the pilot reactor for methanol degradation ....................98

7-7 Photo of pilot reactor. ................... ..... .. .......... ...... .... ..... .............99

8-1 Process flow diagram for mercury recovery pilot studies. ................... ........ ...........110









8-2 Schematic of pilot reactor for mercury recovery. ............. ....................... ..................111

8-3 Photo of pilot reactor with UV lights illuminated. .......................................112

8-4 Influent and effluent mercury concentrations for the pilot reactor packed with virgin
STC pellets (Test N o. 1). ........................................ ............... .. ...... .. .. 113

8-5 Influent and effluent mercury concentrations of the pilot reactor packed with virgin
pellets (Chamber A) and regenerated pellets (Chamber B). ................ .................114

8-6 Process flow diagram for full-scale installation of mercury recovery units..................14

8-7 Influent and effluent concentrations for the full-scale reactor during its second
adsorption cycle. ......... .... .......... .......... ..... ...... ..................................115

8-8 Comparison of cost per pound of mercury removed for activated carbon and STC as
a function of influent mercury loading (for systems designed to treat up to 2765
g /d ay ) ................... .......................................................................... 1 16









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

PHOTOCATALYTIC OXIDATION OF HAZARDOUS AIR POLLUTANTS USING SILICA-
TITANIA COMPOSITES IN A PACKED-BED REACTOR

By

Jennifer Morgan Stokke

May 2008

Chair: David W. Mazyck
Major: Environmental Engineering Sciences

This work centered on the optimization and scale-up of photocatalytic reactors employing

silica-titania composites (STC) for two applications involving the abatement of hazardous air

pollutants (HAPs) from industrial facilities: (1) degradation of HAPs, particularly methanol,

emitted from pulp and paper mills; and (2) recovery of mercury emitted from chlor-alkali plants.

STC were synthesized with varying pore sizes (50 A, 120 A, and 260 A) and TiO2 loadings

(0-60%) and were tested for the removal of methanol from a humid air stream. The efficiency of

methanol oxidation was dependent on the surface area of the STC and the space time of the gas

in the reactor. For 120 A 12% and 260 A 12% STC irradiated with UVA light, a lag time of 1.0 s

and 1.2 s, respectively, was observed before mineralization began. After this lag time, which was

zero for the 50 A 12% STC, the data followed pseudo-first order reaction kinetics and the rate

constant, k, was 0.40 s-1 for all pore sizes. Using the 50 A STC, the efficiency was further

improved by using a 4% TiO2 loading and UVC lamp, which generated a higher photon flux

compared to a UVA lamp. The presence of H2S in the gas stream decreased methanol removal

efficiency and resulted in SO2 and SO42- oxidation byproducts. When compared to other catalyst

supports, the STC was more efficient in a low-humidity gas stream with a relative humidity (RH)

of less than 0.22% at 230C. In a high humidity gas stream (RH = 95% at 230C), the efficiency of









the STC was inhibited by water vapor due to its surface chemistry and performed similarly to

TiO2-coated activated carbon. When compared to TiO2-glass spheres, the use of an adsorbent

catalyst support resulted in higher degradation efficiencies. Based on the promising bench-scale

results, a 40 ACFM pilot reactor was fabricated employing a packed bed of STC and a 4.3 s

space time through the packed bed. The pilot reactor achieved methanol removal rates up to 66

+ 7% with less than 1 ppmv formaldehyde production at steady state.

A pilot-scale photocatalytic reactor packed with STC was tested at a chlor-alkali facility

over a three-month period. This pilot reactor treated up to 10 ACFM of end-box exhaust and

achieved 95% mercury removal. The pilot reactor was able to maintain excellent removal

efficiency even with large fluctuations in influent mercury concentration (400-1600 ug/ft3). The

STC pellets were regenerated ex-situ with hydrochloric acid and performed similarly to virgin

STC pellets when returned to service. Based on these promising results, two full-scale reactors

with in-situ regeneration capabilities were installed and operated. After optimization, these

reactors performed similarly to the pilot reactor. A cost analysis was performed comparing the

treatment costs (i.e., cost per pound of mercury removed) for sulfur-impregnated activated

carbon and the STC system. The STC proved to be both technologically and economically

feasible for this installation.









CHAPTER 1
INTRODUCTION

Industrialized nations are faced with environmental problems related to the remediation of

hazardous wastes, treatment of contaminated water, and control of air pollution from industries,

military installations, and the civilian sector (Hoffmann et al., 1995). Hazardous air pollutants

(HAPs) pose particular threats to human health or the environment due to their toxicity. The

Clean Air Act requires the Environmental Protection Agency (EPA) to regulate emissions of 188

HAPs from industrial sources. Although a variety of technologies exist for the removal of

HAPs, heterogeneous photocatalysis has the potential to offer significant benefits with respect to

energy, size, and reliability compared to traditional technologies. Photocatalysis has proven

effective at the bench-scale for the oxidation of a variety of organic compounds (e.g., alcohols,

aliphatic compounds, aromatic compounds) to inert byproducts, such as CO2 and H20

(Hoffmann et al., 1995). In addition, inorganic compounds can undergo photocatalytic

transformation, which can be manipulated to enhance removal by adsorption onto the catalyst

surface or subsequent treatment processes (Huang et al., 1996; Nguyen et al., 2004; Pitoniak et

al., 2003; Pitoniak et al., 2005; Ferguson et al., 2005).

Although much research has been devoted to photocatalysis for environmental

applications, the testing and application of these systems beyond the bench-scale is limited.

Photocatalytic systems for treatment of gas-phase contaminants typically employ a thin film of

titania (TiO2) on, for example, reactor walls, wire mesh, or glass beads (Alberici and Jardim,

1997; Dijkstra et al., 2002; Kim and Hong, 2002; Peral et al., 1997; Chang et al., 2000). These

systems may not lend themselves well to scale up due to their low adsorption capacity, mass

transfer limitations, and problems with catalyst immobilization or durability (Devilliers, 2006).

In order to overcome these issues, a nano-structured silica-titania composite (STC) was









developed and has been tested at the bench-scale for the removal of mercury from synthetic flue

gas (Pitoniak et al., 2003; Pitoniak et al., 2005; Li and Wu, 2006). The STC has also been tested

for the degradation of organic compounds in water (Londeree, 2002; Holmes, 2003; Ludwig et

al., 2008) and deactivation of pathogens (Garton, 2005). Although the results of these studies

showed that the STC was capable of capturing mercury and completely degrading organic

compounds, the STC had not been developed past the laboratory scale.

This work centered on the optimization and scale-up of STC for two new applications

involving the abatement of HAPs emitted from industrial facilities: (1) degradation of HAPs,

particularly methanol, emitted from pulp and paper mills; and (2) recovery of mercury emitted

from chlor-alkali plants. These two applications are discussed in further detail below.

Degradation of HAPs Emitted from Pulp and Paper Mills

Forest products provide essential resources such as energy and materials. Compared to

fossil fuels (e.g., coal and oil) the resources from forest products are more sustainable and

diverse. However, the processing of forest products creates unwanted by-products including

volatile organic compounds (VOCs), some of which are HAPs (Someshwar, 1994). VOCs can

directly affect human health and are one of the main precursors of tropospheric ozone, a

respiratory irritant in humans and a major contributor to smog formation. Maximum Achievable

Control Technology (MACT) standards, as part of the 1998 Cluster Rules promulgated by the

EPA, require that the concentration of HAPs in high volume low concentration (HVLC) gases

emitted from pulp and paper mills be decreased by at least 90%.

Thermal oxidation is the most commonly applied technique for the control of VOC/HAP

emissions from the forest products industry sources (Varma, 2003; Garner, 2001). While

effective, these measures require a constant fuel supply to support the thermal energy

requirements and create other unfavorable byproducts (such as NOx). In addition, the fans and









ductwork required for the transport of gases are costly since gas streams are usually directed to

existing boilers (Varma, 2003). Thus, a cost-effective technique for in-situ treatment of these

pollutants is needed.

This work focuses on optimizing the STC and reactor operating parameters for the

treatment of HVLC gases emitted from pulp and paper mills, namely brown stock washers.

Typical HVLC gases are composed of air saturated with water vapor and contain HAPs (such as

methanol, acetaldehyde, formaldehyde), other VOCs (such as acetone and methyl ethyl ketone)

and total reduced sulfur (TRS) species (Someshwar, 1994; Varma, 2003). Methanol is the

primary constituent in HVLC gases and contributes to over 90% of the total HAP emissions from

brown stock washer vent gases (Someshwar, 1994). A study conducted by the Someshwar

(1994) at the National Council of Air and Stream Improvement found that the methanol

concentration in vent gases at 16 pulp and paper mills ranged from 32 to 2,263 ppmv. The range

of concentrations was attributed to the type of washers, vent gas flow rates, and methanol

concentration of the shower water. Vent gas flow rates were also found to be highly variable,

ranging from 162 to 16,808 scfm (Someshwar, 1994).

This work presents the optimization of the STC properties (e.g, surface area, pore size,

TiO2 loading) and reactor parameters controlling mass transfer and photon flux to achieve 90%

removal and complete oxidation of methanol, which was used as the target HAP. The

competitive effects of water vapor and H2S (a target TRS species) on methanol removal were

studied since both constituents are present in HVLC gases. In addition, the removal of H2S was

investigated to understand if the STC can provide a co-benefit for odor removal. The

performance of the STC was compared to that of TiO2 coated on other catalyst supports (i.e.,









activated carbon and nonporous glass spheres). Finally, a 40 ft3/min (ACFM) pilot reactor was

designed and tested for methanol removal from a humid air stream.

The following hypotheses were investigated in this study:

* As the available total surface area of the STC increases, the rate of methanol oxidation will
increase proportionally.

* The rate of methanol oxidation is limited by the resistance to mass transfer of the
contaminant to the STC surface and the space time of the gas in the reactor.

* The distribution of ultraviolet (UV) light in the reactor will dictate the rate of methanol
oxidation.

* The rate of methanol oxidation will be dependent on the transparency of the catalyst
support, with more transparent supports achieving greater oxidation rates.

* TRS present in HVLC gases will be oxidized to sulfate on the catalyst surface, which will
accumulate over time and occupy adsorption/oxidation sites thereby decreasing the rate of
methanol oxidation.

The objectives of this research were as follows:

* Synthesize and characterize STC with various properties (i.e., pore size, surface area, and
TiO2 loading).

* Develop and optimize a bench-scale photocatalytic reactor using STC for the removal of
VOCs/HAPs, using methanol as the target pollutant.

* Compare the silica gel to other catalyst supports with varying transparency and surface
area.

* Assess the effects of water vapor and TRS on the removal of methanol.

* Design, fabricate and test a pilot-scale reactor for the optimization of VOCs/HAP removal
using methanol as the target constituent.

Recovery of Mercury Vapor Emitted in End-Box Exhaust at Chlor-alkali Facilities

The release of mercury to the environment is of particular concern due to its volatility,

persistence and tendency to bioaccumulate. Methylation of mercury by microbes in the

environment results in the formation of methylmercury, which rapidly bioaccumulates in the

cells of higher organisms in aquatic systems. Methyl mercury is magnified at each trophic level









of the food chain thereby threatening ecosystems as well as human health. The primary exposure

route for humans is through the consumption of fish containing methylmercury (D'Itri et al.,

1978). This exposure can lead to adverse neurological effects, particularly in the developing

fetus and during early childhood (Schettler, 2001).

Mercury is released into the environment from anthropogenic sources such as coal-fired

power plants, cement plants, waste incinerators, and manufacturing processes such as the chlor-

alkali industry. The coal-fired power industry is the largest source of mercury in the United

States, emitting about 43 tons of mercury in 1999 (Pavlish et al., 2003). Although emissions

from coal-fired power plants are significant, other industries, such as the mercury-cell chlor-

alkali industry, can emit more mercury on a per facility basis (EPA, 1997a). Mercury abatement

from these sources is also crucial in order to protect human health and the environment.

Therefore, this work focuses on the development and implementation of a photocatalytic

technology for mercury recovery in the chlor-alkali industry.

The chlor-alkali industry produces valuable chemicals such as chlorine, hydrogen, and

caustic soda. In 2001, between 150 and 200 chlor-alkali facilities throughout the world used the

mercury-cell process. Although this process uses the mercury in a closed-loop system, mercury

is released to the environment through entrainment in by-product streams, end-box ventilation

systems and fugitive emissions (EPA, 1997b). According to Southworth et al. (2004), an average

of about three tons of mercury per year must be added to the production process at each mercury-

cell facility in the US to account for losses. The National Emissions Standard for Hazardous Air

Pollutants (NESHAP) for mercury cell chlor-alkali plants requires these facilities to drastically

reduce mercury emissions from their gas-phase emissions.









Due to these stricter standards, chlor-alkali facilities must either implement new control

technologies or eliminate mercury from their process. A commercially-available control

technology for reducing mercury emissions is activated carbon adsorption (EPA, 1997b;

Anastas, 1976). While sulfur- or iodine-impregnated activated carbon may reduce mercury

emissions to acceptable levels, it has a finite adsorption capacity and must be replaced and

properly disposed as hazardous waste. Therefore, the mercury problem is being transferred from

the air phase to the solid phase. The cost and risk associated with the continuous replacement and

disposal of mercury-laden activated carbon can be limiting, as many mercury-cell facilities have

chosen to convert to a mercury-free (e.g., membrane) process for manufacturing (EPA, 1997b).

However, conversion to another process requires a large capital investment. Thus, a cost

effective and regenerable solution is required for reducing mercury emissions.

This study focuses on the scale-up of the STC for use in the chlor-alkali industry,

particularly for the recovery of mercury from end-box exhaust. The STC was originally

developed at the bench-scale for mercury removal from flue gas emitted from coal-fired power

plants. Pitoniak (2004) found that the adsorption capacity of the STC increased after periods of

photocatalytic oxidation due to the formation of HgO on the composite surface. The total

mercury (i.e., elemental and oxidized) adsorption capacity was 3%wt, as determined by thermo

gravimetric analysis. The STC could be regenerated by rinsing with acid.

The composition of end-box exhaust consists of air containing trace levels of hydrogen (<

0.02%), water vapor (saturated at 6 8 OC), entrained water droplets, and elemental mercury.

The mercury concentration in end-box exhaust is typically two to three orders of magnitude

higher than that in flue gas. The STC has the potential to be both technically and economically

advantageous for the removal of mercury from end-box exhaust in chlor-alkali facilities because









of its high mercury adsorption capacity and ability to be regenerated with acid and re-used in

bench-scale studies. The mercury-laden acid used for regeneration could be recycled into the

mercury-cell process, thus closing the loop on mercury emissions. This work summarizes the

design and performance of pilot- and full-scale reactors used to recover mercury from the end-

box exhaust at a chlor-alkali facility. In addition, an economic analysis, which compares the

costs of implementing this technology versus using activated carbon at the facility, is presented.

The following hypotheses were investigated in this study:

* The research involving mercury removal from simulated flue gas can be translated to
mercury removal from caustic exhaust.

* Mass transfer and UV light distribution are the limiting factors for mercury removal from
caustic exhaust.

* STC will be economically favorable when compared to activated carbon for the removal of
mercury from end-box exhaust due to the ability of the STC to be regenerated and reused.

The objectives of this study were as follows:

* Confirm the efficacy of the technology for mercury removal from caustic exhaust and
recovery of the sorbed mercury by regeneration with HC1 in pilot and prototype studies at a
chlor-alkali facility.

* Determine the factors that may limit mercury removal efficiency (e.g., residence time,
mass transfer, and UV light distribution within the packed bed) in pilot-scale studies.

* Compare the cost of full-scale treatment units employing STC and activated carbon.









CHAPTER 2
LITERATURE REVIEW

Photocatalysis

When a semiconductor (e.g., TiO2) is irradiated with UV light that has an energy equal to

or greater than the band gap energy, an electron (e-) is promoted from the valence band to the

conduction band, leaving behind a positively charged hole (h ) in the valence band. This

reaction is shown in Equation 1 (Serpone, 1995; Hoffmann et al., 1995).

TiO2 + hv e CB+ h VB (2-1)

These electron-hole pairs can then recombine to generate heat or migrate to the surface and

participate in redox reactions. The positively charged holes are powerful oxidants while the

electrons in the conduction band can participate in reduction reactions.

Electron-hole pairs can oxidize organic pollutants directly via the electron hole or

indirectly via the formation of other powerful oxidants (e.g., hydroxyl radicals). The hydroxyl

radical (*OH), which has been identified as the primary oxidant in the photocatalytic oxidation

of organic compounds (Turchi and Ollis, 1990) and inorganic compounds such as mercury

(Pitoniak, et al. 2003; Pitoniak et al., 2005), can be generated via several pathways shown in

Equations 2-2 through 2-8 (Al-Ekabi and Serpone, 1988).

H20 + hvB *OH + H+ (2-2)

OHf + hvB *OH (2-3)

02 + 2H+ + 2e- H202 (2-4)

H202 + e -1 *OH + OH- (2-5)

H202 2 *OH (2-6)

02 + e- 02- (2-7)

02- + H202 *OH + OH- + 02 (2-8)









Water or hydroxide ions adsorbed at the TiO2 surface can trap electron holes (thereby

preventing recombination) and form hydroxyl radicals (Equations 2-2 and 2-3). Thus, in gas-

phase systems, relative humidity (RH) is often necessary for efficient oxidation of contaminants.

However, some studies have shown that excess RH can decrease oxidation rates for some

compounds due to the competitive adsorption between water and the target pollutant on the

photocatalytic surface (Kim and Hong, 2002; Obee and Brown, 1995; Li and Wu, 2007). As

shown in Equations 2-4 through 2-8, oxygen can serve as an electron acceptor (thus preventing

electron-hole recombination) and create oxidative species such as the superoxide radical (02-),

which can also react to create hydroxyl radicals.

The photocatalytic oxidation of organic pollutants yields a series of intermediate

byproducts of progressively higher carbon to oxygen ratios. If mineralization is achieved, inert

byproducts, such as carbon dioxide, water, and dilute mineral acids (in the case of halogenated

compounds) are formed (Ollis et al., 1991). Metals can be either oxidized or reduced via

photocatalysis. This is advantageous for separation of metals from gas or liquid streams if the

photocatalytically treated form of the metal is more easily removed by either sorption/deposition

onto the catalyst surface or in subsequent treatment processes. For example, water entering a

potable water treatment plant can be pretreated so that As(III) is oxidized to As(V), which is

more easily removed by conventional water treatment processes (Ferguson et al., 2005).

Removal of oxidized metals from water streams via photoreduction and subsequent deposition

onto the surface of TiO2 has proven effective for a variety of other metals such as silver (Huang

et al., 1996), cadmium (Nguyen et al., 2003), and selenium (Nguyen et al., 2004). The

photocatalytic oxidation of gas-phase elemental mercury (Hg) to HgO has resulted in enhanced









removal (Pitoniak et al., 2003). This enhanced removal of mercury via photocatalytic oxidation

will be discussed in more detail later in this chapter.

Photocatalytic reactions can be catalyzed by a variety of semiconductor materials (e.g.,

TiO2, ZnO, CdS, ZnS and Fe203) due to their electronic structure, which is comprised of a full

valence band and empty conduction band (Hoffmann et al., 1995). TiO2 is often used as a

photocatalyst since it is commercially available, non-toxic, relatively inexpensive compared to

heavy metal catalysts, and resistant to catalyst poisoning (Hurum et al., 2003). The principal

catalytic phases of TiO2 are anatase and rutile. Generally, anatase phase TiO2 is regarded as the

preferred phase due to its greater adsorption affinity for organic compounds and lower rate of

electron-hole recombination (Hurum et al., 2003).

The inactivity of rutile TiO2 is attributed to the rapid rate of electron-hole recombination.

However, TiO2 with a combination of anatase and rutile phases (such as Degussa P25) has been

shown to have enhanced photocatalytic activity compared to pure anatase phase TiO2. In mixed-

phase systems, electrons generated by the rutile-phase TiO2 can be transferred and trapped in

lower energy anatase lattice sites. Degussa P25 contains unusually small nanoclusters of rutile

crystallites dispersed within anatase crystallites. This morphology, which is responsible for the

enhanced activity of Degussa P25, allows for the rapid transfer of electrons from rutile to

anatase, resulting in catalytic "hot spots" at the rutile-anatase interface (Hurum et al., 2003).

Due to its enhanced activity and commercial availability, Degussa P25 (70% anatase, 30% rutile)

is frequently used in photocatalytic systems (Turchi and Ollis, 1989; Chen et al., 1999; Alberici

and Jardim, 1997; Minero et al., 1992; Obee and Brown 1995).

TiO2 is often immobilized as a thin film on various surfaces (e.g., beads, woven mesh,

reactor walls) or on/in adsorbent materials (e.g., silica gel, alumina, activated carbon), which act









as catalyst supports. When TiO2 is immobilized as a thin film, the photocatalytic reaction rate

can be affected by mass transfer limitations (Ollis et al., 1991). The use of TiO2 for practical

applications is dependent on the immobilization of the TiO2 particles onto a support such that the

composite is durable, provides a reasonably high surface area, and allows accessibility to the

immobilized catalyst (Shul et al., 2003). The use of adsorbent materials as catalyst supports is

discussed in further detail below.

Use of Adsorbents as Catalyst Supports

By using an adsorbent catalyst support, the contaminant is concentrated around the

photocatalyst. This has been shown to result in an increase in the photocatalytic reaction rate

(Anderson and Bard, 1997; Tsumura et al., 2002; Vohra and Tanaka, 2003; Torimoto et al.,

1996). In addition, organic contaminants are more likely to be mineralized since the

intermediates, which can be toxic, can be retained and further oxidized (Torimoto et al., 1996;

Lu et al., 1999).

Activated carbon is a commercially-available adsorbent material that has been studied for

use as a catalyst support (Yong et al., 2005; Arana et al., 2004; Torimoto et al., 1996; Lu et al.,

1999; Tsumura et al., 2002). Activated carbon is made by heating a carbonaceous precursor,

usually wood or coal, in the presence of an activating agent (e.g., steam, carbon dioxide,

oxygen). Properties of activated carbon such as surface area and pore size distribution are

affected by the activation method (i.e., temperature, time, and activating agent). TiO2 can be

deposited onto the activated carbon by various methods such as boil deposition (Arana et al.,

2004) or chemical impregnation (Yong et al., 2005). The use of activated carbon as a catalyst

support has proven effective. For example, Arana et al. (2004) found that TiO2-coated activated

carbon was more efficient than bare TiO2 for the mineralization of methanol. Lu et al. (1999)

showed that granular activated carbon (GAC) was the most efficient catalyst support for the









degradation of propoxur when compared to other catalyst supports such as zeolite, brick, quartz,

and glass beads. Bare TiO2 achieved higher oxidation rates compared to the GAC/TiO2. This

was attributed to the blockage of photons by the GAC, which decreased the number of photons

which reached the surface of the TiO2 to promote oxidation reactions. Although the overall

oxidation rate decreased, the mineralization rate of the TiO2/GAC was greater than that of the

bare TiO2. The superior performance of the GAC compared to other catalyst supports and its

ability to increase mineralization rates compared to bare TiO2 were attributed to its adsorption

properties.

Mixed oxide materials containing TiO2 supported on silica (SiO2) have also proven

effective for a variety of photocatalytic reactions. SiO2 is advantageous as a catalyst support

since it has high thermal stability and excellent mechanical strength (Gao and Wachs, 1999). In

addition, SiO2 is transparent to UV light. SiO2-TiO2 mixed oxide materials are generally

prepared by sol-gel and co-precipitation methods. The sol-gel hydrolysis route is most widely

used due to the ability to control textural and surface properties of the resulting composite

material (Gao and Wachs, 1999).

SiO2- TiO2 materials have shown superior catalytic efficiency due to quantum particle

effects and the formation of Ti-O-Si linkages (Anderson and Bard, 1997; Gao and Wachs, 1999).

Ti-O-Si linkages have resulted in enhanced adsorption of some contaminants (such as phenol and

R6-G) via strong Bronsted acid sites, which led to enhanced degradation rates (Anderson and

Bard, 1997; Yang and Chen, 2005). Yang and Chen (2005) also suggested that SiO2 prevented

electron-hole recombination by accepting electrons and providing hydroxyls for hole capture at

the SiO2-TiO2 interface, thereby further increasing photocatalytic efficiency of SiO2-TiO2

composites compared to TiO2 alone.









Shul et al. (2003) showed that the presence of SiO2 as a catalyst support for TiO2 helped to

promote the efficiency of acetaldehyde degradation because the SiO2 enhanced the effective

surface area of the TiO2 and increased the adsorption capacity of the composite compared to

unsupported TiO2. The photocatalytic activity of the composite increased with increasing surface

area of the SiO2 support. It was suggested that the increase in photocatalytic degradation rate was

due to an increase in the concentration of reactants and intermediates near the TiO2. Uchiyama et

al. (2005) found that the incorporation of SiO2 was also beneficial for the degradation of

acetylacetone. Composites with Ti/Si ratios of 0 (pure SiO2), 0.05, 0.25, 0.5, and 1.0 (pure TiO2)

were tested for acetylacetone adsorption and degradation. The amount of adsorbed acetaldehyde

was proportional to the specific surface area of the material, whether it be unsupported TiO2,

SiO2, or a composite of SiO2 and TiO2. The composite with Ti/Si = 0.05 achieved the highest

rate of degradation, likely due to its high surface area (1103 m2/g) and increased transparency

compared to the other composites containing TiO2. During the photocatalytic oxidation of

acetylacetone, molecules that were adsorbed on the SiO2 transferred to the photocatalyst, due to

the concentration gradient on the surface of the composite, where they were subsequently

oxidized.

Xu et al. (1997) also showed that TiO2 supported onto a silica gel substrate resulted in

higher degradation efficiencies for the oxidation of acetophenone in water compared to

unsupported TiO2. The supported TiO2 showed higher photoactivity when supported on smaller

SiO2 particles. This suggests that the dispersing effect of the SiO2 support on the TiO2 particles

was operative in enhancing its photoactivity.

Photocatalytic Degradation of Methanol

The photocatalytic oxidation of methanol can proceed via two pathways: (1) direct

oxidation (Pathway 1), where methanol and its organic byproducts are oxidized by the electron









hole and mineralized to C02; and (2) indirect oxidation (Pathway 2), where methanol and its

organic byproducts are oxidized by adsorbed *OH and mineralized to CO2 and H20 (Chen et al.,

1999).

The reactions for Pathway 1 (direct oxidation) are shown in Equations 2-9 through 2-15.

CH3OH + -s CH30-s + H+ (2-9)

CH30s + h CH3Os* (2-10)

CH3Os*+ h CH20s+H+ (2-11)

CH2Os+ h CHOs*+H+ (2-12)

CHOs* +h + H20 CHOOHs + H+ (2-13)

CHOOHs + h+ CHOOs* + H+ (2-14)

CHOOs* +h+ CO2s+H+ (2-15)

The reactions for Pathway 2 (indirect oxidation) are shown in Equations 2-16 through 2-

22.

H20s + h+ *OHs + H+ (2-16)

CH30s + *OHs CH3Os* + OH-s (2-17)

CH30s*+*OHs CH20 + H20s (2-18)

CH20 + *OHs CHOs* + H20s (2-19)

CHO*+ *OHs CHOOHs (2-20)

CHOOHs + *OHs CHOOs* + H20s (2-21)

CHOO* +*OHs CO2s+ H20s (2-22)

The H+ product from the above reactions can react with 02 to form H20 via the reduction

reaction shown in Equation 2-23 (Noguchi et al., 1998).

02+ 4e- + 4H+ 2H20 (2-23)









It is generally reported that the degradation of organic molecules via photocatalytic

oxidation proceeds by the formation and subsequent reaction with hydroxyl radicals (Pathway 2)

(Turchi and Ollis, 1990; Hoffmann et al., 1995; Obee and Brown, 1996). According to Kim and

Hong (2002), methanol degradation was achieved using a TiO2 thin film in the absence of water

vapor. They suggested that methanol was oxidized by hydroxyl radicals that were formed from

the hydroxyl groups of the methanol. The degradation rate reached an optimum at a water vapor

concentration of 0.383 mol/m3. At water vapor concentrations higher than the optimum, the

degradation efficiency decreased suggesting that water molecules competitively adsorbed to

catalyst surface thereby decreasing the degradation rate.

Yamakata et al. (2003) suggested an alternative mechanism for the photocatalytic

oxidation of methanol, whereby methanol is oxidized directly by the electron hole, even in the

presence of water vapor. In their study, the adsorbed water vapor was responsible for the

electron-consuming reaction while methanol was responsible for the hole consuming reaction.

The results suggested that the hole directly reacted with methanol, which adsorbed on the TiO2

surface as the methoxy species (CH30-), and that the hydroxyl radicals did not play a role in the

photocatalytic oxidation since the highest activity of hole consuming reactions was achieved in

the absence of water vapor. They suggested that water vapor enhanced the photocatalytic

oxidation of methanol by preventing electron accumulation, which would otherwise cause

defective sites on the TiO2 surface, and electron-hole recombination.

Literature generally suggests that methanol is degraded first to formaldehyde, then to

formic acid, and finally to carbon dioxide and water (Lichtin et al., 1994; Tsuru et al., 2003;

Chen et al., 1999), which corresponds to both Pathway 1 and Pathway 2 described above. Some

have suggested alternative pathways that result in the formation of formates, particularly methyl









format (Tsuru et al., 2003; Sadeghi et al., 1996; Arana et al., 2004). Tsuru et al. (2003) found

low levels of methyl format in the effluent when degrading methanol using TiO2 membranes.

They suggested that the presence of methyl format was a result of esterification reactions

between methanol and formic acid. Arana et al. (2004) studied methanol degradation using a

flow-through column (4 mm diameter, 15 mm height) packed with Degussa P25 TiO2. They

observed no mineralization of methanol and concluded from FTIR studies that bare TiO2 did not

degrade methanol due to the fast production of formates that resulted from methoxide (CH30-),

which poisoned the active centers of the photocatalyst.

The results from Arana et al. (2004) contradict those reported by Alberici and Jardim

(1997) and Kim and Hong (2002). Alberici and Jardim (1997) found that a TiO2 thin film

achieved 98% degradation of methanol with no catalyst deactivation. In addition, no intermediate

oxidation byproducts were present in the effluent, which is likely a result of the long residence

time (approximately 2 min) used in the experiments. Kim and Hong (2002) also observed that

methanol degradation was achieved using a TiO2 thin film.

For studies in which methanol degradation was achieved, the degradation rate of methanol

depended on variables such as the influent methanol concentration (Kim and Hong, 2002; Tsuru

et al., 2003), water vapor concentration (Kim and Hong, 2002), residence time (Tsuru et al.,

2003) and photon flux of the UV lamp (Kim and Hong, 2002; Alberici and Jardim, 1997). Tsuru

et al. (2003) reported that a decrease in flow rate yielded greater mineralization due to the

resulting increase in residence time. The degradation rate of methanol generally followed

Langmuir-Hinshelwood kinetics such that the reaction was first order at low concentration and

zero order at high concentrations (Kim and Hong, 2002; Tsuru et al., 2003). This transition

occurred at influent concentrations greater than 500 ppmv for Kim and Hong (2002) and greater









than 4000 ppmv for Tsuru et al. (2003). The concentration in which the reaction order transitions

from first order to zero order varied based on the reactor system and experimental conditions.

Tsuru et al. (2003) also found that the conversion of degraded methanol to carbon dioxide and

water was not dependent on influent methanol concentration, indicating that higher levels of

intermediate byproducts were formed as a result of increased influent methanol concentration.

The intrinsic rate of photocatalytic reactions is limited by the photon flux of the UV light

source (Hoffmann et al., 1995). For illumination levels above one sun equivalent, the oxidation

rate generally increases with the square root of the light intensity. For levels below one sun

equivalent, the oxidation rate increases linearly with light intensity. One sun equivalent is about

1 to 2 mW/cm2 (Obee and Brown, 1995). Kim and Hong (2002) reported that the degradation

rate of methanol was dependent on the photon flux of UV light such that the photocatalytic

degradation rate increased with the square root of the photon flux. When the degradation

efficiency of methanol was tested using a black lamp (peak wavelength = 352 nm) and a

germicidal lamp (peak wavelength = 254 nm), the germicidal lamp resulted in higher degradation

rates. Since the band gap energy of anatase phase TiO2 is 3.2 eV, photons with wavelengths less

than 385 nm are required to excite the photocatalyst. Thus, both the black lamp and germicidal

lamps can be used to promote photocatalytic reactions. It should be noted that more energetic

irradiation, such as that provided by a germicidal lamp, may affect the degradation efficiency by

direct photolysis of the organic compound or the formation of radicals that may alter conversion

yields (Alberici and Jardim, 1997). Kim and Hong (2002) attributed the higher degradation rates

observed with the germicidal lamp to the greater photon flux emitted from this lamp compared to

the black lamp. They reported that the photon flux from the germicidal lamp was about two

times that emitted from the black lamp. These results contrast with those reported by Alberici









and Jardim (1997), who showed that a germicidal lamp did not increase degradation rates of any

VOC (including methanol) compared to a black lamp. In these studies, both the germicidal and

black lamp had a nominal power rating of 30W; however, the radiant power output of the

germicidal lamp was 25% greater than that of the black lamp.

Photocatalytic Oxidation of H2S

While many studies have been performed using TiO2 for the degradation of VOCs, few

of these studies have investigated the oxidation and removal of inorganic or sulfur-containing

compounds. The photocatalytic reaction of H2S with TiO2 has the potential to form byproducts

such as SO2, S042-, and elemental sulfur. In order to completely oxidize H2S to S042-, an eight

electron transfer is required. The formation of S042- would result in deposition onto the surface

of the TiO2, since this species does not exist in the gas phase (Kataoka et al., 2005). Likewise,

the elemental sulfur that could be formed as a byproduct would also deposit on the surface of the

catalyst. This sulfur/sulfate deposition could potentially deactivate the catalyst over time by

blocking active sites. Alternatively, SO2 that could be formed as a result of the oxidation of H2S

could desorb back into this gas stream. Since SO2 is a toxic gas and regulated by the EPA,

subsequent removal processes (e.g., wet scrubbing) would be required to remove this gas.

Canela et al. (1998) and Kataoka et al. (2005) investigated the oxidation of H2S using a

TiO2 thin film and found that S042- was formed and accumulated on the TiO2 surface without

producing any gaseous intermediates (e.g., SO2). Portela et al. (2007) studied the oxidation of

H2S using TiO2 thin films coated on polymeric materials and found that both S042- and SO2 were

produced as oxidation products. The accumulation of S042- on the TiO2 surface resulted in a

decrease in oxidation efficiency over the 15 hour duration of the experiments. Canela et al.

(1998) experienced no catalyst deactivation over a 20 hour period when the influent H2S

concentration was 217 ppmv. However, when the inlet concentration was increased to 600 ppmy,









catalyst deactivation occurred, decreasing the removal efficiency from 99% to 36% after 2 hours.

However, it was not clear whether the decrease in efficiency was a result of the accumulation of

byproducts or the reactor temperature increasing from ambient (about 220C) to a working

temperature of 520C as a result of turning on the lamp. This increase in temperature could

decrease adsorption of H2S onto the TiO2, which could subsequently decrease the oxidation rate.

Washing the catalyst with deionized water was sufficient to remove the majority of the

accumulated S042- (Portela et al., 2007; Canela et al., 1998). Canela et al. (1998) recovered 95%

of the S042- that accumulated on the catalyst. Panela et al. (2007) removed the majority of the

SO42- from the catalyst after the first rinse and tested the catalysts for their ability to oxidize H2S

after multiple regenerations. They found that the water wash restored most of the initial activity

of the photocatalyst.

Variables such as residence time, humidity, and initial H2S concentration have been found

to effect the degradation efficiency of H2S using a TiO2 thin film. Canela et al. (1998) studied the

effect of residence time on H2S conversion efficiency. The results indicated that increasing

space time between 0.27 min and 2.46 min resulted in an increase in the conversion efficiency

from about 35% to 95%. In the same study, they found that the reaction exhibited a pseudo-first

order dependence on H2S concentration at influent concentrations between 30 ppmv and 855

ppmv.

Water vapor concentration is often an important variable when studying the efficiency of

photocatalytic reactions. Portela et al. (2007) found that the optimal humidity for oxidation of

H2S was 20% when studying air containing 35 ppmv of H2S and various RHs between 0 and 70%

(at 400C). It was hypothesized that water vapor played a key role in the reaction due to hole









trapping and hydroxyl radical formation. However, above 20% RH, the water vapor was thought

to hinder photocatalytic oxidation due to its competition with H2S for adsorption sites.

In a study by Kato et al. (2005), the incorporation of Ag nanoparticles onto a TiO2 filter

resulted in photocatalytic oxidation rates seven times higher than that of the un-modified TiO2.

X-ray photoelectron spectroscopy showed that H2S, elemental sulfur, AgS, and SO2 were not

present on the Ag-TiO2 film after use. Additionally, no byproducts were identified in the effluent

gas stream. S042-, which was trapped on the photocatalyst, was the only byproduct present.

However, the efficiency of H2S removal did not degrade over the duration of the experiment

(approximately 9 hours). It was hypothesized that the deposition of a noble metal, such as Ag,

would enhance photocatalytic reaction rates by increasing the charge separation efficiency and

inhibiting electron-hole recombination. In addition, the Ag enhanced the adsorption capacity of

the Ag- TiO2 film, which could have also led to increased oxidation rates. There was no

measurable adsorption of H2S onto the un-modified TiO2. The adsorption of compounds onto

TiO2, or other hydrous oxides, is most often attributed to hydrogen bonding. Hydrogen bonds

form between two functional groups, one which serves as a Bronsted acid and the other as a

Lewis base. According to Sopyan (2007), H2S does not readily form bonds with the hydroxyl

groups on TiO2 surfaces and has been shown to form hydrogen bonds in only strong basic

environments. In the same study, ammonia, which is a strong Lewis base, showed adsorption

capacity ten times greater than that of H2S as a result of its ability to participate in hydrogen

bonding with the TiO2 surface.

Enhanced Mercury Recovery using Photocatalysis

The removal of mercury via photocatalysis has been investigated for application in flue

gases at coal-fired power plants. Under UV irradiation, TiO2 converts Hg to HgO (as shown in

Equations 2-24 and 2-25), which is retained on the TiO2 surface (Rodriguez et al., 2004).









Hg+ *OH HgOH (2-24)

HgOH HgO+H+ (2-25)

In general, two approaches have been studied for Hg removal using TiO2: (1) Hgo

oxidation and capture on in-situ generated TiO2 particles (Wu et al., 1998; Lee et al., 2004; Lee

et al., 2001; Rodriguez et al., 2004); and (2) synergistic adsorption and oxidation of Hgo using

STC pellets in a packed bed (Pitoniak et al., 2003; Pitoniak et al., 2005; Li and Wu, 2006; Li and

Wu, 2007).

In-situ generated TiO2 particles have been created by the injection of a TiO2 sorbent

precursor into the combustor system. The precursor injection conditions were manipulated such

that agglomerated nano-sized (20-30 nm) TiO2 particles were formed. These agglomerates had a

high surface area and open structure, which would allow effective UV irradiation and minimal

resistance to mass transfer. The in-situ generated TiO2 particles showed no Hg capture in the

absence of UV irradiation, indicating that physical adsorption was not an effective pathway for

removal (Lee et al., 2004). However, in the presence of UV irradiation, the TiO2 particles

removed greater than 98% of the influent mercury (Lee et al., 2001). Water vapor enhanced

mercury removal at low concentrations by increasing hydroxyl radical generation on the TiO2

surface, which increased the number of active sites for oxidation and capture of Hgo. At very

high water vapor concentrations, it was expected that the water vapor would inhibit Hg removal

due to competition for adsorption sites, which would reduce the number of available sites for the

oxidation of Hgo (Rodriguez et al., 2004).

A novel, high surface area silica-gel impregnated with TiO2 nanoparticles for mercury

vapor control from flue gas was developed by Pitoniak et al. (2003, 2005) and has been further

developed by Li and Wu (2006, 2007). In this work, this material is referred to as STC. The









STC exhibited synergistic adsorption and photocatalytic oxidation for enhanced mercury

removal. Mercury vapor was adsorbed onto the STC surface and subsequently oxidized and

retained after irradiation with UV light. Mercury capture was achieved with continuous or

intermittent UV radiation. When intermittent UV irradiation was applied, Hg adsorption

increased after the periods of UV irradiation.

The STC demonstrated a high capacity for mercury (10 30 mg/g) (Pitoniak et al., 2005)

and achieved high levels of Hg removal (greater than 99%) when continuously irradiated with

UV light using a 0.78 s residence time (Pitoniak et al., 2003). Studies investigating the impact

on residence time of Hg removal revealed that the removal efficiency decreased as the residence

time decreased from 0.78 s to 0.16 s. It was determined that adsorption was the rate limiting

factor and that mass transfer should be improved to achieve better removal (Pitoniak et al.,

2003).

Li and Wu (2007) determined that mercury removal using the STC followed Langmuir-

Hinshelwood kinetics. These results suggested that the STC has a great potential for the removal

of mercury from gas streams containing high levels of Hg0. However, it was also found that

water vapor significantly inhibited the capture of mercury (Li and Wu, 2006; Li and Wu, 2007).

The STC achieved the highest level of mercury removal in the absence of water vapor. Li and

Wu (2007) postulated that the source of hydroxyl radicals in the absence of water vapor may be

the silanol groups on the SiO2 surface, which may serve as a the source for hydroxyl radical

production. The decrease in capture efficiency as the water vapor concentration increased was

likely due to competitive adsorption (Li and Wu, 2007). In addition, water vapor resulted in the

re-emission of captured mercury from the STC due to the repellent effect of water and the









photocatalytic reduction of HgO to Hgo, which subsequently desorbed from the STC surface.

The proposed mechanism for the reduction of HgO is show in Equation 2-26 (Li and Wu, 2006).

HgO+H20+2e Hg +2 OH- (2-26)

Although the re-emission of mercury affected the overall capture efficiency of the STC, Li

and Wu (2006) concluded that re-emission could be minimized by the appropriate application of

UV irradiation.

An advantage of the STC over other adsorbents is that it can be regenerated by rinsing with

acid (Pitoniak et al., 2003). The acid wash removed the mercury from the surface of the STC by

transferring it into the acid solution. Although the majority of the captured mercury was

recovered by the acid rinse, the performance of the regenerated pellets was not reported.









CHAPTER 3
EXPERIMENTAL

Synthesis of Photocatalytic Materials

Silica-Titania Composite (STC) Pellets

The STC pellets were prepared using an acid-catalyzed, sol-gel technique with tetraethyl

orthosilicate (TEOS) as the silica precursor. Degussa P25 TiO2 was used as the TiO2 source and

was mixed into the liquid precursors before gelation. Nitric acid and hydrofluoric acid (HF)

were used to catalyze hydrolysis and condensation reactions, thereby decreasing the time to

gelation. The pore size of the STC was manipulated by varying the amount of HF used during

synthesis.

For bench-scale studies, TEOS (Fisher Scientific, reagent grade) was added to a solution of

deionized water and ethanol (Aaper Alcohol, 200 proof) using a TEOS:water:ethanol volume

ratio of 7:5:10. To catalyze the reaction, 1 N nitric acid, prepared from 15.8 N nitric acid (Fisher

Scientific, certified A.C.S.), and 3%w HF, prepared from 48%w HF (Fisher Scientific, certified

A.C.S.), were added to the solution. In order to vary the pore size of the STC, either 2, 4, or 8

mL of 3%w HF per 220 mL of TEOS/ethanol/water solution. For every 100 mL of TEOS, 1 to

60 g of Degussa P25 TiO2 (Majemac Enterprises) were mixed into the solution. The ingredients

were mixed via a magnetic stir plate before being transferred into 96-well assay plates, where the

solution gelled as cylindrical pellets. The assay plates were sealed and the pellets aged at room

temperature for 48 hours and then at 650C for 48 hours. The pellets were then transferred to

Teflon containers, whose lids had a pin-sized hole, and dried for 18 hours at 1030C followed by

6 hours at 1800C. Each lid had a small hole to allow the liquid expelled from the pores of the

STC pellets to slowly escape as vapor during the drying process, thereby preventing collapse of









the pores. Finally, the STC pellets were calcined at 4500C for 2 hours. The resulting STC pellets

were about 3 mm in diameter and 5 mm in length.

In order to produce a sufficient quantity of STC pellets for the pilot and full-scale studies,

the bench-scale synthesis method was modified to increase production efficiency while

producing composites with similar characteristics (i.e., surface area and pore size). TEOS

(Silbond Condensed) was added to water, ethanol (Spectrum Chemicals), 1 N nitric acid,

prepared from 15.8 N nitric acid (Fisher Scientific, certified A.C.S.), and 3%w HF, prepared from

48%w HF (Fisher Scientific, certified A.C.S.). A known mass ofDegussa P25 TiO2 (Majemac

Enterprises) was mixed into the solution based on a ratio of 4 g of TiO2 per 100 mL of TEOS.

The ingredients were stirred using a paddle mixer and then transferred to molds, which were

made from 5.1 cm thick polyethylene sheets drilled with 0.8 cm diameter holes. Each mold was

approximately 40.6 cm by 61.0 cm and contained 2,750 holes. The molds were filled by pouring

the liquid sol into the molds, which were sealed on the bottom and top with sheets of solid

polyethylene. The gels were aged at 650C for 48 hours. The lids were then loosened and the

pellets dried in the molds at 1030C. The pellets were removed from the molds, transferred to

Pyrex containers, and then heated to 1800C. After aging and drying, pellets were approximately

3 mm in diameter and 20 mm in length.

TiO2-Coated Activated Carbon (AC)

TiO2-coated AC was synthesized by coating granular BioNucharl20 (MeadWestvaco) with

Degussa P25 TiO2 via a boil deposition method. A TiO2 slurry was made by adding 3 g of

Degussa P25 TiO2 to 200 mL of DI water. Next, 30 g of AC were added to the slurry and heated

on a hot plate until all of the water evaporated, leaving the TiO2 coated onto the outer surface of

the AC. The actual quantity of TiO2 deposited on the AC was determined by taking the

difference of the measured ash content of the as-received and TiO2-coated AC. To determine the









ash content of an AC sample, about 1 g of dry material was heated to 5500C for 24 hours to

remove the carbonaceous portion of the AC. The measured TiO2 loading on the AC was 5.0 +

0.4 %wt (error given as the standard deviation of triplicate measurements).

TiO2-Coated Glass Spheres

Solid glass spheres (5 mm diameter) were coated with a TiO2 slurry (20%wt Degussa P25

TiO2 dispersed in water) and then dried at 1100C. After drying, excess TiO2 was separated from

the TiO2-coated glass spheres by gently shaking the beads on a sieve with 4 mm openings. The

mass of the spheres was measured before and after the TiO2 coating was applied. The resulting

mass of TiO2 on each glass sphere was 0.925 mg, which equates to 11.75 g of TiO2 per m2 of

glass spheres. The BET surface area of the TiO2 (as measured by a Quantachrome NOVA 2200e,

Boynton Beach, FL) was approximately 50 m2/g.

Characterization of Photocatalytic Materials

The STC pellets and TiO2-coated AC were analyzed for surface area, total pore volume,

and average pore size using a Quantachrome NOVA 2200e (Boynton Beach, FL). A

Quantachrome Autosorb was used for nitrogen adsorption/desorption isotherms. The samples

were vacuum outgassed at 1800C for 24 hours. The surface area was determined using the

Brunauer-Emmett-Teller (BET) model with the nitrogen adsorption data (P/Po = 0.1 to 0.3). The

total pore volume was calculated based on nitrogen adsorption at P/Po = 0.995. The average pore

size was calculated using Equation 3-1, assuming non-intersecting, cylindrical pores:

d = 4*Vp/S (3-1)

where d is the average pore diameter, S is the surface area, and Vp is the total pore volume.

For the analysis of pore size distribution, the desorption isotherm was analyzed using the Barrett,

Joyner, and Halenda (BJH) method.









Bench-scale Reactor for Methanol and H2S Removal Studies

The adsorption and photocatalytic oxidation of HAPs in a simulated HVLC gas emitted

from pulp and paper mills were tested using a bench-scale annular reactor. The reactor had an 8

mm annulus and contained an eight-watt UV bulb (Spectronics Corporation) with a peak

wavelength of either 365 nm (UVA) or 254 nm (UVC). The UV bulb was surrounded by a quartz

tube, which had an outside diameter of 25 mm. A schematic of the reactor is shown in Figure 3a.

UV intensity measurements inside of the reactor were taken using chemical (potassium

ferrioxalate) actinometry as described by Murov et al. (1993).

A known quantity of photocatalytic material (STC, TiO2-coated AC, or TiO2-coated glass

spheres) varying between 30 and 90 cm3 (bulk volume) was packed into the annulus of the

reactor. Where specified, an annular reactor with a 25 mm annulus was used for preliminary

scale-up studies. For these studies, 132 cm3 of photocatalytic material were packed into the

annulus. A schematic of the 25 mm annulus reactor is shown in Figure 3b.

The materials were tested for adsorption capacity in the dark and for simultaneous

adsorption and photocatalytic activity when irradiated with UV light. The reactor was kept in the

ambient atmosphere and, when irradiated with UV light, the temperature of the packed bed rose

to about 500C. In order to achieve this temperature during adsorption studies, the reactor was

wrapped with heat tape and controlled using a Variac variable voltage transformer.

Compressed air containing 1000 ppmv of methanol or H2S was diluted with air to obtain an

influent gas stream containing 50 ppmv of methanol and/or H2S and a RH less than 0.22%. For

the experiments requiring a high humidity, the dilution air was passed through a water bubbler to

obtain an influent gas stream with an RH of about 95% at 230C. The gas stream flowed

continuously through the reactor in a single pass configuration. A schematic of the reactor set-up

is shown in Figure 3-1. Initial studies performed with an empty reactor showed no photolysis of









methanol or H2S in the presence of UVA or UVC light. Similarly, adsorption of methanol and

H2S to the reactor and its appurtenances was negligible.

Various experiments were conducted to study the effects of space time and face velocity on

methanol removal using STC pellets packed in the 8 mm reactor. Space time (') is the time

required to process one bed volume of gas in an empty reactor and was calculated using Equation

3-2:

c = V/Q (3-2)

where Q is the gas flow rate and V is the reactor volume occupied by the packed bed. The

face velocity, or superficial velocity (v), was calculated using Equation 3-3:

v = Q/A (3-3)

where A is the cross-sectional area of the packed bed.

Space times were varied between 1.1 s and 4.3 s in the 8 mm reactor by varying the gas

flow rate (0.42 1.68 L/min) passing through a 30 cm3 packed bed. The effect of face velocity

on methanol removal was studied by flowing 0.42, 0.84, and 1.26 L/min of gas through STC bed

volumes of 30, 60, and 90 cm3, respectively, to achieve a constant space time of 4.3 s and face

velocities of 0.093 m/s, 0.19 m/s, and 0.28 m/s.

Pilot Reactor for Methanol Removal

A 40 ACFM pilot reactor was designed and fabricated based on optimization studies

conducted at the bench-scale. A water/methanol mixture was vaporized and injected into the gas

stream so that the influent methanol concentration was about 50 ppmv and RH between 95 and

99%. The static pressure and velocity pressure were monitored before the inlet to the pilot

reactor. Thermocouples were used to measure inlet and outlet temperatures.









A UVP radiometer (Upland, CA) was used to measure the UV intensity through an

observation window on the side of the reactor. Initial studies performed with an empty reactor

showed no photolysis of methanol and that adsorption of methanol to the reactor and its

appurtenances was negligible.

Analysis of Methanol and Oxidation Byproducts

Both influent and effluent methanol concentrations for bench- and pilot-scale studies

were tested using the National Council for Air and Stream Improvement (NCASI) Chilled

Impinger Method (NCASI, 1995). Methanol concentrations were quantified using a

PerkinElmer Clams 500 GC/FID (Wellesly, MA) with a 50 m x 320 |im polyethylene glycol

column. A 1 p.L volume of sample was used for analysis. The injector temperature was set to

1100C. The oven temperature was initially 400C (1 min hold) and was ramped to 540C at

2C/min (2 min hold) and then to 2200C at 210C/min (5 min hold). The minimum detection limit

(MDL) for methanol in the gas phase was 0.6 ppmv. Effluent formaldehyde concentrations were

quantified colorimetrically using a Hach DR/4000U spectrophotometer (Loveland, CO) as

described in the NCASI Chilled Impinger Method (NCASI, 1995). The MDL for formaldehyde

was 0.014 ppmv. In order to quantify total byproduct formation, total organic carbon (TOC)

concentrations of the impinger samples were determined using a Tekmar Dohrmann Apollo 9000

TOC analyzer (Mason, Ohio). In addition to impinger measurements, real-time influent and

effluent measurements were taken during the pilot studies using a ThermoElectron TVA-1000B

portable FID detector (Waltham, MA).

Analysis of H2S and Oxidation Byproducts

Air-phase concentrations of H2S were measured by passing 100 mL of gas through pre-

calibrated, direct-read Nextteq gas detection tubes (0.25 120 ppmv). Air-phase concentrations









of SO2 were measured using a Varian 2100T GC/MS (Palo Alto, CA) equipped with a Supel-Q

Plot, 30 m x 0.32 mm column. Samples were collected in 1 L Tedlar bags and stored in the dark

for no longer than 24 hours. A 75 um Carboxen-PDMS SPME fiber was injected into the Tedlar

bag through a silicon septum for 10 min. The fiber was injected into the GC-MS for 5 min at an

injection temperature of 200 C. The oven temperature was initially 450C and ramped to 2500C a

rate of 250C/min after an initial 0.75 min hold. The instrument detection limit for SO2 was 1

ppmv.

Sulfate loading on the STC pellets was measured by soaking 30 cm3 of pellets in 500 mL

of nanopure water (18.2 MQ-cm) with gentle stirring using a magnetic stirrer for 24 hours. The

water was filtered using a 0.45 |tm vacuum filter. A 10 mL sample of filtered water was analyzed

for sulfate using the EPA Method 9038 with a Hach DR/4000U spectrophotometer (Loveland,

CO). The MDL for sulfate measurements was 1 mg/L.

Mercury Analysis

Influent and effluent mercury concentrations for pilot and full-scale mercury removal units

were determined using EPA Method 101. The MDL for mercury corresponded to an air-phase

concentration of 0.1 [tg/ft3.

For pilot-scale studies, STC pellets were regenerated ex-situ to remove the mercury. The

pellets were removed from the reactor, placed in a bath of 37%wt HC1 and mixed gently for 1

hour. The pellets were then rinsed with water and placed in an oven at about 700C for 12 hours.

The mercury concentration of the pellets was determined before and after regeneration by

digesting 1 g of pellets in 50 mL of aqua regia solution. The mercury concentration of this

solution was measured using a Hydra AA Spectrophotometer (Leeman Labs), which has an









MDL of 0.2 [tg/L. The moisture content of the pellets after heat treatment at 700C was about

60%, as determined by gravimetric analysis.

For full-scale studies, regeneration was performed in-situ by soaking the pellets inside of

the reactor with 37%wt HC1 for 30 min. After draining the mercury-laden acid from the reactor,

the pellets were soaked with water for 30 min and the water was subsequently drained.

Colburn j-factor

The Colbum j-factor (jd) is a dimensionless parameter for mass transfer based on

Reynold's number (Re) and the shape, size and packing characteristics of the STC pellets. This

parameter was used to assess the effects of mass transfer on the performance of the pilot reactor

and as a design parameter for the full-scale reactors. The j-factor was calculated using Equations

3-4 through 3-7, where Y is the shape factor (0.91 for cylindrical pellets), v is the superficial

velocity of the gas, a is the surface area per volume of the packing media, dp is the particle

diameter (see equation 3-6 for non-spherical particles), s is the bed porosity, and p and [t are the

density and viscosity of the gas (Knudsen et al., 1999).

Re = v*p/(Qt*Y*a) (3-4)

a =6(1-s)/dp (3-5)

dp = 0.567 *sqrt (part. surf. area) (3-6)

jd =0.91*Y*Re-.51 (3-7)












































Figure 3-1. Bench-scale reactor set-up used for methanol/H2S
oxidation studies.


30.5cm.


Removable top
assembly

/ as Inlet and
Catalyst Fill Fort


Quartz Sleeve
425 mm OD)

uV Lamp


- 5 mm Glass
Support Spheres

- Pyrex Reactor
(41 mm ID)
Porous Glass Frit

0.635 cm Gas Outlet


Figure 3-2. Reactor drawings. A) 8 mm annulus reactor. B) 25 mm annulus reactor.


UV Lamp
[itania-coated
Media


adsorption and photocatalytic









CHAPTER 4
CHARACTERIZATION OF STC

STC pellets with three different average pore sizes and varying TiO2 loadings were

synthesized by varying the quantity of HF and TiO2 added to the liquid precursors before

gelation. BET surface area, total pore volume, and average hydraulic pore size are shown in

Table 4-1. The error represents the standard deviation of grab samples taken from at least three

batches of pellets. The STC pellets were labeled according to their approximate hydraulic pore

size and TiO2 loading (% mass TiO2 per 100 mL of TEOS). Based on a mass balance of the STC

synthesis from raw ingredients to the final product (i.e., dried pellets), the estimated TiO2 mass

loadings for STC labeled as 0, 1, 4, 12, and 60% were 0, 3, 12, 30, and 70%w, respectively.

The surface area and total pore volume of the STC were dependent on the type of STC.

The various STC labeled as "50 A" were synthesized with a constant HF concentration and

varying concentrations of TiO2. The addition of the non-porous TiO2 (ca. 50 m2/g) to the porous

silica gel resulted in an overall decrease in surface area and pore volume of the composite. The

HF concentration also had an effect on both surface area and pore volume. For the STC

synthesized with 12% TiO2 and varying concentrations of HF (50 A 12%, 120 A 12%, 260 A

12%), an increase in HF resulted in a decrease in surface area and an increase in pore volume

due to the widening of the pores. The data do not suggest that significant pore blockage occurred

as a result of TiO2 addition. The change in specific surface area and pore volume was due to the

addition of the non-porous TiO2 to a porous silica gel, creating a composite material with

specific surface area and pore volume values representative of a mixture of the two materials.

To illustrate this, Figure 4-1 shows the actual and expected specific surface area values of the

STCs. The expected values were calculated based on the mass percentage of silica in the

composite multiplied by the surface area of silica gel synthesized with the same HF









concentration and no TiO2 plus the mass percentage of TiO2 multiplied by the surface area of

TiO2 (50 m2/g). As shown in Figure 4-1, the measured specific surface areas of the STC are

similar to the expected values. The error bars shown for the measured values were based on the

measured error between multiple batches, as shown in Table 4-1. The error bars shown for the

expected values were based on the error associated with the measured value of the silica gel

synthesized with the same HF concentration and no TiO2.

The nitrogen adsorption/desorption isotherms for the STC are shown in Figure 4-2. All of

the STC exhibit a Type IV isotherm with H1 and H2 hysteresis loops, which is characteristic of

mesoporous materials. The 50 A STC mainly exhibited H2 hysteresis, while the 120 A and 260

A STC exhibited H1 hysteresis.

Although the effect of various factors on hysteresis is not fully understood, hysteresis

shapes have been associated with specific pore structures. The H1 hysteresis loop, which was

exhibited by the 120 A and 260 A STC, indicates that the material is comprised of uniform

spheres arranged in a fairly regular array with a narrow distribution of pore sizes (Sing et al.,

1984). H2 hysteresis, which was exhibited by the 50 A STC, has been attributed to pores with an

ink bottle shape (i.e., narrow necks and wide bodies) (Sing et al., 1984). However, poorly

defined pore shape or distribution in a disordered material, which results in pore blocking and

percolation, could also cause the H2 loop (Thommes, 2004). Although the 50 A STC generally

exhibited H2 hysteresis, the 50 A 60% had a hysteresis shape more representative of Hi. The 50

A 60% hysteresis loop narrowed at a relative pressure of about 0.8, rather than 0.5 to 0.6 for the

50 A STC of lower TiO2 loadings (0 12%). This suggests that the 60% TiO2 loading may

increase the size of the pore neck, allowing the pore body to empty at higher relative pressures,

or result in a less extensive pore network, as evidenced by the decrease in total pore volume.









The presence of approximately uniform spherical primary particles was confirmed by SEM

images taken by Byrne et al. (2008). The SEM images showed that the spherical particles were

present in all of the STC, regardless of pore size or TiO2 loading. The size of the primary

particles was smallest for the 260 A STC and grew larger for STC synthesized with lower HF

concentrations. Thus, the 50 A STC were comprised of the largest primary particles. This was

attributed to the shorter gelation times for STC with higher HF concentrations, which resulted in

a shorter time for particle growth and, hence, a smaller primary particle size. The gelation time

was dependent on the HF concentration used during synthesis since it catalyzes hydrolysis and

condensation reactions. The 50 A 12% STC gelled after about 12 hours while the 120 A 12% and

260 A 12% gelled after approximately 2.5 hours and 10 minutes, respectively. The SEM images

also showed that for the 260 A 12% and 260 A 60% STC, the change in TiO2 concentration did

not noticeably change primary particle size or morphology.

The pore size distribution of the STC is important since it can affect the diffusivity of

methanol and oxidation byproducts into and out of the STC (Satterfield, 1970). The pore size

distributions of STC synthesized with various HF and TiO2 concentrations are shown as the

differential pore volume as a function of pore diameter in Figure 4-3. The hydraulic pore sizes

shown in Table 4-1 are sometimes different than the average pore sizes shown in Figure 4-3. If

the pores were truly cylindrical in shape, which is the assumption made for the calculation of the

hydraulic pore size, the average pore size associated with the desorption isotherm would equal

the hydraulic pore size. The disparity between the desorption pore size and the hydraulic pore

size can be attributed to the actual shape of the pores and/or network effects.

The pore size distributions for the 50 A 12%, 120 A 12%, and 260 A 12% STC, as shown

in Figure 4-3a, were unimodal and the area of the peak was relative to the total pore volume of









the STC. The 50 A 12% STC had the narrowest range of pore sizes, with about 95% of the pore

volume resulting from pores between 19 A and 56 A. As the pore size of the STC increased due

to the increased quantity of HF added to the liquid sol during synthesis, the total pore volume

increased and the pore size distribution became broader. For the 120 A 12% STC, 95% of the

pore volume resulted from pores with diameters between 49 A and 172 A. The 260 A 12% STC

had the broadest pore size distribution, with 94% of the pore volume resulting from pores with

diameters between 155 A and 336 A.

Pore size distributions for 50 A STC with varying TiO2 loadings (0-60%) are shown in

Figurer 4-3b. For TiO2 loadings between 0 and 12%, differences between the pore size

distributions were mainly a result of the decrease in pore volume associated with increased TiO2

loadings. The distributions for these STC were unimodal and the shapes of the distribution

curves were similar. The 50 A 60% STC had a much broader distribution and drastically smaller

differential pore volume at the 50 A pore diameter. The pore volume at 50 A was approximately

0.005 cc/A/g compared to 0.037 cc/A/g for the 50 A 4% STC. Although the peak differential

pore volume remained at about 50 A, pore sizes between approximately 20 and 250 A

contributed to the total pore volume for the 50 A 60% STC.

The 50 A 60% STC was comprised mostly of TiO2 (i.e., about 70%wt of the dried product

was TiO2). Thus, the amount of shrinkage that occurred during the aging and drying process was

reduced. The total volume of a dried 50 A 60% STC pellet was about 61% less than its volume

immediately after gelation. The volume of a 50 A 12% STC pellet, for example, was reduced by

74%. This indicates that the syneresis of the SiO2, which results in the formation of additional Si-

O-Si linkages resulting in the shrinkage of the STC during aging and drying, was inhibited. This

may have resulted in the formation of a broader pore size distribution by disallowing the









shrinking of the silica network and formation of additional linkages to create the smaller pores

(i.e., those around 50 A in diameter).

Table 4-1. BET surface area, total pore volume, and calculated pore size for the STC synthesized
with varying concentrations of HF and TiO2.
STC BET surface area (m2/g) Total pore volume (cc/g) Pore size (A)
50 A 0% 662 54 0.86 0.03 52 3
50 A 1% 636 27 0.88 0.03 55 1
50 A 4% 617 19 0.76 0.04 49 4
50 A 12% 532 43 0.60 + 0.06 46 4
50 A 60% 297 24 0.48 0.05 65 5
120 A 12% 290 23 0.69 0.11 123 7
260 A 12% 154 18 0.99 0.10 259 11


" 600
S500
S400

= 300
- 200


Figure 4-1. Measured and expected surface area data for STC.


50A 0% 50A 1% 50A 4% 50A 12% 50A 60% 120A 12% 260A 12%










800

700 -...--50A 12%
700
120A 12%
S600 -260A 12%

500

S400

300

200 ---




0 0.2 0.4 0.6 0.8 1
P/Po

B
700

n600 -50A 0%
50A 4%
0
S. ---50A 12%
500
50A 60%

S400
o J// -----^^-T
I 300

200




0
0 0.2 0.4 0.6 0.8 1
P/Po


Figure 4-2. Nitrogen adsorption/desorption isotherms. A) STC with varying pore sizes and
constant TiO2 loading (12%). B) 50 A STC with varying TiO2 loadings (0-60%).



























I I


Pore Diameter (Angstroms)



B


50A 0%
-50A 1%
- 50A 4%

-50A 12%
- 50A 60%


0 50 100 150 200 250 300
Pore Diameter (Angstroms)


Figure 4-3. Pore size distributions. A) STC with varying pore sizes and constant TiO2 loading
(12%). B) 50 A STC with varying TiO2 loadings (0-60%).


0.040

0.035

0.030

0.025

0.020

0.015

0.010


0.005 --


0.000


0.045

0.040

0.035

0.030

0.025

0.020

0.015

0.010

0.005

0.000


--50A 12%
120A12%

260A 12%



,I
i I'

iI


0.005









CHAPTER 5
OPTIMIZATION OF METHANOL DEGRADATION USING STC PELLETS IN A BENCH-
SCALE REACTOR

The use of STC pellets for VOC abatement is not widely studied since photocatalytic

reactors designed to treat VOCs typically employ a TiO2 or SiO2-TiO2 thin film. The objective of

the work presented in this chapter was to investigate the use of the STC pellets for the removal

and degradation of methanol from a high humidity air stream, specifically to determine the

effects of STC pore size, TiO2 loading, face velocity and space time on methanol degradation.

The conditions for the experiments described in this chapter are summarized in Table 5-1.

An influent methanol concentration of 50 ppmv was used for all studies. This concentration was

picked based on discussions with industry. Study No. 1 represents the baseline conditions. The

space time was manipulated in Study No. 2, 3, and 4 by varying the flow rate of gas containing

50 ppmv of methanol with an influent relative humidity of about 95% (at 230C) through the

reactor packed with 30 cm3 of pellets. For space times of 8.6 s, 2.1 s, and 1.1 s, the total flow

rate through the reactor was 0.21, 0.84, and 1.68 L/min, respectively. Since the flow rate was

adjusted to achieve the various space times, the face velocity, methanol loading rate (mg of

methanol entering the reactor per minute) and water vapor loading rate also varied with space

time, as shown in Table 5-1. Study No. 1, 5, and 6 were performed to assess the effect of varying

face velocity at a constant 4.3 s space time. The bed depth was adjusted to achieve the constant

space time at the varying face velocities. Flow rates of 0.42, 0.84, and 1.26 L/min containing 50

ppmv of methanol were passed through packed beds with volumes of 30 cm3, 60 cm3, and 90

cm3, respectively. This resulted in a constant 4.3 s space time and face velocities of 0.093 m/s,

0.19 m/s, and 0.28 m/s. All error bars shown in this chapter represent the standard deviation of

at least triplicate measurements, unless otherwise stated.









Adsorption

STC pellets of varying pore sizes (50 A, 120 A, 260 A) and constant TiO2 loading (12%)

were tested in the dark to determine methanol adsorption capacity using conditions for Study No.

1. A 12% TiO2 loading was chosen since previous studies for the degradation of organic

compounds from water have found that this TiO2 loading was optimal (Ludwig et al., 2008;

Londeree, 2002). The temperature of the reactor was kept at 500C to simulate the reactor

temperature when irradiated with UV light. As shown in Figure 5-1, the time required to reach

exhaustion was dependent on the pore size of the STC pellets, with the smallest pore size (i.e., 50

A) having the greatest methanol adsorption capacity. The total amount of methanol adsorbed by

the various STC, as determined by integration of the adsorption breakthrough curves, indicated

that methanol adsorption occurred via monolayer coverage versus pore filling since the

adsorption trends correlated with the BET surface area data, but not with the total pore volume.

Simultaneous Adsorption and Oxidation of Methanol

The photocatalytic oxidation of methanol was tested by repeating the experiments

described above in the presence of UVA light. Figure 5-2 shows the methanol removal

capabilities of the 50 A, 120 A, and 260 A STC pellets. Breakthrough of methanol (i.e., effluent

concentration greater than the detection limit of 0.6 ppmv or C/Co = 0.012) occurred for all STC.

The time to breakthrough for the 50 A pellets occurred after about 180 minutes in the presence of

UV light, whereas in the dark the initial breakthrough occurred immediately. For the 120 A and

260 A STC, breakthrough occurred during the first sample period. However, the effluent

concentration was lower in the presence of UV light than in the dark. For example, the

normalized effluent concentration (C/Co) of the first sample from the reactor packed with 260 A

STC was about 0.65 in the dark and 0.07 when illuminated. Thus, it does not appear that a

minimum surface coverage of water vapor or methanol was required for degradation to occur.









After a period of time (e.g., approximately 400 min for the 50 A STC pellets), the system

appeared to reach a pseudo-steady state (i.e., the effluent concentration was relatively constant)

such that the adsorption rate was equal to the oxidation rate.

The removal efficiency of the 120 A and 260 A pellets at steady state was similar (about

80%). The 50 A pellets removed greater than 90% of the methanol at steady state. The enhanced

performance of the 50 A pellets shown in Figure 5-2a is a result of the high internal surface area

of these pellets. Although the UV probably does not reach the very center of each STC pellet due

to the opacity of the TiO2, these results suggest that photons do penetrate past the external

surfaces of the pellets; otherwise, the performance of the pellets at steady state would be similar

regardless of internal surface area.

To assess whether the system in Figure 5-2a was truly at steady state, an extended study

was conducted over a period of four days using the 50 A pellets. This study (results shown in

Figure 5-2b) confirmed that the system was at steady state, achieving approximately 90%

methanol removal for the duration of the experiment.

Formation of Photocatalytic Byproducts

Formaldehyde, a byproduct of the photocatalytic oxidation of methanol (Tsuru et al., 2003;

Peral et al., 1997), was identified in the effluent of all studies conducted in the presence of UV

light. Formaldehyde was not detected in the influent gas. The effluent formaldehyde

concentration was dependent on the pore size of the STC pellets, as shown in Figure 5-3. The

effluent formaldehyde concentration increased over time until a steady state was reached. The

effluent formaldehyde concentration produced at steady state by the 260 A, 120 A, and 50 A

pellets was approximately 7, 4, and 2 ppmy, respectively. STC pellets with a higher surface area

possessed a greater number of active sites for adsorption and subsequent photocatalytic oxidation

to occur. In addition, diffusion of formaldehyde out of the pellets may have depended on the pore









size of the composite (i.e., a composite with a smaller pore size had greater resistance to pore

diffusion) (Chang et al., 2000). Restricted diffusion may have retained byproducts for further

oxidation. The effects of mass transfer will be discussed in further detail later in this chapter.

In addition to testing effluent streams for methanol and formaldehyde, TOC analysis was

performed in order to identify additional organic byproducts. For all studies, TOC analysis

proved that other organic byproducts were not measurable regardless of the test parameters. For

example, when 50 A 12% STC pellets were irradiated with UVA light, methanol and

formaldehyde were released in the effluent at a rate of about 0.003 mg of methanol (as carbon)

per min and about 0.004 mg of formaldehyde (as carbon) per min at steady state. The results of

TOC analysis showed the total rate of carbon released in the effluent was 0.007 mg/min. This

mass balance shows that the TOC present in the effluent (0.007 mg C/min) was accounted for by

the methanol (0.004 mg C/min) and formaldehyde (0.003 mg C/min). In addition, the absence of

methyl format, which is another byproduct that has been identified in the effluent of reactor

systems that photooxidize methanol (Tsuru et al., 2003; Sadeghi et al., 1996; Arana et al., 2004),

was confirmed by GC analysis.

Formic acid is a known intermediate byproduct in the oxidation from formaldehyde to

carbon dioxide and water. The lack of the presence of measurable amounts of formic acid in the

effluent can be attributed to the following reasons: (1) formic acid formed from the

photocatalytic oxidation of methanol is strongly adsorbed to the TiO2 surface (Lichtin et al.,

1994) and (2) only one electron hole is necessary for the total degradation of formic acid, which,

according to other literature values, results in a high apparent quantum yield (0.45) compared to

other organic compounds (0.06-0.001) (Dijkstra et al., 2002). Since formic acid is degraded









directly to carbon dioxide and water, the balance of the effluent carbon should be present as

carbon dioxide (Dijkstra et al., 2002).

Effect of Space Time on Methanol Degradation

The effect of space time on methanol removal using the 50 A 12% STC was tested using

experimental conditions 2, 3, and 4, which resulted in space times of 8.6 s, 2.1 s, and 1.1 s. The

space time was controlled by adjusting the air flow rate, although the influent methanol

concentration (50 ppm,) and relative humidity (95%) remained the same. Thus, the face velocity

of the air entering the packed bed and methanol and water vapor loading rate (mg/min) increased

as the space time decreased. The normalized effluent methanol concentration (C/Co) is shown in

Figure 5-4 for the duration of each study. Initially, the effluent methanol concentration was low,

due to adsorption onto the STC. For all space times, the effluent methanol concentration

increased over time until steady state removal was achieved. The effluent methanol

concentration at steady state was dependent on the space time. The shortest space time (1.1 s)

had the highest normalized effluent methanol concentration of about 0.52. The normalized

effluent methanol concentration for the 2.1 s and 8.6 s space times were about 0.38 and 0.025,

respectively. The error bars shown in Figure 5-4 represent the range of duplicate tests. Error bars

are not shown for the 1.1 s residence time since the experiment was not replicated.

Formaldehyde, which is an intermediate byproduct of methanol oxidation, was identified

in the effluent for all studies. Figure 5-5 shows the effluent formaldehyde concentrations for the

duration of the experiments for the various space times. The effluent formaldehyde

concentrations increased from time zero until reaching a steady state concentration. This

concentration was also dependent on the space time. The effluent formaldehyde concentrations

for the 8.6 s, 2.1 s, and 1.1 s residence times were about 0.3 ppmy, 3 ppmy, and 8 ppmy,

respectively. No other organic byproducts were identified in the effluent. Since one mole of









methanol is oxidized to one mole of formaldehyde and no other intermediate byproducts were

identified, the difference between the influent molar flux and the total effluent molar flux of

methanol and formaldehyde represents the mineralization rate of methanol.

The 120 A 12% and 260 A 12% STC were also tested for methanol removal using space

times of 1.1, 2.1, and 8.6 s (Study Nos. 2-4). For all STC, the effluent methanol and

formaldehyde concentrations at steady state are shown in Figure 5-6 for the various space times.

For both the 120 A and 260 A STC, the effluent methanol and formaldehyde concentrations were

dependent on the space time. The effluent formaldehyde and methanol concentrations for the 120

A were less than that of the 260 A STC for all space times studied. At a 1.1 s space time, the 260

A STC showed no mineralization of methanol, only oxidation of methanol to formaldehyde,

which then desorbed into the effluent gas stream rather than remain at the reaction site for

subsequent oxidation. Thus, at the 1.1 s space time, the conversion of methanol to formaldehyde

was the dominant reaction mechanism. For space times greater than 1.1 s, this mechanism no

longer dominated since both methanol and formaldehyde were oxidized. The longer space times

were achieved by lowering the gas flow rates, which resulted in lower methanol and water vapor

loading rates (as shown in Table 5-1). The lower loading rates presumably decreased the

competition for adsorption/reaction sites and the higher space times allowed more time for the

subsequent adsorption and oxidation of desorbed formaldehyde to occur, thus allowing the

oxidation of formaldehyde to proceed.

The greatest difference in performance between the various STC was observed at the

shortest residence time studied (1.1 s). The total normalized effluent concentration (1-XA), which

accounts for the concentration of both methanol and formaldehyde in the effluent, was 0.68 for

50 A, 0.91 for 120 A, and 1.0 for 260 A. As the space time increased, the performance of the









various STC converged so that the performance was more similar at the 8.6 s space time (l-XA

was 0.035 for 50 A, 0.045 for 120 A, and 0.049 for 260 A). The difference in performance

between the various pore sizes was likely a result of the surface area of the STC, with the higher

surface area STC (i.e., 50 A) resulting in better performance due to the greater number of

available adsorption and reaction sites. This was most pronounced at the 1.1 s space time, where

the ratio of influent methanol loading to surface area was the greatest and the most significant

difference in performance was observed. In addition to surface area effects, performance may

also be affected by mass transfer, which is directly effected by the pore size distribution of the

STC. For example, the STC with a small pore size may resist the transfer of methanol into the

STC or constrain byproducts from leaving the pellet, thereby allowing for complete oxidation.

This hypothesis is discussed further below.

Mass Transfer

The kinetics of the system can be limited by the effects of external and internal mass

transfer of methanol into the STC and byproducts out of the STC. In order to assess the effects

of external mass transfer on the methanol oxidation rate and the effluent byproduct concentration

(i.e., formaldehyde concentration), Study Nos. 5-7 were conducted with 50 A 12% STC. The 50

A STC was chosen for these studies since it had the smallest pore size and narrowest pore size

distribution. Thus, the 50 A STC would experience the greatest resistance to mass transfer. For

these studies, the face velocity was varied between 0.093 and 0.28 m/s while the space time was

held constant at 4.3 s by changing the volume of the packed bed. The steady state methanol and

formaldehyde concentrations in the effluent are shown in Figure 5-7. The effluent methanol

concentration decreased as a result of an increase in face velocity while the effluent

formaldehyde concentration was similar for all cases. This methanol removal trend was likely a

result of the decrease in resistance to mass transfer from the gas stream to the media as the face









velocity increased (Satterfield, 1970). Compared to the influent concentration (50 ppmv), the

difference in methanol removal from about 5.8 ppmv for the 0.093 m/s face velocity to 4.2 ppmv

for the 0.28 m/s face velocity was about 1.6 ppmv. Since this difference represents only 3% of

the influent concentration, the kinetic analysis presented later in this paper will neglect the

effects of external mass transfer.

Using the experimental methanol mineralization rates observed at steady state, the Weisz

modulus (Mw), or modified Thiele modulus, was calculated for Study Nos. 1-4, where space time

was varied. The Weisz modulus is used to estimate kinetic limitations as a result of resistance to

pore diffusion using the experimentally-determined rate of methanol removal at steady state.

Modulus values less than 0.15 indicate that there are no kinetic limitations as a result of pore

diffusion. The expression for the Weisz Modulus is shown in Equation 5-1(Levenspiel, 1999).

Mw = L2r/(Co*De) (5-1)

where L is the characteristic length (radius/2 for cylindrical pellets); r is the mean rate of

methanol oxidation per unit volume of STC; and De is the effective diffusion coefficient of

methanol within the STC.

The effective diffusion coefficient (De) is a function of the diffusion coefficient of

methanol (D), catalyst grain porosity (se) and tortuosity of the STC ('c), as shown in Equation 5-

2. The diffusion coefficient of methanol in air is about 15x10-6 and the tortuosity of the pores

was assumed to be 3, which is a typical value for mesoporous silica gels (Doucet et al. 2006).

De = D* Sc/ Zc (5-2)

The calculated Weisz modulus values for the various STC are shown in Table 5-2 for

Study No. 1-4, where space time was varied. For each of the space times, the Weisz modulus

indicated that there were no kinetic limitations as a result of pore diffusion. Although the 50 A









12% STC exhibited the most narrow pore size distribution and smallest average pore diameter of

the STC, the pore size was large enough to allow pore diffusion with very little resistance.

Therefore, the rates measured in the experiments are a result of the chemical (or intrinsic)

kinetics (Doucet et al., 2006).

Kinetics

The rate of reaction for gas-solid phase photocatalytic reactions can often be described by

the Langmuir-Hinshelwood (L-H) model, where the rate of reaction is equal to the rate constant

(k) times the surface coverage of the contaminant. Since the mineralization rate of methanol can

be described by the rate equation for a single molecular reaction and the influent concentration

was low, the L-H model can be simplified to the pseudo first-order equation shown in Equation

5-3 to describe the mineralization rate of methanol (Chen et al., 1999; Zou et al., 2006).

In(1-XA)= k + A (5-3)

where A is a constant. Although initial rate of photocatalytic degradation is typically used

for the L-H model, in this study the initial rate of photocatalytic degradation could not be easily

differentiated from the removal rate due to adsorption. In addition, the inhibitory effects of water

vapor may initially vary over time as the STC surface reaches its adsorption capacity for water

vapor. It should be noted that water vapor adsorption tests proved that capillary condensation of

water vapor did not occur in the pores of any STC studied. This data is discussed in further detail

in Chapter 6. At steady state conditions, the methanol removal rate is directly related to the

photocatalytic reaction rate and inhibitory effects of water vapor can be assumed to be constant

for a given system. Therefore, the reaction rate was modeled using the steady state

mineralization rates at the various space times studied (Study No. 1-4). The rate constant (k) was









determined for STC of the various pore sizes by a linear regression of -ln(1-XA) versus c, as

shown in Figure 5-8.

The use of the L-H model resulted in a good fit of the data (R2 = 0.99 for all STC). The

rate constant was 0.40 s-1 for all the STC pore sizes studied. The difference in conversion rates

was due to a lag time before mineralization occurred, which was observed with the 120 A and

260 A STC. Oxidation of methanol to intermediate byproducts occurred during the lag phase;

however further oxidation required for mineralization did not proceed. A lag phase was also

described by Chu and Wong (2004) in their study of the oxidation of alachlor in water, where the

dominant mechanism during the lag phase was the oxidation of alachlor to intermediate

byproducts with no mineralization. In this study, the lag time was expressed by the constant (A),

which was 0, -0.41, and -0.49 for the 50 A, 120 A, and 260 A STC, respectively. The lag times

for the 120 A and 260 A STC are equivalent to the intercept of the regression lines in Figure 5-8

with the abscissa, which are approximately 1.0 s for 120 A and 1.2 s for 260 A. Thus, the L-H

model is applicable to describe the mineralization of methanol for the 120 A and 260 A STC at

times greater than these lag times. In the case of the 50 A STC, the constant A is zero, indicating

that the L-H model is applicable for all space times. The zero lag time experienced by the 50 A

STC was likely due to its high surface area, which provides more adsorption and reaction sites

for the higher influent loadings associated with the low space times.

Effect of TiO2 Loading on Methanol Degradation

STC with an average pore size of about 50 A were tested with various TiO2 loadings (1%,

4%, 12%, and 60%) using Study No. 1 conditions. As shown in Figure 5-9, all of the STC

pellets, regardless of TiO2 loading, removed similar amounts of methanol (ca. 90%) at steady

state when continuously illuminated with UVA light. The time to initial breakthrough (i.e., time









when the effluent methanol concentration was detectable) was similar for STC pellets loaded

with 1-12% TiO2. Initial breakthrough for the pellets loaded with 60% TiO2 occurred

immediately due to the lower specific surface area of the composite (297 m2/g) compared to that

of the STC with lower TiO2 loadings (e.g., 617 m2/g for the 50 A 4% STC).

TiO2 loading did affect effluent formaldehyde concentration, as shown in Figure 5-10.

This graph shows that a 4% TiO2 loading was optimum, resulting in steady state effluent

formaldehyde concentrations below 0.5 ppmv. According to Byrne et al. (2008), the total

available TiO2 surface area for the 4%, 12%, and 60% TiO2 loadings was 3.4 m2/g, 8.1 m2/g, and

13 m2/g, respectively. Therefore, the total surface area of TiO2 in the systems containing the STC

with 4%, 12%, and 60% was 51 m2, 137.7 m2, and 234 m2. Although Byrne et al. (2008) did not

measure the available TiO2 surface area for the 50 A 1% STC, it can be assumed that the total

available surface area would be lower than that for the 4% system based on the trend observed

for TiO2 loadings between 4 and 60%. Thus, the total quantity of active sites on the 1% STC was

likely less than that for the 4% STC, leading to greater overall oxidation of methanol and its

byproducts by the 4% STC.

Based on the total amount of available TiO2, one may expect that the system containing the

60% STC would achieve the greatest rate of mineralization (i.e., lowest formaldehyde

concentration) since it had the greatest TiO2 surface area available. However, the overall surface

area of the composite was 297 m2/g, which was lower than that of the 50 A STC with lower TiO2

loadings (e.g., 50 A 4% STC had a surface area of 617 m2/g). Therefore, the higher total surface

area of the 50 A 4% STC may have enhanced the adsorption of the methanol such that the

available TiO2 was more efficient. In addition, increased TiO2 loadings may have resulted in a









decrease in transparency such that total amount of irradiated TiO2 was greater for the 4% STC

compared to the 12% and 60% STC.

In order to test this hypothesis, UV transparency tests were conducted using 50 A STC

pellets of the varying TiO2 loadings. The pellets were packed tightly into a 10 mm quartz cuvette

and placed in a Hach DR4000 spectrophotometer. The % transmittance of the UV light through

the packed bed was measured at a wavelength of 365 nm and the results are shown in Figure 5-

11.

The % transmittance of the 50 A 0% STC was about 8%. The 50 A 0% STC was

comprised solely of SiO2, which is known to be highly transparent to UVA light. The low %

transmittance shown in these results was likely due to the scattering of UVA light through the

packed bed of pellets since the detector of the spectrophotometer measured only the UV light

transmitted in a straight path through the cuvette (i.e., UV light that was scattered rather than

absorbed was not measured). No significant difference was measured between the STC of

varying TiO2 loadings. The quantity of UV light transmitted through the pellets containing TiO2

was likely a result of the light penetrating through only the interstitial space of the pellets. Based

on physical observations of the pellets, the 1% STC appeared to be semi-transparent while the

60% STC appeared to be opaque. Therefore, it can be expected that the distance that the light

could penetrate through the STC would be different. However, the 10 mm thickness used in this

experiment was likely too large to be able to measure differences in transparency through the

STC since the light would have to pass through 2-3 layers of pellets and scattering would occur.

Thus, it is possible that the UV could penetrate farther into the 1% STC, but still be completely

attenuated or scattered over a distance of 10 mm. Thus, the hypothesis stated above can be

neither confirmed nor denied based on these data. In order to evaluate the penetration of UV









light through the STC, individual pellets or thinner layers of STC made with varying TiO2

loadings should be measured using an integrating sphere to capture both transmitted and

scattered light.

Effect of UV Wavelength on Methanol Degradation

The effect of UV wavelength on methanol oxidation was evaluated by using 50 A 12%

STC irradiated with UVA and UVC lamps using Study No. 1 conditions. The initial methanol

breakthrough time (i.e., the time when the effluent concentration is measurable) was about 308 +

28 min when the TiO2 was activated using the UVC lamp. This time was greater than the initial

breakthrough time (135 25 min) when the UVA lamp was used. The UVA and UVC lamps

performed similarly with respect to methanol removal when the reactor reached steady state,

resulting in the oxidation of about 85% of the influent methanol. Figure 5-12 shows that the use

of the UVC lamp resulted in lower effluent formaldehyde concentrations (i.e., more complete

oxidation) compared to the UVA lamp. The total degradation efficiency was about 88% for the

UVA lamp and 90% for the UVC lamp.

The UVC lamp enhanced reactor performance because the photocatalytic reaction rate is

proportional to the rate of generation of electrons and holes on the TiO2 surface, which is in turn

proportional to the photon flux, or light intensity (Dijkstra, 2002). The intensity of the UVA and

UVC light inside of the reactor system was 8.71E-5 and 1.13E-4 mEinsteins/s/cm3, as

determined by potassium ferroxalate actinometry. Thus, the intensity emitted by the UVC light

was about 29% greater than that emitted by the UVA lamp. The increase in degradation as a

result of increased light intensity can be predicted using the pseudo first order model shown in

Equation 5-4.

C/Co = k*T (5-4)









where k = kol for variable light intensity (I). As previously determined, the k value for

Study No. 1-4 conditions when illuminated with UV light was 0.4s-1. The predicted normalized

effluent concentration as a function of adsorbed light flux is shown in Figure 5-13. According to

this model, the normalized effluent concentration of the system irradiated with UVC light should

be approximately 0.90. This assumed that 90% of the light emitted from the UV lamp was

absorbed by the STC pellets and the remainder passed through the packed bed, which was

confirmed by actinometry measurements. The actual normalized effluent concentration was

indeed 0.90.

Effect of H2S on Methanol Degradation

TRS species are often present in HVLC gases and are a major contributor to the

characteristic odor associated with pulp and paper mills. The effect of H2S, which was used as

the representative TRS species, on methanol removal was studied. In addition, the removal of

H2S was investigated to understand if the STC can provide a co-benefit for odor removal.

Studies were conducted using 50 A 4% STC and Study No. 1 conditions with 50 ppmv H2S

added to the gas stream.

In the dark, there was no measurable adsorption of H2S onto the STC at 500C after 10

minutes. Other studies have shown that H2S had a weak adsorption affinity for the TiO2 surface

(Kato et al., 2005; Portela et al., 2007; Sopyan 2007). Kato et al. (2005) did not observe

adsorption of H2S onto TiO2 coated on an alumina substrate. Portela et al. (2007) tested gases

containing an influent H2S concentration of 35 ppmv and RH between 0 and 70%. They observed

an initial decrease in effluent H2S concentration due to adsorption for a short period of time (i.e.,

less than an hour); however the extent of adsorption was not quantified. The duration ofH2S

adsorption was dependent on the humidity of the gas stream, since water vapor competed with

H2S for adsorption sites. In this work, water vapor competition would be expected to further









inhibit H2S adsorption since the influent RH was about 95%. Sopyan (2007) attributed the weak

adsorption affinity of H2S for the TiO2 surface to the difficulty of forming hydrogen bonds

between H2S and hydroxyl groups of the TiO2 surface, which is the primary pathway for

adsorption onto TiO2 surfaces. Since adsorption via hydrogen bonding is also the primary

pathway for adsorption onto the surface of SiO2 (Travert et al., 2002), the use of a composite

material containing SiO2 and TiO2, such as the STC, would not be expected to enhance H2S

adsorption. Sopyan (2007) concluded that the extent of adsorption of contaminants onto the TiO2

surface was related to their ability to serve as an electron donor (i.e., Lewis basic group).

Ammonia, which was reported to be a strong Lewis base, showed adsorption capacities an order

of magnitude greater than that of H2S, which has lower electron-donor ability and has been

shown to adsorb via hydrogen bonding only in strongly basic environments (Sopyan, 2007).

The STC pellets were tested for oxidation of methanol and H2S by irradiating the STC with

UVC light. Note that there was no photolysis of methanol or H2S when irradiated with UVC

light alone. As shown in Figure 5-14, photocatalytic oxidation of both methanol and H2S was

achieved. The effluent H2S concentration was about 28.4 + 2.4 ppmv for the duration of the

experiment, which corresponds to an oxidation efficiency of approximately 43%. Effluent

methanol concentrations were initially low and increased until a steady state was reached. Initial

breakthrough occurred during the first sample (as opposed to occurring after about 300 minutes

for the non-H2S system). The effluent methanol concentration at steady state was about 10 ppmy,

which corresponds to a removal efficiency of about 80%. This corresponds to a decrease in

removal efficiency from about 95% when H2S was absent from the gas stream (results shown in

Chapter 6). In the presence of H2S, the decrease in methanol removal efficiency was likely due to

the competition for oxidation sites on the surface of the STC. Since both the methanol and H2S









concentrations reached a steady state, the formation of sulfur containing byproducts did not

likely contribute to the deactivation of the catalyst during the experiment.

Formaldehyde, an oxidation byproduct of methanol, and SO2, an oxidation byproduct of

H2S, were identified in the effluent. Effluent formaldehyde concentrations were about 0.9 ppmv

at steady state. This concentration was about two times greater than effluent formaldehyde

concentrations at steady state when H2S was absent from the influent gas (results shown in

Chapter 6). Analysis of the STC pellets after the completion of the experiment showed that about

2.8 mg/g of S042- was loaded onto the surface of the pellets after approximately 3000 min. A

mass balance of the reactor confirmed that S042- and SO2 were the only byproducts of H2S

conversion. The SO2 concentration increased over the duration of the experiment. This may be

caused by (1) the saturation of water and methanol on the surface of the STC, which are more

strongly adsorbed than the SO2, would promote desorption of the SO2 and inhibit its subsequent

adsorption onto the STC for further oxidation and retention; and (2) the accumulation of S042- on

the surface of the STC over time may inhibit the oxidation of SO2 to SO42- such that the SO2

desorbed from the STC into the effluent before being oxidized further.

The 50 A 4% STC were tested for H2S removal (Co = 50 ppmv) in the absence of methanol

using Study No. 1 conditions. Both H2S and SO2 were present in the effluent, as shown in

Figure 5-15. The effluent H2S concentration was 27.9 + 3.5 ppmv for the duration of the study.

This was similar to the previous study in which methanol was also present in the gas stream. In

addition, the effluent SO2 concentration also followed a trend similar to that in Figure 5-14 when

methanol was present in the influent gas. Thus, oxidation of H2S was not significantly affected

by the presence of methanol. This was likely because the competition from water vapor

dominated since the influent concentration of water vapor was about 28,000 ppmv while that of









methanol was 50 ppmv. S042- was identified on the surface of the STC and a mass balance


showed that SO2 and S042- were the only oxidation byproducts.


Table 5-1. Summary of experimental conditions.
Study No. V Q v T Co Methanol Loading Water Vapor
(cm3) (L/min) (m/s) (s) (ppmv) Rate (mg/min) Loading Rate
(mg/min)
1 30 0.42 0.093 4.3 50 0.028 5.4
2 30 0.21 0.046 8.6 50 0.014 2.7
3 30 0.84 0.19 2.1 50 0.056 10.9
4 30 1.68 0.37 1.1 50 0.11 22.8
5 60 0.84 0.19 4.3 50 0.056 10.9
6 90 1.26 0.28 4.3 50 0.084 16.3


Table 5-2. Weisz modulus values for variable space time experiments.


Pore size (A)
Study No. 1
50
120
260
Study No. 2
50
120
260
Study No. 3
50
120
260
Study No. 4
50
120
260


TC (s)


Weisz modulus (Mw)


0.058
0.042
0.036

0.028
0.028
0.028

0.077
0.035
0.033

0.090
0.009
0.000










I .L

1.0

0.8


0.6

0.4 -- 50A 12% (532 m2/g)
S- 120A 12% (290 m2/g)

0 2 -- 260A 12% (154 m2g)

0.0 1


0 50 100


150 200 250 300 350 400


Time (min)


Figure 5-1. Adsorption breakthrough curves for STC pellets of varying pore sizes (50 A, 120 A,
and 260 A) and constant TiO2 loading (12%).

1.0


^c-~--*tf*------^C
1 2 3 4 5
Time (days)


A


-- 50 A 12%
-- 120 A 12%
- 260 A 12%


Time (min)

Figure 5-2. Methanol removal using STC pellets. A) Methanol removal using STC of varying
pore sizes illuminated with UVA light. B) Extended study for 50 A 12% STC pellets.


1000


T_, -- "^___











- 50 A 12%


- 120 A 12%
- -260A 12%


600


900


1200


1500


1800


Time (min)
Effluent formaldehyde concentrations from STC pellets of varying pore sizes (50 A,
120 A, and 260 A) when illuminated with UVA light.


- 1.1 s
-- 2.1 s
-- 8.6 s


1000


1500


2000


2500


3000


Time (min)
Figure 5-4. Normalized effluent methanol concentration for 50 A 12% STC illuminated with
UVA light at various space times.


300


Figure 5-3.


(


(M) I u E U I -- -












m-1.1 s


1[ 2 1 s
S8.6 s









Jo ooooo-


1000


1500


2000


2500


500


3000


Time (min)

Figure 5-5. Effluent formaldehyde concentration for 50 A 12% STC illuminated with UVA light
at various space times.


T (S)
Figure 5-6. Effluent methanol and formaldehyde concentrations at steady state at varying space
times. A) 50 A. B) 120 A. C) 260 A STC.


* Formaldehyde
E Methanol


hF-










60

50

40

S30

S20

10

0


1.1 2.1 4.3 8.6


T (S)


Figure 5-6. Continued.


Formaldehyde
SMethanol










4.3 8.6
T (s)


60

50

40

S30

S20
1
0
0






















0.093 m/s


* Formaldehyde
O Methanol


0.19 m/s
Face Velocity


0.28 m/s


Figure 5-7. Effluent methanol and formaldehyde concentrations at steady state for variable face
velocities and constant space time (4.3 s) for 50 A 12% STC.

A
4.0
3.5
-In(1-XA) = 0.401
3.0 2
3.0 R2 = 0.99
2.5
2.0
1.5
1.0
0.5
0.0
0 2 4 6 8 10
T (S)
Figure 5-8. Linear regression of L-H model using mineralization rates achieved at various space
times. A) 50 A 12% STC. B) 120 A 12% STC. C) 260 A 12% STC.


^I












3.3
-ln(l-XA)= 0.40c 0.41 -ln(l-XA)= 0.40T 0.49
2 23.0
R2 = 0.99 R2 = 0.99
2.5

X 2.0

s 1.5

1.0

0.5

0.0 -
0 2 4 6 8 10 0 2 4 6 8 10
'C (s) C (s)

Figure 5-8. Continued.


0.4


-- 1%

-*-4%

-- 12%

6 60%


0.3



0.2



0.1



0.0


0 200 400 600 800 1000

Time (min)

Figure 5-9. Effect of TiO2 loading in 50 A STC on methanol removal when illuminated with
UVA light.











1.8
1.6 T

1.4 -







0.4
0.2


50A 1% 50A 4% 50A 12% 50A 60%

Figure 5-10. Effect of TiO2 loading on formaldehyde production at steady state using 50 A STC
illuminated with UVA light.


10
9
8
7
6
5
;- 4
3
2T
1-
0
50A 0% 50A 1% 50A 4% 50A 12% 50A 60%

Figure 5-11. Percent transmittance of UVA light through 50 A STC with various TiO2 loadings.




























1000


1500


2000


Time (min)

Figure 5-12. Effluent formaldehyde concentrations using 50A 12% STC irradiated with UVA
and UVC light.




1.0
0.9
0.8
0.7
0.6
S0.5
0.4
0.3
0.2
0.1
0.0
O.OE+00 2.0E-05 4.0E-05 6.0E-05 8.0E-05 1.0E-04 1.2E-04 1.4E-04


I (mEinsteins/s/cm 3)

Figure 5-13. Predicted C/Co versus absorbed light flux for 50 A 12% STC, 4.3 s residence time,
Co = 50 ppm, and 95% relative humidity.


--UVA
UVC











Hydrogen Sulfide
E Sulfur Dioxide
Methanol
S* o Formaldehyde




E-3
[]0

C) 0


0 500 1000 1500 2000 2500 3000 3500
Time (min)


Figure 5-14. Effluent concentrations of H2S, methanol and oxidation byproducts from 50 A 4%
STC illuminated with UVC light (Co methanol = 50 ppmv, Co H2S = 50 ppmv).


500


1000


1500


Time (min)


Figure 5-15. Effluent concentrations ofH2S and SO2 from 50 A 4% STC illuminated with UVC
light (Co H2S = 50 ppmv).


Hydrogen Sulfide
o Sulfur Dioxide










n
-mO









CHAPTER 6
EFFECT OF CATALYST SUPPORT ON THE PHOTOCATALYTIC DEGRADATION OF
METHANOL IN A PACKED-BED REACTOR

The objective of this study was to investigate the effect of the catalyst support on methanol

oxidation efficiency using STC, TiO2 coated on an opaque adsorbent material (i.e., activated

carbon (AC)), and TiO2 coated on a non-adsorbent material (i.e., 5 mm glass spheres).

Preliminary scale-up was performed at the bench-scale by increasing the annulus size of the

reactor from 8 mm to 25 mm in order to evaluate the efficacy of these materials for full-scale

applications. In addition, the competitive effects of water vapor on methanol adsorption capacity

and oxidation efficiency were evaluated.

The STC and TiO2-coated AC were analyzed for BET surface area and total pore volume.

The results of this analysis are shown in Table 6-1. A comparison of the total surface area and

TiO2 mass loading of the materials normalized for the reactor bed volume are shown in Table 6-

2. Although the total specific surface area of the STC was less than half of that of the TiO2-

coated AC (617 m2/g versus 1424 m2/g), the surface area per reactor volume was more similar

(308 m2/cm3 versus 380 m2/cm3) due to the higher bulk density of the STC. The bulk density of

the STC and TiO2-coated AC was about 0.53 and 0.26 g/cm3, respectively. The TiO2-coated

glass spheres had the smallest total surface area per bed volume (0.43 m2/cm3) since the glass

spheres were non-porous and only external surface area contributed to the total available surface

area.

The TiO2 mass loading per reactor volume was similar for the TiO2-coated AC and TiO2-

coated glass spheres (9.9 and 10.5 mg/cm3, respectively). The 50 A 4% STC had a total TiO2

mass loading of 26.7 mg/cm3, which was about 2.5 times greater than the TiO2-coated AC and

TiO2-coated glass spheres. The TiO2 loading was calculated based on the amount of TiO2

contained in each material and does not indicate the amount of TiO2 that was available for









adsorption and photocatalytic reactions. For example, based on studies by Byrne et al. (2008),

about 32% of the TiO2 in the 50 A 4% STC was inaccessible due to the formation of the SiO2

matrix around the TiO2 during synthesis. Therefore, the effective TiO2 loading per bed volume

would be 18.2 mg/cm3. Based on the synthesis method for the TiO2 coated AC and glass

spheres, the majority of the TiO2 should be accessible since TiO2 particles were deposited on the

external surface of the materials.

Methanol Adsorption and Oxidation in a Low Humidity Gas Stream

The TiO2-coated materials were tested for methanol removal via adsorption (i.e., in the

dark) at room temperature and simultaneous adsorption and oxidation (i.e., in the presence of

UVC light) using a 4.3 s residence time, an influent methanol concentration of 50 ppmv and a

water vapor concentration of less than 65 ppmv (typically 10 ppmv), which was specified by the

supplier of the compressed air used as the feed gas. A water vapor concentration of 65 ppmv is

equivalent to an RH of about 0.22% at 23 C. As shown in Figure 6-1, the adsorption capacity

of the TiO2-coated glass spheres was low (ca. 0.1 mg/g) and initial breakthrough (i.e., effluent

concentration greater than the MDL of 0.6 ppmv or C/Co = 0.012) was achieved during the first

sample. The TiO2-doped adsorbent materials achieved extended periods of methanol removal

(i.e., 1000 min for the AC and 3200 min for the STC) before breakthrough. The total methanol

capacity of the STC (11 mg/g) was greater than that of the TiO2-coated AC (6 mg/g). The

adsorption capacity of the silica gel is a result of its surface chemistry, which is dominated by

silanol functional groups (Si-OH) that serve as good adsorption sites for alcohols (e.g., methanol)

via hydrogen bonding (Parida et al., 2006). In the presence of UV light, the effluent methanol

concentrations for the TiO2-coated glass spheres and STC were below the MDL (0.6 ppmv or

C/Co = 0.012). The effluent concentration of the TiO2-coated AC reached a steady state

concentration of about 2.5 ppmv (C/Co = 0.05) after about 2500 min.









The effluent was tested for the presence of intermediate organic byproducts. As shown in

Figure 6-2, TiO2-coated glass spheres and TiO2-coated AC created formaldehyde in the effluent,

whereas the STC did not. According to TOC and GC analysis, other oxidation byproducts (e.g.,

formic acid, formates) were not identified in the effluent gas stream.

The formaldehyde concentration for the TiO2-coated AC was low for the first 1000 min

and then steadily increased until reaching a steady state concentration of about 6.3 ppmv after

2000 min. The production of formaldehyde as an intermediate oxidation byproduct exceeded the

adsorption capacity of the AC, thus causing the breakthrough profile shown in Figure 6-2. The

effluent formaldehyde concentration for the TiO2-coated glass spheres was about 0.3 ppmv at

steady state, which was achieved after about 500 min. For the STC system, effluent methanol

concentrations were below the MDL and effluent formaldehyde concentrations were less than

0.1 ppmv for the duration of the experiment. Thus, the STC resulted in the most efficient system

for the mineralization of methanol. Several factors may have attributed to the enhanced

efficiency of the STC system:

* The STC system had twice the amount of available TiO2 surface area, which likely resulted
in more active sites for oxidation leading to an increase in the overall efficiency.

* The STC exhibited a greater adsorption capacity for methanol compared to both the TiO2-
coated AC and the TiO2-coated glass spheres. This enhanced adsorption capacity would
concentrate the pollutant near the photocatalyst surface, potentially increasing the
oxidation rate. Studies have shown that enhanced adsorption capacity has resulted in
higher oxidation rates for organic compounds (Anderson and Bard, 1997; Tsumura et al.,
2002; Vohra and Tanaka, 2003; Torimoto et al., 1996).

* The STC may have been less affected by the low water vapor concentration than both the
TiO2-coated AC and glass spheres. It has been suggested that Si02-TiO2 materials use
hydroxyl groups at the Si02-TiO2 interface to produce hydroxyl radicals for the
degradation of organic compounds thus enhancing the efficiency of the material (Yang and
Chen, 2005).

Compared to the TiO2-coated AC, both the STC and TiO2-coated glass spheres were more

efficient for methanol mineralization. This was likely due to the transparency of the silica gel and









glass spheres, which would allow for more efficient use of the UV radiation within the reactor.

AC, which is opaque, does not allow transmission of UV light into or through the carbon matrix.

For the TiO2-coated AC system at steady state, the effluent concentration of formaldehyde

was about 6.3 ppmv at steady state while the methanol concentration was about 1 ppmv. This

indicates that methanol was preferentially oxidized over formaldehyde in the TiO2-coated AC

system. The effluent concentration of formaldehyde from the TiO2-coated glass spheres was 0.3

ppmv while the effluent methanol concentration was non-detectable. Since the MDL for

methanol was 0.6 ppmy, it was unclear if methanol was preferentially oxidized in this case. The

oxidation of methanol to formaldehyde can proceed by the direct oxidation of methanol in the

electron hole in the absence of water vapor (Equations 2-9 through 2-11). However, the

subsequent reaction from formaldehyde to formic acid requires a water molecule (Equation 2-

13). The oxidation of formic acid to carbon dioxide and water can be accomplished by reaction

with the electron hole in the absence of water vapor (Equations 2-14 and 2-15), which supports

the absence of formic acid in the reactor effluent. Thus, the rate limiting step in the

mineralization of methanol using TiO2-coated AC was the oxidation of formaldehyde to formic

acid. In the case of the STC, no formaldehyde was present in the effluent likely due to its higher

overall efficiency (as discussed above). The availability of hydroxyl radicals at the Si02-TiO2

interface may serve to increase the efficiency of formaldehyde oxidation to formic acid by

allowing the indirect oxidation mechanism to proceed. In addition, the STC may have been able

to adsorb the water vapor that was present in the gas stream better than the TiO2-AC due to its

surface functionality, which consists of silanol groups (Si-OH) that easily adsorb water via

hydrogen bonding (Nawrocki, 1997; Morrow and Gay, 2000). This adsorbed water vapor could

function to participate in direct oxidation or form hydroxyl radicals for indirect oxidation.









Methanol Adsorption and Oxidation in a High Humidity Gas Stream

In order to further elucidate the competitive effects of water vapor on the TiO2-coated

materials, they were tested for methanol adsorption and oxidation in a high humidity gas stream

(RHi = 95% at 23 C). As shown in Figure 6-3, the adsorption breakthrough profiles were steeper

and the adsorption capacities of the materials were reduced (1.2 mg/g for STC and 1.9 mg/g for

AC). Although the AC broke through earlier than the STC, its adsorption capacity on a mg/g

basis was greater than that of the STC due to the difference in bulk density between the two

materials (i.e., 0.26 g/cm3 for AC and 0.53 g/cm3 for STC). The humid gas stream negatively

affected the STC adsorption capacity more than that of the AC, indicating that the competitive

effect of water vapor was more pronounced for STC than AC. This can be expected since the

STC is dominated by silanol functional groups, which readily form hydrogen bonds with water,

while the AC has a more heterogeneous surface chemistry (Puri, 1970).

In the presence of UV, the methanol removal efficiencies of the materials were about 95%

for both the STC and TiO2-coated AC. The systems reached steady state after approximately

250 min for AC and 400 min for STC. The TiO2-coated glass spheres immediately achieved

steady-state removal efficiency, which was between 80 and 90%. Steady state was quickly

reached due to the low surface area and resulting low adsorption capacity of the TiO2-coated

glass spheres.

As shown in Figure 6-4, the increase in RH resulted in decreased formaldehyde production

in the TiO2-coated AC system (i.e., about 0.8 ppmv at steady state). The reaction of

formaldehyde to formic acid was no longer limited by the lack of water vapor, which may have

resulted in lower formaldehyde concentrations in the effluent compared to the low RH case

shown in Figure 6-2. In addition, Yamakata et al. (2003) showed that water vapor enhanced the









photocatalytic oxidation of methanol by preventing electron accumulation, which would

otherwise cause defective sites on the TiO2 surface and faster electron-hole recombination.

Although the water vapor increased the overall efficiency of the TiO2-AC system, it

resulted in a decrease in efficiency for the TiO2-coated glass spheres and STC. The humid gas

stream resulted in an increase in effluent formaldehyde concentrations for the TiO2-coated glass

spheres (about 1 ppmv) and STC (0.5 ppmv). The decrease in overall oxidation efficiency

compared to that in dry conditions was likely a result of water competition for adsorption and

reaction sites. The presence of higher effluent methanol concentrations in these systems indicates

that the competition with water vapor was prevalent. It was hypothesized that the STC was more

negatively affected by the increase in water vapor than the TiO2-coated AC because the surface

coverage of water vapor on the STC was greater than that on the AC. Water vapor adsorption

studies confirmed this hypothesis and are discussed in more detail below.

Water Vapor Adsorption

In order to elucidate the effects of water vapor on the STC and TiO2-coated AC systems,

the water vapor adsorption profile was determined by measuring the effluent relative humidity of

humid gas (influent RH = 95% at 23 C) after passing through the reactor packed with TiO2-

coated AC or STC and illuminated with UV light. The results of these studies are shown in

Figure 6-5. Note that the MDL for the hygrometer was 12% RH, so the lowest RH value plotted

on the graph is 12% even though the actual effluent RH may have been lower. The total amount

of water adsorbed by the STC and TiO2-coated AC was 59 and 25 mg/g, respectively. For the

STC system, this indicates that there were 3.7 water molecules adsorbed per nm2 of STC, giving

a surface coverage of about 0.26 m2 of water per m2 of STC (26%). For the TiO2-coated AC,

there were 0.58 water molecules adsorbed per nm2, giving a surface coverage of about 0.05 m2 of









water per m2 (5%). Thus, the competition between water vapor and methanol on the STC surface

would be much greater than that on the TiO2-coated AC.

Capillary condensation within the pores was not evident for either the STC or TiO2-coated

AC. The Kelvin equation predicts that at 500C, which is the approximate temperature of the bed

with the UV light on, capillary condensation should only occur in pores less than 12 A in

diameter. For both the STC and TiO2-coated AC, the majority of the pores were greater than 12

A in diameter. For the STC, pores greater than 12 A in diameter contributed to more than 95% of

the total pore volume. According to Khan (2006), approximately 90% of the pore volume in

BioNuchar AC, which was used in this study, resulted from pores greater than 12 A in diameter.

In order to elucidate the effects UV light may have on the adsorption of water onto TiO2-

doped materials, the STC was tested for water adsorption in the dark at 50 OC. The results, also

shown in Figure 6-5, indicate that there was no difference between the breakthrough profiles of

the STC tested in the dark at 500C or in the presence of UV light. Thus, the effect of UV light on

water vapor adsorption was primarily a function of the heat generated by the lamp.

Reactor Scale-up using TiO2-doped Materials

The TiO2-doped materials were tested in a large annulus reactor (25 mm annulus) in order

to evaluate the TiO2-doped materials for potential use in full-scale applications, particularly for

the treatment of HVLC gases emitted from pulp and paper mills. The TiO2-doped sorbents were

tested for methanol removal in a high humidity gas stream (RH = 95% at 23 C). The high

humidity case was chosen since HVLC gases are saturated with water vapor. The flow rate was

adjusted such that the mass transfer characteristics (i.e., residence time and face velocity) were

the same as those from the 8 mm annulus studies. As shown in Figure 6-6, the TiO2-coated glass

spheres achieved about 40% methanol removal over the duration of the study. The breakthrough

profile for the STC was shallower than that of the TiO2-coated AC, likely due to the greater









capacity for methanol adsorption by the STC, resulting in higher removal rates for the STC

between about 100 and 1300 min. Both TiO2-coated adsorbents achieved about 50% methanol

removal at steady state. Note that error bars are present only for the experiments using STC in

Figures 6-6 and 6-7 since experiments testing the TiO2-coated AC and glass spheres were not

replicated.

As shown in Figure 6-7, the effluent formaldehyde concentrations for the STC and TiO2-

coated AC were similar and increased steadily until reaching a steady state concentration of

between 1.5 and 2.0 ppmv. The TiO2-coated glass spheres achieved steady state production of

formaldehyde at about 1.6 ppmv. According to TOC analysis, other oxidation byproducts (e.g.,

formic acid, formates) were not identified in the effluent gas stream.

The decrease in methanol oxidation efficiency in the large annulus reactor compared to

that of the small annulus reactor was likely a result of inadequate UV light exposure within the

entire packed bed. In the large annulus reactor, the TiO2-doped adsorbents performed better than

the TiO2-coated glass spheres, indicating that adsorption of methanol onto a high surface area

catalyst support resulted in higher oxidation rates. However, no difference was discerned

between the silica gel, which is transparent, and AC, which is opaque, when used as the catalyst

support in the large annulus reactor. The similar performance between the STC and the AC may

be a function of the STC's affinity for water vapor, which resulted in a decrease in oxidation

efficiency in the high humidity gas stream. It is expected that at lower RH, the efficiency of the

STC system would improve as the competition of water vapor decreases.

Table 6-1. BET surface area and average pore volume of TiO2-doped sorbents.
Material BET surface area (m2/g) Pore volume (cc/g)
STC 616 0.79
TiO2-coated AC 1424 1.06









Table 6-2. Surface area and TiO2 loading per reactor volume.
Material Total surface area per bed volume TiO2 loading per bed volume
(m2/cm3) (mg/cm3)
50 A 4% STC 308 26.7
TiO2-coated AC 380 10.5
TiO2-coated glass spheres 0.43 9.9


1.0 ,'
0.9 ,--
0.8
0.7 -,'"
0.6 0
S0.5 -
0.4 ---.-
0.3 i -
0.2
I '
0.1
o. o -----------------------


0 1000 2000 3000 4000 5000 6000
Time (min)

---TiO2-AC: UV -0--- TiO2-AC: Dark
A STC: UV ----- STC: Dark
TiO2-Glass Spheres: UV ----- TiO2-Glass Spheres: Dark
Figure 6-1. Normalized effluent methanol concentration for titania-doped materials used in the
dark (adsorption only) and with UV light (adsorption and oxidation).












7

6

E 5

4

S3

2

E 1


0 1000 2000 3000 4000 5000 6000

Time (min)

-- TiO2-AC: UV -- STC: UV TiO2-Glass Spheres: UV

Figure 6-2. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials.





0.8 -
0.7 i
0.7 ---, -,--------------

0.6/
0.5 -A-
0.4 /
0.3 i/
0.2 -
0.1


0 500 1000 1500 2000

Time (min)

STiO2-AC: UV --T---Ti02-AC: Dark
A STC: UV --A--- STC: Dark
TiO2-Glass Spheres: UV ------ TiO2-Glass Spheres:Dark


Figure 6-3. Normalized effluent methanol concentration for TiO2-doped materials used in the
dark and with UV light in a high humidity gas stream (RH = 95%).















1.0




0.5




0.0 m
0 500 1

Tin

-- TiO2-AC: UV -A- STC: UV


000

ie (min


1500 2000


2Gla Sphere: UV
TiO2-Glass Spheres: UV


Figure 6-4. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a
high-humidity gas stream (RH = 95%).


0 50


100 150 200 250 300 350


400


Time (min)
Figure 6-5. Water vapor adsorption breakthrough profile in a high humidity gas stream (RH
95%).


..... Influent

-A- STC: UVC

-A- STC: Dark at 50 deg. C

-- TiO02-AC: UVC


















100 1500 2000


L/>


Time (min)


-- TiO2-AC: UV -A- STC: UV


TiO2-Glass Spheres: UV


Figure 6-6. Normalized effluent methanol concentration for TiO2-doped materials used in a large
annulus reactor (25 mm) and high humidity gas stream (RHi = 95%).


0 500 1000 1500
Time (min)


2000


-- TiO2-AC: UV -A- STC: UV


TiO2-Glass Spheres: UV


Figure 6-7. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a
25 mm annular reactor and high humidity gas stream.


1000


1500


2000









CHAPTER 7
PILOT STUDIES FOR METHANOL DEGRADATION

Based on the promising results and knowledge gained from bench-scale studies, a pilot-

scale reactor was designed to treat 40 ACFM using 50 A 4% STC pellets and UVC lamps. The

laboratory-scale synthesis process for the STC pellets was scaled up to make the approximately

100 pounds of pellets needed for the studies. This scale-up process represented a substantial

increase in production compared to the laboratory scale synthesis technique, which yielded

several grams of pellets per batch. A novel test apparatus was designed and built to study the UV

light distribution through a packed bed of STC pellets. Using this information as well as that

learned during bench-scale studies, a pilot reactor was designed, fabricated, and tested for

methanol degradation from an air stream saturated with water vapor.

STC Synthesis for Pilot-Scale Studies

In order to produce a sufficient quantity of STC pellets (50 A 4%) for the pilot studies, the

bench-scale synthesis method was modified to increase production efficiency while producing

composites with similar characteristics (i.e., surface area and pore size). Silbond Condensed

TEOS was chosen as the silica source. Silbond Condensed was chosen since it contained 94%

TEOS with the balance comprised of ethanol and polysilicates. This TEOS source was chosen

since it was economical and the balance of ingredients would not likely negatively affect the

characteristics of the final STC product. The TEOS was added to water, ethanol, 1 N nitric acid,

3%wt hydrofluoric acid, and Degussa P25 TiO2. The ingredients were stirred using a paddle

mixer. Figure 7-1 shows an image of the mixing assembly used for blending the STC raw

materials (mixing container is not shown to allow a view of the mixing paddle bars).

After mixing, the liquid STC was then transferred to molds, which were made from 5.1 cm

thick polyethylene sheets drilled with 0.8 cm diameter holes. Each mold was approximately 40.6









cm by 61.0 cm and contained 2,750 holes. An image of one mold is shown in Figure 7-2. The

molds were filled by pouring the liquid sol into the molds, which were sealed on the bottom and

top with sheets of solid polyethylene. The STC pellets were aged at 650C for 48 hours. The lids

were then loosened and the pellets dried in the molds at 1030C in dual thermal chambers. The

pellets were removed from the molds, transferred to Pyrex containers, and then heated to 1800C

in the dual thermal chambers. After aging and drying, pellets were approximately 3 mm in

diameter and 20 mm in length.

Since no heat transfer system was commercially available that was suitable for the STC

preparation, a double-chamber heat transfer system was designed. This heat chamber provided a

pneumatically sealed heat-generating source that was isolated from the STC materials. The

system allowed the necessary temperature variations within the inner-chamber, where the pellet

trays resided during aging and drying. The inner-chamber was equipped with a ventilation

system that allowed any volatile gases to discharge away from any source of heat or flame. This

gas could be passed through a heat exchanger for recovery of the ethanol. A photo of the heat

chamber is shown in Figure 7-3.

The STC pellets were made in a series of batches, each of which yielded 7-8 pounds of

pellets. A grab sample was taken from each batch and analyzed for surface area, total pore

volume, and pore size. The results of this analysis are shown in Table 7-1. The STC synthesized

for pilot studies used the same ratio of raw ingredients as the pellets used for bench-scale testing.

However, the dried STC pellet had a higher BET surface area and smaller pore size than those

synthesized in the lab (surface area was 723 m2/g versus 617 m2/g and pore size was 38 A versus

49 A). The differences between these measurements can be attributed to slight differences in the

synthesis procedure that were inherent with scale-up. Since the bench-scale studies indicated









that STC with higher surface area and smaller pore sizes achieved better performance, there was

no attempt to adjust the pilot-scale procedure to achieve STC with different characteristics.

UV Light Distribution in a Packed Bed of STC

In order to determine the spacing of UV lamps for the pilot reactor, a unique test system

was developed to measure the UV light penetration through a packed bed of STC. A series of

square boxes of the same height and varying length/width dimensions were manufactured from

Alzak reflective material. A support stand was made to support a UVC lamp in the center

surrounded by any one of the boxes. Each box had a 14 inch port for UV measurements. A

schematic of the test system (provided by the fabricator) is shown in Figure 7-4.

Each box was filled with the STC pellets and the UV lamp was illuminated. UV

measurements were taken through the port on the side of the box to measure the amount of light

penetrating through the packed bed. This was repeated for each box so that the UV light

penetration through various bed depths could be measured. The results are shown in Table 7-2.

The UV intensity decreased as the packed bed depth increased for two reasons: (1) the intensity

of UV light decreases as a function of the inverse square route of the distance from the lamp; and

(2) a deeper bed of STC would also serve to attenuate more light. No UV light was measured

through a packed bed depth of 1 inch. Therefore, the maximum distance between any portion of

the packed bed and a UV lamp should be less than 1 inch in order to avoid "dark spots" in the

reactor.

Pilot Studies

Figure 7-5 shows a process flow diagram for the pilot studies. A variable frequency drive

fan was used to push the gas stream through the pilot reactor. The space time was about 4.3 s,

similar to that used in bench-scale studies. A water/methanol mixture was vaporized and injected

into the gas stream so that the influent methanol concentration was about 50 ppmv and relative









humidity between 95 and 99%. The static pressure and velocity pressure were monitored before

the inlet to the pilot reactor. Thermocouples were used to measure inlet and outlet temperature.

The pilot reactor was fabricated by MicroEnergy Systems, Inc. (Oakland, MD). A general

arrangement drawing for the pilot reactor, which was provided by MicroEnergy Systems, Inc., is

shown in Figure 7-6. A photo of the reactor is shown in Figure 7-7. The dimensions of the

reactor were approximately 17 in by 17 in by 8 ft tall. The gas flowed upwards into the bottom of

the reactor and through a plenum, which supported the packed bed of STC pellets. The STC

pellets were packed between 46 UVC lamps (American Ultraviolet Corporation), which were

oriented in the reactor parallel to the air flow. The lamps were spaced 2.375 in. center to center

such that 79% of the pellets were within 0.6875 in. of a UV lamp and no pellet was more than 1

in. from a UV lamp. The UV lamps were housed in 1 in. quartz sleeves, each of which included

electrical connectors and wires, the latter of which passed through sealed stoppers. Top and

bottom support racks (shown in section A-A of Figure 7-6) were designed to support each UV

lamp assembly. A UV-Technik Electronic Ballast Power Center was installed in conjunction

with the 46 UV bulbs, thus giving the capability to vary UV radiation in response to a variable

voltage power input, which was controlled by the potentiometer setting. A UVP radiometer

(Upland, CA) was used to measure the UV intensity through an observation window on the side

of the reactor. The inside of the reactor was lined with Alzak UV reflective material. Initial

studies performed with an empty reactor showed no photolysis of methanol and that adsorption

of methanol to the reactor and its appurtenances was negligible.

Two tests were conducted with variable UV intensities controlled by varying the

potentiometer setting, which controlled the voltage input. For the first test, the potentiometer was

set to 22%. The UV intensity (as measured through an observation port located on the side of the









reactor) was about 6.5 mW/cm2. For the second test, the potentiometer was set to 100% and the

corresponding UV intensity was about 11 mW/cm2. In both cases the breakthrough profile was

similar to that seen in bench-scale studies. The effluent methanol concentration was non-

detectable for the first 4-6 hours, due to adsorption on the virgin STC surface, and then increased

until the steady state was reached (i.e., the rate of adsorption equaled the rate of oxidation). The

results of the pilot tests are shown in Table 7-3. The methanol oxidation efficiency was

dependent on the UV intensity at steady state, with the higher UV intensity resulting in about

66% removal while the lower UV intensity resulted in about 27.5% removal efficiency. The

effluent formaldehyde concentration was minimal (i.e., less than 1 ppmv) for both studies, thus

indicating that the majority of methanol was oxidized to inert byproducts (i.e., carbon dioxide

and water vapor). The effluent temperatures were significantly higher than the influent

temperatures for both tests due to the heat generated by the high intensity lamps. When operated

at their maximum intensity, the lamps require 60 W and output 16.2 W as UVC. Therefore, most

of the input energy was lost as heat, which resulted in the elevated effluent temperatures.

The effluent temperatures suggest that the reactor operated under elevated temperatures

compared to the temperature tested at the bench-scale, which was about 50C (122F). Elevated

temperatures can result in a decrease in UV intensity since UV lamps operate most efficiently at

1040F. Based on manufacturer provided data, the UV lamps operate at about 70% efficiency at

139.30F (low intensity setting) and 50% efficiency at 157.60F (high intensity setting). This

decrease in photon flux could yield a decrease in oxidation rate. In addition, a decrease in the

adsorption affinity of methanol for the STC surface due to the higher temperatures could result in

decreased oxidation rates. Therefore, improved results may be achieved by adding a cooling

system to the UV lamps to transfer heat away from the lamps and packed bed.









In addition to temperature effects, the efficiency of the pilot reactor may have been

dictated by the UV light distribution in the packed bed. The maximum distance between an STC

pellet and a UV lamp in the reactor was 1 in. The reactor was designed so that the UV intensity

approached zero as the distance away from the UV lamp increased to 1 in. However, the UV

intensities achieved in the portions of the bed approaching 1 in. from the UV lamp may not have

been sufficient to achieve higher oxidation rates. Thus, a minimum UV intensity across the entire

bed may be necessary to increase oxidation efficiency.

Table 7-1. BET surface area, pore volume, and pore size analysis for pilot STC.
BET surface area (m2/g) Total pore volume (cc/g) Actual pore size (A)
723 67 0.75 0.12 38 3

Table 7-2. UV intensity measurements through packed beds of STC of varying depths.
Box No. Distance (in.) UV intensity (iW/cm2)
5 0.25 42
4 0.5 16
3 0.75 3
2 1.0 0
1 2.0 0

Table 7-3. Results of pilot studies with variable potentiometer settings.

UV Steady state Formaldehyde
Inlet temp Outlet temp
Experiment intensity t methanol production
(mW/cm2) (OF) (OF) removal (ppmv)
LowIntensity 6.5 75.6 5.4 139.3 3.0 27.5 5.5% 0.28 0.26
(Setting = 22%)
High Intensity
High Intensity 11.5 72.6 4.0 157.6 6.0 66 7% 0.74 0.10
(Setting = 100%)
























Figure 7-1. Mixing assembly for blending the raw ingredients for pilot-scale STC synthesis.


Figure 7-2. Pilot-scale molds for STC synthesis.


Figure 7-3. Specialty heat chamber for pilot-scale STC synthesis.















































f .


UV Bulb connector


SFloor Support Plate




Figure 7-4. Alzak box test system for measuring UV light penetration through packed beds of
various depths.

































Figure 7-5. Process flow diagram for methanol degradation pilot studies.


t


17 in


15 i a a a a r
a s a a S
15in *eases


2" PVC half-sections
46: 1" OD Quartz tubes
46: 24" UVC bulbs (red)
SS-rod support matrix (gray)


SECTION A-A


Figure 7-6. General arrangement drawing of the pilot reactor for methanol degradation.


I I




= 4/".


---I


,r


I













-- UV-Technik Electronic
Ballast Power Center



Observation Window


Figure 7-7. Photo of pilot reactor.









CHAPTER 8
DEVELOPMENT OF A REGENERABLE SYSTEM EMPLOYING STC PELLETS FOR
MERCURY REMOVAL FROM END-BOX EXHAUST AT A CHLOR-ALKALI FACILITY

The STC technology may be advantageous for the removal of mercury from end-box

exhaust in chlor-alkali facilities due to its high mercury adsorption capacity and ability to be

regenerated with acid. The mercury-laden acid used for regeneration can be recycled into the

mercury-cell process, thus closing the loop on mercury emissions. This chapter summarizes the

design and performance of pilot- and full-scale reactors used to recover mercury from the end-

box exhaust at a chlor-alkali facility. In addition, an economic analysis, which compares the

costs of implementing this technology versus using activated carbon at the facility, is presented.

Pilot-Scale Packed Bed Reactor

A pilot reactor was fabricated and tested for mercury removal from end-box exhaust,

which consisted of air and trace quantities of hydrogen (0.02%), contained elemental mercury

vapor, and was saturated with water vapor. The total flow rate of the caustic exhaust system was

about 350 ACFM during the pilot studies. The exhaust passed through a heat exchanger, which

reduced the temperature to between 6 and 8 C, and a series of knock-out pots to remove

entrained water and mercury droplets. A slipstream of the exhaust was taken after the knock-out

pots and passed through a blower, into the pilot reactor, and back into the vent for release into

the atmosphere. A process flow diagram for this process is shown in Figure 8-1.

The pilot reactor contained a fixed bed of STC pellets packed around UV lamps. The STC

pellets had an average BET surface area of 351 m2/g, pore volume of 0.95 cm3/g, and pore

diameter of 109 A. The reactor contained two chambers (Chamber A and Chamber B), which

could be operated and sampled independently of each other. Two Teflon-coated UV lamps

(American Ultraviolet Corporation), were positioned vertically in each chamber and on two

external reactor walls (one wall of each chamber). The reactor walls were made up of Lucite









UVT material, which is transparent to UV light. The outer shell of the reactor was made of Alzak

reflective metal in order to reflect light into the packed bed. The exhaust flowed upwards through

the reactor while the STC pellets were continuously irradiated with the UV lamps. Each lamp

was controlled individually such that any combination of lamps could be illuminated at one time.

A schematic and photo of the pilot reactor are shown in Figures 8-2 and 8-3.

Pilot Study Results

The pilot-scale study treating mercury in a slipstream from caustic exhaust was conducted

from February to May of 2005 at a chlor-alkali facility in the US. For all pilot studies, the UV

lamps were operated continuously in order to maximize mercury removal efficiency to help meet

this facility's goal of zero emissions. The objective of the pilot experiments was two-fold: (1)

confirm the efficacy of the technology for mercury removal and recovery by regeneration with

HC1; and (2) determine the factors that may limit mercury removal efficiency (e.g., residence

time, mass transfer, and UV light distribution within the packed bed).

In order to confirm the efficacy of the technology and test the effectiveness of regeneration

using HC1, a series of two experiments were conducted. In the first experiment, the reactor was

packed with virgin STC pellets and continuously operated for 10 days using a flow rate of 10

ACFM and a space time of 0.53 s. As shown in Figure 8-4, the influent concentrations ranged

between 400 and 1600 ug/ft3 (or 1795 and 7180 ppbv). Although the influent mercury

concentration was highly variable, the effluent mercury concentration remained low (26 + 17

ug/ft3 for Chamber A and 31 + 16 ug/ft3 for Chamber B). The mercury removal rate achieved by

the reactor over this 10 day period was 96 3 %. Although the reactor ran continuously for the

10 day period, no samples were collected on days 3-5 since the chlor-alkali facility employees

were not available to collect samples. Table 8-1, which refers to this experiment as "Test No. 1",

shows a summary of results for the pilot data. The error is represented as the standard deviation.









After the 10 days of operation, the pellets were removed from the reactor and regenerated

in HC1. The mercury concentration on the pellets was determined by digesting the pellets before

and after regeneration. About 99% of the mercury was removed from the pellets using this

regeneration procedure. In order to test the effect of regeneration on the performance of the STC

pellets, a second experiment was conducted in which virgin pellets were packed in Chamber A

and the regenerated pellets were packed in Chamber B. The reactor was operated using the same

flow rate and space time (10 ACFM and 0.53 s) as the previous experiment. As shown in Figure

8-5, both the new and regenerated pellets performed similarly and achieved greater than 90%

removal for at least 21 days. After 35 days of operation, the effluent mercury concentration

approached that of the influent; thus, breakthrough was achieved. The mercury loading on the

pellets was calculated by performing a mass balance on the reactor. After the first 21 days of

operation (i.e., before breakthrough) the mercury loading on the virgin pellets was about 260

mg/g and the loading on the regenerated pellets was about 247 mg/g. Since these mass loadings

are similar, this confirms that the regeneration procedure did not impact the performance of the

STC pellets.

A series of tests were conducted in which the effects of space time, mass transfer, and UV

irradiation were studied. For each experiment (shown as Test Nos. 2 through 4 in Table 8-1),

virgin STC pellets were used in the pilot reactor. In order to hypothetically improve upon the

performance of Test No. 1, the flow rate was decreased for Test No. 2, resulting in a longer space

time and higher j-factor (i.e., improved mass transfer). The average effluent concentration for

Test No. 2 was 11.5 ug/ft3, which is significantly lower than the effluent concentrations from

Test No. 1 (P = 0.0248 based on two-tailed t-test with 95% confidence interval). To test whether

this improvement was a result of better mass transfer or the increase in space time, the bed height









was increased from 10 in. to 31 in. and approximately the same flow rate as Test No. 2 was used

for Test No. 3A. This increase in space time from 0.67 to 2.3 s resulted in lower mercury effluent

concentrations, which averaged 2.9 ug/ft3. This suggests that the mercury removal rate is limited

by the rate of photocatalytic oxidation, rather than mass transfer.

A factor that may limit the rate of photocatalytic oxidation in a packed bed is ineffective

UV light distribution within the bed. Therefore, in Test No. 3B, the outer UV lamps positioned

behind the transparent side wall of the reactor were illuminated to enhance the distribution and

increase the intensity of UV light in the bed. This increased irradiation did not result in a

decrease in effluent mercury concentration. Thus, the rate of photocatalytic oxidation did not

increase with the increase in UV radiation within the bed. Therefore, the space time through the

bed was further increased in attempts to achieve lower effluent mercury concentrations.

In Test No. 4, the flow rate was decreased to 3.6 ACFM in order to increase the space time

to 4.6 s. The effluent mercury concentration from this test averaged 3.6 ug/ft3, which was not

significantly different than Test No. 3A (P = 0.1646). Thus, the increase in space time from 2.3

to 4.6 s did not enhance mercury removal as was seen with the increase in space times between

0.53 and 2.3 s. At space times greater than 2.3 s, the removal rate through the reactor was limited

by factors) other than photocatalytic oxidation rate. For example, at low effluent mercury

concentrations (i.e., 3-6 ug/ft3), there may not be sufficient driving force for mass transfer and

subsequent adsorption. Additionally, photocatalytic reduction of sorbed mercuric oxide (HgO) to

elemental mercury may occur by the free electrons on the TiO2 surface, which are generated by

irradiation of this surface with UV light (Li and Wu, 2006). Although the photocatalytic

oxidation of elemental mercury is the primary reaction pathway since oxygen serves as a

stronger electron trap than HgO, a photocatalytic reduction effect was seen in the work of Li and









Wu (2006) while studying the use of STC for mercury removal from flue gas. Elemental

mercury has a lower adsorption affinity for the STC surface; thus, desorption or restrained

adsorption of elemental mercury from the gas stream may occur. Thus, photocatalytic reduction

may inhibit the attainment of lower effluent mercury concentrations.

Full-scale Reactor

Based on the promising results from the pilot studies, two full-scale reactors were designed

and installed. Each of these reactors was designed to treat up to 1200 ACFM and 1600 ug/ft3 and

have in-situ regeneration capabilities. Each unit was designed to handle the entire flow rate so

that the other could be offline for regeneration. The footprint of each reactor was about 3 feet by

6 feet. Before entering the reactor, the exhaust passed through a heat exchanger and demister,

which removed entrained liquid droplets greater than 4 |tm in diameter. The temperature entering

the online reactor was typically between 10 and 150C. Regeneration was performed in-situ by

soaking the pellets inside of the reactor with 37%wt HC1 for 30 min. The mercury-laden acid was

then drained from the reactor and the pellets were rinsed with water. Both the HC1 and water

were introduced into the reactor by a spray nozzle located inside the top of the reactor above the

packed bed. A process flow diagram for this full-scale system is shown in Figure 8-6.

The design of the pilot reactor was scaled-up so that the space time and Colburn j-factor in

the full-scale units were 0.63 s and 0.10, respectively, when treating the maximum design flow

rate of 1200 ACFM. The UV light spacing was similar to that used in the pilot scale reactor. The

reactors were filled with STC pellets with an average BET surface area of 453 m2/g, pore volume

of 1.04 cm3/g, and pore diameter of 92 A. Two full-scale reactors were installed in parallel so

that one could be in operation while the other was regenerated. This also provides a level of









redundancy since regeneration requires only a short period of time (i.e., 4 hours) compared to the

time that the other reactor will be online before regeneration is required (i.e., 30 days).

Data from the first three operation cycles for the reactors are shown in Table 8-2. The error

is represented as the standard deviation of measurements taken over the operation time. For all

studies, the effluent mercury concentration remained steady throughout the run (i.e., error is not

associated with breakthrough). The cycles are designated using a letter (either A or B) followed

by a number. The letter indicates which reactor was in operation and the number indicates the

cycle number. For example, Run No. A-i indicates the first operation cycle for Reactor A. For

the three operation cycles shown, the average flow rate decreased with each run for unknown

reasons. The flow rate in subsequent runs (results not shown) varied between 400 and 700

ACFM.

Run No. A-i averaged slightly more than 50% removal over a 29 day period. This poor

performance was due to the lack of water entering the bed. The influent water loading into the

pilot was about 3 lbs of water/1000 ft3 of air; however, the loading into the full-scale reactor was

about 0.45 lbs/1000 ft3. This is a result of changes made to the exhaust system (including the

addition of a new heat exchanger and demister) during the time between the pilot reactor testing

and the installation of the full-scale reactors. During Run No. A-2, the demister was initially

bypassed in order to increase the influent water loading and test this hypothesis. After 15 days of

operation, the demister was brought online and water was added into the reactor through the

nozzle in the top of the reactor above the packed bed. About 60 gallons of water were added to

the reactor once per day for the remainder of the cycle.

Figure 8-7 shows the influent and effluent mercury concentrations for this run (No. A-2).

During the first 15 days, while the demister was offline, the effluent mercury concentrations









were low. However, the influent mercury concentrations were highly variable, ranging between

about 400 and 1800 ug/ft3. Since the demister typically removed some of the elemental mercury

that was entrained as small droplets in the exhaust, it was brought back online on Day 15 in order

to decrease the influent mercury loading to the bed, thus increasing bed life and time between

regenerations. Therefore, water was added through a nozzle positioned in the top of the reactor

after the demister was brought back online. This proved equally as effective at reducing the

effluent mercury concentration. During the entire 43-day cycle, the actual flow rate through the

reactor was 560 140 ACFM. The average effluent concentration was 21.5 23.8 ug/ft3, which

corresponds to a mercury removal rate of 95 5%. The mercury loading on the pellets at the end

of this operational cycle was about 190 mg/g.

Some of the water, which was added through the nozzle to saturate the bed during Run No.

A-2, passed through the packed bed and was drained from the reactor. During Run No. B-l, the

demister was online and water was added through the top nozzle. However, the water loading

rate was optimized so that 15 gallons of water were added every 2 hours. This water application

rate did not result in excess water passing through the packed bed and yielded low effluent

mercury concentrations, which averaged 10.8 ug/ft3.

The addition of water to the full-scale reactor was necessary to provide sufficient hydroxyl

radicals on the TiO2 surface. This indicates that mercury was not directly oxidized by the

electron holes on the TiO2 surface, but was indirectly oxidized by hydroxyl radicals, which are

powerful oxidants. These hydroxyl radicals are created by the oxidation of adsorbed water vapor.

Due to the low influent water concentration and the rise in temperature through the reactor

caused by heat generated by the UV lamps, adsorption of water vapor to the STC surface was









low. Therefore, constant water addition was necessary to maintain a sufficient quantity of water

on the surface of the STC.

Economic Analysis

An economic analysis was performed to assess the feasibility of the STC technology

compared to treated (e.g., sulfur or iodine impregnated) activated carbon. The design for both

technologies was based on the maximum design flow rate and influent mercury concentration of

1200 ACFM and 1600 ug/ft3, respectively. This corresponds to an influent mercury loading of

2765 g/day. The cost per pound of mercury removed was based on capital and O&M (operation

and maintenance) costs over a 20 year period. This analysis assumes a mercury removal rate of

95% and continuous operation for both technologies and is solely based on economics. No

attempt was made to assign a value to the risk associated with the handling, transport, and offsite

disposal of mercury-laden waste for the case of activated carbon.

The costs associated with the STC technology are based on the actual full-scale installation

described above. The costs associated with the activated carbon installation include a pre-heater

and two beds of sulfur-impregnated activated carbon. The function of the pre-heater is to

increase the temperature of the exhaust to about 100 F in order to enhance mercury adsorption.

The beds should be installed in series with provisions for bypass or reversal of flow to provide

maximum utilization and reliability of the adsorption system (Anastas, 1976). The design of the

carbon beds was based on a face velocity of 1 ft/s and space time of 8 s (Anastas, 1976; Klett,

2002). Thus, each carbon bed would be approximately 5 ft in diameter and 8 feet in height.

Capital Costs

The capital cost associated with the STC technology was based on the actual fabrication

cost (in 2006 dollars) of the two full-scale reactors. This cost included support stands for the









reactors, associated electrical equipment (e.g., ballasts for UV lamps), and a PLC system, which

allows the control room to remotely monitor the status of the reactor.

The capital cost for the carbon installation was based on the modular approach (Guthrie,

1974). This approach enables estimation of components of a chemical processing unit from

known equipment costs using multipliers to adjust for design variations such as equipment

configuration, material of construction and design pressure. A Marshall and Swift Index of 1250

was used to scale the cost data to 2006 dollars. Using this approach, the capital cost for the pre-

heater and two carbon beds were estimated to be about $100,000.

O&M Costs

The operation and maintenance costs include those incurred annually for the following: (1)

utilities (i.e., electricity), (2) operating labor, (2) maintenance materials and labor, and (3) taxes

and insurance. The only utility needed for both technologies was electricity ($0.10/KWh). Labor

cost was based on the 2006 Construction Cost Index wage of $3 1/hour. Maintenance materials

and labor for operation and maintenance were based on reasonable assumptions for both

technologies. Taxes and insurance were estimated to be 2.5% of the capital cost (Anastas, 1976).

The maintenance material costs for the STC technology were based on the cost to replace:

(1) UV lamps and gaskets annually and (2) ballasts and STC pellets once every five years. Forty

labor hours per year were included for maintenance. Operation costs included energy consumed

by the UV lamps and blower. The number of regenerations per year, which varies based on

influent mercury loading, was based on an STC capacity of 200 mg/g. The cost of acid and water

was considered negligible since both are readily available at the facility. Four labor hours per

regeneration were included in the operation costs. Costs associated with recycling the mercury

back into the process was not incorporated into the analysis. Similarly, credit for mercury

recovery was not included.









Replacement frequency of the carbon was based on an operational capacity of 100 mg/g

(EPA, 1997b). Carbon replacement costs were estimated at $6.43 per pound and disposal was

estimated to be $500 per ton (Klett et al., 2002). Operation costs for the carbon included that for

the energy to power the blower. Maintenance materials and labor was estimated to be 2% of the

capital cost per year (Anastas, 1976).

Economic Feasibility

The cost per pound of mercury removed was estimated for both technologies as a function

of influent mercury loading, as shown in Figure 8-8. The cost per pound for the STC technology

was lower than that for the activated carbon technology at influent mercury loadings greater than

149 g/day, which is the loading at which the STC and activated carbon cost curves intersect. Due

to the uncertainty of the actual cost and performance associated with activated carbon

installations, the error associated with the activated carbon costs may be as high as + 30%

(Anastas, 1976). Therefore, the difference in cost between the two technologies can only be

considered significant for influent mercury loadings greater than about 470 g/day (i.e., the

influent mercury loading at which the cost for the STC technology is more than 30% different

that the cost for the activated carbon technology). At the design influent mercury loading of 2765

g/day, the cost per pound of mercury removed is about $20 for the STC technology and $84 for

activated carbon. Thus, the STC technology is economically feasible at the design influent

mercury loading. For the full-scale operation data presented in Table 8-2, the average influent

mercury loading of the three runs was 357 g/day. At this loading rate, the economically favorable

technology cannot be determined due to the uncertainty associated with the estimated cost of the

activated carbon installation.

The shape of the cost curves is similar for the STC and activated carbon technologies.

However, the dominate cost for the STC technology is that for the capital while, at influent










mercury loadings greater than about 300 g/day, O&M costs comprise a majority of the activated

carbon costs. The cost of activated carbon replacement (material replacement and disposal fees

only) is constant at $70 per pound of mercury removed. At low influent mercury concentrations,

the activated carbon costs increase rapidly. This is because the fixed costs are not diluted with

high mercury removal rates. Thus, the total cost per pound of mercury removed becomes highly

dependent on the capital cost at lower influent loading rates.

Table 8-1. Summary of pilot experiments.
Test No. Flow (ACFM) Colbum j Space time (s) Influent (ug/ft3) Effluent (ug/ft3)
1 10 0.106 0.53 762 286 28 16.3
2 8 0.122 0.67 652 328 11.5 7
3A 7.2 0.129 2.3 817 203 2.9 0.3
3B 7.2 0.129 2.3 817 203 3.5 + 1.2
4 3.6 0.183 4.6 550 208 5.8 4.5


Table 8-2. Full-scale performance data for three operation cycles.
Run No. Operation Time Flow Rate Influent (ug/ft3) Effluent (ug/ft3)
(days) (ACFM)
A-i 29 682 22 445 75 204 +132
A-2 43 560 140 595 451 21.5 23.8
B-1 28 400 64 234 116 10.8 + 5.6


Velocity
Pressure Static
Static


End-box
Exhaust
Temp.

Figure 8-1. Process flow diagram for mercury recovery pilot studies.


To Stack













Dual outlet ducts
& globe valves


3 UVT Lucite windows

Alzak reflective plates
er(movable to view elletIs)


Dual or single chambers
for housing STC pellets
Inlet ports for pellets

4 External UV Bulbs


A_ [O 'O* A
1 I I


Center UVT Lucite, 0
Center UVT Lucite 1 ,,, I..,- 4 Internal UV Bulbs
"sandwiched" between llanges 4 In l UV
to allow removal to create larger
chamber, or remain in place for
dual-chamber or single-smaller 2 Inlet ducts
chamber during test operations B

PLAN VIEW


Dual outlet
ducts & globe
valves






































SECTION B-B


Figure 8-2. Schematic of pilot reactor for mercury recovery.


Approx 54 in





































Figure 8-3. Photo of pilot reactor with UV lights illuminated.











1800

1600 -
Influent
1400- E Effluent A
A Effluent B
S1200 -
io +



S800

S600

C 400

200

0 %4aAiAArAi '*AiAA4AAi a

0 2 4 6 8 10

Time (Days)


Figure 8-4. Influent and effluent mercury concentrations for the pilot reactor packed with virgin
STC pellets (Test No. 1).












1600


1400


1200


1000


800


600


400


200



0 5 10 15 20 25 30 35 40 45

Time (days)



Figure 8-5. Influent and effluent mercury concentrations of the pilot reactor packed with virgin
pellets (Chamber A) and regenerated pellets (Chamber B).


End-box
Exhaust


To Stack


Temp.


Temp.


HCI or Water
For Regeneration


Hg-Laden HCI
Or Water Recycled to Hg-Cell


Figure 8-6. Process flow diagram for full-scale installation of mercury recovery units.













2000

1800

1600

S1400

1200

| 1000

U 800

3 600

400

200

0


0 5 10 15


20 25
Time (days)


Figure 8-7. Influent and effluent concentrations for the full-scale reactor during its second
adsorption cycle.


Influent

Effluent




*++ *

**

*m

t
"- ^
J *- w
ii-- -ll ---- l ----- -. ... .--- *. i--- l-


30 35


40 45










$1,000


$900 STC Technology

S $800 -- STC: Capital

$700
-- Activated Carbon
B $600
s E Activated Carbon
$500 Replacement

3 $400

o $300

$200

$100 -------El- -~ - -----
so P_ -- .- -


0 500 1000 1500 2000 2500 3000

Influent Mercury Loading (g/day)


Figure 8-8. Comparison of cost per pound of mercury removed for activated carbon and STC as
a function of influent mercury loading (for systems designed to treat up to 2765
g/day).









CHAPTER 9
CONCLUSIONS

STC pellets were synthesized using a sol-gel method with varying concentrations of HF

and TiO2. Mesoporous STC with varying pore volumes and surface areas were created. The

performance of these STC for methanol removal was found to be a function of the surface area

of the STC and space time of the gas in the reactor. The reaction kinetics were not limited by

external or internal resistances to mass transfer. The 120 A 12% and 260 A 12% STC exhibited a

lag time before achieving methanol mineralization, while the 50 A 12% STC did not. All STC

exhibited pseudo-first order reaction kinetics with a similar rate constant of 0.40 s1. The 50 A

STC were synthesized with varying TiO2 loadings and the optimum loading was found to be 4%.

Methanol oxidation was enhanced with the higher photon flux associated with using a UVC lamp

rather than a UVA lamp. The bench-scale system reached steady state removal and was able to

remove about 90% of the influent methanol with little byproduct formation (less than 1 ppmv

formaldehyde). When H2S was introduced into the system, the H2S competed with methanol

resulting in a decrease in methanol removal efficiency to about 80%. Although the H2S did not

adsorb well to the STC surface, the H2S was oxidized by the STC when illuminated with UV

light to SO2 and SO42-. A pilot-scale reactor was designed and fabricated to treat 40 ACFM of

humid air laden with 50 ppmv of methanol. Pilot-scale studies showed about 66% methanol

removal efficiency at steady state when the space time of the gas through the packed bed was 4.3

s. The methanol removal efficiency achieved in the pilot studies was less than that in the bench-

scale studies because the UV light distribution within the packed bed was limited and elevated

reaction temperatures likely inhibited the oxidation rate by decreasing the adsorption of methanol

onto the STC and the efficiency of the UV lamps.









STC and TiO2-coated AC were compared to TiO2-coated glass spheres for the removal of

methanol. In a low humidity environment (RH = 0.22%), the adsorption capacity of the STC (11

mg/g) was greater than that of TiO2-coated AC (6 mg/g) and TiO2-coated glass spheres. The

silanol groups (Si-OH) on the STC surface promoted methanol adsorption in a low humidity gas

stream, where competition with water vapor was low. When STC was irradiated with UV light,

no methanol or oxidation byproducts were detected in the effluent. Formaldehyde, an oxidation

byproduct of methanol, was detected in the effluent using TiO2-coated AC and TiO2-coated glass

spheres. The TiO2-coated AC showed that the limiting reaction in the mineralization of methanol

was the oxidation of formaldehyde to formic acid. In a high humidity gas stream (RH = 95%),

the adsorption capacity of the STC (1.2 mg/g) and TiO2-coated AC (1.9 mg/g) was reduced due

to the competition with water vapor for adsorption sites. The overall efficiency of the TiO2-

coated AC increased likely due to the presence of water vapor, which is required for the

oxidation of formaldehyde to formic acid. The methanol adsorption capacity of the AC was

greater than the STC in the high humidity gas stream since the surface chemistry of the AC was

more heterogeneous than that of the STC, which was dominated by silanol groups that strongly

adsorb water via hydrogen bonding. In the presence of UV light, the STC and TiO2-coated AC

reached a steady state methanol removal efficiency of 95%. Water vapor adsorption studies

found that the surface coverage of water on the STC surface was greater than that on the TiO2

surface and that the adsorption of water vapor on the STC was solely affected by the heat

generated from the UV lamp. In a high humidity gas stream with a large annulus reactor, the

STC and TiO2-coated AC performed similarly, achieving 50% methanol removal compared to

the TiO2-coated glass spheres, which achieved 40% methanol removal. Thus, using an adsorbent

material as a catalyst support in larger scale systems was beneficial. However, the use of silica









gel, which was transparent, versus AC, which was opaque, did not result in a difference in the

photocatalytic oxidation rate in the 25 mm annulus reactor. This was likely due to the decrease in

degradation efficiency associated with water competition on the STC surface, which was not as

prevalent for the TiO2-coated AC.

The series of pilot experiments performed a chlor-alkali facility confirmed the efficacy of

the scale up of the STC technology for the removal of elemental mercury from end-box exhaust.

The experiments showed that the reactor was able to consistently achieve 96% mercury removal

in a variety of system conditions, including highly variable influent conditions (400 to 1600

ug/ft3). In the range of flow rates tested, mercury removal rate was limited by space time and not

mass transfer. An increase in space time between 0.53 s and 2.3 s yielded slight increases in

mercury removal. However, longer space times did not result in a change in mercury removal.

Regeneration with concentrated HC1 was performed and proved to be effective. The regenerated

pellets were successfully brought back online and performed similarly to virgin pellets. Two full-

scale reactors were installed and performed similarly to the pilot-scale reactors. An economic

analysis showed that the cost per pound of mercury removed was less for the STC technology

than for sulfur impregnated activated carbon at influent mercury loading rates greater than 470

g/day. At lower influent loading rates, neither technology could be deemed economically

favorable due to the uncertainty associated with the activated carbon cost estimate. Thus, the

STC proved to be both technologically and economically feasible for this application.

Recommendations for future work are listed below:

S The competitive effects of adsorption and oxidation of VOCs in a multi-component system
employing STC should be modeled since these systems will be more indicative of real-
world applications;









* The transparency of STC containing various TiO2 loadings should be quantified to better
understand the relationship between the quantity of TiO2 that is irradiated and oxidation
efficiency;

* The effects of adsorption on H2S oxidation efficiency should be studied in order engineer a
composite that promotes the oxidation of H2S to S042- rather than SO2, which would need
to be scrubbed from the gas before being released to the atmosphere.

As a result of this work, the following contributions to science were made:

* First to demonstrate that the oxidation efficiency of photocatalyst pellets for the
degradation of a gaseous organic compound (methanol) was affected by material properties
of the pellets (i.e., pore size and TiO2 loading), photon flux, and space time of the gas
through the reactor;

* First to model the rate of photocatalytic oxidation of gaseous methanol by photocatalyst
pellets of various pore sizes using pseudo-first order equations and demonstrate that a lag
time existed before mineralization proceeded, which was dependent on the internal surface
area of the pellets;

* First to demonstrate that the limiting step in the mineralization of methanol differed
depending on both the type of catalyst support and the humidity of the gas entering the
reactor;

* Supported evidence that both SO2 and S042- were formed as byproducts as a result of the
photocatalytic oxidation of H2S, which contrasts some studies that reported the formation
of only S042-;

* First to demonstrate that the oxidation efficiency of methanol decreased as a result of the
competitive oxidation of H2S in a humid gas stream and that the oxidation of H2S was
unaffected by the presence of methanol;

* First to implement a full-scale technology employing photocatalysis for the removal of
mercury from gas-phase emissions in the chlor-alkali industry.









LIST OF REFERENCES


Al-Ekabi, H., Serpone, N., 1988. Kinetic studies in heterogeneous photocatalysis. 1.
Photocatalytic degradation of chlorinated phenols in aerated aqueous solutions over TiO2
supported glass matrix. J. Phys. Chem. 92, 5726-5731.

Alberici, R.M., Jardim, W.F., 1997. Photocatalytic destruction of VOCs in the as-phase using
titanium dioxide. Appl. Catal. B Environ. 14, 55-68.

Anastas, Y., 1976. Molecular sieve mercury control processes in chlor-alkali plants, EPA-600/2-
76-014. U.S. Government Printing Office, Washington D.C.

Anderson, C., Bard, A., 1997. Improved photocatalytic activity and characterization of mixed
TiO2/SiO2 and TiO2/A1203 Materials. J. Phys. Chem. 101, 2611-2616.

Arana, J., Dona-Rodriguez, J., Cabo, C., Gonzalez-Diaz, O., Herrara-Melian, J., Perez-Pena, J.,
2004. FTIR study of gas-phase alcohols photocatalytic degradation with TiO2 and AC-
TiO2. Appl. Cataly. B -Environ. 53, 221-232.

Byrne, H., Kostedt, W., Stokke, J., Mazyck, D., 2008. Characterization of HF-catalyzed silica
gels doped with Degussa P25 Titanium Dioxide. Submitted to J. Non-Cryst. Solids.

Canela, M., Alberici, R., Jardim, W., 1998. Gas-phase destruction of H2S using TiO2/UV-Vis. J.
Photochem. Photobiol. A 112, 73-80.

Chang, H., Wu, N., Zhu, F., 2000. A kinetic model for photocatalytic degradation of organic
contaminants in a thin-film TiO2-catalyst. Wat. Res. 34, 407-416.

Chen, J., Ollis, D., Rulkens, W., Bruning, H., 1999. Photocatalyzed oxidation of alcohols and
organochlorides in the presence of native TiO2 and metallized TiO2 suspensions. Part (II):
Photocatalytic mechanisms. Wat. Res. 33, 669-676.

Chu, W., Wong, C., 2004. Study of herbicide alachlor removal in a photocatalytic process
through the examination of the reaction mechanism. Ind. Eng. Chem. Res. 43, 5027.

D'Itri, F., Andren, A., Doherty, R., Wood, J., 1978. An Assessment of Mercury in the
Environment. National Academy of the Sciences, Washington, D.C.

Devilliers, D., 2006. Semiconductor photocatalysis: Still an active research area despite barriers
to commercialization. Energia, 17, 1-3.

Dijkstra, M., Panneman, H., Winkelman, J., Kelly, J., Beenackers, A., 2002. Modeling the
photocatalytic degradation of formic acid in a reactor with immobilized catalyst. Chem.
Eng. Sci. 57, 4895-4907.

Doucet, N., Bocquillon, F., Zahraa, O., Bouchy, M., 2006. Kinetics of photocatalytic VOCs
abatement in a standardized reactor. Chemosphere 65, 1188.









Ferguson, M., Hoffmann, M., Hering J., 2005. TiO2-photocatalyzed As(III) oxidation in aqueous
suspensions: reaction kinetics and effects of adsorption. Environ. Sci. Technol. 39, 1880-
1886.

Gao, X., Wachs, I., 1999. Titania-silica as catalysts: molecular structural characteristics and
physico-chemical properties. Catal. Today 51, 233-254.

Garner, J., 2001. Air emission control regulations pose new challenges for mills. Pulp and Paper
75, 44-46.

Garton, M.J., 2005. Photocatalytic oxidation of selected organic contaminants and inactivation of
microorganisms in a continuous flow reactor packed with titania-doped silica. Masters
Thesis, University of Florida.

Guthrie, K., 1974. Process plant estimating, evaluation, and control. Craftsman Book Co. of
America, Solana Beach, CA. pp 1-604.

Hoffmann, M.R., Martin, S.T., Choi, W., Bahnemann, D.F., 1995. Environmental applications of
semiconductor photocatalysis. Chem. Rev. 95, 69-96.

Holmes, F., 2003. The performance of a reactor using photocatalysis to degrade a mixture of
organic contaminants in aqueous solution. Masters Thesis, University of Florida.

Huang, M., Tso, E., Datye, E., Prairie, M., Stange, B., 1996. Removal of silver in photographic
processing waste by TiO2-based photocatalysis. Environ. Sci. Technol. 30, 3084-3088.

Hurum, D., Agrios, A., Gray, K., Rajh, T., Thurnauer, M., 2003. Explaining the enhanced
photocatalytic activity of Degussa P25 mixed-phase TiO2 using EPR. J. Phys. Chem. B
107, 4545-4549.

Kataoka, S., Lee, E., Tejedor-Tejedor, I., Anderson, M., 2005. Photocatalytic degradation of
hydrogen sulfide and in situ FT-IR analysis of reaction products on surface TiO2. Appl.
Catal. B -Environ. 61, 159-163.

Kato, S., Hirano, Y., Iwata, M., Sano, T., Takeuchi, K., Matsuzawa, S., 2005. Photocatalytic
degradation of gaseous sulfur-deposited titanium dioxide. Appl. Catal. B Environ. 57,
109-115.

Khan, A., 2006. Modification of activated carbon to improve aqueous manganese removal. PhD
Dissertation, University of Florida.

Kim, B.S., Hong, C.S., 2002. Kinetic study for photocatalytic degradation of volatile organic
compounds in air using thin film TiO2 photocatalyst. Appl. Catal. B Environ. 35, 305-
315.

Klett, M., Maxwell, R., Rutkowski, M., 2002. The cost of mercury in an IGCC Plant (final
report). Prepared for The United States Department of Energy National Energy
Technology Laboratory.









Knudsen, J., Hottel, H., Sarofim, A., Wankat, P., Knaebel, K., 1999. Mass transfer. In: Green,
D., Maloney, J., Perry, R. (Eds.), Perry's Chemical Engineer's Handbook. McGraw Hill,
New York. pp 72-74.

Lee, T., Biswas, P., Hedrick, E., 2001. Comparison of Hg capture efficiencies of three in situ
generated sorbents. Env. Energy Eng. 47, 954-961.

Lee, T., Biswas, P., Hedrick, E., 2004. Overall kinetics of heterogeneous elemental mercury
reactions on TiO2 sorbent particles with UV irradiation. Ind. Eng. Chem. Res. 43, 1411-
1417.

Levenspiel, O., 1999. Pore diffusion resistance. In: Chemical Reaction Engineering, third ed.
John Wiley & Sons, New York, pp. 470-477.

Li, Y., Wu, C.Y., 2007. Kinetic study for photocatalytic oxidation of elemental mercury on a
SiO2-TiO2 nanocomposite. Environ. Eng. Sci. 24, 3-12.

Li, Y., Wu, C.Y., 2006. Role of moisture in adsorption, photocatalytic oxidation, and reemission
of elemental mercury on a SiO2-TiO2 nanocomposite. Environ. Sci. Technol. 40, 6444-
6448.

Lichtin, N., Avudaithai, M., Berman, E., Dong, J., 1994. Photocatalytic oxidative degradation of
vapors of some organic compounds over TiO2. Res. Chem. Intermed. 20, 755-781.

Londeree, D., 2002. Silica-titania composites for water treatment. Masters Thesis, University of
Florida.

Lu, M., Chen, J., Chang, K., 1999. Effect of adsorbents coated with titanium dioxide on the
photocatalytic degradation of propoxur. Chemosphere 38, 617-627.

Ludwig, C., Mazyck, D., Chadik, P., Stokke, J., 2008. The performance of silica-titania carbon
composites for photocatalytic degradation of gray water. Submitted to J. Env. Eng.

Minero C., Catozzo, F., Pelizzetti, E., 1992. Role of adsorption in photocatalyzed reactions of
organic molecules in aqueous TiO2 suspensions. Langmuir 8, 481-486.

National Council for Air and Stream Improvement, Inc. (NCASI), 1998. Method CI/SG/PULP
94.02 chilled impinger/silica gel tube test method at pulp mill sources for methanol,
acetone, acetaldehyde, methyl ethyl ketone and formaldehyde, methods manual
(03.B.003). NCASI, Research Triangle Park, NC.

Nguyen, V., Amal, R., Beydoun, D., 2003. Effect of format and methanol on photoreduction/
removal of toxic cadmium ions using TiO2 semiconductor as photocatalyst. Chem. Eng.
Sci. 58, 4429-4439.

Nguyen, V., Beydoun, D., Amal, R., 2004. Photocatalytic reduction of selenite and selenate
using TiO2 photocatalyst. J. Photochem. Photobiol. A 171, 117-124.









Morrow, B., Gay, I., 2000. Infrared and MNR characterization of the silica surface. In: Papirer,
E. (ed.), Adsorption on Silica Surfaces. Marcel Dekker, Inc., New York, pp. 9-61.

Murov, S., Carmichael, I., Hug, M., 1993. Potassium ferrioxalate actinometry. In: Handbook of
Photochemistry. Mercel Dekker, New York, pp. 299-305.

Noguchi, T., Fujishima, A., Sawunyama, P., Hashimoto, K., 1998. Photocatalytic degradation of
gaseous formaldehyde using TiO2 film. Environ. Sci. Technol. 32, 3831-3833.

Nawrocki, J., 1997. The silanol group and its role in liquid chromatography. J. Chromatogr. A
779, 29-71.

Obee, T., Brown, R., 1995. TiO2 photocatalysis for indoor air applications: Effect of humidity
and trace contaminant levels on oxidation rates of formaldehyde, toluene, and 1,3-
butadiene. Environ. Sci. Technol. 29, 1223-1232.

Ollis, D., Pelizzetti, E., Serpone, N., 1991. Destruction of water contaminants. Environ. Sci.
Technol. 25, 1523-29.

Parida, S., Dash, S., Patel, S., Mishra, B., 2006. Adsorption of organic molecules on the silica
surface. Adv. Colloid Interfac. 121, 77-110.

Pavlish, J., Sondreal, E., Mann, M., Olson, E., Galbreath, K., Laudal, D., Benson, S., 2003.
Status Review of Mercury Control Options for Coal-fired Power Plants. Fuel Process.
Technol. 82, 89-165.

Peral, J., Domenech, X., Ollis, D., 1997. Heterogeneous photocatalysis for purification,
decontamination and deodorization of air. J. Chem. Technol. Biotechnol. 70, 117-140.

Pitoniak, E., Wu, C. Y., Londeree, D., Mazyck, D., Bonzongo, J.D., Powers, K., Sigmund, W.,
2003. Nanostructured silica-gel doped with TiO2 for Hg vapor control. J. Nanopart. Res. 5,
282-292.

Pitoniak, E., 2004. Evaluation of nanostructured silica-titania composites in an adsorption/
photocatalytic oxidation system for elemental mercury vapor control. Master of
Engineering Thesis, University of Florida.

Pitoniak, E., Wu, C. Y., Powers, K. W., Sigmund, W., 2005. Adsorption enhancement
mechanisms of silica-titania nanocomposites for elemental mercury vapor removal.
Environ. Sci. Technol. 39, 1269-1274.

Portela, R. Sanchez, B., Coronado, J., Candal, R., Suarez, S., 2007. Selection of TiO2-support:
UV transparent alternatives and long-term use limitations for H2S removal. Catal. Today
129, 223-230.

Puri, B., 1970. Surface complexes on carbons. In: Walker, P. (Ed.), Chemistry and Physics of
Carbon. Marcel Dekker, New York, pp. 194-247.









Rodriguez, S., Almquist, C., Lee, T., Furuuchi, M., Hedrick, E., Biswas, P., 2004. A mechanistic
model for mercury capture with in situ-generated titania particles: Role of water vapor. J.
Air & Waste Manage. Assoc. 54, 149-156.

Satterfield, C., 1970. Mass transfer in heterogeneous catalysis. MIT Press, Cambridge, MA.

Schettler, T., 2001. Toxic threats to neurological development of children. Environ. Health
Persp. Supplements 109, 813-817.

Serpone, N., 1995. Brief introductory remarks on heterogeneous photocatalysis. Sol. Energ.
Mat. Sol. C. 38, 369-379.

Shul, Y., Kim, H., Haam, S., Han, H., 2003. Photocatalytic characteristics of TiO2 supported on
Si02. Res. Chem. Intermed. 29, 849-859.

Sing, K., Everett, D., Haul, R., Moscou, L., Pierotti, R., Rouquerol, J., Siemieniewska, T., 1984.
Reporting physisorption data for gas/solid systems. Pure Appl. Chem. 57, 603-619.

Someshwar, A., 1994. NCASI technical bulletin 678: Volatile organic emissions from pulp and
paper mill sources part IV kraft brownstock washing, screening, and rejects refining
sources. National Council for Air and Stream Improvement, Research Triangle Park, NC.

Sopyan, I., 2007. Kinetic analysis on photocatalytic degradation of gaseous acetaldehyde,
ammonia and hydrogen sulfide on nanosized porous TiO2 films. Sci. Technol. Adv. Mat. 8,
33-39.

Southworth, G.R., Lindberg, S.E., Zhang, H., Anscombe, F.R., 2004. Fugitive mercury
emissions from a chlor-alkali factory: sources and fluxes to the atmosphere. Atmos.
Environ. 38, 597-611.

Thommes, M., 2004. Physical adsorption characterization of ordered and amorphous mesoporous
materials. In: Lu, G., Zha, X. (Eds), Nanoporous Materials: Science and Engineering.
Imperial College Press, London, pp. 317 364.

Torimoto T, Ito S, Kuwabata S, Yoneyama H., 1996. Effects of adsorbents used as supports for
titanium dioxide loading on photocatalytic degradation of propyzamide. Environ. Sci.
Technol. 30, 1275-1281.

Travert, A., Manoilova, O., Tsyganenko, A., Mauge, F., Lavalley, J., 2002. Effect of hydrogen
sulfide and methanethiol adsorption on acidic properties of metal oxides: An infrared
study. J. Phys. Chem. B 106, 1350-1362.

Tsumura, T., Kojitani, N., Umemura, H., Toyoda, M., Inagaki, M., 2002. Composites between
photoactive anatase-type TiO2 and adsorptive carbon. Appl. Surface Sci. 196, 429-436.

Tsuru, T., Kan-no, T., Yoshioka, T., Asaeda, M., 2003. A photocatalytic membrane reactor for
gas-phase reactions using porous titanium oxide membranes. Catal. Today 82, 41-48.









Turchi, C., Ollis, D., 1989. Mixed reactant photocatalysis: Intermediates and mutual rate
inhibition. J. Catal. 119, 483-496.

Turchi, C., Ollis, D., 1990. Photocatalytic degradation of organic water contaminants:
Mechanisms involving hydroxyl radical attack. J. Catal. 122, 178-192.

Uchiyama, H., Suzuki, K., Oaki, Y., Imai, H., 2005. A novel adsorbent photocatalyst consisting
of titania and mesoporous silica nanoparticles. Mat. Sci. Eng. B Solid 123, 248-251.

U.S. Environmental Protection Agency, 1997a. Mercury Study Report to Congress Volume IV:
An Assessment to Exposure to Mercury in the United States, EPA-452/R-97-006. U.S.
Government Printing Office, Washington D.C.

U.S. Environmental Protection Agency, 1997b. Mercury Study Report to Congress Volume VIII:
Evaluation of Mercury Control Technologies and Cost, EPA-452/R-97-010. U.S.
Government Printing Office, Washington D.C.

Varma, V., 2003. Experience with the collection, transport, and burning of kraft mill high
volume low concentration gases, Technical bulletin 03-03. National Council for Air and
Stream Improvement, Gainesville, FL.

Vohra, M., Tanaka, K., 2003. Photocatalytic degradation of aqueous pollutants using silica-
modified TiO2. Water Res. 37, 3992-3996.

Wu, C.Y., Lee, T.G., Tyree, G., Arar, E., Biswas, P., 1998. Capture of mercury in combustion
systems by in situ generated titania particles with UV irradiation. Environ. Eng. Sci. 15,
137-148.

Xu, Y., Zheng, W., Liu, W., 1997. Enhanced photocatalytic activity of supported TiO2:
dispersing effect of SiO2. J. Photochem. Photobiol. A 122, 57-60.

Yang, C., Chen, C., 2005. Synthesis and characterization of silica-capped titania nanorods: An
enhanced photocatalyst. Appl. Catal. A General 294, 40-48.

Yamakata, A., Ishibashi, T., Onishi, H., 2003. Effects of water addition on the methanol
oxidation on Pt/TiO2 photocatalyst studied by time-resolved infrared absorption
spectroscopy. J. Phys. Chem. B 107, 9820-9823.

Yong, T., Schwartz, S., Wu, C.Y., Mazyck, D.W., 2005. Development of TiO2/AC composite by
dry impregnation for the treatment of methanol from humid air streams. Ind. Eng. Chem.
Res. 44, 7366-7372.









BIOGRAPHICAL SKETCH

Jennifer Stokke was born to Susan and Wayne Stokke on August 26, 1981. She grew up in

Tarpon Springs, FL and graduated from East Lake High School in 1999. She then moved to

Gainesville in 1999 to attend the University of Florida, where she studied environmental

engineering and graduated with highest honors in December of 2003. During her last year as an

undergraduate, Jennifer began a research project focusing on the purification of gaseous

emissions. She continued this work during her graduate studies at the University of Florida

under the guidance of Dr. David Mazyck.





PAGE 1

1 PHOTOCATALYTIC OXIDATION OF HAZARD OUS AIR POLLUTANT S USING SILICATITANIA COMPOSITES IN A PACKED-BED REACTOR By JENNIFER MORGAN STOKKE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Jennifer M. Stokke

PAGE 3

3 To my parents, Wayne and Susan

PAGE 4

4 ACKNOWLEDGMENTS I would first like to thank my parents and fa mily for the love, guidance, and support that they have provided through the ye ars. I would also like to tha nk my advisor, Dr. David Mazyck, for the guidance and opportunities that he has provided during my graduate education. I will certainly take away many lifelong lessons from th is experience. I am grateful for the guidance and suggestions provided by my advisory committ ee, Dr. Paul Chadik, Dr. Chang-Yu Wu, and Dr. Hassan El-Shall, as well as t hose of Dr. Angela Lindner. I would like to thank Rick Sheahan and Joe Sines of MicroEnergy Systems, Inc (Oakland, MD) for services related to pilot reactor fabrication and testing, Jim Stainfield from NCASI (Gainesville, FL) for providing analytical suppor t, and Christina Akly, Heather Byrne, Teri Lierman, Paloma Rohrbaugh, Brendon Blum, Miguel Morales, Vanessa Pineda, Gustavo Avila, and Aly Byrne for their assistan ce in conducting laboratory studies. I am appreciative of the Department of Energy and American Forest Product Association for sponsoring a portion of this research via grant number DE-FC36-03ID14437. Finally, I would like to thank past and pres ent members of my research group for their support: Ameena Khan, Heather Byrne, Je nnifer McElroy, Matthew Tennant, Thomas Chestnutt, Morgana Bach, Jack Drwiega, Mi guel Morales, Vivek Shyamasundar, William Kostedt, IV, Christina Ludwig, and Alec Gruss.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................11 CHAP TER 1 INTRODUCTION..................................................................................................................13 Degradation of HAPs Emitted from Pulp and Paper Mills..................................................... 14 Recovery of Mercury Vapor Emitted in End-Box Exhaust at C hlor-alkali Facilities............ 16 2 LITERATURE REVIEW.......................................................................................................20 Photocatalysis.........................................................................................................................20 Use of Adsorbents as Catalyst Supports.................................................................................23 Photocatalytic Degradation of Methanol................................................................................25 Photocatalytic Oxidation of H2S.............................................................................................30 Enhanced Mercury Recovery using Photocatalysis................................................................32 3 EXPERIMENTAL.................................................................................................................. 36 Synthesis of Photocatalytic Materials..................................................................................... 36 Silica-Titania Compos ite (STC) Pellets .......................................................................... 36 TiO2-Coated Activated Carbon (AC).............................................................................. 37 TiO2-Coated Glass Spheres.............................................................................................38 Characterization of Phot ocatalytic Materials ..........................................................................38 Bench-scale Reactor for Methanol and H2S Removal Studies...............................................39 Pilot Reactor for Methanol Removal...................................................................................... 40 Analysis of Methanol and Oxidation Byproducts.................................................................. 41 Analysis of H2S and Oxidation Byproducts...........................................................................41 Mercury Analysis............................................................................................................... .....42 Colburn j-factor......................................................................................................................43 4 CHARACTERIZATION OF STC.........................................................................................45 5 OPTIMIZATION OF METHANOL DEGRA DATION USING STC PELLETS IN A BENCH-SCALE REACTOR .................................................................................................52 Adsorption..............................................................................................................................53 Simultaneous Adsorption and Oxidation of Methanol...........................................................53

PAGE 6

6 Formation of Photocatalytic Byproducts................................................................................ 54 Effect of Space Time on Methanol Degradation.................................................................... 56 Mass Transfer.................................................................................................................. 58 Kinetics............................................................................................................................60 Effect of TiO2 Loading on Methanol Degradation.................................................................61 Effect of UV Wavelength on Methanol Degradation .............................................................64 Effect of H2S on Methanol Degradation.................................................................................65 6 EFFECT OF CATALYST SUPPORT ON THE PHOTOCATALYTIC DEGRADATION OF M ETHANOL IN A PACKED-BED REACTOR...............................78 Methanol Adsorption and Oxidation in a Low Hum idity Gas Stream................................... 79 Methanol Adsorption and Oxidation in a High Hum idity Gas Stream.................................. 82 Water Vapor Adsorption......................................................................................................... 83 Reactor Scale-up using TiO2-doped Materials....................................................................... 84 7 PILOT STUDIES FOR METHANOL DEGRADATION..................................................... 90 STC Synthesis for Pilot-Scale Studies ....................................................................................90 UV Light Distribution in a Packed Bed of STC ..................................................................... 92 Pilot Studies.................................................................................................................. ..........92 8 DEVELOPMENT OF A REGENERABLE SYSTEM EMPLOYING STC PELLETS FOR MER CURY REMOVAL FROM END-BO X EXHAUST AT A CHLOR-ALKALI FACILITY............................................................................................................................100 Pilot-Scale Packed Bed Reactor........................................................................................... 100 Pilot Study Results............................................................................................................ ....101 Full-scale Reactor............................................................................................................. ....104 Economic Analysis.............................................................................................................. .107 Capital Costs..................................................................................................................107 O&M Costs....................................................................................................................108 Economic Feasibility..................................................................................................... 109 9 CONCLUSIONS.................................................................................................................. 117 LIST OF REFERENCES.............................................................................................................121 BIOGRAPHICAL SKETCH.......................................................................................................127

PAGE 7

7 LIST OF TABLES Table page 4-1 BET surface area, total pore volume, and calcu lated pore size for the STC synthesized with varying con centrations of HF and TiO2.................................................49 5-1 Summary of experi m ental conditions................................................................................68 5-2 Weisz modulus values for variable sp ace time experiments............................................. 68 6-1 BET surface area and average pore volume of TiO2-doped sorbents................................85 6-2 Surface area and TiO2 loading per reactor volume............................................................ 86 7-1 BET surface area, pore volume, and pore size analysis for pilot STC.............................. 95 7-2 UV intensity measurements through packed beds of STC of varying depths. .................. 95 7-3 Results of pilot studies with variable potentiom eter settings............................................. 95 8-1 Summary of pilot experiments.........................................................................................110 8-2 Full-scale performance data for three operation cycles. .................................................. 110

PAGE 8

8 LIST OF FIGURES Figure page 3-1 Bench-scale reactor set-up used for methanol/H2S adsorption and photocatalytic oxidation studies................................................................................................................44 3-2 Reactor drawings. A) 8 mm annulus r eactor. B) 25 mm annulus reactor......................... 44 4-1 Measured and expected surface area data for STC. ........................................................... 49 4-2 Nitrogen adsorption/desorption isotherm s. A) STC with varying pore sizes and constant TiO2 loading (12%). B) 50 STC with varying TiO2 loadings (0-60%)........... 50 4-3 Pore size distributions. A) STC with varying pore sizes and constant T iO2 loading (12%). B) 50 STC with varying TiO2 loadings (0-60%)............................................... 51 5-1 Adsorption breakthrough curves for STC pe llets of varying pore sizes (50 120 and 260 ) and constant TiO2 loading (12%).................................................................... 69 5-2 Methanol removal using STC pellets. A) Methanol rem oval using STC of varying pore sizes illuminated with UVA light. B) Extended study for 50 12% STC pellets.... 69 5-3 Effluent formaldehyde concentrations from STC pellets of varying pore sizes (50 120 and 260 ) when illum inated with UVA light.......................................................70 5-4 Normalized effluent methanol concentration for 50 12% STC illuminated with UVA light at various space tim es...................................................................................... 70 5-5 Effluent formaldehyde concentration for 50 12% STC illum inated with UVA light at various space times........................................................................................................ 71 5-6 Effluent methanol and formaldehyde con centratio ns at steady state at varying space times. A) 50 B) 120 C) 260 STC.........................................................................71 5-7 Effluent methanol and formaldehyde concen tratio ns at steady st ate for variable face velocities and constant space time (4.3 s) for 50 12% STC........................................... 73 5-8 Linear regression of L-H m odel using mineralization rate s achieved at various space times. A) 50 12% STC. B) 120 12% STC. C) 260 12% STC................................73 5-9 Effect of TiO2 loading in 50 STC on methanol removal when illuminated with UVA light...........................................................................................................................74 5-10 Effect of TiO2 loading on formaldehyde production at steady state using 50 STC illuminated with UVA light............................................................................................... 75 5-11 Percent transmittance of UVA light through 50 STC with various TiO2 loadings........75

PAGE 9

9 5-12 Effluent formaldehyde concentrations using 50 12% STC irradiated with UVA and UVC light. ..........................................................................................................................76 5-13 Predicted C/Co versus absorbed light flux for 50 12% STC, 4.3 s residence time, Co = 50 ppmv and 95% relative humidity.......................................................................... 76 5-14 Effluent concentrations of H2S, methanol and oxidati on byproducts from 50 4% STC illuminated with UVC light (Co methanol = 50 ppmv, Co H2S = 50 ppmv)............. 77 5-15 Effluent concentrations of H2S and SO2 from 50 4% STC illuminated with UVC light (Co H2S = 50 ppmv)...................................................................................................77 6-1 Normalized effluent meth anol concentration for titani a-doped m aterials used in the dark (adsorption only) and with UV light (adsorption and oxidation)..............................86 6-2 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials.................. 87 6-3 Normalized effluent meth anol concentration for TiO2-doped materials used in the dark and with UV light in a high humidity gas stream (RH = 95%)................................. 87 6-4 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a high-humidity gas stream (RH = 95%).............................................................................. 88 6-5 Water vapor adsorption breakthrough prof ile in a high hum idity gas stream (RH = 95%)...................................................................................................................................88 6-6 Normalized effluent meth anol concentration for TiO2-doped materials used in a large annulus reactor (25 mm) and high humidity gas stream (RHi = 95%).............................. 89 6-7 Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a 25 mm annular reactor and high humidity gas stream.......................................................89 7-1 Mixing assembly for blending the raw i ngredients for pilot-scale STC synthesis. ........... 96 7-2 Pilot-scale molds for STC synthesis.................................................................................. 96 7-3 Specialty heat chamber fo r pilot-scale STC synthesis. ...................................................... 96 7-4 Alzak box test system for measuring UV light penetration through packed beds of various depths. ...................................................................................................................97 7-5 Process flow diagram for meth anol degradation pilot studies. .......................................... 98 7-6 General arrangement drawing of the p ilot reactor for methanol degradation.................... 98 7-7 Photo of pilot reactor..................................................................................................... ....99 8-1 Process flow diagram for mercury recovery pilot studies............................................... 110

PAGE 10

10 8-2 Schematic of pilot reactor for mercury recovery............................................................. 111 8-3 Photo of pilot reactor with UV lights illuminated........................................................... 112 8-4 Influent and effluent merc ury concentrations for the pilo t reactor packed with virgin STC pellets (Test No. 1). ................................................................................................. 113 8-5 Influent and effluent merc ury concentrations of the pilo t rea ctor packed with virgin pellets (Chamber A) and regene rated pellets (Chamber B).............................................114 8-6 Process flow diagram for full-scale in stallation of mercur y recovery units....................114 8-7 Influent and effluent concentrations for the full-scale reacto r during its second adsorption cycle...............................................................................................................115 8-8 Comparison of cost per pound of mercury rem oved for activated carbon and STC as a function of influent mercury loading (for systems designed to treat up to 2765 g/day)...............................................................................................................................116

PAGE 11

11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PHOTOCATALYTIC OXIDATION OF HAZARD OUS AIR POLLUTANT S USING SILICATITANIA COMPOSITES IN A PACKED-BED REACTOR By Jennifer Morgan Stokke May 2008 Chair: David W. Mazyck Major: Environmental Engineering Sciences This work centered on the optimization and scale-up of photocatalyt ic reactors employing silica-titania composites (STC) for two applicati ons involving the abatement of hazardous air pollutants (HAPs) from industria l facilities: (1) degradation of HAPs, particularly methanol, emitted from pulp and paper mills; and (2) recovery of mercury emitted from chlor-alkali plants. STC were synthesized with varying pore sizes (50 120 and 260 ) and TiO2 loadings (0-60%) and were tested for the removal of metha nol from a humid air stream. The efficiency of methanol oxidation was dependent on the surface area of the STC and the space time of the gas in the reactor. For 120 12% and 260 12% STC irradiated with UVA light, a lag time of 1.0 s and 1.2 s, respectively, was observed before mineralization began. After this lag time, which was zero for the 50 12% STC, the data followed pse udo-first order reaction ki netics and the rate constant, k, was 0.40 s-1 for all pore sizes. Using the 50 STC, the efficiency was further improved by using a 4% TiO2 loading and UVC lamp, which generated a higher photon flux compared to a UVA lamp. The presence of H2S in the gas stream decreased methanol removal efficiency and resulted in SO2 and SO4 2oxidation byproducts. When compared to other catalyst supports, the STC was more efficien t in a low-humidity gas stream with a relative humidity (RH) of less than 0.22% at 23C. In a high humidity ga s stream (RH = 95% at 23C), the efficiency of

PAGE 12

12 the STC was inhibited by water vapor due to it s surface chemistry and performed similarly to TiO2-coated activated carbon. When compared to TiO2-glass spheres, the use of an adsorbent catalyst support resulted in highe r degradation efficiencies. Ba sed on the promising bench-scale results, a 40 ACFM pilot reactor was fabricat ed employing a packed bed of STC and a 4.3 s space time through the packed bed. The pilot reactor achieved methanol removal rates up to 66 7% with less than 1 ppmv formaldehyde production at steady state. A pilot-scale photocatalytic reac tor packed with STC was tested at a chlor-alk ali facility over a three-month period. This pilot reactor treated up to 10 ACFM of end-box exhaust and achieved 95% mercury removal. The pilot reactor was able to maintain excellent removal efficiency even with large fluctuations in influent mercury concentration (400-1600 ug/ft3). The STC pellets were regenerated ex -situ with hydrochloric acid and performed similarly to virgin STC pellets when returned to service. Based on these promising results, two full-scale reactors with in-situ regeneration capab ilities were installed and ope rated. After optimization, these reactors performed similarly to the pilot reactor. A cost analysis was performed comparing the treatment costs (i.e., cost per pound of mercur y removed) for sulfur-impregnated activated carbon and the STC system. The STC proved to be both technologica lly and economically feasible for this installation.

PAGE 13

13 CHAPTER 1 INTRODUCTION Industrialized nations are faced with environmental problems re lated to the remediation of hazardous wastes, treatment of contaminated water, and control of air pollution from industries, military installations, and the civilian sector (Hoffmann et al., 1995). Hazardous air pollutants (HAPs) pose particular threats to human health or the environment due to their toxicity. The Clean Air Act requires the Environmental Protec tion Agency (EPA) to regulate emissions of 188 HAPs from industrial sources. Although a variety of technologies exist for the removal of HAPs, heterogeneous photocatalysis has the potential to offer significant benefits with respect to energy, size, and reliability compared to tradit ional technologies. Photocatalysis has proven effective at the bench-scale for the oxidation of a variety of organi c compounds (e.g., alcohols, aliphatic compounds, aromatic compounds ) to inert byproducts, such as CO2 and H2O (Hoffmann et al., 1995). In addition, inorganic co mpounds can undergo photocatalytic transformation, which can be manipulated to enhance removal by adsorption onto the catalyst surface or subsequent treatment processes (Huang et al., 1996; Nguyen et al., 2004; Pitoniak et al., 2003; Pitoniak et al., 2005; Ferguson et al., 2005). Although much research has been devoted to photocatalysis for environmental applications, the testing and application of these systems beyond the bench-scale is limited. Photocatalytic systems for treatment of gas-phase contaminants typically employ a thin film of titania (TiO2) on, for example, reactor wa lls, wire mesh, or glass be ads (Alberici and Jardim, 1997; Dijkstra et al., 2002; Ki m and Hong, 2002; Peral et al., 1997; Chang et al., 2000). These systems may not lend themselves well to scale up due to their low adsorption capacity, mass transfer limitations, and problems with catalyst immobilization or durabi lity (Devilliers, 2006). In order to overcome these issues, a nano-st ructured silica-titania composite (STC) was

PAGE 14

14 developed and has been tested at the bench-scale for the removal of mercury from synthetic flue gas (Pitoniak et al., 2003; Pitoniak et al., 2005; Li and Wu, 2006). The STC has also been tested for the degradation of organic compounds in water (Londeree, 2002; Holmes, 2003; Ludwig et al., 2008) and deactivation of pathogens (Gart on, 2005). Although the results of these studies showed that the STC was capable of capturi ng mercury and completely degrading organic compounds, the STC had not been develope d past the laboratory scale. This work centered on the optimization and scale-up of STC for two new applications involving the abatement of HAPs emitted from in dustrial facilities: (1) degradation of HAPs, particularly methanol, emitted from pulp and paper mills; and (2) recovery of mercury emitted from chlor-alkali plants. These two applications are discussed in furt her detail below. Degradation of HAPs Emitted from Pulp and Paper Mills Forest products provide essentia l resources such as energy an d materials. Compared to fossil fuels (e.g., coal and oil) the resources fr om forest products are more sustainable and diverse. However, the processing of forest products creates unwanted by-products including volatile organic compounds (VOCs), some of which are HAPs (Someshwar, 1994). VOCs can directly affect human health and are one of the main precursors of tropospheric ozone, a respiratory irritant in humans and a major contributor to smog formation. Maximum Achievable Control Technology (MACT) standards, as part of the 1998 Cluster Rules promulgated by the EPA, require that the concentration of HAPs in high volume low concentration (HVLC) gases emitted from pulp and paper mills be decreased by at least 90% Thermal oxidation is the most commonly appl ied technique for the control of VOC/HAP emissions from the forest products industr y sources (Varma, 2003; Garner, 2001). While effective, these measures require a consta nt fuel supply to support the thermal energy requirements and create other unf avorable byproducts (such as NOx ). In addition, the fans and

PAGE 15

15 ductwork required for the transport of gases are co stly since gas streams are usually directed to existing boilers (Varma, 2003). Thus, a cost-effec tive technique for in-situ treatment of these pollutants is needed. This work focuses on optimizing the STC and reactor operating parameters for the treatment of HVLC gases emitted from pulp and paper mills, namely brown stock washers. Typical HVLC gases are composed of air saturate d with water vapor and contain HAPs (such as methanol, acetaldehyde, formaldehyde), other VOCs (such as acetone and methyl ethyl ketone) and total reduced sulfur (TRS) species (Som eshwar, 1994; Varma, 2003). Methanol is the primary constituent in HVLC gases and contributes to over 90% of the total HAP emissions from brown stock washer vent gases (Someshwar 1994). A study conducted by the Someshwar (1994) at the National Council of Air and Stream Improvement found that the methanol concentration in vent gases at 16 pulp and paper mills ranged from 32 to 2,263 ppmv. The range of concentrations was attributed to the type of washers, vent gas flow rates, and methanol concentration of the shower water. Vent gas fl ow rates were also found to be highly variable, ranging from 162 to 16,808 scfm (Someshwar, 1994). This work presents the optimization of th e STC properties (e.g, surface area, pore size, TiO2 loading) and reactor paramete rs controlling mass transfer and photon flux to achieve 90% removal and complete oxidation of methanol, which was used as the target HAP. The competitive effects of water vapor and H2S (a target TRS species) on methanol removal were studied since both constituents are present in HVLC gases. In addition, the removal of H2S was investigated to understand if the STC can provide a co-benefit for odor removal. The performance of the STC was compared to that of TiO2 coated on other catalyst supports (i.e.,

PAGE 16

16 activated carbon and nonporous gl ass spheres). Finally, a 40 ft3/min (ACFM) pilot reactor was designed and tested for methanol re moval from a humid air stream. The following hypotheses were investigated in this study: As the available total surface area of the STC increases, the rate of methanol oxidation will increase proportionally. The rate of methanol oxidation is limited by the resistance to mass transfer of the contaminant to the STC surface and the space time of the gas in the reactor. The distribution of ultraviolet (UV) light in the reactor will dictate the rate of methanol oxidation. The rate of methanol oxidation will be depe ndent on the transparency of the catalyst support, with more transparent supports achieving greater oxidation rates. TRS present in HVLC gases will be oxidized to sulfate on the catalyst surface, which will accumulate over time and occupy adsorption/oxida tion sites thereby decr easing the rate of methanol oxidation. The objectives of this research were as follows: Synthesize and characterize STC with various properties (i.e., pore size, surface area, and TiO2 loading). Develop and optimize a bench-scale photocatalyt ic reactor using STC for the removal of VOCs/HAPs, using methanol as the target pollutant. Compare the silica gel to other catalyst suppor ts with varying transparency and surface area. Assess the effects of water vapor an d TRS on the removal of methanol. Design, fabricate and test a pilot-scale reac tor for the optimization of VOCs/HAP removal using methanol as the target constituent. Recovery of Mercury Vapor Emitted in EndBox Exhaust at Chlor-alkali Facilities The release of mercury to the environment is of particular concern due to its volatility, persistence and tendency to bioaccumulate. Methylation of mercury by microbes in the environment results in the formation of met hylmercury, which rapidly bioaccumulates in the cells of higher organisms in aquatic systems. Me thyl mercury is magnified at each trophic level

PAGE 17

17 of the food chain thereby threatening ecosystems as well as human health. The primary exposure route for humans is through the consumption of fish containing methylme rcury (DItri et al., 1978). This exposure can lead to adverse neurological effects, pa rticularly in the developing fetus and during early childhood (Schettler, 2001). Mercury is released into the environment fr om anthropogenic sources such as coal-fired power plants, cement plants, waste incinerators, and manufacturing proce sses such as the chloralkali industry. The coal-fired power industry is the largest so urce of mercury in the United States, emitting about 43 tons of mercury in 1999 (Pavlish et al., 2003). Although emissions from coal-fired power plants ar e significant, other industries, such as the mercury-cell chloralkali industry, can emit more mercury on a per f acility basis (EPA, 1997a). Mercury abatement from these sources is also crucial in order to protect human health and the environment. Therefore, this work focuses on the devel opment and implementation of a photocatalytic technology for mercury recovery in the chlor-alkali industry. The chlor-alkali industry produces valuable chemicals such as chlorine, hydrogen, and caustic soda. In 2001, between 150 and 200 chlor-alka li facilities throughout the world used the mercury-cell process. Although this process uses the mercury in a closed-loop system, mercury is released to the environment through entrainment in by-produc t streams, end-box ventilation systems and fugitive emissions (EPA, 1997b). Accord ing to Southworth et al. (2004), an average of about three tons of mercury per year must be added to the production pr ocess at each mercurycell facility in the US to account for losses. The National Emissions Standard for Hazardous Air Pollutants (NESHAP) for mercury cell chlor-alkali plants requires these facilities to drastically reduce mercury emissions from their gas-phase emissions.

PAGE 18

18 Due to these stricter standards, chlor-alkali facilities must either implement new control technologies or eliminate mercury from thei r process. A commercially-available control technology for reducing mercury emissions is activated carbon adsorption (EPA, 1997b; Anastas, 1976). While sulfuror iodine-impregnated activated carbon may reduce mercury emissions to acceptable levels, it has a finite adsorption capacity and must be replaced and properly disposed as hazardous waste. Therefore, the mercury problem is being transferred from the air phase to the solid phase. The cost and risk associated with the continuous replacement and disposal of mercury-laden activated carbon can be limiting, as many mercury-cell facilities have chosen to convert to a mercury-free (e.g., me mbrane) process for manufacturing (EPA, 1997b). However, conversion to another process requir es a large capital investment. Thus, a cost effective and regenerable solution is required for reducing mercury emissions. This study focuses on the scale-up of the STC for use in the chlor-alkali industry, particularly for the recovery of mercury fr om end-box exhaust. The STC was originally developed at the bench-scale for mercury removal from flue gas emitted from coal-fired power plants. Pitoniak (2004) found that the adsorption capacity of the STC incr eased after periods of photocatalytic oxidation due to the formation of HgO on the composite surface. The total mercury (i.e., elemental and oxi dized) adsorption capacity was 3%wt, as determined by thermo gravimetric analysis. The STC could be regenerated by rins ing with acid. The composition of end-box exhaust consists of air containing trace levels of hydrogen (< 0.02%), water vapor (saturated at 6 8 C), entrained water dr oplets, and elemental mercury. The mercury concentration in end-box exhaust is typically two to three orders of magnitude higher than that in flue gas. The STC has the po tential to be both technically and economically advantageous for the removal of mercury from end-box exhaust in chlor-alkali facilities because

PAGE 19

19 of its high mercury adsorption capacity and ability to be regenerated with acid and re-used in bench-scale studies. The mercuryladen acid used for regeneration could be recycled into the mercury-cell process, thus clos ing the loop on mercury emissions. This work summarizes the design and performance of pilotand full-scale reactors used to recover mercury from the endbox exhaust at a chlor-al kali facility. In addition, an econom ic analysis, which compares the costs of implementing this technol ogy versus using activated carbon at the facility, is presented. The following hypotheses were investigated in this study: The research involving mercury removal from simulated flue gas can be translated to mercury removal from caustic exhaust. Mass transfer and UV light distribution are th e limiting factors for mercury removal from caustic exhaust. STC will be economically favorable when compar ed to activated carbon for the removal of mercury from end-box exhaust due to the abilit y of the STC to be regenerated and reused. The objectives of this study were as follows: Confirm the efficacy of the technology for mercury removal from caustic exhaust and recovery of the sorbed mercur y by regeneration with HCl in p ilot and prototype studies at a chlor-alkali facility. Determine the factors that may limit mercur y removal efficiency (e.g., residence time, mass transfer, and UV light distribution within the packed bed) in pilot-scale studies. Compare the cost of full-scale treatment units employing STC and activated carbon.

PAGE 20

20 CHAPTER 2 LITERATURE REVIEW Photocatalysis When a semiconductor (e.g., TiO2) is irradiated with UV light that has an energy equal to or greater than the band gap energy, an electron (e-) is promoted from the valence band to the conduction band, leaving behind a positively charged hole (h+) in the valence band. This reaction is shown in Equation 1 (S erpone, 1995; Hoffmann et al., 1995). TiO2 + hv eCB + h+ VB (2-1) These electron-hole pairs can then recombine to generate heat or migrate to the surface and participate in redox reactions. The positively charged holes are powerful oxidants while the electrons in the conducti on band can participate in reduction reactions. Electron-hole pairs can oxidize organic pollu tants directly via the electron hole or indirectly via the formation of other power ful oxidants (e.g., hydroxyl radicals). The hydroxyl radical (*OH), which has been identified as the primary oxidant in the photocatalytic oxidation of organic compounds (Turchi and Ollis, 1990) and inorganic compounds such as mercury (Pitoniak, et al. 2003; Pitoniak et al., 2005), can be generated vi a several pathways shown in Equations 2-2 through 2-8 (Al-Ekabi and Serpone, 1988). H2O + hVB + *OH + H+ (2-2) OH+ hVB + *OH (2-3) O2 + 2H+ + 2eH2O2 (2-4) H2O2 + e*OH + OH(2-5) H2O2 hv 2 *OH (2-6) O2 + eO2 (2-7) O2 + H2O2 *OH + OH+ O2 (2-8)

PAGE 21

21 Water or hydroxide ions adsorbed at the TiO2 surface can trap electron holes (thereby preventing recombination) and form hydroxyl ra dicals (Equations 2-2 an d 2-3). Thus, in gasphase systems, relative humidity (RH) is often n ecessary for efficient oxidation of contaminants. However, some studies have shown that excess RH can decrease oxidation rates for some compounds due to the competitive adsorption betw een water and the target pollutant on the photocatalytic surface (Kim and Hong, 2002; Ob ee and Brown, 1995; Li and Wu, 2007). As shown in Equations 2-4 through 2-8, oxygen can se rve as an electron acceptor (thus preventing electron-hole recombination) a nd create oxidative species such as the superoxide radical (O2 -), which can also react to cr eate hydroxyl radicals. The photocatalytic oxidation of organic pollutants yields a series of intermediate byproducts of progressively higher carbon to oxygen ratios. If mi neralization is achieved, inert byproducts, such as carbon dioxide, water, and dilute mineral acids (in the case of halogenated compounds) are formed (Ollis et al., 1991). Metals can be either oxidized or reduced via photocatalysis. This is advantageous for separation of metals from gas or liquid streams if the photocatalytically treated form of the metal is more easily remove d by either sorption/deposition onto the catalyst surface or in subsequent treatme nt processes. For example, water entering a potable water treatment plant can be pretreated so that As(III) is oxidized to As(V), which is more easily removed by conventional water tr eatment processes (Ferguson et al., 2005). Removal of oxidized metals from water stre ams via photoreduction and subsequent deposition onto the surface of TiO2 has proven effective for a variety of other metals such as silver (Huang et al., 1996), cadmium (Nguyen et al., 2003), and selenium (Nguyen et al., 2004). The photocatalytic oxidation of gasphase elemental mercury (Hgo) to HgO has resulted in enhanced

PAGE 22

22 removal (Pitoniak et al., 2003). This enhanced re moval of mercury via ph otocatalytic oxidation will be discussed in more detail later in this chapter. Photocatalytic reactions can be catalyzed by a variety of semiconductor materials (e.g., TiO2, ZnO, CdS, ZnS and Fe2O3) due to their electronic structur e, which is comprised of a full valence band and empty conduction band (Hoffmann et al., 1995). TiO2 is often used as a photocatalyst since it is commercia lly available, non-toxic, relatively inexpensive compared to heavy metal catalysts, and resist ant to catalyst poisoning (Hurum et al., 2003). The principal catalytic phases of TiO2 are anatase and rutile. Generally, anatase phase TiO2 is regarded as the preferred phase due to its greater adsorption affinity for organi c compounds and lower rate of electron-hole recombination (Hurum et al., 2003). The inactivity of rutile TiO2 is attributed to the rapid rate of electron-hole recombination. However, TiO2 with a combination of anatase and rutile phases (such as Degussa P25) has been shown to have enhanced photocatalytic ac tivity compared to pur e anatase phase TiO2. In mixedphase systems, electrons gene rated by the rutile-phase TiO2 can be transferred and trapped in lower energy anatase lattice sites. Degussa P25 contains unusually small nanoclusters of rutile crystallites dispersed within anat ase crystallites. This morphol ogy, which is responsible for the enhanced activity of Degussa P25, allows for th e rapid transfer of electrons from rutile to anatase, resulting in catalytic hot spots at the rutile-anatase interf ace (Hurum et al., 2003). Due to its enhanced activity and commercial avai lability, Degussa P25 (70% anatase, 30% rutile) is frequently used in photocatal ytic systems (Turchi and Ollis, 1989; Chen et al., 1999; Alberici and Jardim, 1997; Minero et al., 1992; Obee and Brown 1995). TiO2 is often immobilized as a thin film on various surfaces (e.g., beads, woven mesh, reactor walls) or on/in adsorbent materials (e.g., s ilica gel, alumina, activ ated carbon), which act

PAGE 23

23 as catalyst supports. When TiO2 is immobilized as a thin film, the photocatalytic reaction rate can be affected by mass transfer limitations (Ollis et al., 1991). The use of TiO2 for practical applications is dependent on the immobilization of the TiO2 particles onto a support such that the composite is durable, provides a reasonably hi gh surface area, and allows accessibility to the immobilized catalyst (Shul et al., 2003). The use of adsorbent mate rials as catalyst supports is discussed in further detail below. Use of Adsorbents as Catalyst Supports By using an adsorbent catalyst support, the contaminant is concentrated around the photocatalyst. This has been s hown to result in an increase in the photocatalytic reaction rate (Anderson and Bard, 1997; Tsumura et al., 2002; Vohra and Tanaka, 2003; Torimoto et al., 1996). In addition, organic contaminants are more likely to be mineralized since the intermediates, which can be toxi c, can be retained and furthe r oxidized (Torimoto et al., 1996; Lu et al., 1999). Activated carbon is a commercially -available adsorbent material that has been studied for use as a catalyst support (Yong et al., 2005; Arana et al., 2004; To rimoto et al., 1996; Lu et al., 1999; Tsumura et al., 2002). Activated carbon is made by heating a carbonaceous precursor, usually wood or coal, in the presence of an activating agent (e .g., steam, carbon dioxide, oxygen). Properties of activated carbon such as surface area and pore size distribution are affected by the activation method (i.e., temp erature, time, and activating agent). TiO2 can be deposited onto the activated carbon by various methods such as boil deposition (Arana et al., 2004) or chemical impregnation (Yong et al., 2005) The use of activated carbon as a catalyst support has proven effective. For exampl e, Arana et al. (2004) found that TiO2-coated activated carbon was more efficient than bare TiO2 for the mineralization of methanol. Lu et al. (1999) showed that granular activated carbon (GAC) was the most efficient catalyst support for the

PAGE 24

24 degradation of propoxur when compared to other catalyst supports such as zeolite, brick, quartz, and glass beads. Bare TiO2 achieved higher oxidation ra tes compared to the GAC/TiO2. This was attributed to the blockage of photons by the GAC, which decreased the number of photons which reached the surface of the TiO2 to promote oxidation reac tions. Although the overall oxidation rate decreased, the mi neralization rate of the TiO2/GAC was greater than that of the bare TiO2. The superior performance of the GAC comp ared to other catalyst supports and its ability to increase mineralization rates compared to bare TiO2 were attributed to its adsorption properties. Mixed oxide materials containing TiO2 supported on silica (SiO2) have also proven effective for a variety of photocatalytic reactions. SiO2 is advantageous as a catalyst support since it has high thermal stabilit y and excellent mechanical stre ngth (Gao and Wachs, 1999). In addition, SiO2 is transparent to UV light. SiO2-TiO2 mixed oxide materials are generally prepared by sol-gel and co-precipitation methods. The sol-gel hydrolysis route is most widely used due to the ability to control textural and surface properties of the resulting composite material (Gao and Wachs, 1999). SiO2TiO2 materials have shown supe rior catalytic efficiency due to quantum particle effects and the formation of Ti-O-Si linkage s (Anderson and Bard, 1997; Gao and Wachs, 1999). Ti-O-Si linkages have resulted in enhanced adsorption of some c ontaminants (such as phenol and R6-G) via strong Bronsted acid sites, which led to enhanced degradation rates (Anderson and Bard, 1997; Yang and Chen, 2005) Yang and Chen (2005) al so suggested that SiO2 prevented electron-hole recombination by accepting electrons and providi ng hydroxyls for hole capture at the SiO2-TiO2 interface, thereby further increasi ng photocatalytic efficiency of SiO2-TiO2 composites compared to TiO2 alone.

PAGE 25

25 Shul et al. (2003) showed that the presence of SiO2 as a catalyst support for TiO2 helped to promote the efficiency of acetaldehyde degradation because the SiO2 enhanced the effective surface area of the TiO2 and increased the adsorption capacity of the composite compared to unsupported TiO2. The photocatalytic activity of the com posite increased with increasing surface area of the SiO2 support. It was suggested th at the increase in photocat alytic degradation rate was due to an increase in the concentration of reactants and intermediates near the TiO2. Uchiyama et al. (2005) found that the incorporation of SiO2 was also beneficial for the degradation of acetylacetone. Composites with Ti/Si ratios of 0 (pure SiO2), 0.05, 0.25, 0.5, and 1.0 (pure TiO2) were tested for acetylacetone adsorption and degradation. The amount of adsorbed acetaldehyde was proportional to the specific surface area of the material, whether it be unsupported TiO2, SiO2, or a composite of SiO2 and TiO2. The composite with Ti/Si = 0.05 achieved the highest rate of degradation, likely due to its high surface area (1103 m2/g) and increased transparency compared to the other composites containing TiO2. During the photocatalytic oxidation of acetylacetone, molecules that were adsorbed on the SiO2 transferred to the photocatalyst, due to the concentration gradient on the surface of th e composite, where they were subsequently oxidized. Xu et al. (1997) also showed that TiO2 supported onto a silica gel substrate resulted in higher degradation efficiencies for the oxi dation of acetophenone in water compared to unsupported TiO2. The supported TiO2 showed higher photoactivity when supported on smaller SiO2 particles. This suggests that the dispersing effect of the SiO2 support on the TiO2 particles was operative in enhanc ing its photoactivity. Photocatalytic Degrad ation of Methanol The photocatalytic oxidation of methanol can proceed via two pathways: (1) direct oxidation (Pathway 1), where methanol and its organic byproducts are oxidized by the electron

PAGE 26

26 hole and mineralized to CO2; and (2) indirect oxidation (Pat hway 2), where methanol and its organic byproducts are oxidized by ad sorbed *OH and mineralized to CO2 and H2O (Chen et al., 1999). The reactions for Pathway 1 (direct oxidati on) are shown in Equations 2-9 through 2-15. CH3OH + -s CH3Os + H+ (2-9) CH3Os + h+ CH3Os (2-10) CH3Os + h+ CH2Os + H+ (2-11) CH2Os + h+ CHOs + H+ (2-12) CHOs + h+ + H2O CHOOHs + H+ (2-13) CHOOHs + h+ CHOOs + H+ (2-14) CHOOs + h+ CO2s + H+ (2-15) The reactions for Pathway 2 (indirect oxida tion) are shown in Equations 2-16 through 222. H2Os + h+ *OHs + H+ (2-16) CH3Os + *OHs CH3Os + OHs (2-17) CH3Os + *OHs CH2Os + H2Os (2-18) CH2Os + *OHs CHOs + H2Os (2-19) CHOs + *OHs CHOOHs (2-20) CHOOHs + *OHs CHOOs + H2Os (2-21) CHOOs + *OHs CO2s + H2Os (2-22) The H+ product from the above reactions can react with O2 to form H2O via the reduction reaction shown in Equation 2-23 (Noguchi et al., 1998). O2 + 4e+ 4H+ 2H2O (2-23)

PAGE 27

27 It is generally reported that the degrada tion of organic molecules via photocatalytic oxidation proceeds by the formation and subsequent reaction with hydroxyl radicals (Pathway 2) (Turchi and Ollis, 1990; Hoffmann et al., 1995; Obee and Brown, 1996). According to Kim and Hong (2002), methanol degradation was achieved using a TiO2 thin film in the absence of water vapor. They suggested that methanol was oxidized by hydroxyl radicals that were formed from the hydroxyl groups of the methanol. The degrada tion rate reached an optimum at a water vapor concentration of 0.383 mol/m3. At water vapor concentrations higher than the optimum, the degradation efficiency decreased suggesting that water molecules competitively adsorbed to catalyst surface thereby decreasing the degradation rate. Yamakata et al. (2003) suggested an alte rnative mechanism for the photocatalytic oxidation of methanol, whereby methanol is oxidi zed directly by the elect ron hole, even in the presence of water vapor. In their study, the adsorbed water vapor was responsible for the electron-consuming reaction while methanol wa s responsible for the hole consuming reaction. The results suggested that the ho le directly reacted with methanol, which adsorbed on the TiO2 surface as the methoxy species (CH3O-), and that the hydroxyl radicals did not play a role in the photocatalytic oxidation since the highest activity of hole consuming reactions was achieved in the absence of water vapor. They suggested that water vapor enhan ced the photocatalytic oxidation of methanol by preventing electron accumulation, which would otherwise cause defective sites on the TiO2 surface, and electron-hole recombination. Literature generally suggests that methanol is degraded first to formaldehyde, then to formic acid, and finally to carbon dioxide and wa ter (Lichtin et al., 1994; Tsuru et al., 2003; Chen et al., 1999), which corresponds to both Pa thway 1 and Pathway 2 described above. Some have suggested alternative pathways that result in the formation of formates, particularly methyl

PAGE 28

28 formate (Tsuru et al., 2003; Sade ghi et al., 1996; Arana et al., 2004). Tsuru et al. (2003) found low levels of methyl formate in the effl uent when degrading methanol using TiO2 membranes. They suggested that the presence of methyl formate was a result of esterification reactions between methanol and formic acid. Arana et al (2004) studied methanol degradation using a flow-through column (4 mm diameter, 15 mm height) packed with Degussa P25 TiO2. They observed no mineralization of methanol and c oncluded from FTIR studies that bare TiO2 did not degrade methanol due to the fast production of formates that resulted from methoxide (CH3O-), which poisoned the active centers of the photocatalyst. The results from Arana et al. (2004) contra dict those reported by Alberici and Jardim (1997) and Kim and Hong ( 2002). Alberici and Jardim (1997) found that a TiO2 thin film achieved 98% degradation of methanol with no catalyst deactiv ation. In addition, no intermediate oxidation byproducts were present in the effluent, which is likely a result of the long residence time (approximately 2 min) used in the experime nts. Kim and Hong (2002) also observed that methanol degradation was achieved using a TiO2 thin film. For studies in which methanol degradation was achieved, the degradation rate of methanol depended on variables such as the influent methanol concentr ation (Kim and Hong, 2002; Tsuru et al., 2003), water vapor concentration (Kim and Hong, 2002), residence time (Tsuru et al., 2003) and photon flux of the UV lamp (Kim and Hong, 2002; Alberici and Jardim, 1997). Tsuru et al. (2003) reported that a decr ease in flow rate yielded gr eater mineralization due to the resulting increase in residence time. The degradation rate of methanol generally followed Langmuir-Hinshelwood kinetics such that the reac tion was first order at low concentration and zero order at high concentrations (Kim and H ong, 2002; Tsuru et al., 2003). This transition occurred at influent concentr ations greater than 500 ppmv for Kim and Hong (2002) and greater

PAGE 29

29 than 4000 ppmv for Tsuru et al. (2003). The concentrati on in which the reaction order transitions from first order to zero order varied based on the reactor system and experimental conditions. Tsuru et al. (2003) also found that the conversio n of degraded methanol to carbon dioxide and water was not dependent on influent methanol c oncentration, indicating that higher levels of intermediate byproducts were formed as a result of increased influent methanol concentration. The intrinsic rate of photocatalytic reactions is limited by the photon flux of the UV light source (Hoffmann et al., 1995). For illumination le vels above one sun equivalent, the oxidation rate generally increases with the square root of the light intensity. For levels below one sun equivalent, the oxidation rate in creases linearly with light intens ity. One sun equivalent is about 1 to 2 mW/cm2 (Obee and Brown, 1995). Kim and Hong (2002) reported that the degradation rate of methanol was dependent on the photon fl ux of UV light such th at the photocatalytic degradation rate increased with the square root of the phot on flux. When the degradation efficiency of methanol was tested using a black lamp (peak wavelength = 352 nm) and a germicidal lamp (peak wavelength = 254 nm), the germicidal lamp resulte d in higher degradation rates. Since the band gap energy of anatase phase TiO2 is 3.2 eV, photons with wavelengths less than 385 nm are required to excite the photocatalys t. Thus, both the black lamp and germicidal lamps can be used to promote photocatalytic reac tions. It should be noted that more energetic irradiation, such as that provided by a germicidal lamp, may affect the degradation efficiency by direct photolysis of the organic compound or the formation of radi cals that may alter conversion yields (Alberici and Jardim, 1997) Kim and Hong (2002) attributed the higher degr adation rates observed with the germicidal lamp to the greate r photon flux emitted from this lamp compared to the black lamp. They reported that the photon fl ux from the germicidal lamp was about two times that emitted from the black lamp. These re sults contrast with those reported by Alberici

PAGE 30

30 and Jardim (1997), who showed that a germicidal lamp did not increase degradation rates of any VOC (including methanol) compared to a black la mp. In these studies, both the germicidal and black lamp had a nominal power rating of 30 W; however, the radian t power output of the germicidal lamp was 25% greater than that of the black lamp. Photocatalytic Oxidation of H2S While many studies have been performed using TiO2 for the degradation of VOCs, few of these studies have investigat ed the oxidation and removal of inorganic or sulfur-containing compounds. The photocatal ytic reaction of H2S with TiO2 has the potential to form byproducts such as SO2, SO4 2-, and elemental sulfur. In order to completely oxidize H2S to SO4 2-, an eight electron transfer is required. The formation of SO4 2would result in deposition onto the surface of the TiO2, since this species does no t exist in the gas phase (Kat aoka et al., 2005) Likewise, the elemental sulfur that could be formed as a byproduct would also deposit on the surface of the catalyst. This sulfur/sulfate deposition could potentially deacti vate the catalyst over time by blocking active sites. Alternatively, SO2 that could be formed as a result of the oxidation of H2S could desorb back into this gas stream. Since SO2 is a toxic gas and regulated by the EPA, subsequent removal processes (e .g., wet scrubbing) would be required to remove this gas. Canela et al. (1998) and Ka taoka et al. (2005) inves tigated the oxidation of H2S using a TiO2 thin film and found that SO4 2was formed and accumulated on the TiO2 surface without producing any gaseous intermediates (e.g., SO2). Portela et al. (2007) studied the oxidation of H2S using TiO2 thin films coated on polymeric materials and found that both SO4 2and SO2 were produced as oxidation products. The accumulation of SO4 2on the TiO2 surface resulted in a decrease in oxidation efficiency over the 15 hour duration of the experiments. Canela et al. (1998) experienced no catalyst deactivation over a 20 hour period when the influent H2S concentration was 217 ppmv. However, when the inlet concen tration was increased to 600 ppmv,

PAGE 31

31 catalyst deactivation occurred, decr easing the removal efficiency fr om 99% to 36% after 2 hours. However, it was not clear whether the decrease in efficiency was a result of the accumulation of byproducts or the reactor temper ature increasing from ambien t (about 22C) to a working temperature of 52C as a result of turning on the lamp. This increase in temperature could decrease adsorption of H2S onto the TiO2, which could subsequently decrease the oxi dation rate. Washing the catalyst with dei onized water was sufficient to remove the majority of the accumulated SO4 2(Portela et al., 2007; Canela et al., 1998). Canela et al. (1998) recovered 95% of the SO4 2that accumulated on the catalyst. Panela et al. (2007) removed the majority of the SO4 2from the catalyst after the first rinse and test ed the catalysts for their ability to oxidize H2S after multiple regenerations. They found that the wa ter wash restored most of the initial activity of the photocatalyst. Variables such as residence time, humidity, and initial H2S concentration have been found to effect the degradation efficiency of H2S using a TiO2 thin film. Canela et al. (1998) studied the effect of residence time on H2S conversion efficiency. The re sults indicated that increasing space time between 0.27 min and 2.46 min resulted in an increase in the conversion efficiency from about 35% to 95%. In the same study, they found that the reaction ex hibited a pseudo-first order dependence on H2S concentration at influent concentrations between 30 ppmv and 855 ppmv. Water vapor concentration is often an importa nt variable when studying the efficiency of photocatalytic reactions. Portela et al. (2007) found that the optim al humidity for oxidation of H2S was 20% when studyi ng air containing 35 ppmv of H2S and various RHs between 0 and 70% (at 40C). It was hypothesized that water vapor played a key role in the reaction due to hole

PAGE 32

32 trapping and hydroxyl radical formation. However, above 20% RH, the water vapor was thought to hinder photocatalytic oxidation due to its competition with H2S for adsorption sites. In a study by Kato et al. (2005), the incor poration of Ag nanopa rticles onto a TiO2 filter resulted in photocatalytic oxidation rates seven times higher than that of the un-modified TiO2. X-ray photoelectron spectroscopy showed that H2S, elemental sulfur, AgS, and SO2 were not present on the Ag-TiO2 film after use. Additionally, no byproducts were identified in the effluent gas stream. SO4 2-, which was trapped on the photocatalyst, was the only byproduct present. However, the efficiency of H2S removal did not degrade over th e duration of the experiment (approximately 9 hours). It was hypothesized that the deposition of a noble metal, such as Ag, would enhance photocatalytic reaction rates by in creasing the charge sepa ration efficiency and inhibiting electron-hole recombination. In additi on, the Ag enhanced the adsorption capacity of the AgTiO2 film, which could have also led to increased oxidation ra tes. There was no measurable adsorption of H2S onto the un-modified TiO2. The adsorption of compounds onto TiO2, or other hydrous oxides, is most often attributed to hydrogen bonding. Hydrogen bonds form between two functional groups, one which se rves as a Bronsted acid and the other as a Lewis base. According to Sopyan (2007), H2S does not readily form bonds with the hydroxyl groups on TiO2 surfaces and has been shown to fo rm hydrogen bonds in only strong basic environments. In the same study, ammonia, whic h is a strong Lewis base, showed adsorption capacity ten times greater than that of H2S as a result of its ability to participate in hydrogen bonding with the TiO2 surface. Enhanced Mercury Recovery using Photocatalysis The removal of mercury via photocatalysis has been investigated for application in flue gases at coal-fired power pl ants. Under UV irradiation, TiO2 converts Hgo to HgO (as shown in Equations 2-24 and 2-25), whic h is retained on the TiO2 surface (Rodriguez et al., 2004).

PAGE 33

33 Hgo + *OH HgOH (2-24) HgOH HgO + H+ (2-25) In general, two approaches have be en studied for Hg removal using TiO2: (1) Hgo oxidation and capture on in-situ generated TiO2 particles (Wu et al., 1998 ; Lee et al., 2004; Lee et al., 2001; Rodriguez et al., 2004); and (2) synergistic ad sorption and oxidation of Hgo using STC pellets in a packed bed (Pitoniak et al., 200 3; Pitoniak et al., 2005; Li and Wu, 2006; Li and Wu, 2007). In-situ generated TiO2 particles have been create d by the injection of a TiO2 sorbent precursor into the combustor system. The precurs or injection conditions were manipulated such that agglomerated nano-sized (20-30 nm) TiO2 particles were formed. These agglomerates had a high surface area and open structure, which w ould allow effective UV irradiation and minimal resistance to mass transfer The in-situ generated TiO2 particles showed no Hg capture in the absence of UV irradiation, indicating that physical adso rption was not an e ffective pathway for removal (Lee et al., 2004). However, in the presence of UV irradiation, the TiO2 particles removed greater than 98% of the influent me rcury (Lee et al., 2001). Water vapor enhanced mercury removal at low concentrations by increasing hydroxyl radical generation on the TiO2 surface, which increased the number of ac tive sites for oxidation and capture of Hgo. At very high water vapor concentrations, it was expected that the water vapor would inhibit Hg removal due to competition for adsorption sites, which w ould reduce the number of available sites for the oxidation of Hgo (Rodriguez et al., 2004). A novel, high surface area sili ca-gel impregnated with TiO2 nanoparticles for mercury vapor control from flue gas was developed by P itoniak et al. (2003, 2005) and has been further developed by Li and Wu (2006, 2007). In this work, this material is referred to as STC. The

PAGE 34

34 STC exhibited synergistic adsorption and phot ocatalytic oxidation for enhanced mercury removal. Mercury vapor was adsorbed onto the STC surface and subsequently oxidized and retained after irradia tion with UV light. Mercury capture was achieved with continuous or intermittent UV radiation. When intermittent UV irradiation was applied, Hg adsorption increased after the periods of UV irradiation. The STC demonstrated a high capacity for merc ury (10 30 mg/g) (Pitoniak et al., 2005) and achieved high levels of Hg re moval (greater than 99%) when continuously irradiated with UV light using a 0.78 s residence time (Pitoniak et al., 2003). Studies investigating the impact on residence time of Hg removal revealed that th e removal efficiency decreased as the residence time decreased from 0.78 s to 0.16 s. It was determined that adsorption was the rate limiting factor and that mass transfer should be improved to achieve better removal (Pitoniak et al., 2003). Li and Wu (2007) determined that mercury removal using the STC followed LangmuirHinshelwood kinetics. These results suggested that the STC has a great potential for the removal of mercury from gas streams containing high levels of Hgo. However, it was also found that water vapor significantly inhibited the capture of mercury (Li and Wu, 2006; Li and Wu, 2007). The STC achieved the highest leve l of mercury removal in the absence of water vapor. Li and Wu (2007) postulated that the source of hydroxyl ra dicals in the absence of water vapor may be the silanol groups on the SiO2 surface, which may serve as a the source for hydroxyl radical production. The decrease in capture efficiency as the water vapor concentration increased was likely due to competitive adsorption (Li and Wu, 2007). In addition, water vapor resulted in the re-emission of captured mercury from the STC du e to the repellent effect of water and the

PAGE 35

35 photocatalytic reduction of HgO to Hgo, which subsequently desorbed from the STC surface. The proposed mechanism for the reduction of Hg O is show in Equation 2-26 (Li and Wu, 2006). HgO + H2O + 2eHgo + 2 OH(2-26) Although the re-emission of mercur y affected the overall capture efficiency of the STC, Li and Wu (2006) concluded that re-emission could be minimized by th e appropriate application of UV irradiation. An advantage of the STC over other adsorbents is that it can be rege nerated by rinsing with acid (Pitoniak et al., 2003). The acid wash removed the mercury from the surface of the STC by transferring it into th e acid solution. Although the majority of the captured mercury was recovered by the acid rinse, the performance of the regenerated pellets was not reported.

PAGE 36

36 CHAPTER 3 EXPERIMENTAL Synthesis of Photocatalytic Materials Silica-Titania Composite (STC) Pellets The STC pellets were prepared using an acidcatalyzed, sol-gel technique with tetraethyl orthosilicate (TEOS) as the si lica precursor. Degussa P25 TiO2 was used as the TiO2 source and was mixed into the liquid precursors before gela tion. Nitric acid and hydrofluoric acid (HF) were used to catalyze hydrolysis and condens ation reactions, thereby decreasing the time to gelation. The pore size of the STC was manipulat ed by varying the amount of HF used during synthesis. For bench-scale studies, TEOS (F isher Scientific, reagent grade) was added to a solution of deionized water and ethanol (A aper Alcohol, 200 proof) using a TEOS:water:ethanol volume ratio of 7:5:10. To catalyze the reaction, 1 N nitric acid, prepared from 15.8 N nitric acid (Fisher Scientific, certified A.C.S.), and 3%wt HF, prepared from 48%wt HF (Fisher Scientific, certified A.C.S.), were added to the solution. In order to vary the pore size of the STC, either 2, 4, or 8 mL of 3%wt HF per 220 mL of TEOS/ethanol/water solution. For every 100 mL of TEOS, 1 to 60 g of Degussa P25 TiO2 (Majemac Enterprises) were mixed into the solution. The ingredients were mixed via a magnetic stir pl ate before being transferred into 96-well assay plates, where the solution gelled as cylindrical pellet s. The assay plates were sealed and the pellets aged at room temperature for 48 hours and then at 65C for 48 hours. The pellets were then transferred to Teflon containers, whose lids had a pin-sized hole, and dried for 18 hours at 103C followed by 6 hours at 180C. Each lid had a small hole to allow the liquid expelled from the pores of the STC pellets to slowly escape as vapor during the drying process, thereby preventing collapse of

PAGE 37

37 the pores. Finally, the STC pellets were calcined at 450C for 2 hours. The resulting STC pellets were about 3 mm in diameter and 5 mm in length. In order to produce a sufficient quantity of ST C pellets for the pilot and full-scale studies, the bench-scale synthesis met hod was modified to increase production efficiency while producing composites with similar characterist ics (i.e., surface area and pore size). TEOS (Silbond Condensed) was added to water, ethano l (Spectrum Chemicals), 1 N nitric acid, prepared from 15.8 N nitric acid (Fishe r Scientific, certified A.C.S.), and 3%w HF, prepared from 48%w HF (Fisher Scientific, certified A.C. S.). A known mass of Degussa P25 TiO2 (Majemac Enterprises) was mixed into the solu tion based on a ratio of 4 g of TiO2 per 100 mL of TEOS. The ingredients were stirred us ing a paddle mixer and then tran sferred to molds, which were made from 5.1 cm thick polyethylene sheets drille d with 0.8 cm diameter holes. Each mold was approximately 40.6 cm by 61.0 cm and containe d 2,750 holes. The molds were filled by pouring the liquid sol into the molds, which were seal ed on the bottom and top with sheets of solid polyethylene. The gels were aged at 65C for 48 hours. The lids were then loosened and the pellets dried in the molds at 103C. The pellets were removed from the molds, transferred to Pyrex containers, and then heated to 180C. After aging and dr ying, pellets were approximately 3 mm in diameter and 20 mm in length. TiO2-Coated Activated Carbon (AC) TiO2-coated AC was synthesized by coating gr anular BioNuchar120 (MeadWestvaco) with Degussa P25 TiO2 via a boil deposition method. A TiO2 slurry was made by adding 3 g of Degussa P25 TiO2 to 200 mL of DI water. Next, 30 g of AC were added to the slurry and heated on a hot plate until all of the wa ter evaporated, leaving the TiO2 coated onto the outer surface of the AC. The actual quantity of TiO2 deposited on the AC was determined by taking the difference of the measured ash content of the as-received and TiO2-coated AC. To determine the

PAGE 38

38 ash content of an AC sample, a bout 1 g of dry material was h eated to 550C for 24 hours to remove the carbonaceous portion of the AC. The measured TiO2 loading on the AC was 5.0 0.4 %wt (error given as the standard deviat ion of triplicate measurements). TiO2-Coated Glass Spheres Solid glass spheres (5 mm diameter) were coated with a TiO2 slurry (20%wt Degussa P25 TiO2 dispersed in water) and then dried at 110C. After drying, excess TiO2 was separated from the TiO2-coated glass spheres by gently shaking th e beads on a sieve with 4 mm openings. The mass of the spheres was measured before and after the TiO2 coating was applied. The resulting mass of TiO2 on each glass sphere was 0.925 mg, which equates to 11.75 g of TiO2 per m2 of glass spheres. The BET surface area of the TiO2 (as measured by a Quantachrome NOVA 2200e, Boynton Beach, FL) was approximately 50 m2/g. Characterization of Phot ocatalytic Materials The STC pellets and TiO2-coated AC were analyzed for surface area, total pore volume, and average pore size using a Quantachrome NOVA 2200e (Boynton Beach, FL). A Quantachrome Autosorb was used for nitrogen adsorption/desorption isotherms. The samples were vacuum outgassed at 180C for 24 hours. The surface area was determined using the Brunauer-Emmett-Teller (BET) model with the nitr ogen adsorption data (P/Po = 0.1 to 0.3). The total pore volume was calculated based on nitrogen adsorption at P/Po = 0.995. The average pore size was calculated using Equation 3-1, a ssuming non-intersecting, cylindrical pores: d = 4*Vp/S (3-1) where d is the average pore diameter, S is the surface area, and Vp is the total pore volume. For the analysis of pore size dist ribution, the desorption isotherm was analyzed using the Barrett, Joyner, and Halenda (BJH) method.

PAGE 39

39 Bench-scale Reactor for Methanol and H2S Removal Studies The adsorption and photocatalytic oxidation of HAPs in a simulated HVLC gas emitted from pulp and paper mills were tested using a bench-scale annular reactor. The reactor had an 8 mm annulus and contained an eight-watt UV bulb (Spectronics Corporation) with a peak wavelength of either 365 nm (UVA) or 254 nm (UVC). The UV bulb was surrounded by a quartz tube, which had an outside diamet er of 25 mm. A schematic of the reactor is shown in Figure 3a. UV intensity measurements inside of the reac tor were taken using chemical (potassium ferrioxalate) actinometry as de scribed by Murov et al. (1993). A known quantity of photocatal ytic material (STC, TiO2-coated AC, or TiO2-coated glass spheres) varying between 30 and 90 cm3 (bulk volume) was packed into the annulus of the reactor. Where specified, an annular reacto r with a 25 mm annulus was used for preliminary scale-up studies. For these studies, 132 cm3 of photocatalytic material were packed into the annulus. A schematic of the 25 mm annulus reactor is shown in Figure 3b. The materials were tested for adsorption cap acity in the dark and for simultaneous adsorption and photocatalytic activit y when irradiated with UV light The reactor was kept in the ambient atmosphere and, when irradiated with UV light, the temperature of the packed bed rose to about 50C. In order to achieve this temp erature during adsorption st udies, the reactor was wrapped with heat tape and controlled using a Variac variable vo ltage transformer. Compressed air containing 1000 ppmv of methanol or H2S was diluted with air to obtain an influent gas stream containing 50 ppmv of methanol and/or H2S and a RH less than 0.22%. For the experiments requiring a high humidity, the dilution air was passed through a water bubbler to obtain an influent gas stream with an RH of about 95% at 23C. The gas stream flowed continuously through the reactor in a single pass configuration. A schematic of the reactor set-up is shown in Figure 3-1. Initial studies performed with an empty reactor showed no photolysis of

PAGE 40

40 methanol or H2S in the presence of UVA or UVC light. Similarly, adsorption of methanol and H2S to the reactor and its appurtenances was negligible. Various experiments were conducted to study the effects of space time and face velocity on methanol removal using STC pellets packed in the 8 mm reactor. Space time ( ) is the time required to process one bed volume of gas in an empty reactor and was calculated using Equation 3-2: = V/Q (3-2) where Q is the gas flow rate and V is the reactor volume occupied by the packed bed. The face velocity, or superficial velocity (v ), was calculated using Equation 3-3: v = Q/A (3-3) where A is the cross-sectional area of the packed bed. Space times were varied between 1.1 s and 4.3 s in the 8 mm reactor by varying the gas flow rate (0.42 1.68 L/min) passing through a 30 cm3 packed bed. The effect of face velocity on methanol removal was studied by flowing 0.42, 0.84, and 1.26 L/min of gas through STC bed volumes of 30, 60, and 90 cm3, respectively, to achieve a cons tant space time of 4.3 s and face velocities of 0.093 m/s, 0.19 m/s, and 0.28 m/s. Pilot Reactor for Methanol Removal A 40 ACFM pilot reactor was designed and fabricated based on optimization studies conducted at the bench-scale. A water/methanol mixture was vaporized and injected into the gas stream so that the influent meth anol concentration was about 50 ppmv and RH between 95 and 99%. The static pressure and velocity pressure were monitored before the inlet to the pilot reactor. Thermocouples were used to measure inlet and outlet temperatures.

PAGE 41

41 A UVP radiometer (Upland, CA) was used to measure the UV intensity through an observation window on the side of the reactor. Initial studies performed with an empty reactor showed no photolysis of methanol and that ad sorption of methanol to the reactor and its appurtenances was negligible. Analysis of Methanol and Oxidation Byproducts Both influent and effluent methanol concen trations for benchand pilot-scale studies were tested using the National Council for Air and Stream Improveme nt (NCASI) Chilled Impinger Method (NCASI, 1995). Methanol co ncentrations were quantified using a PerkinElmer Clarus 500 GC/FID (Wellesly, MA) with a 50 m x 320 m polyethylene glycol column. A 1 L volume of sample was used for analysis The injector temperature was set to 110C. The oven temperature was initially 40C (1 min hold) and was ramped to 54C at 2C/min (2 min hold) and then to 220C at 21 C/min (5 min hold). The minimum detection limit (MDL) for methanol in the gas phase was 0.6 ppmv. Effluent formaldehyde concentrations were quantified colorimetrically using a Hach DR/4000U spectrophotometer (Loveland, CO) as described in the NCASI Chilled Impinger Meth od (NCASI, 1995). The MDL for formaldehyde was 0.014 ppmv. In order to quantify to tal byproduct formation, to tal organic carbon (TOC) concentrations of the impinger samples were determined using a Tekmar Dohrmann Apollo 9000 TOC analyzer (Mason, Ohio). In addition to impinger measurements, real-time influent and effluent measurements were taken during the pilot studies using a ThermoElectron TVA-1000B portable FID detector (Waltham, MA). Analysis of H2S and Oxidation Byproducts Air-phase concentrations of H2S were measured by passing 100 mL of gas through precalibrated, direct-read Nextteq ga s detection tubes (0.25 120 ppmv). Air-phase concentrations

PAGE 42

42 of SO2 were measured using a Varian 2100T GC/M S (Palo Alto, CA) equipped with a Supel-Q Plot, 30 m x 0.32 mm column. Sample s were collected in 1 L Tedlar bags and stored in the dark for no longer than 24 hours. A 75 um Carboxen-PDMS SPME fiber was injected into the Tedlar bag through a silicon septum for 10 min. The fiber was injected in to the GC-MS for 5 min at an injection temperature of 200 C. The oven temper ature was initially 45C and ramped to 250C a rate of 25C/min after an initial 0.75 min hold. The instrument detection limit for SO2 was 1 ppmv. Sulfate loading on the STC pellets was measured by soaking 30 cm3 of pellets in 500 mL of nanopure water (18.2 M -cm) with gentle stirring using a magnetic stirrer for 24 hours. The water was filtered using a 0.45 m vacuum filter. A 10 mL sample of filtered water was analyzed for sulfate using the EPA Method 9038 with a Hach DR/4000U spectrophotometer (Loveland, CO). The MDL for sulfate measurements was 1 mg/L. Mercury Analysis Influent and effluent mercury concentrations for pilot and fu ll-scale mercury removal units were determined using EPA Method 101. The MDL for mercury corresponded to an air-phase concentration of 0.1 g/ft3. For pilot-scale studies, STC pellets were re generated ex-situ to remove the mercury. The pellets were removed from the r eactor, placed in a bath of 37%wt HCl and mixed gently for 1 hour. The pellets were then rinsed with water and placed in an oven at about 70C for 12 hours. The mercury concentration of the pellets was determined before and after regeneration by digesting 1 g of pellets in 50 mL of aqua re gia solution. The mercury concentration of this solution was measured using a Hydra AA Spect rophotometer (Leeman Labs), which has an

PAGE 43

43 MDL of 0.2 g/L. The moisture content of the pellets after heat treatment at 70C was about 60%, as determined by gravimetric analysis. For full-scale studies, regenerati on was performed in-situ by so aking the pellets inside of the reactor with 37%wt HCl for 30 min. After draining the me rcury-laden acid from the reactor, the pellets were soaked with water for 30 min and the water was subsequently drained. Colburn j-factor The Colburn j-factor (jd) is a dimensionless paramete r for mass transfer based on Reynolds number (Re) and the shap e, size and packing characterist ics of the STC pellets. This parameter was used to assess the effects of mass transfer on the performance of the pilot reactor and as a design parameter for the full-scale reactor s. The j-factor was calc ulated using Equations 3-4 through 3-7, where is the shape factor (0.91 for cylindr ical pellets), v is the superficial velocity of the gas, a is the surface ar ea per volume of the packing media, dp is the particle diameter (see equation 3-6 for non-spherical particles), is the bed porosity, and and are the density and viscosity of the gas (Knudsen et al., 1999). Re = v* /( *a) (3-4) a = 6(1)/dp (3-5) dp = 0.567 *sqrt (part. surf. area) (3-6) jd = 0.91* *Re-0.51 (3-7)

PAGE 44

44 Figure 3-1. Bench-scale reactor set-up used for methanol/H2S adsorption and photocatalytic oxidation studies. Figure 3-2. Reactor drawings. A) 8 mm annulus reactor. B) 25 mm annulus reactor.

PAGE 45

45 CHAPTER 4 CHARACTERIZATION OF STC STC pellets with three different average pore sizes and varying TiO2 loadings were synthesized by varying the quantity of HF and TiO2 added to the liquid precursors before gelation. BET surface area, total pore volume, a nd average hydraulic pore size are shown in Table 4-1. The error represents the standard devia tion of grab samples taken from at least three batches of pellets. The STC pellets were labeled according to their ap proximate hydraulic pore size and TiO2 loading (% mass TiO2 per 100 mL of TEOS). Based on a mass balance of the STC synthesis from raw ingredients to the final product (i.e., drie d pellets), the estimated TiO2 mass loadings for STC labeled as 0, 1, 4, 12, and 60% were 0, 3, 12, 30, and 70%w, respectively. The surface area and total pore vo lume of the STC were dependent on the type of STC. The various STC labeled as were synthesized with a c onstant HF concentration and varying concentrations of TiO2. The addition of the non-porous TiO2 (ca. 50 m2/g) to the porous silica gel resulted in an overall decrease in su rface area and pore volume of the composite. The HF concentration also had an effect on both surface area and pore volume. For the STC synthesized with 12% TiO2 and varying concentrations of HF (50 12%, 120 12%, 260 12%), an increase in HF resulted in a decrease in surface area and an increase in pore volume due to the widening of the pores. The data do not suggest that significant po re blockage occurred as a result of TiO2 addition. The change in specific su rface area and pore volume was due to the addition of the non-porous TiO2 to a porous silica gel, creatin g a composite material with specific surface area and pore volume values represen tative of a mixture of the two materials. To illustrate this, Figure 4-1 shows the actual and expected specific surf ace area values of the STCs. The expected values were calculated based on the mass percentage of silica in the composite multiplied by the surface area of silica gel synthesized with the same HF

PAGE 46

46 concentration and no TiO2 plus the mass percentage of TiO2 multiplied by the surface area of TiO2 (50 m2/g). As shown in Figure 4-1, the measured specific surface areas of the STC are similar to the expected values. The error bars shown for the measured values were based on the measured error between multiple batches, as shown in Table 4-1. The error bars shown for the expected values were based on the error associated with the measured value of the silica gel synthesized with the same HF concentration and no TiO2. The nitrogen adsorption/desor ption isotherms for the STC are shown in Figure 4-2. All of the STC exhibit a Type IV isotherm with H1 and H2 hysteresis loops, which is characteristic of mesoporous materials. The 50 STC mainly exhibited H2 hysteresis, while the 120 and 260 STC exhibited H1 hysteresis. Although the effect of various factors on hysteresis is not fully understood, hysteresis shapes have been associated with specific por e structures. The H1 hysteresis loop, which was exhibited by the 120 and 260 STC, indicates that the material is comprised of uniform spheres arranged in a fairly regu lar array with a narrow distribu tion of pore sizes (Sing et al., 1984). H2 hysteresis, which was exhibited by the 50 STC, has been attributed to pores with an ink bottle shape (i.e., narrow necks and wide bodies) (Sing et al., 1984). However, poorly defined pore shape or distributi on in a disordered material, which results in pore blocking and percolation, could also cause the H2 loop (Thommes, 2004). Although th e 50 STC generally exhibited H2 hysteresis, the 50 60% had a hystere sis shape more representative of H1. The 50 60% hysteresis loop narrowed at a relative pressu re of about 0.8, rather than 0.5 to 0.6 for the 50 STC of lower TiO2 loadings (0 12%). This suggests that the 60% TiO2 loading may increase the size of the pore neck, allowing the pore body to empty at highe r relative pressures, or result in a less extensive pore network, as evidenced by the de crease in total pore volume.

PAGE 47

47 The presence of approximately uniform sphe rical primary particle s was confirmed by SEM images taken by Byrne et al. (2008). The SEM imag es showed that the s pherical particles were present in all of the STC, regardless of pore size or TiO2 loading. The size of the primary particles was smallest for the 260 STC and grew larger for STC synthesized with lower HF concentrations. Thus, the 50 STC were comprise d of the largest primary particles. This was attributed to the shorter gelation times for STC w ith higher HF concentrations, which resulted in a shorter time for particle growth and, hence, a smaller primary particle size. The gelation time was dependent on the HF concentration used dur ing synthesis since it catalyzes hydrolysis and condensation reactions. The 50 12% STC gelle d after about 12 hours while the 120 12% and 260 12% gelled after approximately 2.5 hours and 10 minutes, respectively. The SEM images also showed that for the 260 12% a nd 260 60% STC, the change in TiO2 concentration did not noticeably change primary particle size or morphology. The pore size distribution of th e STC is important since it ca n affect the diffusivity of methanol and oxidation byproducts into and out of the STC (Sa tterfield, 1970). The pore size distributions of STC synthesi zed with various HF and TiO2 concentrations are shown as the differential pore volume as a func tion of pore diameter in Figur e 4-3. The hydraulic pore sizes shown in Table 4-1 are sometimes different than the average pore sizes shown in Figure 4-3. If the pores were truly cylindrical in shape, which is the assumption made for the calculation of the hydraulic pore size, the average pore size associated with the desorption isotherm would equal the hydraulic pore size. The dispar ity between the desorption pore size and the hydraulic pore size can be attributed to the actual shape of the pores and/or network effects. The pore size distributions for the 50 12%, 120 12%, and 260 12% STC, as shown in Figure 4-3a, were unimodal and the area of the peak was relative to the total pore volume of

PAGE 48

48 the STC. The 50 12% STC had the narrowest range of pore sizes, with about 95% of the pore volume resulting from pores between 19 and 56 As the pore size of the STC increased due to the increased quantity of HF added to the liquid sol during s ynthesis, the total pore volume increased and the pore size distribution became broader. For the 120 12% STC, 95% of the pore volume resulted from pores with diamet ers between 49 and 172 The 260 12% STC had the broadest pore size distri bution, with 94% of the pore volu me resulting from pores with diameters between 155 and 336 Pore size distributions for 50 STC with varying TiO2 loadings (0-60% ) are shown in Figurer 4-3b. For TiO2 loadings between 0 and 12%, differences between the pore size distributions were mainly a result of the decreas e in pore volume associated with increased TiO2 loadings. The distributions for these STC were unimodal and the shapes of the distribution curves were similar. The 50 60% STC had a mu ch broader distribution and drastically smaller differential pore volume at the 50 pore diamet er. The pore volume at 50 was approximately 0.005 cc//g compared to 0.037 cc//g for the 50 4% STC. Although the peak differential pore volume remained at about 50 pore sizes between approximately 20 and 250 contributed to the total pore volume for the 50 60% STC. The 50 60% STC was comprised mostly of TiO2 (i.e., about 70%wt of the dried product was TiO2). Thus, the amount of shrinkage that occurr ed during the aging and drying process was reduced. The total volume of a dried 50 60% STC pellet was about 61% less than its volume immediately after gelation. The volume of a 50 12% STC pellet, for example, was reduced by 74%. This indicates that the syneresis of the SiO2, which results in the formation of additional SiO-Si linkages resulting in the shrinkage of th e STC during aging and drying, was inhibited. This may have resulted in the formation of a broa der pore size distri bution by disallowing the

PAGE 49

49 shrinking of the silica network and formation of additional linkages to create the smaller pores (i.e., those around 50 in diameter). Table 4-1. BET surface area, total pore volume, and calculated pore size for the STC synthesized with varying concentrations of HF and TiO2. STC BET surface area (m2/g) Total pore volume (cc/g) Pore size () 50 0% 662 54 0.86 0.03 52 3 50 1% 636 27 0.88 0.03 55 1 50 4% 617 19 0.76 0.04 49 4 50 12% 532 43 0.60 0.06 46 4 50 60% 297 24 0.48 0.05 65 5 120 12% 290 23 0.69 0.11 123 7 260 12% 154 18 0.99 0.10 259 11 0 100 200 300 400 500 600 700 800 50A 0%50A 1%50A 4%50A 12%50A 60%120A 12%260A 12%BET Surface Area (m2/g) Measured Ex p ected Figure 4-1. Measured and expected surface area data for STC.

PAGE 50

50 0 100 200 300 400 500 600 700 800 00.20.40.60.81P/PoVolume (cm3/g *103) 50A 12% 120A 12% 260A 12% 0 100 200 300 400 500 600 700 00.20.40.60.81P/PoVolume (cm3/g *103) 50A 0% 50A 4% 50A 12% 50A 60% Figure 4-2. Nitrogen adsorption/de sorption isotherms. A) STC with varying pore sizes and constant TiO2 loading (12%). B) 50 STC with varying TiO2 loadings (0-60%). A B

PAGE 51

51 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0100200300400500600 Pore Diameter (Angstroms)Dv(d) (cc/Angstrom/g) 50A 12% 120A 12% 260A 12% 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 050100150200250300 Pore Diameter (Angstroms)Dv(d) (cc/Angstrom/g) 50A 0% 50A 1% 50A 4% 50A 12% 50A 60% Figure 4-3. Pore size distribu tions. A) STC with varying pore sizes and constant TiO2 loading (12%). B) 50 STC with varying TiO2 loadings (0-60%). B A

PAGE 52

52 CHAPTER 5 OPTIMIZATION OF METHANOL DEGRADAT ION USING STC PELLETS IN A BENCHSCALE REACTOR The use of STC pellets for VOC abatement is not widely studied since photocatalytic reactors designed to treat VOCs typically employ a TiO2 or SiO2-TiO2 thin film. The objective of the work presented in this chapter was to inves tigate the use of the STC pellets for the removal and degradation of methanol from a high humid ity air stream, specifically to determine the effects of STC pore size, TiO2 loading, face velocity and space time on methanol degradation. The conditions for the experiments described in this chapter are summarized in Table 5-1. An influent methanol concentration of 50 ppmv was used for all studies. This concentration was picked based on discussions with industry. Stud y No. 1 represents the baseline conditions. The space time was manipulated in Study No. 2, 3, and 4 by varying the flow rate of gas containing 50 ppmv of methanol with an influent relative humidity of about 95% (at 23C) through the reactor packed with 30 cm3 of pellets. For space times of 8.6 s, 2.1 s, and 1.1 s, the total flow rate through the reactor was 0.21, 0.84, and 1.68 L/min, respectively. Since the flow rate was adjusted to achieve the various space times, th e face velocity, methanol loading rate (mg of methanol entering the reactor per minute) and wate r vapor loading rate also varied with space time, as shown in Table 5-1. Study No. 1, 5, and 6 we re performed to assess the effect of varying face velocity at a constant 4.3 s space time. The bed depth was adjusted to achieve the constant space time at the varying face velocities. Flow rates of 0.42, 0.84, and 1.26 L/min containing 50 ppmv of methanol were passed through packed beds with volumes of 30 cm3, 60 cm3, and 90 cm3, respectively. This resulted in a constant 4.3 s space time and face velocities of 0.093 m/s, 0.19 m/s, and 0.28 m/s. All error ba rs shown in this chapter repres ent the standard deviation of at least triplicate measuremen ts, unless otherwise stated.

PAGE 53

53 Adsorption STC pellets of varying pore sizes (50 120 260 ) and constant TiO2 loading (12%) were tested in the dark to determine methanol adsorption capacity using conditions for Study No. 1. A 12% TiO2 loading was chosen since previous st udies for the degradation of organic compounds from water have found that this TiO2 loading was optimal (Ludwig et al., 2008; Londeree, 2002). The temperature of the reacto r was kept at 50C to simulate the reactor temperature when irradiated with UV light. As shown in Figure 5-1, the time required to reach exhaustion was dependent on the pore size of the ST C pellets, with the sma llest pore size (i.e., 50 ) having the greatest methanol adsorption capacity. The total amount of me thanol adsorbed by the various STC, as determined by integration of the adsorption breakth rough curves, indicated that methanol adsorption occurred via monolayer coverage versus pore filling since the adsorption trends correlated with the BET surface area data, but not with the total pore volume. Simultaneous Adsorption and Oxidation of Methanol The photocatalytic oxidation of methanol was tested by repeating the experiments described above in the presence of UVA light. Figure 5-2 shows the methanol removal capabilities of the 50 120 and 260 STC pelle ts. Breakthrough of meth anol (i.e., effluent concentration greater than the detection limit of 0.6 ppmv or C/Co = 0.012) occurred for all STC. The time to breakthrough for the 50 pellets occu rred after about 180 minutes in the presence of UV light, whereas in the dark th e initial breakthrough occurred immediately. For the 120 and 260 STC, breakthrough occurred during the first sample period. However, the effluent concentration was lower in the presence of UV light than in the dark. For example, the normalized effluent concentration (C/Co) of the first sample from the reactor packed with 260 STC was about 0.65 in the dark and 0.07 when il luminated. Thus, it does not appear that a minimum surface coverage of water vapor or meth anol was required for degradation to occur.

PAGE 54

54 After a period of time (e.g., approximately 400 min for the 50 STC pellets), the system appeared to reach a pseudo-steady state (i.e., the effluent concentration was relatively constant) such that the adsorption rate was equal to the oxidation rate. The removal efficiency of the 120 and 260 pellets at steady state was similar (about 80%). The 50 pellets removed greater than 90% of the methanol at steady state. The enhanced performance of the 50 pellets shown in Figure 5-2a is a result of the high internal surface area of these pellets. Although the UV probably does not reach the very center of each STC pellet due to the opacity of the TiO2, these results suggest that photons do penetrate past the external surfaces of the pellets; otherwise, the performance of the pellets at steady state would be similar regardless of internal surface area. To assess whether the system in Figure 5-2a was truly at steady stat e, an extended study was conducted over a period of four days using th e 50 pellets. This study (results shown in Figure 5-2b) confirmed that the system was at steady state, achieving approximately 90% methanol removal for the duration of the experiment. Formation of Photocatalytic Byproducts Formaldehyde, a byproduct of the photocatalytic oxidation of me thanol (Tsuru et al., 2003; Peral et al., 1997), was identified in the effluent of all studies conducte d in the presence of UV light. Formaldehyde was not detected in the influent gas. The effluent formaldehyde concentration was dependent on the pore size of the STC pellets, as shown in Figure 5-3. The effluent formaldehyde concentration increased over time until a steady state was reached. The effluent formaldehyde concentration produced at steady state by the 260 120 and 50 pellets was approximately 7, 4, and 2 ppmv, respectively. STC pellets with a higher surface area possessed a greater number of activ e sites for adsorption and subsequent photocatalytic oxidation to occur. In addition, diffusion of formaldehyde out of the pellets may have depended on the pore

PAGE 55

55 size of the composite (i.e., a composite with a smaller pore size had greater resistance to pore diffusion) (Chang et al., 2000). Restricted diffusion may have retained byproducts for further oxidation. The effects of mass transfer will be discussed in further detail later in this chapter. In addition to testing effluent streams for methanol and formaldehyde, TOC analysis was performed in order to identify additional organic byproducts. Fo r all studies, TOC analysis proved that other organic byproducts were not measurable regardless of the test parameters. For example, when 50 12% STC pellets were irradiated with UVA light, methanol and formaldehyde were released in the effluent at a rate of about 0.003 mg of methanol (as carbon) per min and about 0.004 mg of formaldehyde (as car bon) per min at steady state. The results of TOC analysis showed the total rate of carbon released in the effluent was 0.007 mg/min. This mass balance shows that the TOC present in the effluent (0.007 mg C/min) was accounted for by the methanol (0.004 mg C/min) and formaldehyde (0.003 mg C/min). In addition, the absence of methyl formate, which is another byproduct that has been identified in the effluent of reactor systems that photooxidize methanol (Tsuru et al ., 2003; Sadeghi et al., 199 6; Arana et al., 2004), was confirmed by GC analysis. Formic acid is a known intermediate byproduc t in the oxidation from formaldehyde to carbon dioxide and water. The lack of the presence of measurable amounts of formic acid in the effluent can be attributed to the following reasons: (1) formic acid formed from the photocatalytic oxidation of methanol is strongly adsorbed to the TiO2 surface (Lichtin et al., 1994) and (2) only one electron hole is necessary for the total degradation of formic acid, which, according to other literature values, results in a high apparent quantum yield (0.45) compared to other organic compounds (0.06-0.001) (Dijkstra et al., 2002). Sin ce formic acid is degraded

PAGE 56

56 directly to carbon dioxide and water, the balance of the effl uent carbon should be present as carbon dioxide (Dijkstra et al., 2002). Effect of Space Time on Methanol Degradation The effect of space time on methanol remova l using the 50 12% STC was tested using experimental conditions 2, 3, and 4, which resulted in space times of 8.6 s, 2.1 s, and 1.1 s. The space time was controlled by adjusting the air flow rate, although the influent methanol concentration (50 ppmv) and relative humidity (95%) remained the same. Thus, the face velocity of the air entering the packed be d and methanol and water vapor lo ading rate (mg/min) increased as the space time decreased. The normalized efflue nt methanol concentrati on (C/Co) is shown in Figure 5-4 for the duration of each study. Initially, the effluent methanol concentration was low, due to adsorption onto the STC. For all space times, the effluent methanol concentration increased over time until steady state removal was achieved. The effluent methanol concentration at steady state was dependent on the space time. The shortest space time (1.1 s) had the highest normalized effluent methanol concentration of about 0.52. The normalized effluent methanol concentration for the 2.1 s and 8.6 s space times were about 0.38 and 0.025, respectively. The error bars show n in Figure 5-4 represent the range of duplicate tests. Error bars are not shown for the 1.1 s residence time si nce the experiment was not replicated. Formaldehyde, which is an intermediate byprodu ct of methanol oxidation, was identified in the effluent for all studies. Figure 5-5 show s the effluent formaldehyde concentrations for the duration of the experiments for the various space times. The effluent formaldehyde concentrations increased from time zero until reaching a steady state concentration. This concentration was also dependent on the space ti me. The effluent formaldehyde concentrations for the 8.6 s, 2.1 s, and 1.1 s residence times were about 0.3 ppmv, 3 ppmv, and 8 ppmv, respectively. No other organic byproducts were identified in the effluent. Since one mole of

PAGE 57

57 methanol is oxidized to one mole of formal dehyde and no other intermediate byproducts were identified, the difference between the influent molar flux and the total effluent molar flux of methanol and formaldehyde represents the mineralization rate of methanol. The 120 12% and 260 12% STC were also te sted for methanol removal using space times of 1.1, 2.1, and 8.6 s (Study Nos. 2-4). For all STC, the effluent methanol and formaldehyde concentrations at st eady state are shown in Figure 5-6 for the various space times. For both the 120 and 260 STC, the effluent me thanol and formaldehyde concentrations were dependent on the space time. The effluent form aldehyde and methanol concentrations for the 120 were less than that of the 260 STC for all sp ace times studied. At a 1.1 s space time, the 260 STC showed no mineralization of methanol, only oxidation of methanol to formaldehyde, which then desorbed into the effluent gas stream rather than remain at the reaction site for subsequent oxidation. Thus, at the 1.1 s space time, the conversion of methanol to formaldehyde was the dominant reaction mechan ism. For space times greater than 1.1 s, this mechanism no longer dominated since both methanol and formal dehyde were oxidized. The longer space times were achieved by lowering the gas flow rates, whic h resulted in lower methanol and water vapor loading rates (as shown in Table 5-1). The lower loading rates presumably decreased the competition for adsorption/reaction sites and the higher space times allowed more time for the subsequent adsorption and oxidation of desorbed formaldehyde to occur, thus allowing the oxidation of formaldehyde to proceed. The greatest difference in performance betw een the various STC was observed at the shortest residence time studied (1.1 s). The total normalized effluent concentration (1-XA), which accounts for the concentration of both methanol and formaldehyde in the effluent, was 0.68 for 50 0.91 for 120 and 1.0 for 260 As the space time increased, the performance of the

PAGE 58

58 various STC converged so that the performance was more similar at the 8.6 s space time (1-XA was 0.035 for 50 0.045 for 120 and 0.049 for 260 ). The difference in performance between the various pore sizes was likely a result of the surface area of the STC, with the higher surface area STC (i.e., 50 ) resulting in better performance due to the greater number of available adsorption and reaction sites. This was most pronounced at the 1.1 s space time, where the ratio of influent methanol loading to su rface area was the greatest and the most significant difference in performance was observed. In addition to surface area effects, performance may also be affected by mass transfer, which is dire ctly effected by the pore size distribution of the STC. For example, the STC with a small pore size may resist the transfer of methanol into the STC or constrain byproducts from leaving the pellet, thereby allowing for complete oxidation. This hypothesis is discussed further below. Mass Transfer The kinetics of the system can be limited by the effects of external and internal mass transfer of methanol into the STC and byproducts out of the STC. In orde r to assess the effects of external mass transfer on the methanol oxidation rate and the efflue nt byproduct concentration (i.e., formaldehyde concentration), Study Nos. 57 were conducted with 50 12% STC. The 50 STC was chosen for these studies since it had the smallest pore size and narrowest pore size distribution. Thus, the 50 STC would experience the greatest re sistance to mass transfer. For these studies, the face velocity was varied between 0.093 and 0.28 m/s while the space time was held constant at 4.3 s by changing the volume of the packed bed. The steady state methanol and formaldehyde concentrations in the effluent are shown in Figure 5-7. The effluent methanol concentration decreased as a result of an in crease in face velocity while the effluent formaldehyde concentration was similar for all ca ses. This methanol removal trend was likely a result of the decrease in resist ance to mass transfer from the gas stream to the media as the face

PAGE 59

59 velocity increased (Satterfie ld, 1970). Compared to the infl uent concentration (50 ppmv), the difference in methanol re moval from about 5.8 ppmv for the 0.093 m/s face velocity to 4.2 ppmv for the 0.28 m/s face velocity was about 1.6 ppmv. Since this difference represents only 3% of the influent concentration, the kinetic analysis presented later in this paper will neglect the effects of external mass transfer. Using the experimental methanol mineralizati on rates observed at steady state, the Weisz modulus (Mw), or modified Thiele modulus, was calculated for Study Nos. 1-4, where space time was varied. The Weisz modulus is used to estimate kinetic limitations as a result of resistance to pore diffusion using the experimentally-determined rate of methanol removal at steady state. Modulus values less than 0.15 indicate that ther e are no kinetic limitations as a result of pore diffusion. The expression for the Weisz Modulus is shown in Equation 5-1(Levenspiel, 1999). Mw = L2*r/(Co*De) (5-1) where L is the characteristic length (radius/2 for cylindrical pellets); r is the mean rate of methanol oxidation per unit volume of STC; and De is the effective diffusion coefficient of methanol within the STC. The effective diffusion coefficient (De) is a function of the diffusion coefficient of methanol (D), catal yst grain porosity ( c) and tortuosity of the STC ( c), as shown in Equation 52. The diffusion coefficient of methanol in air is about 15x10-6 and the tortuosity of the pores was assumed to be 3, which is a typical value for mesoporous silica gels (Doucet et al. 2006). De = D*cc (5-2) The calculated Weisz modulus values for th e various STC are shown in Table 5-2 for Study No. 1-4, where space time was varied. For each of the space times, the Weisz modulus indicated that there were no kinetic limitations as a result of pore diffusion. Although the 50

PAGE 60

60 12% STC exhibited the most narrow pore size distribution and smallest average pore diameter of the STC, the pore size was large enough to allo w pore diffusion with very little resistance. Therefore, the rates measured in the experiment s are a result of the chemical (or intrinsic) kinetics (Doucet et al., 2006). Kinetics The rate of reaction for gas-solid phase photoc atalytic reactions can often be described by the Langmuir-Hinshelwood (L-H) model, where the rate of reaction is equal to the rate constant (k) times the surface coverage of the contaminant. Since the mineralization rate of methanol can be described by the rate equati on for a single molecular reaction and the influent concentration was low, the L-H model can be simplified to th e pseudo first-order equation shown in Equation 5-3 to describe the mineraliza tion rate of methanol (Chen et al., 1999; Zou et al., 2006). ln (1-XA) = k (5-3) where A is a constant. Although initial rate of photocatalytic degradation is typically used for the L-H model, in this study the initial rate of photocatalytic degradation could not be easily differentiated from the removal rate due to adsorp tion. In addition, the inhibitory effects of water vapor may initially vary over time as the STC surface reaches its adsorption capacity for water vapor. It should be noted that water vapor adsorp tion tests proved that capillary condensation of water vapor did not occur in the pores of any STC studied. This data is discussed in further detail in Chapter 6. At steady state conditions, the meth anol removal rate is directly related to the photocatalytic reaction rate and i nhibitory effects of water vapor can be assumed to be constant for a given system. Therefore, the reacti on rate was modeled us ing the steady state mineralization rates at the various space times studied (Study No. 1-4). The rate constant (k) was

PAGE 61

61 determined for STC of the various pore sizes by a linear regression of -ln(1-XA) versus as shown in Figure 5-8. The use of the L-H model result ed in a good fit of the data (R2 = 0.99 for all STC). The rate constant was 0.40 s-1 for all the STC pore sizes studied. The difference in conversion rates was due to a lag time before mineralization o ccurred, which was observed with the 120 and 260 STC. Oxidation of methanol to interm ediate byproducts occurred during the lag phase; however further oxidation required for mineraliz ation did not proceed. A lag phase was also described by Chu and Wong (2004) in their study of the oxidati on of alachlor in water, where the dominant mechanism during the lag phase was the oxidation of alachlo r to intermediate byproducts with no mineralization. In this study, the lag time was expressed by the constant (A), which was 0, -0.41, and -0.49 for the 50 120 and 260 STC, respectively. The lag times for the 120 and 260 STC are equivalent to the intercept of the regre ssion lines in Figure 5-8 with the abscissa, which are approximately 1.0 s for 120 and 1.2 s for 260 Thus, the L-H model is applicable to descri be the mineralization of methanol for the 120 and 260 STC at times greater than these lag times. In the case of the 50 STC, the constant A is zero, indicating that the L-H model is applicable for all space times. The zero lag time experienced by the 50 STC was likely due to its high surface area, wh ich provides more adsorp tion and reaction sites for the higher influent loadings a ssociated with the low space times. Effect of TiO2 Loading on Methanol Degradation STC with an average pore size of about 50 were tested with various TiO2 loadings (1%, 4%, 12%, and 60%) using Study No. 1 conditions. As shown in Figure 5-9, all of the STC pellets, regardless of TiO2 loading, removed similar amounts of methanol (ca. 90%) at steady state when continuously illuminate d with UVA light. The time to initial breakthrough (i.e., time

PAGE 62

62 when the effluent methanol concentration was detectable) was similar for STC pellets loaded with 1-12% TiO2. Initial breakthrough for the pellets loaded with 60% TiO2 occurred immediately due to the lower specific surface area of the composite (297 m2/g) compared to that of the STC with lower TiO2 loadings (e.g., 617 m2/g for the 50 4% STC). TiO2 loading did affect effluent formaldehyde concentration, as shown in Figure 5-10. This graph shows that a 4% TiO2 loading was optimum, resulting in steady state effluent formaldehyde concentrations below 0.5 ppmv. According to Byrne et al. (2008), the total available TiO2 surface area for the 4%, 12%, and 60% TiO2 loadings was 3.4 m2/g, 8.1 m2/g, and 13 m2/g, respectively. Therefore, th e total surface area of TiO2 in the systems containing the STC with 4%, 12%, and 60% was 51 m2, 137.7 m2, and 234 m2. Although Byrne et al. (2008) did not measure the available TiO2 surface area for the 50 1% STC, it can be assumed that the total available surface area would be lower than that for the 4% system based on the trend observed for TiO2 loadings between 4 and 60%. Thus, the tota l quantity of active sites on the 1% STC was likely less than that for the 4% STC, leading to greater overall oxidati on of methanol and its byproducts by the 4% STC. Based on the total amount of available TiO2, one may expect that th e system containing the 60% STC would achieve the grea test rate of mineralizati on (i.e., lowest formaldehyde concentration) since it had the greatest TiO2 surface area available. Ho wever, the overall surface area of the composite was 297 m2/g, which was lower than that of the 50 STC with lower TiO2 loadings (e.g., 50 4% STC had a surface area of 617 m2/g). Therefore, the higher total surface area of the 50 4% STC may have enhanced th e adsorption of the me thanol such that the available TiO2 was more efficient. In addition, increased TiO2 loadings may have resulted in a

PAGE 63

63 decrease in transparency such th at total amount of irradiated TiO2 was greater for the 4% STC compared to the 12% and 60% STC. In order to test this hypothesis, UV transp arency tests were co nducted using 50 STC pellets of the varying TiO2 loadings. The pellets were packed tightly into a 10 mm quartz cuvette and placed in a Hach DR4000 spectrophotometer. The % transmittance of the UV light through the packed bed was measured at a wavelength of 365 nm and the results are shown in Figure 511. The % transmittance of the 50 0% STC was about 8%. The 50 0% STC was comprised solely of SiO2, which is known to be highly transparent to UVA light. The low % transmittance shown in these results was likely due to the scattering of UVA light through the packed bed of pellets since th e detector of the spectrophotometer measured only the UV light transmitted in a straight path through the cuvette (i.e., UV light that was scattered rather than absorbed was not measured). No significant difference was measured between the STC of varying TiO2 loadings. The quantity of UV light tran smitted through the pellets containing TiO2 was likely a result of the light pe netrating through only the interstitial space of the pellets. Based on physical observations of the pellets, the 1% ST C appeared to be semi-transparent while the 60% STC appeared to be opaque. Therefore, it can be expected that the di stance that the light could penetrate through the STC would be different However, the 10 mm thickness used in this experiment was likely too large to be able to measure differences in transparency through the STC since the light would have to pass through 2-3 layers of pellets and scattering would occur. Thus, it is possible that the UV could penetrate fart her into the 1% STC, but still be completely attenuated or scattered over a distance of 10 mm. Thus, the hypothesi s stated above can be neither confirmed nor denied based on these data In order to evaluate the penetration of UV

PAGE 64

64 light through the STC, individua l pellets or thinner layers of STC made with varying TiO2 loadings should be measured using an integrating sphere to capture both transmitted and scattered light. Effect of UV Wavelength on Methanol Degradation The effect of UV wavelength on methanol oxidation was evaluated by using 50 12% STC irradiated with UVA and UVC lamps using Study No. 1 conditions. The initial methanol breakthrough time (i.e., the time when the effluent concentration is measurable) was about 308 28 min when the TiO2 was activated using the UVC lamp. Th is time was greater than the initial breakthrough time (135 25 min) when the UVA lamp was used. The UVA and UVC lamps performed similarly with respect to methanol removal when the reactor reached steady state, resulting in the oxidation of about 85% of the influent methanol. Figure 5-12 shows that the use of the UVC lamp resulted in lower effluent fo rmaldehyde concentrations (i.e., more complete oxidation) compared to the UVA lamp. The total degradation efficiency was about 88% for the UVA lamp and 90% for the UVC lamp. The UVC lamp enhanced reactor performance b ecause the photocatalyt ic reaction rate is proportional to the rate of generation of electrons and holes on the TiO2 surface, which is in turn proportional to the photon flux, or light intensity (Dijkstra, 2002). The intensity of the UVA and UVC light inside of the reactor system was 8.71E-5 and 1.13E-4 mEinsteins/s/cm3, as determined by potassium ferroxalate actinometr y. Thus, the intensity em itted by the UVC light was about 29% greater than that emitted by th e UVA lamp. The increase in degradation as a result of increased light intens ity can be predicted using the pseudo first order model shown in Equation 5-4. C/Co = k* (5-4)

PAGE 65

65 where k = koI for variable light intensity (I). As previously determined, the k value for Study No. 1-4 conditions when illuminated with UV light was 0.4s-1. The predicted normalized effluent concentration as a function of adsorbed light flux is shown in Figure 5-13. According to this model, the normalized effluent concentration of the system irradiated with UVC light should be approximately 0.90. This assumed that 90 % of the light emitted from the UV lamp was absorbed by the STC pellets and the remainde r passed through the packed bed, which was confirmed by actinometry measurements. The ac tual normalized effluent concentration was indeed 0.90. Effect of H2S on Methanol Degradation TRS species are often present in HVLC gases and are a major contributor to the characteristic odor associated with pulp and paper mills. The effect of H2S, which was used as the representative TRS species, on methanol re moval was studied. In addition, the removal of H2S was investigated to understand if the STC can provide a co-benefit for odor removal. Studies were conducted using 50 4% ST C and Study No. 1 conditions with 50 ppmv H2S added to the gas stream. In the dark, there was no measurable adsorption of H2S onto the STC at 50C after 10 minutes. Other studies have shown that H2S had a weak adsorption affinity for the TiO2 surface (Kato et al., 2005; Portela et al., 2007; Sopyan 2007). Kato et al. (2005) did not observe adsorption of H2S onto TiO2 coated on an alumina substrate. Portela et al. (2007) tested gases containing an influent H2S concentration of 35 ppmv and RH between 0 and 70%. They observed an initial decrease in effluent H2S concentration due to adsorption for a short period of time (i.e., less than an hour); however the extent of adsorption was not quantified. The duration of H2S adsorption was dependent on the humidity of the gas stream, since water vapor competed with H2S for adsorption sites. In this work, water vapor competition would be expected to further

PAGE 66

66 inhibit H2S adsorption since the influent RH was about 95%. Sopyan (2007) attributed the weak adsorption affinity of H2S for the TiO2 surface to the difficulty of forming hydrogen bonds between H2S and hydroxyl groups of the TiO2 surface, which is the primary pathway for adsorption onto TiO2 surfaces. Since adsorption via hydrogen bonding is also the primary pathway for adsorption onto the surface of SiO2 (Travert et al., 2002), the use of a composite material containing SiO2 and TiO2, such as the STC, would not be expected to enhance H2S adsorption. Sopyan (2007) concluded that the extent of adsorption of contaminants onto the TiO2 surface was related to their abil ity to serve as an electron donor (i.e., Lewis basic group). Ammonia, which was reported to be a strong Lewis base, showed adsorption capacities an order of magnitude greater than that of H2S, which has lower electron-donor ability and has been shown to adsorb via hydrogen bonding only in strongly basic environments (Sopyan, 2007). The STC pellets were tested for oxidation of methanol and H2S by irradiating the STC with UVC light. Note that there was no photolysis of methanol or H2S when irradiated with UVC light alone. As shown in Fi gure 5-14, photocatalytic oxidat ion of both methanol and H2S was achieved. The effluent H2S concentration was about 28.4 2.4 ppmv for the duration of the experiment, which corresponds to an oxidation efficiency of approximately 43%. Effluent methanol concentrations were initially low and increased until a steady state was reached. Initial breakthrough occurred during the first sample (a s opposed to occurring af ter about 300 minutes for the non-H2S system). The effluent methanol concentration at steady state was about 10 ppmv, which corresponds to a removal efficiency of about 80%. This corresponds to a decrease in removal efficiency from about 95% when H2S was absent from the gas stream (results shown in Chapter 6). In the presence of H2S, the decrease in methanol removal efficiency was likely due to the competition for oxidation sites on the surface of the STC. Since both the methanol and H2S

PAGE 67

67 concentrations reached a steady state, the form ation of sulfur containing byproducts did not likely contribute to the deactivation of the catalyst duri ng the experiment. Formaldehyde, an oxidation byproduct of methanol, and SO2, an oxidation byproduct of H2S, were identified in the effluent. Effluent formaldehyde concentrations were about 0.9 ppmv at steady state. This concentration was about two times greater than effluent formaldehyde concentrations at steady state when H2S was absent from the influent gas (results shown in Chapter 6). Analysis of the STC pellets after the completion of th e experiment showed that about 2.8 mg/g of SO4 2was loaded onto the surface of the pellets after approximately 3000 min. A mass balance of the reactor confirmed that SO4 2and SO2 were the only byproducts of H2S conversion. The SO2 concentration increased over the dura tion of the experiment. This may be caused by (1) the saturation of water and metha nol on the surface of the STC, which are more strongly adsorbed than the SO2, would promote desorption of the SO2 and inhibit its subsequent adsorption onto the STC for further oxidation and retention; and (2) the accumulation of SO4 2on the surface of the STC over time may inhibit the oxidation of SO2 to SO4 2such that the SO2 desorbed from the STC into the effl uent before being oxidized further. The 50 4% STC were tested for H2S removal (Co = 50 ppmv) in the absence of methanol using Study No. 1 conditions. Both H2S and SO2 were present in the effluent, as shown in Figure 5-15. The effluent H2S concentration was 27.9 3.5 ppmv for the duration of the study. This was similar to the previous study in which methanol was also present in the gas stream. In addition, the effluent SO2 concentration also followed a trend si milar to that in Figure 5-14 when methanol was present in the infl uent gas. Thus, oxidation of H2S was not significantly affected by the presence of methanol. This was likely because the competition from water vapor dominated since the influent concentrat ion of water vapor was about 28,000 ppmv while that of

PAGE 68

68 methanol was 50 ppmv. SO4 2was identified on the surface of the STC and a mass balance showed that SO2 and SO4 2were the only oxidation byproducts. Table 5-1. Summary of experimental conditions. Study No. V (cm3) Q (L/min) v (m/s) (s) Co (ppmv) Methanol Loading Rate (mg/min) Water Vapor Loading Rate (mg/min) 1 30 0.42 0.093 4.3 50 0.028 5.4 2 30 0.21 0.046 8.6 50 0.014 2.7 3 30 0.84 0.19 2.1 50 0.056 10.9 4 30 1.68 0.37 1.1 50 0.11 22.8 5 60 0.84 0.19 4.3 50 0.056 10.9 6 90 1.26 0.28 4.3 50 0.084 16.3 Table 5-2. Weisz modulus values for variable space time experiments. Pore size () (s) Weisz modulus (Mw) Study No. 1 50 4.3 0.058 120 4.3 0.042 260 4.3 0.036 Study No. 2 50 8.6 0.028 120 8.6 0.028 260 8.6 0.028 Study No. 3 50 2.1 0.077 120 2.1 0.035 260 2.1 0.033 Study No. 4 50 1.1 0.090 120 1.1 0.009 260 1.1 0.000

PAGE 69

69 0.0 0.2 0.4 0.6 0.8 1.0 1.2 050100150200250300350400 Time (min)C/Co 50 12% (532 m2/g) 120 12% (290 m2/g) 260A 12% (154 m2/g) Figure 5-1. Adsorption breakthrough curves for STC pellets of va rying pore sizes (50 120 and 260 ) and constant TiO2 loading (12%). Figure 5-2. Methanol removal using STC pellets A) Methanol removal using STC of varying pore sizes illuminated with UVA light. B) Extended study for 50 12% STC pellets. A 50A 12% (532 m2/g) 120A 12% (290 m2/g) 260A 12% (154 m2/g) 0.0 0.2 0.4 0.6 0.8 1.0 0 2004006008001000 Time (min)C/Co 50 A 12% 120 A 12% 260 A 12% 0.0 0.2 0.4 0.6 0.8 1.0 012345 Time (days)C/Co B A

PAGE 70

70 0 2 4 6 8 10 0300600900120015001800 Time (min)Formaldehyde (ppm v) 50 A 12% 120 A 12% 260A 12% Figure 5-3. Effluent formaldehyde concentrations from STC pellets of varying pore sizes (50 120 and 260 ) when illuminated with UVA light. 0.0 0.2 0.4 0.6 0.8 1.0 050010001500200025003000 Time (min)C/Co 1.1 s 2.1 s 8.6 s Figure 5-4. Normalized effluent methanol c oncentration for 50 12% STC illuminated with UVA light at various space times.

PAGE 71

71 0 2 4 6 8 10 050010001500200025003000 Time (min)Formaldehyde (ppm v) 1.1 s 2.1 s 8.6 s Figure 5-5. Effluent formaldehyde concentratio n for 50 12% STC illuminated with UVA light at various space times. 0 10 20 30 40 50 60 1.12.14.38.6 (s)Effluent C (ppm v) Formaldehyde Methanol Figure 5-6. Effluent methanol and formaldehyde concentrations at steady state at varying space times. A) 50 B) 120 C) 260 STC. A

PAGE 72

72 0 10 20 30 40 50 60 1.12.14.38.6 (s)Effluent C (ppm v) Formaldehyde Methanol 0 10 20 30 40 50 60 1.12.14.38.6 (s)Effluent C (ppm v) Formaldehyde Methanol Figure 5-6. Continued. C B

PAGE 73

73 0 1 2 3 4 5 6 7 8 9 0.093 m/s0.19 m/s0.28 m/s Face VelocityEffluent C (ppm v) Formaldehyde Methanol Figure 5-7. Effluent methanol and formaldehyde concentrations at steady state for variable face velocities and constant space time (4.3 s) for 50 12% STC. A Figure 5-8. Linear regression of L-H model using mineralization ra tes achieved at various space times. A) 50 12% STC. B) 120 12% STC. C) 260 12% STC. y = 0.40x R2 = 0.996 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0246810 (s)ln (1-XA) -ln(1-XA) = 0.40 R2 = 0.99

PAGE 74

74 B C Figure 5-8. Continued. 0.0 0.1 0.2 0.3 0.4 0 2004006008001000 Time (min)C/Co 1% 4% 12% 60% Figure 5-9. Effect of TiO2 loading in 50 STC on methanol removal when illuminated with UVA light. y = 0.40x 0.49 R2 = 0.99 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0246810 (s)-ln (1-XA) -ln(1-XA) = 0.40 0.49 R2 = 0.99 y = 0.40x 0.41 R2 = 0.99 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0246810 s -ln (1-XA) -ln(1-XA) = 0.40 0.41 R2 = 0.99

PAGE 75

75 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 50A 1%50A 4%50A 12%50A 60%Formaldehyde (ppm v) Figure 5-10. Effect of TiO2 loading on formaldehyde production at steady state using 50 STC illuminated with UVA light. 0 1 2 3 4 5 6 7 8 9 10 50A 0%50A 1%50A 4%50A 12%50A 60%% Transmitance Figure 5-11. Percent transmittance of UVA light through 50 STC with various TiO2 loadings.

PAGE 76

76 0.0 0.5 1.0 1.5 2.0 2.5 0 500 1000 1500 2000 Time (min)Formaldehyde (ppm v) UVA UVC Figure 5-12. Effluent formaldehyde concentra tions using 50 12% STC irradiated with UVA and UVC light. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0E+002.0E-054.0E-056.0E-058.0E-051.0E-041.2E-041.4E-04 I (mEinsteins/s/cm3)(1-XA) Figure 5-13. Predicted C/Co versus absorbed light flux for 50 12% STC, 4.3 s residence time, Co = 50 ppmv and 95% relative humidity.

PAGE 77

77 0 5 10 15 20 25 30 35 40 45 50 0500100015002000250030003500 Time (min)Concentration (ppm v) Hydrogen Sulfide Sulfur Dioxide Methanol Formaldeh y de Figure 5-14. Effluent concentrations of H2S, methanol and oxidati on byproducts from 50 4% STC illuminated with UVC light (Co methanol = 50 ppmv, Co H2S = 50 ppmv). 0 5 10 15 20 25 30 35 40 45 50 050010001500 Time (min)Concentration (ppm v) Hydrogen Sulfide Sulfur Dioxide Figure 5-15. Effluent concentrations of H2S and SO2 from 50 4% STC illuminated with UVC light (Co H2S = 50 ppmv).

PAGE 78

78 CHAPTER 6 EFFECT OF CATALYST SUPPORT ON TH E PHOTOCATALYTIC DEGRADATION OF METHANOL IN A PACKED-BED REACTOR The objective of this study was to investigate the effect of the catalyst support on methanol oxidation efficiency using STC, TiO2 coated on an opaque adsorbent material (i.e., activated carbon (AC)), and TiO2 coated on a non-adsorbent material (i.e., 5 mm glass spheres). Preliminary scale-up was performed at the benc h-scale by increasing the annulus size of the reactor from 8 mm to 25 mm in order to evaluate the efficacy of these materials for full-scale applications. In addition, the competitive effects of water vapor on metha nol adsorption capacity and oxidation efficien cy were evaluated. The STC and TiO2-coated AC were analyzed for BET surface area and total pore volume. The results of this analysis are shown in Tabl e 6-1. A comparison of the total surface area and TiO2 mass loading of the materials normalized fo r the reactor bed volume are shown in Table 62. Although the total specific su rface area of the STC was less th an half of that of the TiO2coated AC (617 m2/g versus 1424 m2/g), the surface area per reactor volume was more similar (308 m2/cm3 versus 380 m2/cm3) due to the higher bulk density of the STC. The bulk density of the STC and TiO2-coated AC was about 0.53 and 0.26 g/cm3, respectively. The TiO2-coated glass spheres had the smallest tota l surface area per bed volume (0.43 m2/cm3) since the glass spheres were non-porous and only external surface area contributed to th e total available surface area. The TiO2 mass loading per reactor volume was similar for the TiO2-coated AC and TiO2coated glass spheres (9.9 and 10.5 mg/cm3, respectively). The 50 4% STC had a total TiO2 mass loading of 26.7 mg/cm3, which was about 2.5 times greater than the TiO2-coated AC and TiO2-coated glass spheres. The TiO2 loading was calculated based on the amount of TiO2 contained in each material and doe s not indicate the amount of TiO2 that was available for

PAGE 79

79 adsorption and photocatalytic reactions. For ex ample, based on studies by Byrne et al. (2008), about 32% of the TiO2 in the 50 4% STC was inaccessible due to the formation of the SiO2 matrix around the TiO2 during synthesis. Therefore, the effective TiO2 loading per bed volume would be 18.2 mg/cm3. Based on the synthesis method for the TiO2 coated AC and glass spheres, the majority of the TiO2 should be accessible since TiO2 particles were deposited on the external surface of the materials. Methanol Adsorption and Oxidation in a Low Humidity Gas Stream The TiO2-coated materials were tested for metha nol removal via adsorption (i.e., in the dark) at room temperature and simultaneous ad sorption and oxidation (i.e ., in the presence of UVC light) using a 4.3 s residence time, an influent methanol concentration of 50 ppmv and a water vapor concentratio n of less than 65 ppmv (typically 10 ppmv), which was specified by the supplier of the compressed air used as the feed gas. A water vapor concentration of 65 ppmv is equivalent to an RH of about 0.22% at 23 C. As shown in Figure 6-1, the adsorption capacity of the TiO2-coated glass spheres was low (ca. 0.1 mg/g) and initial breakth rough (i.e., effluent concentration greater th an the MDL of 0.6 ppmv or C/Co = 0.012) was achieved during the first sample. The TiO2-doped adsorbent materials achieved ex tended periods of methanol removal (i.e., 1000 min for the AC and 3200 min for the STC) before breakthrough. The total methanol capacity of the STC (11 mg/g) was greater than that of the TiO2-coated AC (6 mg/g). The adsorption capacity of the silica gel is a resu lt of its surface chemistry, which is dominated by silanol functional groups (Si-OH) that serve as go od adsorption sites for alcohols (e.g., methanol) via hydrogen bonding (Parida et al., 2006). In the presence of UV light, the effluent methanol concentrations for the TiO2-coated glass spheres and STC were below the MDL (0.6 ppmv or C/Co = 0.012). The effluent concentration of the TiO2-coated AC reached a steady state concentration of about 2.5 ppmv (C/Co = 0.05) after about 2500 min.

PAGE 80

80 The effluent was tested for the presence of intermediate organic byproducts. As shown in Figure 6-2, TiO2-coated glass spheres and TiO2-coated AC created formaldehyde in the effluent, whereas the STC did not. According to TOC and GC analysis, other oxidation byproducts (e.g., formic acid, formates) were not identi fied in the effluent gas stream. The formaldehyde concentration for the TiO2-coated AC was low for the first 1000 min and then steadily increased until reaching a steady state c oncentration of about 6.3 ppmv after 2000 min. The production of formaldehyde as an intermediate oxidation byproduct exceeded the adsorption capacity of the AC, thus causing th e breakthrough profile shown in Figure 6-2. The effluent formaldehyde concentration for the TiO2-coated glass spheres was about 0.3 ppmv at steady state, which was achieved after about 500 min. For the ST C system, effluent methanol concentrations were below the MDL and effluent formaldehyde concentrations were less than 0.1 ppmv for the duration of the experiment. Thus, th e STC resulted in the most efficient system for the mineralization of methanol. Several factors may have attributed to the enhanced efficiency of the STC system: The STC system had twice the amount of available TiO2 surface area, which likely resulted in more active sites for oxidation leading to an increase in the overall efficiency. The STC exhibited a greater adsorption capacity for methanol compared to both the TiO2coated AC and the TiO2-coated glass spheres. This enha nced adsorption capacity would concentrate the pollutant near the photo catalyst surface, potentially increasing the oxidation rate. Studies have shown that enha nced adsorption capacity has resulted in higher oxidation rates for organic compounds (Anderson and Bard, 1997; Tsumura et al., 2002; Vohra and Tanaka, 2003; Torimoto et al., 1996). The STC may have been less affected by the low water vapor concen tration than both the TiO2-coated AC and glass spheres. It has been suggested that SiO2-TiO2 materials use hydroxyl groups at the SiO2-TiO2 interface to produce hydroxyl radicals for the degradation of organic compounds thus enhancing the efficiency of the material (Yang and Chen, 2005). Compared to the TiO2-coated AC, both the STC and TiO2-coated glass spheres were more efficient for methanol mineralization. This was like ly due to the transparency of the silica gel and

PAGE 81

81 glass spheres, which would allow for more effici ent use of the UV radiation within the reactor. AC, which is opaque, does not allo w transmission of UV light into or through the carbon matrix. For the TiO2-coated AC system at steady state, the effluent concentration of formaldehyde was about 6.3 ppmv at steady state while the metha nol concentration was about 1 ppmv. This indicates that methanol was preferentia lly oxidized over formaldehyde in the TiO2-coated AC system. The effluent concentration of formaldehyde from the TiO2-coated glass spheres was 0.3 ppmv while the effluent methanol concentra tion was non-detectable. Since the MDL for methanol was 0.6 ppmv, it was unclear if methanol was prefer entially oxidized in this case. The oxidation of methanol to formaldehyde can procee d by the direct oxidation of methanol in the electron hole in the absence of water vapor (Equations 2-9 through 2-11). However, the subsequent reaction from formaldehyde to form ic acid requires a water molecule (Equation 213). The oxidation of formic acid to carbon dioxide and water can be accomplished by reaction with the electron hole in the ab sence of water vapor (Equation s 2-14 and 2-15), which supports the absence of formic acid in the reactor e ffluent. Thus, the rate limiting step in the mineralization of methanol using TiO2-coated AC was the oxidation of formaldehyde to formic acid. In the case of the STC, no formaldehyde was pr esent in the effluent likely due to its higher overall efficiency (as discussed above). The availability of hydroxyl radicals at the SiO2-TiO2 interface may serve to increase the efficiency of formaldehyde oxidation to formic acid by allowing the indirect oxidation mechanism to pro ceed. In addition, the STC may have been able to adsorb the water vapor that was presen t in the gas stream better than the TiO2-AC due to its surface functionality, which consists of silanol groups (Si-OH) that easily adsorb water via hydrogen bonding (Nawrocki, 1997; Morrow and Ga y, 2000). This adsorbed water vapor could function to participate in dire ct oxidation or form hydroxyl ra dicals for indirect oxidation.

PAGE 82

82 Methanol Adsorption and Oxidatio n in a High Humidity Gas Stream In order to further elucidate the compet itive effects of water vapor on the TiO2-coated materials, they were tested for methanol adso rption and oxidation in a hi gh humidity gas stream (RHi = 95% at 23 C). As shown in Figure 6-3, the adsorption breakthrough profiles were steeper and the adsorption capaciti es of the materials were reduced (1.2 mg/g for STC and 1.9 mg/g for AC). Although the AC broke through earlier than the STC, its adsorption capacity on a mg/g basis was greater than that of the STC due to the difference in bulk density between the two materials (i.e., 0.26 g/cm3 for AC and 0.53 g/cm3 for STC). The humid gas stream negatively affected the STC adsorption capacity more than th at of the AC, indicati ng that the competitive effect of water vapor was more pronounced for ST C than AC. This can be expected since the STC is dominated by silanol functional groups, which readily form hydrogen bonds with water, while the AC has a more heterogeneous surface chemistry (Puri, 1970). In the presence of UV, the methanol removal ef ficiencies of the mate rials were about 95% for both the STC and TiO2-coated AC. The systems reached steady state after approximately 250 min for AC and 400 min for STC. The TiO2-coated glass spheres immediately achieved steady-state removal efficiency, which was between 80 and 90%. Steady state was quickly reached due to the low surface area and resu lting low adsorption capacity of the TiO2-coated glass spheres. As shown in Figure 6-4, the increase in RH re sulted in decreased formaldehyde production in the TiO2-coated AC system (i.e., about 0.8 ppmv at steady state). The reaction of formaldehyde to formic acid was no longer limite d by the lack of water vapor, which may have resulted in lower formaldehyde concentrations in the effluent compared to the low RH case shown in Figure 6-2. In addition, Yamakata et al (2003) showed that water vapor enhanced the

PAGE 83

83 photocatalytic oxidation of methanol by preventing electron accumulation, which would otherwise cause defective sites on the TiO2 surface and faster electr on-hole recombination. Although the water vapor increased th e overall efficiency of the TiO2-AC system, it resulted in a decrease in efficiency for the TiO2-coated glass spheres and STC. The humid gas stream resulted in an increase in efflue nt formaldehyde concentrations for the TiO2-coated glass spheres (about 1 ppmv) and STC (0.5 ppmv). The decrease in overa ll oxidation efficiency compared to that in dry conditions was likely a result of water competition for adsorption and reaction sites. The presence of hi gher effluent methanol concentra tions in these systems indicates that the competition with water vapor was preval ent. It was hypothesized that the STC was more negatively affected by the increase in water vapor than the TiO2-coated AC because the surface coverage of water vapor on the STC was greater than that on the AC. Water vapor adsorption studies confirmed this hypot hesis and are discussed in more detail below. Water Vapor Adsorption In order to elucidate the effects of water vapor on the STC and TiO2-coated AC systems, the water vapor adsorption profile was determined by measuring the e ffluent relative humidity of humid gas (influent RH = 95% at 23 C) after passi ng through the reactor packed with TiO2coated AC or STC and illuminated with UV light The results of these studies are shown in Figure 6-5. Note that the MDL for the hygrometer was 12% RH, so the lowest RH value plotted on the graph is 12% even though the actual effluent RH may have been lower. The total amount of water adsorbed by the STC and TiO2-coated AC was 59 and 25 mg/g, respectively. For the STC system, this indicates that there were 3.7 water molecules adsorbed per nm2 of STC, giving a surface coverage of about 0.26 m2 of water per m2 of STC (26%). For the TiO2-coated AC, there were 0.58 water molecules adsorbed per nm2, giving a surface coverage of about 0.05 m2 of

PAGE 84

84 water per m2 (5%). Thus, the competition between water vapor and methanol on the STC surface would be much greater than that on the TiO2-coated AC. Capillary condensation within the pores wa s not evident for either the STC or TiO2-coated AC. The Kelvin equation predicts that at 50C, which is the approximate temperature of the bed with the UV light on, capillary condensation sh ould only occur in pores less than 12 in diameter. For both the STC and TiO2-coated AC, the majority of the pores were greater than 12 in diameter. For the STC, pores greater than 12 in diameter contributed to more than 95% of the total pore volume. According to Khan ( 2006), approximately 90% of the pore volume in BioNuchar AC, which was used in this study, resulted from pores greater than 12 in diameter. In order to elucidate the e ffects UV light may have on the adsorption of water onto TiO2doped materials, the STC was tested for water adsorp tion in the dark at 50 C. The results, also shown in Figure 6-5, indicate that there was no difference between the breakthrough profiles of the STC tested in the dark at 50C or in the pr esence of UV light. Thus, the effect of UV light on water vapor adsorption was primarily a func tion of the heat generated by the lamp. Reactor Scale-up using TiO2-doped Materials The TiO2-doped materials were tested in a large annulus reactor (25 mm annulus) in order to evaluate the TiO2-doped materials for potential use in fu ll-scale applications, particularly for the treatment of HVLC gases emitted from pulp and paper mills. The TiO2-doped sorbents were tested for methanol removal in a high humid ity gas stream (RH = 95% at 23 C). The high humidity case was chosen since HVLC gases are sa turated with water vapor. The flow rate was adjusted such that the mass tran sfer characteristics (i.e., reside nce time and face velocity) were the same as those from the 8 mm annulus studies. As shown in Figure 6-6, the TiO2-coated glass spheres achieved about 40% methanol removal over the duration of the study. The breakthrough profile for the STC was shallower than that of the TiO2-coated AC, likely due to the greater

PAGE 85

85 capacity for methanol adsorption by the STC, re sulting in higher removal rates for the STC between about 100 and 1300 min. Both TiO2-coated adsorbents achieved about 50% methanol removal at steady state. Note that error bars are present only for the experiments using STC in Figures 6-6 and 6-7 since experiments testing the TiO2-coated AC and glass spheres were not replicated. As shown in Figure 6-7, the effluent formal dehyde concentrations for the STC and TiO2coated AC were similar and increased steadily until reaching a steady state concentration of between 1.5 and 2.0 ppmv. The TiO2-coated glass spheres achiev ed steady state production of formaldehyde at about 1.6 ppmv. According to TOC analysis, other oxidation byproducts (e.g., formic acid, formates) were not identi fied in the effluent gas stream. The decrease in methanol oxidation efficiency in the large annulus reactor compared to that of the small annulus reactor was likely a re sult of inadequate UV lig ht exposure within the entire packed bed. In the la rge annulus reactor, the TiO2-doped adsorbents pe rformed better than the TiO2-coated glass spheres, indicating that adsorption of methanol onto a high surface area catalyst support resulted in higher oxidation rates. However, no difference was discerned between the silica gel, which is transparent, and AC, which is opa que, when used as the catalyst support in the large annulus reac tor. The similar performance between the STC and the AC may be a function of the STCs affinity for water va por, which resulted in a decrease in oxidation efficiency in the high humidity gas stream. It is expected that at lower RH the efficiency of the STC system would improve as the competition of water vapor decreases. Table 6-1. BET surface area and average pore volume of TiO2-doped sorbents. Material BET surface area (m2/g) Pore volume (cc/g) STC 616 0.79 TiO2-coated AC 1424 1.06

PAGE 86

86 Table 6-2. Surface area and TiO2 loading per reactor volume. Material Total surface area per bed volume (m2/cm3) TiO2 loading per bed volume (mg/cm3) 50 4% STC 308 26.7 TiO2-coated AC 380 10.5 TiO2-coated glass spheres 0.43 9.9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0100020003000400050006000 Time (min)C/Co TiO2-AC: UV TiO2-AC: Dark STC: UV STC: Dark TiO2-Glass Spheres: UV TiO2-Glass Spheres: Dark Figure 6-1. Normalized effluent methanol concentration for tita nia-doped materials used in the dark (adsorption only) and with UV light (adsorption and oxidation).

PAGE 87

87 0 1 2 3 4 5 6 7 8 0100020003000400050006000 Time (min)Formaldehyde (ppm v) TiO2-AC: UV STC: UV TiO2-Glass Spheres: UV Figure 6-2. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0500100015002000 Time (min)C/Co TiO2-AC: UV TiO2-AC: Dark STC: UV STC: Dark TiO2-Glass Spheres: UV TiO2-Glass Spheres:Dark Figure 6-3. Normalized effluent methanol concentration for TiO2-doped materials used in the dark and with UV light in a high humidity gas stream (RH = 95%).

PAGE 88

88 0.0 0.5 1.0 1.5 0 500 1000 1500 2000 Time (min)Formaldehyde (ppm v) TiO2-AC: UV STC: UV TiO2-Glass Spheres: UV Figure 6-4. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a high-humidity gas stream (RH = 95%). 0 10 20 30 40 50 60 70 80 90 100 050100150200250300350400 Time (min)Relative Humidity (%) Influent STC: UVC STC: Dark at 50 deg. C TiO2-AC: UVC Figure 6-5. Water vapor adsorption breakthrough profile in a hi gh humidity gas stream (RH = 95%).

PAGE 89

89 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 500 1000 1500 2000 Time (min)C/Co TiO2-AC: UV STC: UV TiO2-Glass Spheres: UV Figure 6-6. Normalized effluent methanol concentration for TiO2-doped materials used in a large annulus reactor (25 mm) and high hu midity gas stream (RHi = 95%). 0.0 0.5 1.0 1.5 2.0 2.5 0 500 1000 1500 2000 Time (min)Formaldehyde (ppm v) TiO2-AC: UV STC: UV TiO2-Glass Spheres: UV Figure 6-7. Intermediate byproduct (formaldehyde) formation by TiO2-coated materials used in a 25 mm annular reactor and high humidity gas stream.

PAGE 90

90 CHAPTER 7 PILOT STUDIES FOR METHANOL DEGRADATION Based on the promising results and knowledge gained from bench-scale studies, a pilotscale reactor was designed to tr eat 40 ACFM using 50 4% STC pellets and UVC lamps. The laboratory-scale synthesis process for the STC pe llets was scaled up to make the approximately 100 pounds of pellets needed for the studies. This scale-up process represented a substantial increase in production compared to the laborat ory scale synthesis tec hnique, which yielded several grams of pellets per batch. A novel test apparatus was designed and built to study the UV light distribution through a packed bed of STC pe llets. Using this information as well as that learned during bench-scale studies a pilot reactor was designe d, fabricated, and tested for methanol degradation from an air stream saturated with water vapor. STC Synthesis for Pilot-Scale Studies In order to produce a sufficient quantity of ST C pellets (50 4%) for the pilot studies, the bench-scale synthesis method wa s modified to increase produc tion efficiency while producing composites with similar characteristics (i.e ., surface area and pore size). Silbond Condensed TEOS was chosen as the silica source. Silbond Condensed was chosen since it contained 94% TEOS with the balance comprised of ethanol a nd polysilicates. This TEOS source was chosen since it was economical and the ba lance of ingredients would not likely negatively affect the characteristics of the final STC product. The TEOS was added to water, ethanol, 1 N nitric acid, 3%wt hydrofluoric acid, and Degussa P25 TiO2. The ingredients were stirred using a paddle mixer. Figure 7-1 shows an image of the mixing assembly used for blending the STC raw materials (mixing container is not shown to allow a view of th e mixing paddle bars). After mixing, the liquid STC was then transferred to molds, which were made from 5.1 cm thick polyethylene sheets drilled with 0.8 cm di ameter holes. Each mold was approximately 40.6

PAGE 91

91 cm by 61.0 cm and contained 2,750 holes. An image of one mold is shown in Figure 7-2. The molds were filled by pouring the liquid sol into the molds, which were sealed on the bottom and top with sheets of solid polyethylene. The STC pe llets were aged at 65C for 48 hours. The lids were then loosened and the pellets dried in the molds at 103C in dual thermal chambers. The pellets were removed from the mo lds, transferred to Pyrex contai ners, and then heated to 180C in the dual thermal chambers. After aging a nd drying, pellets were approximately 3 mm in diameter and 20 mm in length. Since no heat transfer system was commercially available that was suitable for the STC preparation, a double-chamber heat transfer system was designed. This heat chamber provided a pneumatically sealed heat-generating source that was isolated from the STC materials. The system allowed the necessary temperature variations within the inner-chamber, where the pellet trays resided during aging and drying. The inner-chamber wa s equipped with a ventilation system that allowed any volatile ga ses to discharge away from any source of heat or flame. This gas could be passed through a heat exchanger for recovery of the ethanol. A photo of the heat chamber is shown in Figure 7-3. The STC pellets were made in a series of batches, each of which yielded 7-8 pounds of pellets. A grab sample was taken from each ba tch and analyzed for surface area, total pore volume, and pore size. The results of this analys is are shown in Table 7-1. The STC synthesized for pilot studies used the same ratio of raw ingred ients as the pellets used for bench-scale testing. However, the dried STC pellet had a higher BET surface area and smaller pore size than those synthesized in the lab (surface area was 723 m2/g versus 617 m2/g and pore size was 38 versus 49 ). The differences between these measurements can be attributed to sli ght differences in the synthesis procedure that were i nherent with scale-up. Since th e bench-scale studies indicated

PAGE 92

92 that STC with higher surface area and smaller pore sizes achieved better performance, there was no attempt to adjust the pilot-scale procedure to achieve STC with diffe rent characteristics. UV Light Distribution in a Packed Bed of STC In order to determine the spacing of UV lamps for the pilot reactor, a unique test system was developed to measure the UV light penetratio n through a packed bed of STC. A series of square boxes of the same height and varying length/width dimensions were manufactured from Alzak reflective material. A support stand was made to support a UVC lamp in the center surrounded by any one of the boxes. Each box had a inch port for UV measurements. A schematic of the test system (provided by the fabricator) is sh own in Figure 7-4. Each box was filled with the STC pellets and the UV lamp was illuminated. UV measurements were taken through the port on the side of the box to measure the amount of light penetrating through the packed bed. This was repeated for each box so that the UV light penetration through various bed depths could be measured. The results are shown in Table 7-2. The UV intensity decreased as the packed bed de pth increased for two reasons: (1) the intensity of UV light decreases as a function of the inverse square route of the distance from the lamp; and (2) a deeper bed of STC would also serve to at tenuate more light. No UV light was measured through a packed bed depth of 1 inch. Therefor e, the maximum distance between any portion of the packed bed and a UV lamp should be less than 1 inch in order to avoid dark spots in the reactor. Pilot Studies Figure 7-5 shows a process flow diagram for th e pilot studies. A variable frequency drive fan was used to push the gas stream through th e pilot reactor. The space time was about 4.3 s, similar to that used in bench-scale studies. A wa ter/methanol mixture was vaporized and injected into the gas stream so that the influent methanol c oncentration was about 50 ppmv and relative

PAGE 93

93 humidity between 95 and 99%. The static pressure and velocity pressure were monitored before the inlet to the pilot reactor. Thermocouples were used to measure inlet and outlet temperature. The pilot reactor was fabricated by MicroEner gy Systems, Inc. (Oakland, MD). A general arrangement drawing for the pilot reactor, whic h was provided by MicroEnergy Systems, Inc., is shown in Figure 7-6. A photo of the reactor is shown in Figure 7-7. The dimensions of the reactor were approximately 17 in by 17 in by 8 ft tall. The gas flowed upwards into the bottom of the reactor and through a plenum, which supporte d the packed bed of STC pellets. The STC pellets were packed between 46 UVC lamps (Ame rican Ultraviolet Corporation), which were oriented in the reactor parallel to the air flow The lamps were spaced 2.375 in. center to center such that 79% of the pellets were within 0.6875 in. of a UV lamp and no pellet was more than 1 in. from a UV lamp. The UV lamps were housed in 1 in. quartz sleeves, each of which included electrical connectors and wire s, the latter of which passed through sealed stoppers. Top and bottom support racks (shown in section A-A of Figure 7-6) were desi gned to support each UV lamp assembly. A UV-Technik Electronic Ball ast Power Center was in stalled in conjunction with the 46 UV bulbs, thus giving the capability to vary UV radiation in response to a variable voltage power input, which was controlled by the potentiometer setting. A UVP radiometer (Upland, CA) was used to measure the UV intens ity through an observation window on the side of the reactor. The inside of the reactor was lined with Alzak UV reflective material. Initial studies performed with an empty reactor showed no photolysis of methanol and that adsorption of methanol to the r eactor and its appurtenances was negligible. Two tests were conducted with variable UV intensities controlled by varying the potentiometer setting, which controlled the voltage input. For the first test, the potentiometer was set to 22%. The UV intensity (as measured throu gh an observation port loca ted on the side of the

PAGE 94

94 reactor) was about 6.5 mW/cm2. For the second test, the poten tiometer was set to 100% and the corresponding UV intensity was about 11 mW/cm2. In both cases the breakthrough profile was similar to that seen in bench-scale studies The effluent methanol concentration was nondetectable for the first 4-6 hours, due to adsorption on the virgin STC surface, and then increased until the steady state was reached (i.e., the rate of adsorption equaled the rate of oxidation). The results of the pilot tests ar e shown in Table 7-3. The meth anol oxidation efficiency was dependent on the UV intensity at steady state, with the higher UV intensity resulting in about 66% removal while the lower UV intensity resu lted in about 27.5% removal efficiency. The effluent formaldehyde concentration was minimal (i.e., less than 1 ppmv) for both studies, thus indicating that the majority of methanol was oxidized to iner t byproducts (i.e., carbon dioxide and water vapor). The effluent temperatures were significantly highe r than the influent temperatures for both tests due to the heat gene rated by the high intensity lamps. When operated at their maximum intensity, the lamps require 60 W and output 16.2 W as UVC. Therefore, most of the input energy was lost as heat, which resulted in the el evated effluent temperatures. The effluent temperatures sugge st that the reactor operated under elevated temperatures compared to the temperature tested at the bench-scale, which was about 50C (122F). Elevated temperatures can result in a decrease in UV intensity since UV lamps operate most efficiently at 104F. Based on manufacturer provided data, th e UV lamps operate at about 70% efficiency at 139.3F (low intensity setting) and 50% efficiency at 157.6F (high intens ity setting). This decrease in photon flux could yiel d a decrease in oxidation rate. In addition, a decrease in the adsorption affinity of methanol for the STC surf ace due to the higher temperatures could result in decreased oxidation rates. Therefore, improve d results may be achieved by adding a cooling system to the UV lamps to transfer heat away from the lamps and packed bed.

PAGE 95

95 In addition to temperature effects, the effi ciency of the pilot reactor may have been dictated by the UV light distribution in the packed bed. The maximum distance between an STC pellet and a UV lamp in the reactor was 1 in. The reactor was designed so that the UV intensity approached zero as the distance away from th e UV lamp increased to 1 in. However, the UV intensities achieved in the portions of the bed approaching 1 in. from the UV lamp may not have been sufficient to achieve higher oxidation rates. Thus, a minimum UV intensity across the entire bed may be necessary to incr ease oxidation efficiency. Table 7-1. BET surface area, pore volume, and pore size analysis for pilot STC. BET surface area (m2/g) Total pore volume (cc/ g) Actual pore size () 723 67 0.75 0.12 38 3 Table 7-2. UV intensity measurements through packed beds of STC of varying depths. Box No. Distance (in.) UV intensity ( W/cm2) 5 0.25 42 4 0.5 16 3 0.75 3 2 1.0 0 1 2.0 0 Table 7-3. Results of pilot studies w ith variable potentiometer settings. Experiment UV intensity (mW/cm2) Inlet temp (F) Outlet temp (F) Steady state methanol removal Formaldehyde production (ppmv) Low Intensity (Setting = 22%) 6.5 75.6 5.4 139.3 3.0 27.5 5.5% 0.28 0.26 High Intensity (Setting = 100%) 11.5 72.6 4.0 157.6 6.0 66 7% 0.74 0.10

PAGE 96

96 Figure 7-1. Mixing assembly for blending the raw ingredients fo r pilot-scale STC synthesis. Figure 7-2. Pilot-scale molds for STC synthesis. Figure 7-3. Specialty heat chamber for pilot-scale STC synthesis.

PAGE 97

97 Top support plate Profile Box #2 Box #1 Box #2 Box #3 Box #4 Box #5 14.00in 14.00in 12.02" 16.00in 3.00in 1/4" diam. all-thread support colums Top support plate Register Plate for Box #2 UV Bulb connector Alzak Ceiing Plate 12.00in Floor Support Plate Figure 1: UV-PELLET TEST RIG ASSEMBLY (03/08/05) Figure 7-4. Alzak box test system for measuring UV light penetr ation through packed beds of various depths. Alzak Ceiling Plate

PAGE 98

98 Figure 7-5. Process flow diagram for methanol degradation pilot studies. Figure 7-6. General arrangement drawing of the pilot reactor for methanol degradation. SECTION A-A 15 in 17 in 46: 1" OD Quartz tubes 46: 24" UVC bulbs (red) SS-rod support matrix (gray) 2" PVC half-sections

PAGE 99

99 Figure 7-7. Photo of pilot reactor. Observation Window UV-Technik Electronic Ballast Power Center

PAGE 100

100 CHAPTER 8 DEVELOPMENT OF A REGENERABLE SYSTEM EMPLOYING STC PELLETS FOR MERCURY REMOVAL FROM END-BOX EXHAUS T AT A CHLOR-ALKA LI FACILITY The STC technology may be advantageous fo r the removal of mercury from end-box exhaust in chlor-alkali facilities due to its hi gh mercury adsorption capacity and ability to be regenerated with acid. The mercury-laden acid us ed for regeneration can be recycled into the mercury-cell process, thus clos ing the loop on mercury emissions. This chapter summarizes the design and performance of pilotand full-scale reactors used to recover mercury from the endbox exhaust at a chlor-al kali facility. In addition, an econom ic analysis, which compares the costs of implementing this technol ogy versus using activated carbon at the facility, is presented. Pilot-Scale Packed Bed Reactor A pilot reactor was fabricated and tested for mercury removal from end-box exhaust, which consisted of air and trac e quantities of hydr ogen (0.02%), contained elemental mercury vapor, and was saturated with water vapor. The tota l flow rate of the caustic exhaust system was about 350 ACFM during the pilot studies. The ex haust passed through a heat exchanger, which reduced the temperature to between 6 and 8 C and a series of knock-out pots to remove entrained water and mercury droplets. A slipstre am of the exhaust was taken after the knock-out pots and passed through a blower, into the pilot re actor, and back into the vent for release into the atmosphere. A process flow diagram fo r this process is shown in Figure 8-1. The pilot reactor contained a fixed bed of STC pellets packed around UV lamps. The STC pellets had an average BET surface area of 351 m2/g, pore volume of 0.95 cm3/g, and pore diameter of 109 The reactor contained tw o chambers (Chamber A and Chamber B), which could be operated and sampled independently of each other. Two Teflon-coated UV lamps (American Ultraviolet Corporation), were positioned vertically in each chamber and on two external reactor walls (one wall of each chambe r). The reactor walls we re made up of Lucite

PAGE 101

101 UVT material, which is transparent to UV light. Th e outer shell of the reactor was made of Alzak reflective metal in order to refl ect light into the packed bed. Th e exhaust flowed upwards through the reactor while the STC pellets were continuousl y irradiated with the UV lamps. Each lamp was controlled individually such that any combina tion of lamps could be illuminated at one time. A schematic and photo of the pilot react or are shown in Figures 8-2 and 8-3. Pilot Study Results The pilot-scale study treating mercury in a slip stream from caustic exhaust was conducted from February to May of 2005 at a chlor-alkali fa cility in the US. For all pilot studies, the UV lamps were operated continuously in order to maxi mize mercury removal efficiency to help meet this facilitys goal of zero emissions. The objective of the pilot experiments was two-fold: (1) confirm the efficacy of the technology for merc ury removal and recovery by regeneration with HCl; and (2) determine the factors that may lim it mercury removal efficiency (e.g., residence time, mass transfer, and UV light distri bution within the packed bed). In order to confirm the efficacy of the technolo gy and test the effectiveness of regeneration using HCl, a series of two experiments were c onducted. In the first experiment, the reactor was packed with virgin STC pellets and continuously operated for 10 days using a flow rate of 10 ACFM and a space time of 0.53 s. As shown in Figure 8-4, the influent concentrations ranged between 400 and 1600 ug/ft3 (or 1795 and 7180 ppbv). Although the influent mercury concentration was highly variable, the effluent mercury concentration remained low (26 17 ug/ft3 for Chamber A and 31 16 ug/ft3 for Chamber B). The mercury removal rate achieved by the reactor over this 10 day period was 96 3 %. Although the r eactor ran continuously for the 10 day period, no samples were collected on days 3-5 since the chlor-alkali facility employees were not available to collect samples. Table 8-1, which refers to this experiment as Test No. 1, shows a summary of results for the pilot data. The error is represented as the standard deviation.

PAGE 102

102 After the 10 days of operation, the pellets were removed from the reactor and regenerated in HCl. The mercury concentration on the pellet s was determined by digesting the pellets before and after regeneration. About 99% of the mercur y was removed from the pellets using this regeneration procedure. In order to test the effect of regenera tion on the performance of the STC pellets, a second experiment was conducted in wh ich virgin pellets were packed in Chamber A and the regenerated pellets were packed in Chamber B. The reactor was operated using the same flow rate and space time (10 ACFM and 0.53 s) as the previous experiment. As shown in Figure 8-5, both the new and regenerate d pellets performed similarly and achieved greater than 90% removal for at least 21 days. After 35 days of operation, the effluent mercury concentration approached that of the influent; thus, break through was achieved. The mercury loading on the pellets was calculated by performing a mass balan ce on the reactor. After the first 21 days of operation (i.e., before breakthrough) the mercur y loading on the virgin pellets was about 260 mg/g and the loading on the regenerated pellets was about 247 mg/g. Since these mass loadings are similar, this confirms that the regeneration procedure did not impact the performance of the STC pellets. A series of tests were conducted in which th e effects of space time, mass transfer, and UV irradiation were studied. For each experiment (s hown as Test Nos. 2 through 4 in Table 8-1), virgin STC pellets were used in the pilot reactor. In order to hypothetically improve upon the performance of Test No. 1, the flow rate was de creased for Test No. 2, re sulting in a longer space time and higher j-factor (i.e., improved mass transfer). The average effluent concentration for Test No. 2 was 11.5 ug/ft3, which is significantly lower than the effluent concentrations from Test No. 1 (P = 0.0248 based on two-tailed t-test w ith 95% confidence interv al). To test whether this improvement was a result of better mass transfer or the increase in sp ace time, the bed height

PAGE 103

103 was increased from 10 in. to 31 in. and approximately the same flow rate as Test No. 2 was used for Test No. 3A. This increase in space time from 0.67 to 2.3 s resulted in lower mercury effluent concentrations, which averaged 2.9 ug/ft3. This suggests that the merc ury removal rate is limited by the rate of photocatalytic oxid ation, rather than mass transfer. A factor that may limit the rate of photocatalyt ic oxidation in a packed bed is ineffective UV light distribution within the bed. Therefore, in Test No. 3B, the outer UV lamps positioned behind the transparent side wall of the reactor were illuminated to enhance the distribution and increase the intensity of UV li ght in the bed. This increased irradiation did not result in a decrease in effluent mercury concentration. Thus, the rate of photocat alytic oxidation did not increase with the increase in UV radiation with in the bed. Therefore, the space time through the bed was further increased in attempts to ach ieve lower effluent mercury concentrations. In Test No. 4, the flow rate was decreased to 3.6 ACFM in order to increase the space time to 4.6 s. The effluent mercury concentration from this test averaged 3.6 ug/ft3, which was not significantly different than Test No. 3A (P = 0.1646). Thus, the increase in space time from 2.3 to 4.6 s did not enhance mercury removal as was seen with the increase in space times between 0.53 and 2.3 s. At space times greater than 2.3 s, the removal rate through the reactor was limited by factor(s) other than photocatalytic oxidation rate. For ex ample, at low effluent mercury concentrations (i.e., 3-6 ug/ft3), there may not be sufficient driving force for mass transfer and subsequent adsorption. Additionally, photocatalytic reduction of sorbed mercuric oxide (HgO) to elemental mercury may occur by the free electrons on the TiO2 surface, which are generated by irradiation of this surface with UV light (Li and Wu, 2006). Alt hough the photocatalytic oxidation of elemental mercury is the primary reaction pathway since oxygen serves as a stronger electron trap than HgO, a photocatalytic reduction effect wa s seen in the work of Li and

PAGE 104

104 Wu (2006) while studying the use of STC for me rcury removal from flue gas. Elemental mercury has a lower adsorption affinity for th e STC surface; thus, desorption or restrained adsorption of elemental mercury from the gas st ream may occur. Thus, photocatalytic reduction may inhibit the attainment of lower effluent mercury concentrations. Full-scale Reactor Based on the promising results from the pilot st udies, two full-scale reactors were designed and installed. Each of these reactors wa s designed to treat up to 1200 ACFM and 1600 ug/ft3 and have in-situ regeneration capabilities. Each un it was designed to handle the entire flow rate so that the other could be offline for regeneration. The footprint of each reac tor was about 3 feet by 6 feet. Before entering the reac tor, the exhaust passed through a heat exchanger and demister, which removed entrained liquid droplets greater than 4 m in diameter. The temperature entering the online reactor was typically between 10 and 15C. Regeneration was performed in-situ by soaking the pellets inside of the reactor with 37%wt HCl for 30 min. The mercury-laden acid was then drained from the reactor and the pellets we re rinsed with water. Both the HCl and water were introduced into the reactor by a spray nozzle located inside th e top of the reactor above the packed bed. A process flow diagram for this full-scale system is shown in Figure 8-6. The design of the pilot reactor was scaled-up so that the space time and Colburn j-factor in the full-scale units were 0.63 s and 0.10, respectiv ely, when treating the maximum design flow rate of 1200 ACFM. The UV light spacing was similar to that used in the pilot scale reactor. The reactors were filled with STC pellets w ith an average BET surface area of 453 m2/g, pore volume of 1.04 cm3/g, and pore diameter of 92 Two full-scal e reactors were installed in parallel so that one could be in operation while the other was regenerated. This also provides a level of

PAGE 105

105 redundancy since regeneration requir es only a short period of time (i .e., 4 hours) compared to the time that the other reactor will be online before regeneration is required (i.e., 30 days). Data from the first three operation cycles for the reactors are shown in Table 8-2. The error is represented as the standard deviation of measurements taken over the operation time. For all studies, the effluent mercury c oncentration remained steady thr oughout the run (i.e., error is not associated with breakthrough). The cycles are desi gnated using a letter (e ither A or B) followed by a number. The letter indicates which reactor was in operation and the number indicates the cycle number. For example, Run No. A-1 indicate s the first operation cycle for Reactor A. For the three operation cycles show n, the average flow rate decr eased with each run for unknown reasons. The flow rate in s ubsequent runs (results not s hown) varied between 400 and 700 ACFM. Run No. A-1 averaged slightly more than 50% removal over a 29 day period. This poor performance was due to the lack of water entering the bed. The influent water loading into the pilot was about 3 lbs of water/1000 ft3 of air; however, the loading into the full-scale reactor was about 0.45 lbs/1000 ft3. This is a result of changes made to the exhaust system (including the addition of a new heat exchanger and demister) during the time between the pilot reactor testing and the installation of the full-scale reactors. During Run No. A-2, the demister was initially bypassed in order to increase the in fluent water loading and test th is hypothesis. After 15 days of operation, the demister was brought online and water was added into th e reactor through the nozzle in the top of the reactor above the packed bed. About 60 gallons of water were added to the reactor once per day for th e remainder of the cycle. Figure 8-7 shows the influent and effluent mercury concentrations for this run (No. A-2). During the first 15 days, while the demister was offline, the effluent mercury concentrations

PAGE 106

106 were low. However, the influent mercury concen trations were highly variable, ranging between about 400 and 1800 ug/ft3. Since the demister typically rem oved some of the elemental mercury that was entrained as small drop lets in the exhaust, it was brou ght back online on Day 15 in order to decrease the influent mercury loading to th e bed, thus increasing bed life and time between regenerations. Therefore, water was added through a nozzle positioned in the top of the reactor after the demister was brought back online. This proved equally as eff ective at reducing the effluent mercury concentration. During the entire 43-day cycle, the actual flow rate through the reactor was 560 140 ACFM. The average e ffluent concentration was 21.5 23.8 ug/ft3, which corresponds to a mercury removal rate of 95 5% The mercury loading on the pellets at the end of this operational cycl e was about 190 mg/g. Some of the water, which was added through th e nozzle to saturate the bed during Run No. A-2, passed through the packed bed and was draine d from the reactor. During Run No. B-1, the demister was online and water was added throug h the top nozzle. However, the water loading rate was optimized so that 15 gallons of water were added every 2 hours. This water application rate did not result in excess water passing thr ough the packed bed and yielded low effluent mercury concentrations, which averaged 10.8 ug/ft3. The addition of water to the full-scale reacto r was necessary to provide sufficient hydroxyl radicals on the TiO2 surface. This indicates that mercur y was not directly oxidized by the electron holes on the TiO2 surface, but was indirectly oxidized by hydroxyl radicals, which are powerful oxidants. These hydroxyl radicals are creat ed by the oxidation of ad sorbed water vapor. Due to the low influent water concentration a nd the rise in temperat ure through the reactor caused by heat generated by the UV lamps, adsorp tion of water vapor to the STC surface was

PAGE 107

107 low. Therefore, constant water addition was necessa ry to maintain a sufficient quantity of water on the surface of the STC. Economic Analysis An economic analysis was performed to asse ss the feasibility of the STC technology compared to treated (e.g., sulfur or iodine im pregnated) activated carbon. The design for both technologies was based on the maximum design flow rate and influent mercury concentration of 1200 ACFM and 1600 ug/ft3, respectively. This corresponds to an influent mercury loading of 2765 g/day. The cost per pound of mercury remove d was based on capital and O&M (operation and maintenance) costs over a 20 year period. This analysis assumes a mercury removal rate of 95% and continuous operation for both technolo gies and is solely based on economics. No attempt was made to assign a value to the risk as sociated with the handling, transport, and offsite disposal of mercury-laden waste for the case of activated carbon. The costs associated with the STC technology ar e based on the actual full-scale installation described above. The costs associated with the ac tivated carbon installatio n include a pre-heater and two beds of sulfur-impregnated activated carbon. The function of the pre-heater is to increase the temperature of the exhaust to abou t 100 F in order to enhance mercury adsorption. The beds should be installed in series with provisions for bypass or reversal of flow to provide maximum utilization and reliability of the adsorption system (Anastas, 1976). The design of the carbon beds was based on a face velocity of 1 f t/s and space time of 8 s (Anastas, 1976; Klett, 2002). Thus, each carbon bed would be approximately 5 ft in diameter and 8 feet in height. Capital Costs The capital cost associated with the STC technology was based on the actual fabrication cost (in 2006 dollars) of the two full-scale reactors. This cost included support stands for the

PAGE 108

108 reactors, associated electrical equipment (e.g., ballasts for UV lamps), and a PLC system, which allows the control room to remotely monitor the status of the reactor. The capital cost for the carbon installation was based on the modular approach (Guthrie, 1974). This approach enables estimation of co mponents of a chemical processing unit from known equipment costs using multipliers to adjust for design variations such as equipment configuration, material of cons truction and design pressure. A Marshall and Swift Index of 1250 was used to scale the cost data to 2006 dollars. Using this appr oach, the capital cost for the preheater and two carbon beds were estimated to be about $100,000. O&M Costs The operation and maintenance costs include t hose incurred annually for the following: (1) utilities (i.e., electricity), (2 ) operating labor, (2) maintenance materials and labor, and (3) taxes and insurance. The only utility needed for bot h technologies was electr icity ($0.10/KWh). Labor cost was based on the 2006 Construction Cost Index wage of $31/hour. Maintenance materials and labor for operation and main tenance were based on reasonable assumptions for both technologies. Taxes and insurance were estimated to be 2.5% of the capital cost (Anastas, 1976). The maintenance material costs for the STC t echnology were based on the cost to replace: (1) UV lamps and gaskets annually and (2) ballasts and STC pellets once every five years. Forty labor hours per year were included for maintena nce. Operation costs included energy consumed by the UV lamps and blower. The number of re generations per year, which varies based on influent mercury loading, was based on an STC capacity of 200 mg/g. The cost of acid and water was considered negligible since both are readily available at th e facility. Four labor hours per regeneration were included in the operation costs. Costs associ ated with recycling the mercury back into the process was not incorporated in to the analysis. Similarly, credit for mercury recovery was not included.

PAGE 109

109 Replacement frequency of the carbon was base d on an operational capacity of 100 mg/g (EPA, 1997b). Carbon replacement costs were estim ated at $6.43 per pound and disposal was estimated to be $500 per ton (Klett et al., 2002). Operation costs for the carbon included that for the energy to power the blower. Maintenance materials and labor was estimated to be 2% of the capital cost per year (Anastas, 1976). Economic Feasibility The cost per pound of mercury removed was estimated for both technologies as a function of influent mercury loading, as shown in Fi gure 8-8. The cost per pound for the STC technology was lower than that for the activ ated carbon technology at influent mercury loadings greater than 149 g/day, which is the loading at which the STC a nd activated carbon cost curves intersect. Due to the uncertainty of the actua l cost and performance associ ated with activated carbon installations, the error associated with the activated carbon costs may be as high as 30% (Anastas, 1976). Therefore, the difference in co st between the two tec hnologies can only be considered significant for influent mercury load ings greater than about 470 g/day (i.e., the influent mercury loading at which the cost fo r the STC technology is more than 30% different that the cost for the activated carbon technology). At the design influent mercury loading of 2765 g/day, the cost per pound of mercury removed is about $20 for the STC technology and $84 for activated carbon. Thus, the STC technology is ec onomically feasible at the design influent mercury loading. For the full-scal e operation data presented in Table 8-2, the average influent mercury loading of the three runs was 357 g/day. At this loading rate, the economically favorable technology cannot be determined due to the uncertai nty associated with the estimated cost of the activated carbon installation. The shape of the cost curves is similar for the STC and activated carbon technologies. However, the dominate cost for the STC technolo gy is that for the capital while, at influent

PAGE 110

110 mercury loadings greater than about 300 g/day, O&M costs comprise a majority of the activated carbon costs. The cost of activated carbon replaceme nt (material replacemen t and disposal fees only) is constant at $70 per pound of mercury removed. At low infl uent mercury concentrations, the activated carbon costs increase rapidly. This is because the fixed costs are not diluted with high mercury removal rates. Thus, the total co st per pound of mercury removed becomes highly dependent on the capital cost at lower influent loading rates. Table 8-1. Summary of pilot experiments. Test No. Flow (ACFM) Colburn j Space time (s) Influent (ug/ft3) Effluent (ug/ft3) 1 10 0.106 0.53 762 286 28 16.3 2 8 0.122 0.67 652 328 11.5 7 3A 7.2 0.129 2.3 817 203 2.9 0.3 3B 7.2 0.129 2.3 817 203 3.5 1.2 4 3.6 0.183 4.6 550 208 5.8 4.5 Table 8-2. Full-scale performance data for three operation cycles. Run No. Operation Time (days) Flow Rate (ACFM) Influent (ug/ft3) Effluent (ug/ft3) A-1 29 682 22 445 75 204 132 A-2 43 560 140 595 451 21.5 23.8 B-1 28 400 64 234 116 10.8 5.6 Heat Exchanger Knock-Out Pot Pilot Blower Pilot Reactor Velocity Pressure Static Pressure Temp. Temp. Hg Sampling Temp. To Stack End-box Exhaust Hg Sampling SlipStream Figure 8-1. Process flow diagram fo r mercury recovery pilot studies.

PAGE 111

111 4 External UV Bulbs 4 Internal UV Bulbs 3 UVT Lucite windows 2 Inlet ducts Alzak reflective plates (removable to view pellets) Center UVT Lucite window "sandwiched" between flanges to allow removal to create larger chamber, or remain in place for dual-chamber or single-smaller chamber during test operations Approx. 9 in Approx 6 in Approx 54 in Dual trap doors allow removal of pellets during testing Removable perforated metal-floor inserts (accessible thru trap doors) Dual-channel inlet plenum SECTION A-A SECTION B-B AA B BPLAN VIEW Dual or single chambers for housing STC pellets Dual outlet ducts & globe valves Inlet ports for pellets Inlet ports for pellets Dual inlet ducts External UV Bulbs Dual outlet ducts & globe valves Figure 8-2. Schematic of pilot reactor for mercury recovery.

PAGE 112

112 Figure 8-3. Photo of pilot reactor with UV lights illuminated.

PAGE 113

113 Figure 8-4. Influent and e ffluent mercury concentrations for the pilot reactor packed with virgin STC pellets (Test No. 1). 0 200 400 600 800 1000 1200 1400 1600 1800 0246810Time (Days)Mercury Concentration (ug/ft3) Influent Effluent A Effluent B

PAGE 114

114 Figure 8-5. Influent and e ffluent mercury concentrations of th e pilot reactor packed with virgin pellets (Chamber A) and regene rated pellets (Chamber B). Heat Exchanger Demister Reactor 1 (Online) Temp. Temp. To Stack End-box Exhaust Hg Sampling Temp. Hg Sampling Reactor 2 (Offline) HCl or Water For Regeneration Hg-Laden HCl Or Water Recycled to Hg-Cell Figure 8-6. Process flow diagram for full-scal e installation of merc ury recovery units. 0 200 400 600 800 1000 1200 1400 1600 051015202530354045Time (days)Mercury Concentration (ug/ft3) Influent Effluent A: Virgin STC Effluent B: Regenerated STC

PAGE 115

115 Figure 8-7. Influent and effluent concentrations for the full-scale reactor during its second adsorption cycle. 0 200 400 600 800 1000 1200 1400 1600 1800 2000 051015202530354045Time (days)Mercury Concentration (ug/ft3) Influent Effluent

PAGE 116

116 Figure 8-8. Comparison of cost per pound of mercury removed for activated carbon and STC as a function of influent mercury loading (for systems designed to treat up to 2765 g/day). $0 $100 $200 $300 $400 $500 $600 $700 $800 $900 $1,000 050010001500200025003000Influent Mercury Loading (g/day)Cost per Pound Hg Removed STC Technology STC: Capital Activated Carbon Activated Carbon Replacement

PAGE 117

117 CHAPTER 9 CONCLUSIONS STC pellets were synthesized using a sol-gel method with va rying concentrations of HF and TiO2. Mesoporous STC with varying pore volumes and surface areas were created. The performance of these STC for methanol remova l was found to be a function of the surface area of the STC and space time of the gas in the re actor. The reaction kinetics were not limited by external or internal resistances to mass tran sfer. The 120 12% and 260 12% STC exhibited a lag time before achieving methanol mineraliz ation, while the 50 12% STC did not. All STC exhibited pseudo-first order re action kinetics with a sim ilar rate constant of 0.40 s-1. The 50 STC were synthesized with varying TiO2 loadings and the optimum loading was found to be 4%. Methanol oxidation was enhanced with the highe r photon flux associated with using a UVC lamp rather than a UVA lamp. The bench-scale system reached steady state removal and was able to remove about 90% of the influent methanol wi th little byproduct forma tion (less than 1 ppmv formaldehyde). When H2S was introduced into the system, the H2S competed with methanol resulting in a decrease in methanol rem oval efficiency to about 80%. Although the H2S did not adsorb well to the STC surface, the H2S was oxidized by the STC when illuminated with UV light to SO2 and SO4 2-. A pilot-scale reactor was designed a nd fabricated to treat 40 ACFM of humid air laden with 50 ppmv of methanol. Pilot-scale studi es showed about 66% methanol removal efficiency at steady state when the sp ace time of the gas through the packed bed was 4.3 s. The methanol removal efficiency achieved in the pilot studies was less than that in the benchscale studies because the UV light distribution wi thin the packed bed was limited and elevated reaction temperatures likely inhibited the oxidatio n rate by decreasing the adsorption of methanol onto the STC and the efficiency of the UV lamps.

PAGE 118

118 STC and TiO2-coated AC were compared to TiO2-coated glass spheres for the removal of methanol. In a low humidity environment (RH = 0.22%), the adsorption capacity of the STC (11 mg/g) was greater than that of TiO2-coated AC (6 mg/g) and TiO2-coated glass spheres. The silanol groups (Si-OH) on the STC surface promoted methanol adsorption in a low humidity gas stream, where competition with water vapor was low. When STC was irradiated with UV light, no methanol or oxidation byproducts were detected in the effluent. Formaldehyde, an oxidation byproduct of methanol, was detected in the effluent using TiO2-coated AC and TiO2-coated glass spheres. The TiO2-coated AC showed that the limiting reac tion in the mineralization of methanol was the oxidation of formaldehyde to formic aci d. In a high humidity gas stream (RH = 95%), the adsorption capacity of the STC (1.2 mg/g) and TiO2-coated AC (1.9 mg/g) was reduced due to the competition with water vapor for adsorp tion sites. The overall efficiency of the TiO2coated AC increased likely due to the presen ce of water vapor, which is required for the oxidation of formaldehyde to formic acid. The methanol adsorption capacity of the AC was greater than the STC in the high humidity gas st ream since the surface chemistry of the AC was more heterogeneous than that of the STC, wh ich was dominated by sila nol groups that strongly adsorb water via hydrogen bonding. In the presence of UV light, the STC and TiO2-coated AC reached a steady state methanol removal effici ency of 95%. Water vapor adsorption studies found that the surface coverage of water on the STC surface was greater than that on the TiO2 surface and that the adsorption of water vapor on the STC was solely affected by the heat generated from the UV lamp. In a high humidity ga s stream with a large annulus reactor, the STC and TiO2-coated AC performed similarly, achieving 50% methanol removal compared to the TiO2-coated glass spheres, which achieved 40% me thanol removal. Thus, using an adsorbent material as a catalyst support in larger scale systems was beneficial. However, the use of silica

PAGE 119

119 gel, which was transparent, versus AC, which was opaque, did not result in a difference in the photocatalytic oxidation rate in the 25 mm annulus reactor. This wa s likely due to the decrease in degradation efficiency associated with water competition on the STC surface, which was not as prevalent for the TiO2-coated AC. The series of pilot experiments performed a ch lor-alkali facility confirmed the efficacy of the scale up of the STC technology for the remova l of elemental mercury from end-box exhaust. The experiments showed that the reactor was able to consistently achieve 96% mercury removal in a variety of system conditi ons, including highly variable influent conditions (400 to 1600 ug/ft3). In the range of flow rates tested, merc ury removal rate was limited by space time and not mass transfer. An increase in space time between 0.53 s and 2.3 s yielded slight increases in mercury removal. However, longer space times did not result in a change in mercury removal. Regeneration with concentrated HCl was performe d and proved to be eff ective. The regenerated pellets were successfully brought back online and pe rformed similarly to virgin pellets. Two fullscale reactors were installed and performed sim ilarly to the pilot-scale reactors. An economic analysis showed that the cost per pound of mercury removed was less for the STC technology than for sulfur impregnated activated carbon at influent mercury loading rates greater than 470 g/day. At lower influent loading rates, ne ither technology could be deemed economically favorable due to the uncertainty associated with the activated carbon cost estimate. Thus, the STC proved to be both technologically and econo mically feasible for this application. Recommendations for future work are listed below: The competitive effects of adsorption and oxi dation of VOCs in a multi-component system employing STC should be modeled since these systems will be more indicative of realworld applications;

PAGE 120

120 The transparency of STC containing various TiO2 loadings should be quantified to better understand the relationship be tween the quantity of TiO2 that is irradiat ed and oxidation efficiency; The effects of adsorption on H2S oxidation efficiency should be studied in order engineer a composite that promotes the oxidation of H2S to SO4 2rather than SO2, which would need to be scrubbed from the gas before being released to the atmosphere. As a result of this work, the following contributions to science were made: First to demonstrate that the oxidation efficiency of photocatalyst pellets for the degradation of a gaseous orga nic compound (methanol) was affect ed by material properties of the pellets (i.e., pore size and TiO2 loading), photon flux, and space time of the gas through the reactor; First to model the rate of photocatalytic oxidation of ga seous methanol by photocatalyst pellets of various pore sizes us ing pseudo-first order equations and demonstrate that a lag time existed before mineralization proceeded, which was dependent on the internal surface area of the pellets; First to demonstrate that the limiting step in the mineralization of methanol differed depending on both the type of catalyst suppor t and the humidity of the gas entering the reactor; Supported evidence that both SO2 and SO4 2were formed as byproducts as a result of the photocatalytic oxidation of H2S, which contrasts some studies that reported the formation of only SO4 2-; First to demonstrate that the oxi dation efficiency of methanol decreased as a result of the competitive oxidation of H2S in a humid gas stream and that the oxidation of H2S was unaffected by the presence of methanol; First to implement a full-scale technology employing photocatalysis for the removal of mercury from gas-phase emissions in the chlor-alkali industry.

PAGE 121

121 LIST OF REFERENCES Al-Ekabi, H., Serpone, N., 1988. Kinetic st udies in heterogeneous photocatalysis. 1. Photocatalytic degradation of chlorinated phe nols in aerated aqueous solutions over TiO2 supported glass matrix. J. Phys. Chem. 92, 5726-5731. Alberici, R.M., Jardim, W.F., 1997. Photocatalytic destruction of VOCs in the as-phase using titanium dioxide. Appl. Catal. B Environ. 14, 55-68. Anastas, Y., 1976. Molecular sieve mercury contro l processes in chlor-alkali plants, EPA-600/276-014. U.S. Government Printing Office, Washington D.C. Anderson, C., Bard, A., 1997. Improved photocataly tic activity and charact erization of mixed TiO2/SiO2 and TiO2/Al2O3 Materials. J. Phys. Chem. 101, 2611-2616. Arana, J., Dona-Rodriguez, J., Cabo, C., Gonzal ez-Diaz, O., Herrara-Melia n, J., Perez-Pena, J., 2004. FTIR study of gas-phase alcohols photocatalytic degradation with TiO2 and ACTiO2. Appl. Cataly. B Environ. 53, 221-232. Byrne, H., Kostedt, W., Stokke, J., Mazyck, D ., 2008. Characterization of HF-catalyzed silica gels doped with Degussa P25 Titanium Dioxi de. Submitted to J. Non-Cryst. Solids. Canela, M., Alberici, R., Jardim, W., 1998. Gas-phase destruction of H2S using TiO2/UV-Vis. J. Photochem. Photobiol. A 112, 73-80. Chang, H., Wu, N., Zhu, F., 2000. A kinetic mode l for photocatalytic degradation of organic contaminants in a thin-film TiO2-catalyst. Wat. Res. 34, 407-416. Chen, J., Ollis, D., Rulkens, W., Bruning, H., 1999. Photocatalyzed oxidation of alcohols and organochlorides in the presence of native TiO2 and metallized TiO2 suspensions. Part (II): Photocatalytic mechanisms. Wat. Res. 33, 669-676. Chu, W., Wong, C., 2004. Study of he rbicide alachlor removal in a photocatalytic process through the examination of the reaction mechanism. Ind. Eng. Chem. Res. 43, 5027. DItri, F., Andren, A., Dohe rty, R., Wood, J., 1978. An Assessment of Mercury in the Environment. National Academy of the Sciences, Washington, D.C. Devilliers, D., 2006. Semiconductor photocatalysis: St ill an active research area despite barriers to commercialization. Energia, 17, 1-3. Dijkstra, M., Panneman, H., Winkelman, J., Kelly, J., Beenackers, A., 2002. Modeling the photocatalytic degradation of formic acid in a reactor with immobilized catalyst. Chem. Eng. Sci. 57, 4895-4907. Doucet, N., Bocquillon, F., Zahraa, O., Bouc hy, M., 2006. Kinetics of photocatalytic VOCs abatement in a standardized reactor. Chemosphere 65, 1188.

PAGE 122

122 Ferguson, M., Hoffmann, M., Hering J., 2005. TiO2-photocatalyzed As(III) oxidation in aqueous suspensions: reaction kinetics and effects of adsorption. Environ. Sci. Technol. 39, 18801886. Gao, X., Wachs, I., 1999. Titania-silica as catalysts: molecular structural characteristics and physico-chemical properties. Catal. Today 51, 233-254. Garner, J., 2001. Air emission control regulations pose new challenges for mills. Pulp and Paper 75, 44-46. Garton, M.J., 2005. Photocatalytic oxidation of sele cted organic contaminan ts and inactivation of microorganisms in a continuous flow reactor packed with titania-doped silica. Masters Thesis, University of Florida. Guthrie, K., 1974. Process plant estimating, eval uation, and control. Cr aftsman Book Co. of America, Solana Beach, CA. pp 1-604. Hoffmann, M.R., Martin, S.T., Choi, W., Bahnema nn, D.F., 1995. Environmental applications of semiconductor photocatalysis. Chem. Rev. 95, 69-96. Holmes, F., 2003. The performance of a reactor us ing photocatalysis to degrade a mixture of organic contaminants in aqueous solution. Masters Thesis, University of Florida. Huang, M., Tso, E., Datye, E., Prairie, M., Stan ge, B., 1996. Removal of silver in photographic processing waste by TiO2-based photocatalysis. Environ. Sci. Technol. 30, 3084-3088. Hurum, D., Agrios, A., Gray, K., Rajh, T., Thurnauer, M., 2003. Explaining the enhanced photocatalytic activity of Degussa P25 mixed-phase TiO2 using EPR. J. Phys. Chem. B 107, 4545-4549. Kataoka, S., Lee, E., Tejedor-Tejedor, I., A nderson, M., 2005. Photocatalytic degradation of hydrogen sulfide and in situ FT-IR analys is of reaction products on surface TiO2. Appl. Catal. B Environ. 61, 159-163. Kato, S., Hirano, Y., Iwata, M., Sano, T., Take uchi, K., Matsuzawa, S., 2005. Photocatalytic degradation of gaseous sulfur-deposited tita nium dioxide. Appl. Ca tal. B Environ. 57, 109-115. Khan, A., 2006. Modification of activated carbon to improve a queous manganese removal. PhD Dissertation, University of Florida. Kim, B.S., Hong, C.S., 2002. Kinetic study for phot ocatalytic degradatio n of volatile organic compounds in air using thin film TiO2 photocatalyst. Appl. Catal. B Environ. 35, 305315. Klett, M., Maxwell, R., Rutkowski, M., 2002. The cost of mercury in an IGCC Plant (final report). Prepared for The United States Department of Energy National Energy Technology Laboratory.

PAGE 123

123 Knudsen, J., Hottel, H., Sarofim, A., Wankat, P., Knaebel, K., 1999. Mass transfer. In: Green, D., Maloney, J., Perry, R. (Eds.), Perrys Chemical Engineers Handbook. McGraw Hill, New York. pp 72-74. Lee, T., Biswas, P., Hedrick, E., 2001. Comparison of Hgo capture efficiencies of three in situ generated sorbents. Env. Energy Eng. 47, 954-961. Lee, T., Biswas, P., Hedrick, E., 2004. Overall kinetics of heterogeneous elemental mercury reactions on TiO2 sorbent particles with UV irra diation. Ind. Eng. Chem. Res. 43, 14111417. Levenspiel, O., 1999. Pore diffusion resistance. In: Chemical Reaction Engineering, third ed. John Wiley & Sons, New York, pp. 470-477. Li, Y., Wu, C.Y., 2007. Kinetic st udy for photocatalytic oxidation of elemental mercury on a SiO2-TiO2 nanocomposite. Environ. Eng. Sci. 24, 3-12. Li, Y., Wu, C.Y., 2006. Role of moisture in adso rption, photocatalytic oxidation, and reemission of elemental mercury on a SiO2-TiO2 nanocomposite. Environ. Sci. Technol. 40, 64446448. Lichtin, N., Avudaithai, M., Berman, E., Dong, J., 1994. Photocatalytic oxidative degradation of vapors of some organic compounds over TiO2. Res. Chem. Intermed. 20, 755-781. Londeree, D., 2002. Silica-titania composites for wate r treatment. Masters Thesis, University of Florida. Lu, M., Chen, J., Chang, K., 1999. Effect of adso rbents coated with ti tanium dioxide on the photocatalytic degradation of propoxur. Chemosphere 38, 617-627. Ludwig, C., Mazyck, D., Chadik, P., Stokke, J ., 2008. The performance of silica-titania carbon composites for photocatalytic degradation of gray water. Submitted to J. Env. Eng. Minero C., Catozzo, F., Pelizzetti, E., 1992. Role of adsorption in photocatalyzed reactions of organic molecules in aqueous TiO2 suspensions. Langmuir 8, 481-486. National Council for Air and Stream Improveme nt, Inc. (NCASI), 1998. Method CI/SG/PULP 94.02 chilled impinger/silica gel tube test method at pulp mill sources for methanol, acetone, acetaldehyde, methyl ethyl ketone and formaldehyde, methods manual (03.B.003). NCASI, Research Triangle Park, NC. Nguyen, V., Amal, R., Beydoun, D., 2003. Effect of formate and methanol on photoreduction/ removal of toxic cadmium ions using TiO2 semiconductor as photocatalyst. Chem. Eng. Sci. 58, 4429-4439. Nguyen, V., Beydoun, D., Amal, R., 2004. Photocatalyt ic reduction of selenite and selenate using TiO2 photocatalyst. J. Photochem. Photobiol. A 171, 117-124.

PAGE 124

124 Morrow, B., Gay, I., 2000. Infrared and MNR characte rization of the silica surface. In: Papirer, E. (ed.), Adsorption on Silica Surfaces. Marcel Dekker, Inc., New York, pp. 9-61. Murov, S., Carmichael, I., Hug, M., 1993. Potass ium ferrioxalate actinom etry. In: Handbook of Photochemistry. Mercel De kker, New York, pp. 299-305. Noguchi, T., Fujishima, A., Sawunyama, P., Hashimoto, K., 1998. Photocatal ytic degradation of gaseous formaldehyde using TiO2 film. Environ. Sci. Technol. 32, 3831-3833. Nawrocki, J., 1997. The silanol group and its role in liquid chromatography. J. Chromatogr. A 779, 29-71. Obee, T., Brown, R., 1995. TiO2 photocatalysis for indoor air appl ications: Effect of humidity and trace contaminant levels on oxidation rates of formaldehyde, toluene, and 1,3butadiene. Environ. Sci. Technol. 29, 1223-1232. Ollis, D., Pelizzetti, E., Serpone, N., 1991. Destruction of water contaminants. Environ. Sci. Technol. 25, 1523-29. Parida, S., Dash, S., Patel, S., Mishra, B., 2006. Adsorption of organic molecules on the silica surface. Adv. Colloid Interfac. 121, 77-110. Pavlish, J., Sondreal E., Mann, M., Olson, E ., Galbreath, K., Laudal, D., Benson, S., 2003. Status Review of Mercury Control Options fo r Coal-fired Power Plants. Fuel Process. Technol. 82, 89-165. Peral, J., Domenech, X., Ollis, D., 1997. He terogeneous photocatalysis for purification, decontamination and deodorization of ai r. J. Chem. Technol. Biotechnol. 70, 117-140. Pitoniak, E., Wu, C. Y., Londeree, D., Mazy ck, D., Bonzongo, J.D., Powers, K., Sigmund, W., 2003. Nanostructured silica-gel doped with TiO2 for Hg vapor control. J. Nanopart. Res. 5, 282-292. Pitoniak, E., 2004. Evaluation of nanostructured silica-titania composites in an adsorption/ photocatalytic oxidation system for elemen tal mercury vapor control. Master of Engineering Thesis, University of Florida. Pitoniak, E., Wu, C. Y., Powers, K. W ., Sigmund, W., 2005. Adsorption enhancement mechanisms of silica-titania nanocomposites for elemental mercury vapor removal. Environ. Sci. Technol. 39, 1269-1274. Portela, R. Sanchez, B., Coronado, J., Ca ndal, R., Suarez, S., 2007. Selection of TiO2-support: UV transparent alternatives and long-term use limitations for H2S removal. Catal. Today 129, 223-230. Puri, B., 1970. Surface complexes on carbons. In: Walker, P. (Ed.), Chemistry and Physics of Carbon. Marcel Dekker, New York, pp. 194-247.

PAGE 125

125 Rodriguez, S., Almquist, C., Lee, T., Furuuchi M., Hedrick, E., Biswas, P., 2004. A mechanistic model for mercury capture with in situ-generated titan ia particles: Role of water vapor. J. Air & Waste Manage. Assoc. 54, 149-156. Satterfield, C., 1970. Mass transfer in heteroge neous catalysis. MIT Press, Cambridge, MA. Schettler, T., 2001. Toxic threat s to neurological developmen t of children. Environ. Health Persp. Supplements 109, 813-817. Serpone, N., 1995. Brief introductory remarks on heterogeneous photocatalysis. Sol. Energ. Mat. Sol. C. 38, 369-379. Shul, Y., Kim, H., Haam, S., Han, H., 2003. Photocatalytic char acteristics of TiO2 supported on SiO2. Res. Chem. Intermed. 29, 849-859. Sing, K., Everett, D., Haul, R., Moscou, L., Piero tti, R., Rouquerol, J., Siemieniewska, T., 1984. Reporting physisorption data for gas/solid systems. Pure Appl. Chem. 57, 603-619. Someshwar, A., 1994. NCASI technical bulletin 678: Volatile organic em issions from pulp and paper mill sources part IV kraft brownsto ck washing, screening, and rejects refining sources. National Council for Air and Stream Improvement, Research Triangle Park, NC. Sopyan, I., 2007. Kinetic analysis on photocatalyt ic degradation of gaseous acetaldehyde, ammonia and hydrogen sulfide on nanosized porous TiO2 films. Sci. Technol. Adv. Mat. 8, 33-39. Southworth, G.R., Lindberg, S.E., Zhang, H ., Anscombe, F.R., 2004. Fugitive mercury emissions from a chlor-alkali factory: sour ces and fluxes to the atmosphere. Atmos. Environ. 38, 597-611. Thommes, M., 2004. Physical adsorption characterization of ordered and amorphous mesoporous materials. In: Lu, G., Zha, X. (Eds), Na noporous Materials: Science and Engineering. Imperial College Press, London, pp. 317 364. Torimoto T, Ito S, Kuwabata S, Yoneyama H., 1996. Effects of adsorbents used as supports for titanium dioxide loading on photocatalytic de gradation of propyzamide. Environ. Sci. Technol. 30, 1275-1281. Travert, A., Manoilova, O., Tsyganenko, A., Ma uge, F., Lavalley, J., 2002. Effect of hydrogen sulfide and methanethiol adsorption on acidic propert ies of metal oxides: An infrared study. J. Phys. Chem. B 106, 1350-1362. Tsumura, T., Kojitani, N., Umemura, H., T oyoda, M., Inagaki, M., 2002. Composites between photoactive anatase-type TiO2 and adsorptive carbon. Appl. Surface Sci. 196, 429-436. Tsuru, T., Kan-no, T., Yoshioka, T., Asaeda, M ., 2003. A photocatalytic membrane reactor for gas-phase reactions using porous titanium oxide membranes. Catal. Today 82, 41-48.

PAGE 126

126 Turchi, C., Ollis, D., 1989. Mixed reactant photo catalysis: Intermediates and mutual rate inhibition. J. Catal. 119, 483-496. Turchi, C., Ollis, D., 1990. Photocatalytic de gradation of organic water contaminants: Mechanisms involving hydroxyl radi cal attack. J. Catal. 122, 178-192. Uchiyama, H., Suzuki, K., Oaki Y., Imai, H., 2005. A novel adsorbent photocatal yst consisting of titania and mesoporous silica nanopartic les. Mat. Sci. Eng. B Solid 123, 248-251. U.S. Environmental Protection Agency, 1997a. Mercury Study Report to Congress Volume IV: An Assessment to Exposure to Mercury in the United States, EPA-452/R-97-006. U.S. Government Printing Office, Washington D.C. U.S. Environmental Protection Agency, 1997b. Mercury Study Report to Congress Volume VIII: Evaluation of Mercury Control Technol ogies and Cost, EP A-452/R-97-010. U.S. Government Printing Office, Washington D.C. Varma, V., 2003. Experience with the collection, transport, and burning of kraft mill high volume low concentration gases, Technical bulletin 03-03. Nationa l Council for Air and Stream Improvement, Gainesville, FL. Vohra, M., Tanaka, K., 2003. Photocatalytic degr adation of aqueous pollutants using silicamodified TiO2. Water Res. 37, 3992-3996. Wu, C.Y., Lee, T.G., Tyree, G., Arar, E., Bi swas, P., 1998. Capture of mercury in combustion systems by in situ generated titania particle s with UV irradiation. Environ. Eng. Sci. 15, 137-148. Xu, Y., Zheng, W., Liu, W., 1997. Enhanced photocatalytic activity of supported TiO2: dispersing effect of SiO2. J. Photochem. Photobiol. A 122, 57-60. Yang, C., Chen, C., 2005. Synthesis and characterization of silica-capped titania nanorods: An enhanced photocatalyst. Appl Catal. A General 294, 40-48. Yamakata, A., Ishibashi, T., Onishi, H., 2003. Effects of water addition on the methanol oxidation on Pt/TiO2 photocatalyst studied by time-re solved infrared absorption spectroscopy. J. Phys. Chem. B 107, 9820-9823. Yong, T., Schwartz, S., Wu, C.Y., Mazyck, D.W., 2005. Development of TiO2/AC composite by dry impregnation for the treatment of metha nol from humid air streams. Ind. Eng. Chem. Res. 44, 7366-7372.

PAGE 127

127 BIOGRAPHICAL SKETCH Jennifer Stokke was born to Susan and Wayne Stokke on August 26, 1981. She grew up in Tarpon Springs, FL and graduated from East Lake High School in 1999. She then moved to Gainesville in 1999 to attend the University of Florida, where she studied environmental engineering and graduated with highest honors in December of 2003. During her last year as an undergraduate, Jennifer began a research project focusing on the purification of gaseous emissions. She continued this work during her gr aduate studies at the University of Florida under the guidance of Dr. David Mazyck.