<%BANNER%>

Removal of Elemental Mercury from Flue Gas Using Nanostructured Silica/Titania/Vanadia Composites

Permanent Link: http://ufdc.ufl.edu/UFE0021202/00001

Material Information

Title: Removal of Elemental Mercury from Flue Gas Using Nanostructured Silica/Titania/Vanadia Composites
Physical Description: 1 online resource (125 p.)
Language: english
Creator: Li, Ying
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: air, capture, catalyst, coal, combustion, emission, flue, gas, interference, irradiation, kinetics, measurement, mercury, moisture, nanostructure, oxidation, photocatalyst, silica, titania, ultraviolet, vanadia
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: As a highly toxic pollutant, mercury (Hg) tends to bioaccumulate in the food chain and exerts adverse effects on human health. The U.S. EPA issued the Clean Air Mercury Rule in 2005 to permanently cap and reduce Hg emissions from coal-fired power plants. A low-cost methodology using a SiO2-TiO2 nanocomposite as a photocatalyst has been recently developed to effectively remove elemental Hg (Hg0) under room conditions. In this research, a bench-scale fixed-bed reactor system has been established and Hg0 removal on the SiO2-TiO2 nanocomposite was examined under both room and flue gas conditions. A kinetic study showed that Hg oxidation on the SiO2-TiO2 nanocomposite under UV irradiation followed the Langmuir-Hinshelwood rate expression. The flue gas components were found to have significant effects on Hg0 removal using the SiO2-TiO2 nanocomposite. HCl and SO2 promoted Hg0 oxidation, while water vapor and NO significantly inhibited Hg removal. The mechanisms of these promotional and inhibitory effects were thoroughly explored in this research. The active phase of the selective catalytic reduction (SCR) catalyst, V2O5, was added to the SiO2 or SiO2-TiO2 composites in an effort to improve the catalytic activity for Hg removal. No UV light activation is needed for the V2O5 doped catalysts, which is a great advantage over the SiO2-TiO2 composites. The Hg0 removal efficiency increased as the V2O5 loading increased from 2 to 8% but decreased as it further increased to 10%. The optimal V2O5 loading was found to be somewhere between 5 and 8%. The SiO2-TiO2-V2O5 exhibited a greater ability of oxidizing Hg compared to the SiO2-V2O5. It was suggested that the Hg oxidation on the V2O5 doped catalysts follows an Eley-Rideal mechanism where HCl, NO, and NO2 are first adsorbed on the catalyst surface and then react with gas-phase Hg0. This research also reported that atomic absorption spectrometry based continuous mercury monitors are subject to interferences by ozone due to its strong absorption bands near the Hg absorption line. On the other hand, Hg interferes with ozone measurement which is based UV adsorption. These mutual interferences can consequently affect the risk assessment of human exposure to both Hg and ozone.
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.
Statement of Responsibility: by Ying Li.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Wu, Chang-Yu.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021202:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021202/00001

Material Information

Title: Removal of Elemental Mercury from Flue Gas Using Nanostructured Silica/Titania/Vanadia Composites
Physical Description: 1 online resource (125 p.)
Language: english
Creator: Li, Ying
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: air, capture, catalyst, coal, combustion, emission, flue, gas, interference, irradiation, kinetics, measurement, mercury, moisture, nanostructure, oxidation, photocatalyst, silica, titania, ultraviolet, vanadia
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: As a highly toxic pollutant, mercury (Hg) tends to bioaccumulate in the food chain and exerts adverse effects on human health. The U.S. EPA issued the Clean Air Mercury Rule in 2005 to permanently cap and reduce Hg emissions from coal-fired power plants. A low-cost methodology using a SiO2-TiO2 nanocomposite as a photocatalyst has been recently developed to effectively remove elemental Hg (Hg0) under room conditions. In this research, a bench-scale fixed-bed reactor system has been established and Hg0 removal on the SiO2-TiO2 nanocomposite was examined under both room and flue gas conditions. A kinetic study showed that Hg oxidation on the SiO2-TiO2 nanocomposite under UV irradiation followed the Langmuir-Hinshelwood rate expression. The flue gas components were found to have significant effects on Hg0 removal using the SiO2-TiO2 nanocomposite. HCl and SO2 promoted Hg0 oxidation, while water vapor and NO significantly inhibited Hg removal. The mechanisms of these promotional and inhibitory effects were thoroughly explored in this research. The active phase of the selective catalytic reduction (SCR) catalyst, V2O5, was added to the SiO2 or SiO2-TiO2 composites in an effort to improve the catalytic activity for Hg removal. No UV light activation is needed for the V2O5 doped catalysts, which is a great advantage over the SiO2-TiO2 composites. The Hg0 removal efficiency increased as the V2O5 loading increased from 2 to 8% but decreased as it further increased to 10%. The optimal V2O5 loading was found to be somewhere between 5 and 8%. The SiO2-TiO2-V2O5 exhibited a greater ability of oxidizing Hg compared to the SiO2-V2O5. It was suggested that the Hg oxidation on the V2O5 doped catalysts follows an Eley-Rideal mechanism where HCl, NO, and NO2 are first adsorbed on the catalyst surface and then react with gas-phase Hg0. This research also reported that atomic absorption spectrometry based continuous mercury monitors are subject to interferences by ozone due to its strong absorption bands near the Hg absorption line. On the other hand, Hg interferes with ozone measurement which is based UV adsorption. These mutual interferences can consequently affect the risk assessment of human exposure to both Hg and ozone.
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.
Statement of Responsibility: by Ying Li.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Wu, Chang-Yu.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021202:00001


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 E20101114_AAAABX INGEST_TIME 2010-11-14T09:26:45Z PACKAGE UFE0021202_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 114263 DFID F20101114_AAAVGL ORIGIN DEPOSITOR PATH li_y_Page_115.jp2 GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
0efb351508fb3f3c91f2c6cdd560c429
SHA-1
2f484a3ab54c02686d44910622c75674a9637b20
25271604 F20101114_AAAVHA li_y_Page_006.tif
10b7c0594681bdeb1c5613a0c4167355
da6bd373b20ace1935c47ac645e1415486138828
11622 F20101114_AAAVGM li_y_Page_116.jp2
7c814d8b3ddefd361cf4701e3e648c2b
842186db0945583ddc6bdcae69ad4456508949aa
116729 F20101114_AAAVFX li_y_Page_099.jp2
c49aafe6771cf19a168910423d48ff2f
1e32b119c816b5abb9ef8bba27aba413937db965
F20101114_AAAVHB li_y_Page_007.tif
301d0d168f850e1f7c2224d6bcd3767f
4d8069e464d1c1b54c57617de8eaea5d3834306e
133404 F20101114_AAAVGN li_y_Page_118.jp2
4d68ecb9477488ad3645f476995dd38d
38d2f73caae47d31bc2da384c0a5729f728bead8
119222 F20101114_AAAVFY li_y_Page_100.jp2
065658dff8e9d5b90af5b6d2a258e47e
434c575e93bf4ae1b002c9afa0a8cde098a84303
F20101114_AAAVHC li_y_Page_008.tif
49becba5745ea754aa29009ca9ac765d
25c6df010bad4a1bb0abd2fc19663fa91157d01d
140490 F20101114_AAAVGO li_y_Page_119.jp2
5a6995d1d611dc3bb768bc146116e4de
62539b81ffdfd651e2ff2a612c7a9c2618e7a583
115734 F20101114_AAAVFZ li_y_Page_101.jp2
e506545e62fe6b11a3be52d78ec5aa2e
6b64df60cc552fb10150f8f8d4d36d19b2b8e34b
F20101114_AAAVHD li_y_Page_009.tif
73b6662540afe388d2203d4390271d9b
758884db626be3e8d0544a2724e5e5d4ac418c1c
1051984 F20101114_AAAVGP li_y_Page_120.jp2
492bdada375851db8670fdc8c7a8b954
0d3474f3802603a122eb59b942f36b071f201c30
F20101114_AAAVHE li_y_Page_010.tif
a37b87a6051869498ba86f5b4417df9c
4aec5735b4e85b78880a6751aece46a6a3e471c0
132401 F20101114_AAAVGQ li_y_Page_121.jp2
7dc41ae9eac3256c8d842a08f09c0f7c
b6ed796c02e1d4ea9a18a1f8206aa78f7ecca170
1053954 F20101114_AAAVHF li_y_Page_011.tif
40c5e5c691b489394c8aef6c5d86e2df
ad6e4be39104b11f198b24801a4e24d82f33698f
134381 F20101114_AAAVGR li_y_Page_122.jp2
4c2963e147566b7477933462b58608d3
df6e7015f591241f3eabb6f13ab58c806c650c2b
F20101114_AAAVHG li_y_Page_012.tif
086a45936b4f3273338cb55ed4698da1
c9d773e22573048be64cacf6fd4b3b8d995acc0b
1051954 F20101114_AAAVGS li_y_Page_123.jp2
945395abfa94578255d09d56b21bc43f
47c742cdf6d5e67a6b9ab939b715c4d9286cbb9f
F20101114_AAAVHH li_y_Page_013.tif
786a30bcb599f391999662f66a0db4a6
7c44695b0ceeafdede434ab6359b5534a2c323a3
1051964 F20101114_AAAVGT li_y_Page_124.jp2
40a0300d98237e6913e05b765178eb64
c85188d59c8b22cd58a7b4daf45739a60e9ecfbe
F20101114_AAAVHI li_y_Page_014.tif
773a41c6e373035edf53ce7318c55c0e
a2764c56402b1a6de43ac01870e92776c6a544c3
86331 F20101114_AAAVGU li_y_Page_125.jp2
5262ab09b1655ee0f243c219b42a4f88
4142fb4d5372699968aa396c6a2879af7fb6e38d
F20101114_AAAVHJ li_y_Page_015.tif
4382ed98dd605c58eba615d73cc7e635
872ee581ab334e36c9aa22194f6fa9806fca7f77
F20101114_AAAVGV li_y_Page_001.tif
b1a6f4df5fa8607a70ae843a48567440
67c1b886ac587f3e9b61a3a2c30b9955f7ef62c6
F20101114_AAAVHK li_y_Page_016.tif
051dd9aeac27c45a2cbe2aa0d07d90c2
d0fe7fb2110d610573013f93ec391f6aa3c0a594
F20101114_AAAVGW li_y_Page_002.tif
8a0b0addc5deb5a680357b9c09a8f8a3
3d9c3dac0434ca02219792fae1ab6e124d7336a6
F20101114_AAAVHL li_y_Page_017.tif
b9064c2abd99f8fafba73a51c4ebed26
c06fedb7695b2597c8900136d6d1463e5ee11f29
F20101114_AAAVGX li_y_Page_003.tif
150d62174687dd76264d6dc8c6f90bee
e9e92a9a2b95d1e4cf39522252c409011716d31a
F20101114_AAAVHM li_y_Page_018.tif
90165692965c330b156e60a757a3fbd1
d79f4da0c733990c1b5b798da0d0157912b43bc5
F20101114_AAAVIA li_y_Page_032.tif
50700da922788ceb633cfc1f179a35f8
aa8b90b85f79fcda5f121faf2d40fe64adcd1d74
F20101114_AAAVHN li_y_Page_019.tif
8bd9e1aac5da91cb0729202537d83416
6c7cec313e5671efa555732145a82ec38d740835
F20101114_AAAVGY li_y_Page_004.tif
d12fc6dc78bfc41a8f06d0fda5a8effd
19510489eff07b4b12cb656776e7f0be86c92499
F20101114_AAAVIB li_y_Page_034.tif
c35481e8f7ceb80e358b27b555f9315b
fff70e45d2202774db54e2f088ffd7677fee4c05
F20101114_AAAVGZ li_y_Page_005.tif
1bdf0dd59dae8ac09c9f1c6e364510b9
a19fb41724a1e49fb2fb9dd434a6089af5e0df3b
F20101114_AAAVIC li_y_Page_035.tif
4b10fd2e9afd6b255abc13030cfe2eb0
e7960bb2de28002c2f624377dae713b65f6cfb86
F20101114_AAAVHO li_y_Page_020.tif
7fd866f006671c809afa872a395e69cd
de4073e80c2450effd5875fe8d0000f66a5abc2f
F20101114_AAAVID li_y_Page_036.tif
7e72f0d4585fda88081469859af46bac
2e0cbe99dcfa161afdc16bb39c63162b6a0ad6d4
F20101114_AAAVHP li_y_Page_021.tif
4ecc2207226f475d87655b6acaf4f161
dc2997152458c4b6f74380ab33fae57c9028a182
F20101114_AAAVIE li_y_Page_037.tif
29f54c7688f97d0af4701badbf4e916a
532e3e251ae041300fab7f09ccfee38426fa8e6c
F20101114_AAAVHQ li_y_Page_022.tif
d92b6b12ce8213e2a217bd5d1038517e
13bacc8457092afbb9a42487d1d9427e214725cb
F20101114_AAAVIF li_y_Page_038.tif
ab7803d37d3e2ffae75b177eddb9d0ea
3e08b91f86a7b28ac7e02b3c7d75aa35f3a94b19
F20101114_AAAVHR li_y_Page_023.tif
6592795e74806dd34d767534d717bee2
da0cce578ee9e48950547c98e0e89b92577fc63a
F20101114_AAAVIG li_y_Page_039.tif
024999ca3afe2f1ceb09b3e2849a8256
7d4c32a51927e20d339195dd49a7b744b8979745
F20101114_AAAVHS li_y_Page_024.tif
d10ca3507917f85c5d41348766365347
194842a80351598ae7927b5b1fe523cfa22c1169
F20101114_AAAVIH li_y_Page_040.tif
091f0114670b031c85e8903c4b0cb259
893b31a602201c998320f028721cde01f8bdc8a6
F20101114_AAAVHT li_y_Page_025.tif
804dc2d72b434592ef01d7ddc6f296a1
aa0de4e4e12ae9e29a76d500c0a8dc1cd063ecd1
F20101114_AAAVII li_y_Page_041.tif
c9f8dc168ab41d67ea5c3b7f0d562cc0
5b5cfda89b807fb04acb2102fcdd0db29c7715e0
F20101114_AAAVHU li_y_Page_026.tif
a4e94b8ae7e83122e8b6568094c7fb59
5867118a641e803598465a347e4e223860fdc4b6
F20101114_AAAVIJ li_y_Page_042.tif
ebd55d7ed0da5c6be2a2cd77a27f6b11
4814cec3e4c480b1ae546967b6e0c98b7406a0d3
F20101114_AAAVHV li_y_Page_027.tif
8a218c5a750efbf516fb05a03db25f16
c3123a0273134c688053f7455817b2c36f3a73ba
F20101114_AAAVIK li_y_Page_043.tif
031147f075300492e24238394799797b
273f5e0f5c730dfe9de0356ee20f25038d0f07a7
F20101114_AAAVHW li_y_Page_028.tif
952046b2b70ae4c3e60d5e1d9116c168
cca55fa86bb1299cf4cd5f11f41c681b21ccd04c
F20101114_AAAVIL li_y_Page_044.tif
7344bb0a08e6888c8113d4c8cd4a0042
f58a5a50d8221cf61d9d89a53f7a344d909fdd01
F20101114_AAAVHX li_y_Page_029.tif
5db930d1ea8983b127e6ea3484adfd7f
caa406ad16113b1042447b3a9d092800b2f16edc
F20101114_AAAVJA li_y_Page_060.tif
aa3b4ef163e1c02223b62db0f27c4a79
e26bfb9e0561de7e417512f46a15173d02146388
F20101114_AAAVIM li_y_Page_045.tif
cf7503754c047076e9573a677fc6b7d3
3560c31dfd836357458d398ab1d49299d0b0c4cc
F20101114_AAAVHY li_y_Page_030.tif
b4fda66fdecfc1108beaa33b47461820
40e7ca641ca86100c998f2579451dd4a5417499d
F20101114_AAAVJB li_y_Page_061.tif
7dfe05a97a657a0668f93acd95a7f6fd
b452f9e1af1bc573354b47cd8c1ec747761b1028
F20101114_AAAVIN li_y_Page_046.tif
3ab82b44ebca19003df345f05b80fe16
66124d0884678243f8a9c5ec0d488dcc7a4448e3
F20101114_AAAVJC li_y_Page_062.tif
dc6fdd1e60dc513797b0682e0c6f9d01
a2292d85ad2d80ad8d7939f98dfaf1d077572d3d
F20101114_AAAVIO li_y_Page_047.tif
01f7d78c293153e99959f5a309351d37
2b7d9c148fa7d2d691a4625a9376e5ce0715a9ec
F20101114_AAAVHZ li_y_Page_031.tif
4e563b236e3707b5115fc92eea345d16
f48e84c69b8f416aa261623c317a52bee4acf635
F20101114_AAAVJD li_y_Page_063.tif
fa7a0cd9603518d930a07e5d020b9768
baa27a74b0f9d2ffea90950b8f1e913d0ed7731f
F20101114_AAAVIP li_y_Page_048.tif
e767588097e59a7855bad31d9f701311
ce8721b553af954b28d9ee59dc2df555ecf91a3a
F20101114_AAAVJE li_y_Page_064.tif
29ead3eed27b63372cff398e37b3203e
787bf79797c5c879d82a9ddcd7bcd423f8a44601
F20101114_AAAVIQ li_y_Page_049.tif
32ff0129156b1ee0fd8ad13ba450165e
4ddfb3c9f639631a6153e115948263705003588f
F20101114_AAAVJF li_y_Page_066.tif
f468aaf8505cfc03033a641225cd09db
17f8a78913fc4a201c8ed38745d3d1aa925665a8
F20101114_AAAVIR li_y_Page_050.tif
595609281cbe75a96372f390f7739f02
3627c3865905ecbd8cb3fc6292022223a42e6941
F20101114_AAAVJG li_y_Page_067.tif
5768b0bf2eafff5160bdc6ef497f4915
d4cfee5fb1828c14a1ef93004bdbfc12b854499c
F20101114_AAAVIS li_y_Page_051.tif
8ac8b063aa7485189cd0617533ec9b9e
d31b53fa95751872f0bc0c143412c3a842002ff8
F20101114_AAAVJH li_y_Page_068.tif
a550880c7c9639c3ac3ad1a9abcb3c94
dd292680ae0b17307bb615bb7e5f9988f41dba4c
F20101114_AAAVIT li_y_Page_052.tif
c936d6a8bf39c8881736171cb3c1a4b2
19e62eb29f74d2460be9c24f8096ccf3f9dbe75c
F20101114_AAAVJI li_y_Page_069.tif
7b162995f54c02c8e7577759728d249f
84f475ff22e2c8d73775906b7703344373b69f19
F20101114_AAAVIU li_y_Page_053.tif
4af90ff2f46cbceefad4024f20dabfaf
67c86df2573f879da66c978a256ba3f0081b3676
F20101114_AAAVJJ li_y_Page_070.tif
dcb17b82801ee6529bfa37c4e0c57ada
17231de81a616b3f9f583425ce747676eac90623
F20101114_AAAVIV li_y_Page_054.tif
7ab5f54cbc825973eb16e270db78320b
b3c0b3676bb76ec2415c691a2fcc4e6dab2a77e1
8423998 F20101114_AAAVJK li_y_Page_071.tif
02c00cc9dcb481bd58d93cc41f012d01
f0e3327d09526f29c771543f2f3b38860f29942c
F20101114_AAAVIW li_y_Page_055.tif
4767eb36473061a431cd358195f14993
7d04fe9047cee5c61d89178e36192d281d9e301e
F20101114_AAAVJL li_y_Page_072.tif
c38c37b8a0c41765340b1ecfbae58da9
e41b0ff000379a936ad2449d82c9f29a2d2979f4
F20101114_AAAVIX li_y_Page_056.tif
a7a49a0f8edfd4fdaf139033d0076c6c
9753534bc416c321f737467df20ef2171316222f
F20101114_AAAVJM li_y_Page_073.tif
588a23a52ca7fb101b6b39eeb4920a43
8e2b46c218af93839385b43ba6049adcb763a7db
F20101114_AAAVIY li_y_Page_057.tif
0a4c31a852e2359fc1d5efc21e281fff
f30bd80de2febe0d469d5e315040b4d886e52218
F20101114_AAAVKA li_y_Page_087.tif
7bc1bf5a3df4ae7ab051e9288ca6ebc5
74e8a9322b913df346969cde1ac1a892cc56f2f3
F20101114_AAAVJN li_y_Page_074.tif
85c746f9b8b9a33f8736f04fda158293
e7ddf1cd667fb93c1f7f0073edb66ff58201e77f
F20101114_AAAVIZ li_y_Page_058.tif
0772f2b96db3b31699b98b7018e5ba3e
8dd173422ffcf8fea29bc7df8a0edf68048ef3d9
F20101114_AAAVKB li_y_Page_088.tif
530ad2d6dc6c625d8a861b8871bfa0cf
2c365a17f2aafc24c4072b128d7fbe09d1074cbb
F20101114_AAAVJO li_y_Page_075.tif
26676ea11ba94442b7584d5dc40fa8d7
b4b5e4ac7943b86443fb86561ba2dbd5bf9bbd4a
F20101114_AAAVKC li_y_Page_089.tif
ef5ec421d5a2ef1b5fba1bf78c90deae
e2f68e9ffd2333cf6027b1d72abec147bcbe21a7
F20101114_AAAVJP li_y_Page_076.tif
d63072969b165d5ddeb66d9fe8b8d5e1
3366bc358ca08868d1d27d1726ba6545205221de
F20101114_AAAVKD li_y_Page_090.tif
f8011d471bc54c038515741bbb242fce
55075b8cb3c4d31b10970e492a9fbdea3eb458c3
F20101114_AAAVJQ li_y_Page_077.tif
807cd0f76ebfd97af881fb5062f7268f
d250e384025211ed7298bf63971e2f4fa8cb71d5
F20101114_AAAVKE li_y_Page_091.tif
f47f40695e005907db0006435920d42d
fe040d4bae5b67404c44364eb68f4bd417d3c69e
F20101114_AAAVJR li_y_Page_078.tif
757bde9833007bcdd0d1f6d153f8895d
ab6acd0756b8bde06cff9d9ced9e37e181f2ea1c
F20101114_AAAVKF li_y_Page_092.tif
1985f865679d9868bb15994b76b82ff5
3b8aae969358d2f6317d25ecfbb1a27ff9b0c421
F20101114_AAAVJS li_y_Page_079.tif
818c3891167adbecdfa7c44a8a0a801b
74a7f3d9bfde35cc48e424912591cc1fcbb2ba9f
F20101114_AAAVKG li_y_Page_093.tif
a21d521496a021843b1beb5e1c6e0a48
598ffcae70551f09292c030c15476dcaf7be2a90
F20101114_AAAVJT li_y_Page_080.tif
5c63a323d07faea203b1b209820faf30
6b53a6b9e0fe8e1b9b40a388b9c396af0e0572f3
F20101114_AAAVKH li_y_Page_094.tif
ab13740ce408b18706378138a1310bd2
84647bb4e1bf0aa1b79f0fda32a507d7d4159c06
F20101114_AAAVJU li_y_Page_081.tif
04118526959ae6ae01419dfd65f5aff9
d2dd5caf891b6b7651afea8e55e0ed18f4ade4cf
F20101114_AAAVKI li_y_Page_095.tif
44944259692ffd1fde6d42d297071379
ef37dc39670cc2b5fbede65f2fac1f112c5c31d4
F20101114_AAAVJV li_y_Page_082.tif
3d791eca3b63404619692db7b9eaf1a8
f88d7309eb1d6c88b9e3d3927641e552d9198ca5
F20101114_AAAVKJ li_y_Page_097.tif
e5fcbb2a4f682762eaf7043b2e502639
6b200e450870b4769dfcd23bdd0c4a66423e3baf
F20101114_AAAVJW li_y_Page_083.tif
d6bbd4635a3b902a3fdf65760c105b81
e80ad69e312c7f5835e543ba80b3c3ee60f676c8
F20101114_AAAVKK li_y_Page_098.tif
1cfdb81a1b701733d7e2c34ff03c5664
3fc34ce8feecb73b80fbcdb4c054a821a3e9917d
F20101114_AAAVJX li_y_Page_084.tif
d44e36c74c91f24af10bff3ca3bb03c4
1ba7c0ee761818059a01ec021ae42105775337ac
F20101114_AAAVKL li_y_Page_099.tif
0e9e6f3ad3ff604cd7013591df698af8
8e8364c29eacacfe5e484407d439407339186ad7
F20101114_AAAVJY li_y_Page_085.tif
17e8aa16a6fa193dc35758109a5bfb3d
59718a27c1ba0ae4281bfa641069e90f09bf0f60
F20101114_AAAVLA li_y_Page_114.tif
eaf8265df2197573e5246395a7cd5f74
837f3c005d05c93ee7682cb576e86490fe28cc68
F20101114_AAAVKM li_y_Page_100.tif
8fb0b3f2db934d0653ed78b24eec7fb4
9c536dd803eb49e6d1ae7f03b804f5466f8c6fd0
F20101114_AAAVJZ li_y_Page_086.tif
c9ecad8aa921fed8310ce856b6373b80
b6af06bbf983819035132558199ee457c3fec0b8
F20101114_AAAVLB li_y_Page_115.tif
e5210c566da6b452c1b13ac808c9f9b7
e9f833a88478f09c03adac971b51a9365a7f6862
F20101114_AAAVKN li_y_Page_101.tif
c231ba1f904fc41f483a4c92b0a2a95b
8266c1c07238b525d2c4c8a3bad1b4290c3fd44a
F20101114_AAAVLC li_y_Page_116.tif
daee692ae341107e0f2d286256665f60
93b0fa65d4864aba38274131cfbedb5e4dcf882a
F20101114_AAAVKO li_y_Page_102.tif
8327f2b91c113f5d4b985d57c40872da
c606392ee8428748284e2e6af92f0783a1146018
F20101114_AAAVLD li_y_Page_117.tif
dea77e8d14e7548001902ad219810b05
bde7be75f4f1835da317f99de5fb68e80c3d2ff7
F20101114_AAAVKP li_y_Page_103.tif
a028a9c00f878832ec423229e8a25ad5
58113115b4bab794123efe6a26ebdcd15bd5fc54
F20101114_AAAVLE li_y_Page_118.tif
da899fe18a53ef1f96293888e2c7134c
68b8817edaf2a56465a4796f7400f4b5a9d25861
F20101114_AAAVKQ li_y_Page_104.tif
98c4e9abdb94cd18234f9e570a3a2496
d8c4a85356a1bedb58cd547ee1f89d9c3129cce7
F20101114_AAAVLF li_y_Page_119.tif
ab541a9c662671f02fc470d83a223bdf
81624e01e379200c802818d57b9bfa8024ac2f73
F20101114_AAAVKR li_y_Page_105.tif
8c1649bcabc0b2a195244bd7c2398586
1bb6b01faa57b083e1f193bb76e3de1ee396b92d
F20101114_AAAVLG li_y_Page_120.tif
d081d23870c59250ac89f96092ca02df
d751c982e3618472ceafb6bfbd1065776eaed4d7
F20101114_AAAVKS li_y_Page_106.tif
a11ddc272465bae1e540c4099722a959
98c2a0ae086d74a954fba4f50f07943abff4e953
F20101114_AAAVLH li_y_Page_121.tif
b4edc816895c4709cbc3e28c44d72ffa
ac1f474ae52ad54ab30802dfc206e40a574e09e3
F20101114_AAAVKT li_y_Page_107.tif
13262a8ac970801df622b044b264bbb6
378ad9277feebbc2f75b2a4d263148ad7a304c2b
F20101114_AAAVLI li_y_Page_122.tif
b6df9fbaa4bfe2ec0aacf3d9ca7feac6
47fb962181e31dfe9e97feb1312aab74af3d93ae
F20101114_AAAVKU li_y_Page_108.tif
6a2a961b940aefb42eaf1654d3595da1
0d1549700a283a4eca041a8d43ad70dc07fc4240
F20101114_AAAVLJ li_y_Page_123.tif
c8470fbc5c7d523de865cf6fbe39ab8f
9a322bad561783f4fb445e1bf79d1c194e3368f1
F20101114_AAAVKV li_y_Page_109.tif
2a8f4204b2263929c6ba4c9fca1f1e85
682a249055da14aca45224dea9c6841455b21cce
F20101114_AAAVLK li_y_Page_124.tif
5ee6ddffa1d284bd414b909317cc6e3f
b0b955daef451035ae8249d78ac976ab32c17b5c
F20101114_AAAVKW li_y_Page_110.tif
2702495297d6700eed30e5bf25bb15b8
05a99f663ee498f1f5d047ad8206fa7a2f10a069
F20101114_AAAVLL li_y_Page_125.tif
94f5c956af007ed6baca510045d33e1f
0c04f24c33b21d8af68cb390761a74ed34382bec
F20101114_AAAVKX li_y_Page_111.tif
8c3e6ec88294d6b62aeef3eb5f0bdf5c
dc9911f1c38cdfd3b47e5d21a30b011b45fc607a
54592 F20101114_AAAVMA li_y_Page_016.pro
7f49f32c646d2bac510c0269fc4bbab0
6b1a2b8c1bb113a071993b5a56961df10889c241
8448 F20101114_AAAVLM li_y_Page_001.pro
7159dc3f575f5312a4d3608896796d59
5e39ca7ee144536c76069d801f57a80d00484641
F20101114_AAAVKY li_y_Page_112.tif
fed557c4549e6348d10532c1022b31f1
d380fa29a8951e099c5f3b5f159cfe8ca68e69ca
53943 F20101114_AAAVMB li_y_Page_017.pro
116d16ea6681aec4fe3a907aec4de276
88ae3e986290507650f3169edcce1c3e39a3ee48
644 F20101114_AAAVLN li_y_Page_002.pro
9b0b00e6a1e45f15632fe19fbfe29051
9eb25863a541975d404f9d2142f8feaaed62a8a7
F20101114_AAAVKZ li_y_Page_113.tif
9d5062d3b8317ab107cc9247d0920149
5ed9c384b9269cd154a6e22e600338198888c721
54553 F20101114_AAAVMC li_y_Page_018.pro
e3b87f81357a01f8654aad7578f6876b
ddc196deefd96020e9c5b1e9a0e144df30aff7c5
2308 F20101114_AAAVLO li_y_Page_003.pro
7d804d4142aaa294e1799bd585cb3570
d80ab1bdc2951382f3fc6b4c49c16254e2d01115
45051 F20101114_AAAVMD li_y_Page_019.pro
b5e1f4118fbfc97d8c3a46bfa323f5e7
67338d26edf84bf46d05ee0fd3bc0992fa533443
36346 F20101114_AAAVLP li_y_Page_004.pro
338efa95a63887cb7f7614c940ccec11
b9fcca83a1f27b995d00b11656338a6d25affd8d
55972 F20101114_AAAVME li_y_Page_021.pro
b6d7f67349b7814ed23eb9d9da9ff0a3
786d79ebb45fb46a6a8a98ce3c0b67a02616cd73
61669 F20101114_AAAVLQ li_y_Page_005.pro
4e55ef3d07695c91d3649fef8e2667ad
5368e7c20dd5e220390711ff651a58387fd9c39e
57053 F20101114_AAAVMF li_y_Page_022.pro
28e09a6e1888a95f6cea1e2e5367b613
f1482bc398ab2d189ac383035c87c1a7b7f89bfd
85056 F20101114_AAAVLR li_y_Page_006.pro
f7eb3c9b3f32835df1647022accf9c24
68ce7326743705312f963238d7979233c87fe2ce
51436 F20101114_AAAVMG li_y_Page_023.pro
fdd81eb99b5acabde2873d9d20250fbf
81614970a085753c2a82329716375647ef19c212
14957 F20101114_AAAVLS li_y_Page_007.pro
6cd415087ec557ad05eb0490d570700b
20caed5c15662f319525def0227b82c044b8b5a1
40778 F20101114_AAAVMH li_y_Page_024.pro
56e86673814b696c68fe9d13b97dd841
303c4733df165c856302487a3a9204b1fdcf261b
14413 F20101114_AAAVLT li_y_Page_008.pro
356175809c4c898fbb896c5ab268e38d
82fe8f43da6c1164f4a32d4bd4f6e11493544bb2
43693 F20101114_AAAVMI li_y_Page_025.pro
c138ba1c47978fa972d9149693734e77
3ec93feeb4aeffce0a40d5ab025b6c9fde255801
68982 F20101114_AAAVLU li_y_Page_009.pro
ceab1b1d90004705588c6f5b66069fbb
752ca2869670b5b9ca745a71d6cf4f9758ccd0b9
43743 F20101114_AAAVMJ li_y_Page_026.pro
71bef4b99953abe3171a7a0fb5121f66
34ce3650fac6c2733a0636851c87d6fbbcc0cb4a
27294 F20101114_AAAVLV li_y_Page_010.pro
084bde590267296fe119b6af4a50f3b2
f94840d670d1e3bc703388abec3f55f4cb2fe7cd
51874 F20101114_AAAVMK li_y_Page_027.pro
df46316c321f5a7fef7eacd3cf62dc11
bfb3f577ee4ebdb51391ab26c059b96a608e4dda
47832 F20101114_AAAVLW li_y_Page_011.pro
8335c7a16630f1d927a34fd0ce24e6d9
80024c48bda64249e96fae1295c39361d1528c72
51356 F20101114_AAAVLX li_y_Page_013.pro
94c172204f1da6aa13a4fba8dac41848
ada452971571614ed65a694dca17b9407471dbd4
54560 F20101114_AAAVML li_y_Page_028.pro
97a7a6c76000bca5dd9e8b8d30d54b9e
593f8426661ede6a68b7976bfd936662a26d31fc
54290 F20101114_AAAVLY li_y_Page_014.pro
a73b823481e4f951e728bce8a47b4140
9b83c74f1fa8b466af5bbd991a45e1872dc4cb98
55450 F20101114_AAAVNA li_y_Page_046.pro
b2cdd55d724e5f694f930f2eaa10c3cb
83f1eba04eda0afc5d157c29ebb3bb4673f954cc
55910 F20101114_AAAVMM li_y_Page_030.pro
d75c18721641568cbe22ed3ab7213f94
fe402b0da949f9beccfe6a15b8247902cc109798
54016 F20101114_AAAVLZ li_y_Page_015.pro
ef2b2bd247066fb8a9fbbdd2feb7de00
8844dacc8a0dd2a0a87057ef90de8eba241dbdc7
47590 F20101114_AAAVNB li_y_Page_047.pro
a0c268902342f0e238f27309fb8aa03c
620d5eaffa0ecaf7d4747f632ae69a12b228df46
36343 F20101114_AAAVMN li_y_Page_031.pro
cce9090fa28515d7b7da4050b3cbef73
86130f27a032253fd60fafa0f4977a9f4fb76f13
56193 F20101114_AAAVNC li_y_Page_048.pro
f7b21f405824bf6a88a7919c4d7a1912
a28129528b03cbc362feaf658e59fa4666bef65a
4518 F20101114_AAAVMO li_y_Page_032.pro
76c9f866dbd3926a137b98d4c820b11c
b49f38c83e3f02674279e23fa570ac8ed0cb6832
50091 F20101114_AAAVND li_y_Page_049.pro
88685cfa1fb49ba23a16e2f1d08e3969
146e95dfdbf65f1cf62668b0683fb68692750567
4283 F20101114_AAAVMP li_y_Page_033.pro
4218f1e234792d0fa8a497f5c5a51bcf
29e91b90de28358886ae00c264dd88f16cf7bd2f
51260 F20101114_AAAVNE li_y_Page_050.pro
3f7d71f5bc084671e1f9240898e75c5d
23b10b3aaf161520f3d4c710659c6058016264ad
8493 F20101114_AAAVMQ li_y_Page_034.pro
10b1bb349d35a191fefa05cb339fdae7
c097cb12d327febdb17b4f21e7a9edfcb1ee2982
55443 F20101114_AAAVNF li_y_Page_051.pro
4e6698d2053912e510802fe66c2abc63
618051f994d35d833acd48abffa93d6ee3f71eef
9297 F20101114_AAAVMR li_y_Page_036.pro
81ef59b202e78c878a60b84672e640c5
3c434b44eeb48fbb6d473c1877456c411f195395
4897 F20101114_AAAVNG li_y_Page_052.pro
756572975c2ada1d3b625c7e2daea706
6b648be9a75270c12546807ce45fd647668deb31
10131 F20101114_AAAVMS li_y_Page_037.pro
137afe2d547220be349ee23de86ec70a
e130ebdb05c482bac5f3182fbd4f5dd491ddd1f8
8736 F20101114_AAAVNH li_y_Page_053.pro
6dc58caf343a149a6f059547a7a5b615
c1a71f65685b50545afaaacc01f7b6917a9cd910
7073 F20101114_AAAVMT li_y_Page_038.pro
2e8c3bae0470dce9cf0864b45ff14a69
2332cc7b39d2234eaf474afaf4d26fc1f6a652f7
13909 F20101114_AAAVNI li_y_Page_054.pro
3301e897dfaba8d05fc87fe9f36986d1
d5df7563016acd56d2beaa376b117823dc276277
10479 F20101114_AAAVMU li_y_Page_039.pro
f3b2e5685f11c1a7f065a2c6e0bba0aa
f51f2697b7d4ec08ab31b4746e1249452240ed94
7964 F20101114_AAAVNJ li_y_Page_055.pro
27a1d2a58ed792e70ab8608d0f878d0d
9f259b2a0669c0459ddef3fba9e97e06ac806f66
55293 F20101114_AAAVMV li_y_Page_040.pro
7b3a60030f40b4604f9942394fa1aaae
179242fa11688041b684b5efffcc7de058c615f6
10315 F20101114_AAAVNK li_y_Page_056.pro
d1fbb8bd3352d37401c896125d01b2d3
17e15538e46e22076462d27f6ee4d60c873f1845
48249 F20101114_AAAVMW li_y_Page_041.pro
335440e991d3689e2b345b351e16cbbb
1c1d2051e1c4920f770a0c0fe8257ee5b93e4644
53542 F20101114_AAAVNL li_y_Page_057.pro
d1cf405c95fbb80cf7dc940594c7f7cb
17a6406f3614ee85d96f00a316cd24885bfe3d32
55935 F20101114_AAAVMX li_y_Page_043.pro
265ce9070cf29ef986d8d006b17b8ba0
d8f72d3cacdb993d8765dc50ed7749935da148b7
51399 F20101114_AAAVOA li_y_Page_074.pro
713fd4c5d4b176af70fa31a055b475a1
c6fa38f35aee1bc88b2044158e97493ce233b38a
51633 F20101114_AAAVNM li_y_Page_058.pro
eb2cafc1cec35c376aba0aae38e7a2b3
ec1977e756495b1e77a8395c90625af5e9c2eed1
53495 F20101114_AAAVMY li_y_Page_044.pro
5723227423d9ff1d27e371e21547bd37
e5b9e0ccbb105e081c3e783b3707ab751a59c9d7
55830 F20101114_AAAVOB li_y_Page_075.pro
c76df82551b79e408eda9d83c304d05b
f0ad3a4a05de7d50668a94c50d89801e3804c4da
56674 F20101114_AAAVNN li_y_Page_060.pro
d51cd100be44b213f77b0834577c6176
84d7e6cb068956064800b3847564ca5da6fc61fd
51223 F20101114_AAAVMZ li_y_Page_045.pro
69613dc649e6c11f8f1cf78b66a8db87
d18632dd9a81ab1b7581d1fbd3bd785843a85fb8
55335 F20101114_AAAVOC li_y_Page_076.pro
99c8a091de4a1261594b7cd5e1ba8b42
ae41293678419a7f19ee3073456c647f77f5c308
49784 F20101114_AAAVNO li_y_Page_061.pro
ddca9fef3015c02b912551c36e2ff27e
cebd8b184adaad623b7a453770bc73e7fbbab43e
49286 F20101114_AAAVOD li_y_Page_077.pro
067e7390e6ec09dffa4ae36adcb27f50
ba7d0011f1289503e23f7b505b7048175bccc2f8
48120 F20101114_AAAVNP li_y_Page_062.pro
2b096b76e91ede5e08ad8d24c0f04b82
fd441fc72c3a7c6657ab3cf1f84a2a30dbdd5ebf
55732 F20101114_AAAVOE li_y_Page_078.pro
b578ab3865f065a5172bf5c017686b2e
ddeddf7cf531cb5cb7918eac4d5e64b32ffa01ef
54687 F20101114_AAAVNQ li_y_Page_063.pro
07ec0424aa18b3718096355a81463e36
f4a9c3e76fac0ec73fd953411fc8fba9c4b8dad0
55033 F20101114_AAAVOF li_y_Page_079.pro
0c1b9e4851c664976a3605d876852f44
ce8f9c11658d2502c7ca136bc9869109086a621b
54827 F20101114_AAAVNR li_y_Page_064.pro
b660dd73818999d719eae0ea27207be0
767c7e56023bc43c4988b7e23251195b8330bf22
54542 F20101114_AAAVOG li_y_Page_080.pro
9913ae1e9c2af56c88e9617e006230cf
67e99dedb0638bd621e96797f7a6384ada6e77f7
56579 F20101114_AAAVNS li_y_Page_065.pro
f420dd772abbba9e8366cfcd5c2e8189
6d402f2da93f8894329835e3e1bc46eda3b9132a
56311 F20101114_AAAVOH li_y_Page_081.pro
ef9a77a98405e7648d52370a13226ef3
0d7c9f5870cb198eb8c9b05208c3ca666d2d5940
47926 F20101114_AAAVNT li_y_Page_066.pro
07ae6739e33f37741ea3b95b6e59fa46
ddea834a7dafefa58087e15d2f95917a674d4c10
54809 F20101114_AAAVOI li_y_Page_082.pro
a7636fbd16e9271c273b55bc3af6d8de
cf24551fc7e8d5521e85b3565aadb85ecabda3e6
47368 F20101114_AAAVNU li_y_Page_067.pro
2dc7dc8f6d083f728e470a961ee1bb6d
773a8a57d39b287e5fdb63e754f2264eb4a32b8c
46935 F20101114_AAAVOJ li_y_Page_083.pro
7bedfbb70b54561c312e9207c4a89b2e
baffc54d737d35cd489dcbb531e77b7becbeac38
14381 F20101114_AAAVNV li_y_Page_068.pro
65c29396b7bd80d56e38e0cc633feb7f
e0e784111d67dd2d66044a1a94f11498597c34e7
49985 F20101114_AAAVOK li_y_Page_084.pro
fbad3c9dd030bd0b70fa504eb9d736d5
f301ec97b9bb06f0a84f66d00bd0482ccd887e57
3343 F20101114_AAAVNW li_y_Page_069.pro
24e22010958d0ff4fdca37a3517d1211
7383c93284e6b56657efb6b6ca7314821aff8058
50114 F20101114_AAAVOL li_y_Page_085.pro
00d8e3b7305ff900f50c8e674e9c5370
f387a2589a7743830118a65154c871ea5a905688
7345 F20101114_AAAVNX li_y_Page_070.pro
f6c163353a488a984cae4a44196d63a2
3c6fded1360f7bffeb895ee71d508f7f4d96e7b9
53068 F20101114_AAAVPA li_y_Page_101.pro
d54d0922d568bdc07d161a45fbf01e59
40789fc44c9502093bfc1c65dc5f6c09e43395b5
50414 F20101114_AAAVOM li_y_Page_086.pro
e75fd4fb04d63623a1c91e49b3ce2a7c
d5944d2eb9d35dc46153ab67f0d7ce87382bc564
54675 F20101114_AAAVNY li_y_Page_072.pro
30195d251c1eb6dbbf3521450577bea1
5e665d4d7a077219291034e07110e12b00e7d9b6
52609 F20101114_AAAVON li_y_Page_087.pro
3cf0fa5511fd03dbe1c11f459607b483
42644b8e4ddd11756252dbcdd10fa670047a7061
51498 F20101114_AAAVNZ li_y_Page_073.pro
54becda087d3975be523c9bb5cabc960
6ab31dc1e58a3945e68e72df400587de57c6b379
57144 F20101114_AAAVPB li_y_Page_102.pro
a2516a3c63536093a698f20a9362c37c
9ce1d6f20f6519d96f3b0903998448fe77342bed
56429 F20101114_AAAVOO li_y_Page_088.pro
f00268f569c7098a6a1a36c5eadb2f4a
1821dff597f5e002d3224694bf4c451070233934
54855 F20101114_AAAVPC li_y_Page_103.pro
df65e8bb24405aca2bf37d9568d93ed9
5cebc27f79302baa5d6da20e8d1a6d2496cfa6c1
9648 F20101114_AAAVOP li_y_Page_089.pro
7da463f8ce39fd7dfd885b3ad4d56842
64b29d6726ac1c799960576de751bb01d10a7595
52208 F20101114_AAAVPD li_y_Page_104.pro
e6800d734b6f432d0ba6b54290b30b38
97d952feaa655a7176f78d67d3af834813e37c3f
31040 F20101114_AAAVOQ li_y_Page_090.pro
8303906fe308f5cf58bc746622b0d0cd
904fcd509c61c7c0b3d9e947d5c2dba5ada54f87
50136 F20101114_AAAVPE li_y_Page_105.pro
ab4e296b8152b0f3ad2373631a95141f
b48322741a916158696279e43ddb84d52cc91134
3935 F20101114_AAAVOR li_y_Page_091.pro
05e2b23e5f327aaf6ed342b1fbc64d78
798528b5a22d6721646820ea63896a230b283fb9
52907 F20101114_AAAVPF li_y_Page_106.pro
b299215b6b64158ef94e83b954369e79
938358e7eaf8d66a023470ed886950a24528b21f
5134 F20101114_AAAVOS li_y_Page_092.pro
33a40d3a65d40861c989ae0743330fa0
4ba7ef3f4aec21e3c273f548ca53b9a70ead37ad
53860 F20101114_AAAVPG li_y_Page_107.pro
7c646d06f916b50e79925b7efc411216
c19d1e169d0bdce0745d46486c56c65a1ba3db14
17370 F20101114_AAAVOT li_y_Page_094.pro
a9c3fc0dc0a750a9197d8a5dd316115f
960badfce4c55f8982c6e452d1a6b9deda630e74
F20101114_AAAVPH li_y_Page_108.pro
86b4e3d2bfdd30f50eb71162ff4c5ddb
53494e31f838d178554d35839bf2854ebdcf54ad
10324 F20101114_AAAVOU li_y_Page_095.pro
15fd3160c13db86bcfa710006eaae083
63e9940955d7d065ac88271b5eeba7b7ca4fa18d
7195 F20101114_AAAVPI li_y_Page_109.pro
4a91127a054c1be1eb84fef1e15f8ad7
7ff7b3ff23e210e33ee27af495fa5b15ef14030c
7342 F20101114_AAAVOV li_y_Page_096.pro
603f4f53da0b77eb56b58cd64220d4fa
2370f54dfb6c63e9c4cbfeca67ce3e3e96baf2da
7421 F20101114_AAAVPJ li_y_Page_110.pro
234068277648d2bd4f184ecb2b9975fe
b669e237fa3f69163089df2a96c6ae5818b02488
6009 F20101114_AAAVOW li_y_Page_097.pro
6fdf36f1404ad2132f24043b00995347
34d330bc879c18e1a54bb2a25c8700c7aca60761
7116 F20101114_AAAVPK li_y_Page_111.pro
c58745094edc164c506ba1e99a66fd1d
da32ac6e55eb12fbed7f619e136d415657209882
5664 F20101114_AAAVOX li_y_Page_098.pro
edae6f4c1958e9db69950da3de99588f
a604a2b1c660497cb2958f668ead4824e40deb31
1490 F20101114_AAAVQA li_y_Page_004.txt
11a660524416c5930e5f52fd4addf0bd
bcbc13161e953d18598fc6bbc6523cf51c6bec7a
52114 F20101114_AAAVPL li_y_Page_112.pro
d430775e0f717ec39cf5454629d63033
43b2594849c7a47859fb6aa0c9fdd978f93223f2
53439 F20101114_AAAVOY li_y_Page_099.pro
769d80cb8e585394932b511a9eaa7ba2
de6ca1f6aa90aaa0166b4bee6ebe9b8897c98fe9
2682 F20101114_AAAVQB li_y_Page_005.txt
777e7530b64608facfc41977181b3916
c0af277dffee0dd6997aef3d95c3bdc75035722f
52444 F20101114_AAAVPM li_y_Page_113.pro
1bf9a0b09de7e70766b2d3e7cbe4b4cd
803f3cca53abde2df70a40bc5d99695e33d2ab5d
54361 F20101114_AAAVOZ li_y_Page_100.pro
19dccaade1d04a774cf397e555994346
fccdb7c64e82e8758e97b017deb5a60e20b29729
46893 F20101114_AAAVPN li_y_Page_114.pro
6bcf3d31d186684802aa3ede96157f30
ed13d33c833d19c11bf22930d6ca677e425f241a
3618 F20101114_AAAVQC li_y_Page_006.txt
f32f5b8a588b6f89122152dd3d3b54c9
4ad252888d671f22fd51c02a5edd5283ac0406c4
53164 F20101114_AAAVPO li_y_Page_115.pro
3e112fcc9107ad4b76f15103b0710c84
cc711d6052652918b0b5fffc90c8bf878506f8f1
639 F20101114_AAAVQD li_y_Page_007.txt
5ab0e567567c6313ff3c2cd16829064f
08f471ab6db708011283f5a820f4b30925476471
3779 F20101114_AAAVPP li_y_Page_116.pro
5853bfef67550ab487881cda5a1d40ab
dd997df6252b4f169a699a1fc3ca78e0fe5cdcb3
581 F20101114_AAAVQE li_y_Page_008.txt
cade7958d815c97a924c19625b5b530c
f29bff3cf655933a72d4d9896ba1a336069b23fc
58647 F20101114_AAAVPQ li_y_Page_117.pro
e59b2126e0c811e88609d3007070a16a
ba3ab8dccc6c166873f5f19ab0e64839d27b5d2f
2813 F20101114_AAAVQF li_y_Page_009.txt
fea4b787090d31068391e6a27b7b6a8f
44316f1c864cb9b0b02beeb879cf5562c4307984
64671 F20101114_AAAVPR li_y_Page_118.pro
34ec8c39b62add1eb8261e9451c2c0c9
a9098bb6df86d1d806e0a01e1f92d0a492a20c45
1138 F20101114_AAAVQG li_y_Page_010.txt
1e5cd854764fba97a55a45cfdd8cab2a
3d30e8daf83d380122c0737ef164bb01b3a91c6d
69930 F20101114_AAAVPS li_y_Page_119.pro
d039c2a3cace864816c97a35a5e65ca1
bd6eb2457dff23d4fd21ff580fedbbb60d6da79c
2075 F20101114_AAAVQH li_y_Page_011.txt
377bdb6fab25ee22f63b8161561175c0
b0cb640a03428efcefea329ea61dc8881be7b88b
65312 F20101114_AAAVPT li_y_Page_121.pro
b587d4175885ee85020eb7d5e4efae39
9180bd5bb34c94d0bb208152f3e5a000ff1c6b8a
857 F20101114_AAAVQI li_y_Page_012.txt
f30c70d5fc202fdb7b89c206c7f6b296
494ba51d7568f5072e9f0d0c12d6f11a42ed1d2d
66634 F20101114_AAAVPU li_y_Page_122.pro
dec62ade9ee40be93fdee7a54d490916
9a9c8af1782515478b9b385444cc6344b41ef807
2146 F20101114_AAAVQJ li_y_Page_013.txt
8f134610ded5e4017bc3949c3597ac6d
245aa9491a5f574238eb34d2bbb41bbc4cf1e66a
60926 F20101114_AAAVPV li_y_Page_123.pro
04683a8e6db7fa5102c9cb1675dde67d
c22d45c8a294b8af89faf4063043fb06064d5293
2150 F20101114_AAAVQK li_y_Page_014.txt
90fb5e1167e519a5b08cf50d06ed3bcb
7012b708ac9ca3a66c862de303f70c0dd8ea32d3
37228 F20101114_AAAVPW li_y_Page_125.pro
be5563fc8005cdbe2dc156976280d9f6
5fdebcb852a53aa38916d07b4d28177d156cf51f
2186 F20101114_AAAVRA li_y_Page_030.txt
22bc8e959a3b6abc966336879aab044d
f9488d119a2106f0873754b9dbebf48166ea4091
2128 F20101114_AAAVQL li_y_Page_015.txt
a38289d1c04eced53c02c7436043a000
439ef683966156b6cb6de37d7f2e600f63e79a66
472 F20101114_AAAVPX li_y_Page_001.txt
9cbaf36d8a8b6970d781f17fc132e8ab
f2fc21809d0e1fa025ecd08ca3dedf30ef06c97a
1488 F20101114_AAAVRB li_y_Page_031.txt
2b5d0f6fdc67b6afda9e352cd13e450e
03b651e7e660cf460b2df577564bf796222c4496
2156 F20101114_AAAVQM li_y_Page_016.txt
66b1a7184d8c7bd0e72eaecb4b3731db
22767160a4fa943d0dbc7b041c915b49a772fab2
75 F20101114_AAAVPY li_y_Page_002.txt
f4d278c8dfe6ab5bff445bdf97d7fba3
0bb8c397c524dc10492753f18329ce646dbc7880
224 F20101114_AAAVRC li_y_Page_032.txt
ab30800c23365c4452aa522e368c30a3
b9169d38f9f45566b8a2a0bca0042889b14fb2d0
2158 F20101114_AAAVQN li_y_Page_017.txt
b19e01a8f5f929b2c166bcb0125162f5
c5cbc8c2aad3b07be92aa5e6367d61e4be4a7eb3
147 F20101114_AAAVPZ li_y_Page_003.txt
b512077393ec7c541a873a3a8ec30312
2ad98c126ee1bbd48d75f768d52e68e92bf94143
2173 F20101114_AAAVQO li_y_Page_018.txt
d0db356ece99b232562783fcbfcd1aab
17204982471b13fc3a1973706b74626475aa9c76
282 F20101114_AAAVRD li_y_Page_033.txt
b9abac31bcde9bac0ae9838392a06d47
7b2ee4855525658d4f81d360e8ef79e80c2abe2b
1806 F20101114_AAAVQP li_y_Page_019.txt
6dcf51e222a97fce2d21e6145cd8c235
f6051f2b10fe437ed666eda26fdbf458e1751d4a
544 F20101114_AAAVRE li_y_Page_034.txt
eb19530ab63499f60aba4351045e0fa4
027300efb668f35b8e17f84a597478823816ab65
2062 F20101114_AAAVQQ li_y_Page_020.txt
70f3a7c01b47b621fbdb972ebb11096e
0a5d9fdd4f61d077bf6827fe6b510b3370fb5064
359 F20101114_AAAVRF li_y_Page_035.txt
02867e044670526601e388e810acda5d
79c4be61ad24bb4d30ac280d97e64b1e606b8833
2235 F20101114_AAAVQR li_y_Page_021.txt
e75f45f2bd0e713cd20c98037da5c06d
c95d49ddfe1b2cd4bb828c712d40efaa7f95b0da
662 F20101114_AAAVRG li_y_Page_036.txt
26a179f1322ccdfc3cdea6a56a454d5b
ce6d9c443f1e86be23e5895d60ca4672e858d023
2228 F20101114_AAAVQS li_y_Page_022.txt
a88bc5ca00717cd47837e481903b2cfd
7d75e56c9ffffd04db65fedc102c272cde20be38
603 F20101114_AAAVRH li_y_Page_037.txt
3e57f1b0f1a1c4697a09010d6b629155
c5393599c110cf7587aa6cd9e911424f282899a1
2024 F20101114_AAAVQT li_y_Page_023.txt
16eee0281a2cae315f7a7e626b0b9b5a
bee1b023c5beb8420d2161e4916a0bf43ef07da5
1789 F20101114_AAAVQU li_y_Page_024.txt
c7f760ff208e94c95e207227c0e00e33
1faa61f976ea1320c3ae7dbc24142f77c7a99c2c
393 F20101114_AAAVRI li_y_Page_038.txt
295791146f4eda305636d664967e63ca
fdb148294ae6ad07bec94998bb40cc196998e59d
1807 F20101114_AAAVQV li_y_Page_025.txt
53f6edb19fd8c59d2d963eec2f07ce3f
00e5cceeb37e5578e2643336084eccd52cf49be2
649 F20101114_AAAVRJ li_y_Page_039.txt
20ab9ca0363d68a0f8f01c653745cf74
a70a5ea22fc60702a066d9c26a668c6e77f0e3c0
1840 F20101114_AAAVQW li_y_Page_026.txt
c67732a3abecb0539b91a2c4479e79ab
01a0909a9015c9fbc3ecc137f4ccd815bc2f9c0c
2247 F20101114_AAAVRK li_y_Page_040.txt
af9f89cc7297f492bbf908e0289ad60e
8fd51da40919ba5e23a834a10dc1fb00b550ebc2
2064 F20101114_AAAVQX li_y_Page_027.txt
3bd3ddbbb18252c883d561f7f7c63ab9
9fc2ac484ac7e4b2ed80367472e672047a8a5ccd
2210 F20101114_AAAVSA li_y_Page_057.txt
407cf4f932bc761f0ddbf34b2b153794
8f95bcf36522dd2cc0152a86bd2b858974d9899e
1949 F20101114_AAAVRL li_y_Page_041.txt
89345e87367c3f6a0da451795926d644
0355b1c8fd08dd002b0906b0a5e0afa52b3cb533
2141 F20101114_AAAVQY li_y_Page_028.txt
99a749f07c06a23b754f9a8cc016e775
cdf6fced063ea7d7c045b35a0cad0508fac44e11
2068 F20101114_AAAVSB li_y_Page_058.txt
a892486d4d24997fc915a551e26874b7
4cf75564a0e8c4ed9f8b8df5911b8c7f47945666
2054 F20101114_AAAVRM li_y_Page_042.txt
9f05bc80f828dcc632bb6af645a62732
442ade25864a07b48adf54f5e5b84afc3f0503d9
2165 F20101114_AAAVQZ li_y_Page_029.txt
5e77cd9a36fcc19aa5e122d6d8b8466e
ba055abb2b931cf9e1291e27bf38b9c596967cef
2187 F20101114_AAAVSC li_y_Page_059.txt
cc56d1cba2e25b80d60e955d876a510d
90aac96fff4382a6ae59a9a5de3cec3e0a077f56
2224 F20101114_AAAVRN li_y_Page_043.txt
822e4f0024bddfad4778c3a5d218338d
6006b6b667e2f96887fca6f15b30dedf42db3acc
2218 F20101114_AAAVSD li_y_Page_060.txt
9877299d7ccc544b44ca4e94a2782836
d0c4388c8acca388e82c5ab726fb4a281bdaab91
2111 F20101114_AAAVRO li_y_Page_044.txt
f13793402fdb9495be5686ca145ac3d2
323adab035b9a3849eed1833472042457bc00ea9
F20101114_AAAVRP li_y_Page_046.txt
46f1db19ad47dd618815445901cb5ff6
32c868d1d10f5323fddbcc98d9313c1f25d0776f
1994 F20101114_AAAVSE li_y_Page_061.txt
53cc8a3e45c0c2959f6e2bc7cd492b0f
cc04dc7cbd2ecf97f19255741f2a8c14da4fef0b
1973 F20101114_AAAVRQ li_y_Page_047.txt
9bfe095fc5e0950171d81c7573fc02e4
b0a2281690d7237ec5cb352b2f7f25350f537bee
1911 F20101114_AAAVSF li_y_Page_062.txt
e19d96ca86fb5ded123135a822342106
7fa33b33a77356f7aa2ec874b4a3888e2d53b004
2205 F20101114_AAAVRR li_y_Page_048.txt
43456af90583d3513e80dd95b5e90908
75a4f840fb87818d2c78f850831746edcde49007
2145 F20101114_AAAVSG li_y_Page_063.txt
7be3799aba1bc058e3432b8de7f6ce7c
aabfd65dfc54e48ec7cb0b9390939c00bd2e6cb0
2006 F20101114_AAAVRS li_y_Page_049.txt
7881c23689fb68a5add7b0f5fcbc6868
f5ca75961198dffa4f7288749f584fb543c83970
2154 F20101114_AAAVSH li_y_Page_064.txt
6ce98ab50abc0041ee31375735d82e8b
894291f704b97a7fd669067620fc1ee0848bb119
2025 F20101114_AAAVRT li_y_Page_050.txt
679bb283bdc7da891ff4d355395b06f4
f583ddd448793bde02e161faae83e2f9d4e8a643
2215 F20101114_AAAVSI li_y_Page_065.txt
cf260510da6cc83614e62ebf1703cf46
4e702058f5a92b176e6358b594d0036ce07234b9
F20101114_AAAVRU li_y_Page_051.txt
91536d5b9c8b65dd0c2da9eabe49ed37
91d2afb7ceeb59ac66346f5e1710d6e525165f05
1936 F20101114_AAAVSJ li_y_Page_066.txt
4af41588d9406fcc151be8d259a01fbe
70f3090fdf42642e1769968b693befd9d38c74dd
197 F20101114_AAAVRV li_y_Page_052.txt
3cec82e072868ebd16ae812ecd5e03c4
4df038ade1393ca762ebaad46d7b8c2dc9bdf5c7
1916 F20101114_AAAVSK li_y_Page_067.txt
b7fac6849d8e6f4753cdd75e63441906
f8e0354f52b701bf81d9ca1e079126ab77877cda
708 F20101114_AAAVRW li_y_Page_053.txt
8ae715eb86300402d6babddb0f64cdda
98f0aea22ea4a4e3a7eaa1d4703fbb551bc06d85
2002 F20101114_AAAVTA li_y_Page_086.txt
c6ac47a1e5c5cbb58a94f29ae0decce0
a9fb622f65ca8a77503790579e07585c9913df9f
629 F20101114_AAAVSL li_y_Page_068.txt
613c9140c8a2ce108ee2adb87b963170
5712534b8f2ebf1fe941f61ea0efe6d044a0557e
846 F20101114_AAAVRX li_y_Page_054.txt
6cfc02207ec62ca75e1b954d554fb159
0788a2504017120d725044b7acca879bb4db1f61
2063 F20101114_AAAVTB li_y_Page_087.txt
e6c9299fc5fe627f74cd129b068cb699
94a6b9460633d5d7a6c90e47c46e91a30fe96970
148 F20101114_AAAVSM li_y_Page_069.txt
e6c21a1ca271e3ef0deb39b3c3f2d694
1ed4c4eab552e044a46f0e1569dee5d5dc288f83
381 F20101114_AAAVRY li_y_Page_055.txt
ac8b664a5c4bbd661383c549d79f9493
b8595e44d4efbaca65082976ae0875e3d7f95adf
2248 F20101114_AAAVTC li_y_Page_088.txt
87dad86b8d8cd829446b923b2b6d66b5
cc0e9b9b71ab648356ba7a197262f9a87733940d
441 F20101114_AAAVSN li_y_Page_070.txt
ed5724a5f64f6cd4a1c5f1b9209a079a
b848d65cf108220f1494af77dd67492a1b4a38b2
494 F20101114_AAAVRZ li_y_Page_056.txt
2256227351e400fdb518416ee67ca461
e49618ca03fa589c8126d31728c4ef15658d13ed
385 F20101114_AAAVTD li_y_Page_089.txt
b2234e8237aa49c35095317f55b0f457
3f429d7457a843522d8669d905cc6a125063e75c
1227 F20101114_AAAVSO li_y_Page_071.txt
e2aa2c08c60d66802b3e0e8c671da989
883de9ae906ce069bb21acdf7c6fe523e5818cfc
1245 F20101114_AAAVTE li_y_Page_090.txt
c8ff44dcbebff282e917430f99cfcede
93fc11a37af9c462b63ebb7190debc6a32d8c23e
2242 F20101114_AAAVSP li_y_Page_072.txt
324ac1be33d32ed8c533a05047caf36c
a7383a97a845c47eef2d838c17a9fe486a869853
2033 F20101114_AAAVSQ li_y_Page_073.txt
cc9ac2f52d18af7cc78e9e5d1a5a5ffa
685b38a4bd633aa743ef264322247768354034c8
216 F20101114_AAAVTF li_y_Page_091.txt
0f287847e96c0369341577ca9a630760
3fdadb7d56a3bff0dcceab821b964c075efffa32
2019 F20101114_AAAVSR li_y_Page_074.txt
3e84f3a6154398fd393c99cacc2305e8
60fd12ec75efb935f5647d26e903b77fcddd61f6
283 F20101114_AAAVTG li_y_Page_092.txt
636399f38ade5f637913e45263aa2cd0
961f6a0276bb91f0fc1898c92a659a9ff1839233
2196 F20101114_AAAVSS li_y_Page_075.txt
5f040c0ee2a440d6cf510cf8859a50ba
da005fadf25b4b961f7875f10d13d59262b413bf
320 F20101114_AAAVTH li_y_Page_093.txt
4c27d472dc7df454aee7a3528e263e76
7cc6e45a289dda1d9d107e8f9ba19076efbf0fb2
2204 F20101114_AAAVST li_y_Page_076.txt
33105301cf4b68c2b830197388667313
fd21eb4416e15c23c0f43d2be16a4b92a478deec
1178 F20101114_AAAVTI li_y_Page_094.txt
11832b5fb76e6407405b7a1b9a1fd40a
006c9ec5e4760408da57191f36aeceae39ee38ec
1952 F20101114_AAAVSU li_y_Page_077.txt
c3ea9f2d1804897e4141a23d91d6ebaf
4fb97029e312ba5f433efa5777515d9443749e7c
454 F20101114_AAAVTJ li_y_Page_096.txt
7a336cbdafdb13327b1921061014016a
e58688af90246a03b6bf235ff0d969b4a45c2838
2161 F20101114_AAAVSV li_y_Page_079.txt
ac2a3906259357340b3eb2f0df4230a1
810eb14062af1459dcfbdd7e1686daea5e7e76bf
315 F20101114_AAAVTK li_y_Page_097.txt
3f2f87d3ab8e74b0bf6d702f83395d7d
01d405c620b3ea0906083130274015ad916657b7
2143 F20101114_AAAVSW li_y_Page_080.txt
064df27a42b83cc954cd06a5477dd9de
2372ebe13d2282708141b55fb9f8e5bfa5b2d0c5
293 F20101114_AAAVTL li_y_Page_098.txt
b52b06b38d9baf42bf77a16c06d83cde
6b45c890762ad90b04bc619043b548422fb52dc5
2206 F20101114_AAAVSX li_y_Page_081.txt
e638f4ba198933375d8c57c971f7d82a
8c674677cd7aff2b82b22ba1f62f26b801b6e694
2117 F20101114_AAAVUA li_y_Page_113.txt
e88484e57e59251af3b1b2abb4b2927e
b81af78a0f1ee6bce6515e544714346bdbb7352f
F20101114_AAAVTM li_y_Page_099.txt
f8112165e87f568534e93d2d1545a17d
f5f587c44c788f2653a50bfc5f8f1fae711fccf8
2157 F20101114_AAAVSY li_y_Page_082.txt
82c79ded8a4b61823b306b5c41a43600
f9a5f8142c1e877f81cc348d4b1122eaa6944588
2148 F20101114_AAAVUB li_y_Page_115.txt
df34eb8f1fdede48aac8e819cdab313a
f23f9d71e2f5013557afcef71ac1b3a3c0e6582c
2136 F20101114_AAAVTN li_y_Page_100.txt
67b751adfa38329f8bee6c469e5d357b
54d2514647fb3a39ccb9fb126d28def879eca100
2038 F20101114_AAAVSZ li_y_Page_085.txt
589d6df4a55f7c3758d93830d7bf5755
768aec5daba6e5e831c045b1629622ee2e6bb3b9
154 F20101114_AAAVUC li_y_Page_116.txt
1ab431013b424a280f8f518f3edb9172
155d54a97b48a1a28187af0404964dde77ab0070
2115 F20101114_AAAVTO li_y_Page_101.txt
1f3adb428e94356e84f76ab863d256d8
70f453507072c807903f24fee7705c9d9b3df209
2411 F20101114_AAAVUD li_y_Page_117.txt
af7f61984f441a683e8c1234bab72938
5e54686934edc83afd63a98869f58385eae41afa
F20101114_AAAVTP li_y_Page_102.txt
f9f84e7b21ba6cc18fc9a12438b42a52
73c185821e68032f219f7b3f047cdf7b872679a2
2851 F20101114_AAAVUE li_y_Page_119.txt
341c94d62e342ffd1f70ae5370af07a7
70d31adc6245b09436a02d9824d3b5088150bf38
2195 F20101114_AAAVTQ li_y_Page_103.txt
df73fd0571c0f72e3435ded67126c84f
b7db4c5c9ca9b4a91942f34b9fbdbf699796c146
2525 F20101114_AAAVUF li_y_Page_120.txt
eb6c5ad37e27822d0bd452e1a1b569b3
74ca5ac0de6c703744d3de5ba8bcc043622dadf0
18509 F20101114_AAAWAA li_y_Page_031.QC.jpg
4b3ba3825c53d27aabe3ab86417ad4fc
8a045b4ec4e564c929d66dd37eeecc9adf00492e
F20101114_AAAVTR li_y_Page_104.txt
a4afe3cbfc8fa048ee0e2b3a5babdc3c
ada7506dd09fb65900b1aa86cb8455b78de89a25
3176 F20101114_AAAWAB li_y_Page_032thm.jpg
d30282e044cf8337942eaebb4591872e
f8c0f5fb61fb065e9ce882d33b0b4b072f976275
1979 F20101114_AAAVTS li_y_Page_105.txt
01df0201c6d14148469a3623f213ec25
ec5138227d62237c577112ce96d690802b93ff83
2670 F20101114_AAAVUG li_y_Page_121.txt
aaf14264d0cb6b946f658441934fc372
dc882e63194cb01d460c5c18d3a79a725a3d60ea
2959 F20101114_AAAWAC li_y_Page_033thm.jpg
4532dff0fd5410058049619e2a3b9361
6abd5e89c90dac49362f68b91c8c47a8c166003b
F20101114_AAAVTT li_y_Page_106.txt
5e4d855ee242b5f6443b0f4027f89ec8
4f48c558ec4af5460d664e160f89c109835bafdc
2735 F20101114_AAAVUH li_y_Page_122.txt
e7e7ddb86e9dfe1ff29121a4fca539b4
1676e84a8540e0ecff8e10de6eea8054e74c2a08
12215 F20101114_AAAWAD li_y_Page_034.QC.jpg
75c92628dd15ad0562f5d8ce5702474f
925c23ef5a10dcf8a3f92f6a341e58a9e6409ad9
2110 F20101114_AAAVTU li_y_Page_107.txt
5a7216db8583f1ada902904a13aaa142
ebd14f47248d0b5c39d8946c5888d06ae153437a
2497 F20101114_AAAVUI li_y_Page_123.txt
c1734e9de9cabc5fb165e15b75332290
0458334d8126d722694a003e111394a214654bff
3781 F20101114_AAAWAE li_y_Page_034thm.jpg
88606011515ca9e7717dbd8d41fc38b9
17fc55516d52d00554f98a7ce02be027f795d19c
F20101114_AAAVTV li_y_Page_108.txt
06b862506ec7d1e47443d7a089323c73
7806a85a5dd0d73409a2cb4e2524b6f4a149ec87
1923 F20101114_AAAVUJ li_y_Page_124.txt
bd4fa5107395babebc07f1e363bf4fa2
52c2cefdfd31aa75da1a462d173f6252650e6200
9542 F20101114_AAAWAF li_y_Page_035.QC.jpg
5d404acfb37e3d16bb479668caeb2bf8
7af4858055a1391492425e7c3b498afc3e927199
322 F20101114_AAAVTW li_y_Page_109.txt
70eeca768fbdd46ba21e486023450b7b
84a7d1985f347c90c3eb139e4e1b2174ef79bb4c
F20101114_AAAVUK li_y_Page_001thm.jpg
63b392792327e904e1683d6daf2b1d21
8cb872425493471f06769bc592af3c9ce7c46467
3289 F20101114_AAAWAG li_y_Page_035thm.jpg
35e7898f6426c923b0db70164142c3d9
2c7b389e6e695b0cc6629c9c2e85d083360614bb
410 F20101114_AAAVTX li_y_Page_110.txt
7b9d9f69fd02978821e412ff2dc4bdb6
0ce8aecf727aee1fb8c817a2bd75c6bc4c09ed97
6911 F20101114_AAAVVA li_y_Page_115thm.jpg
26f7a8cf21fb9be8a65210d209348f67
83776edf8cc3cb29ec503403868ae0a5dd287f7f
932522 F20101114_AAAVUL li_y.pdf
58fea61ba025e198803bfe6f86d1a1ab
11a2c38e597c7430929ad066c4bd06fb991a233c
10699 F20101114_AAAWAH li_y_Page_036.QC.jpg
4fa474e781a7f022649f8ba7d903aeed
8519c90b1ff093b4eb897da290477a030be2b968
396 F20101114_AAAVTY li_y_Page_111.txt
7926968321eaca89739e15a791a2c8ff
0b346709338e7f7a67eaa990611b3dba86b89f67
26847 F20101114_AAAVVB li_y_Page_122.QC.jpg
8badc164262238a9a315f0fb8fb9cf27
f05d80bfa10e3007302845d4706d9925e011cfeb
6783 F20101114_AAAVUM li_y_Page_080thm.jpg
ee2098b0e79e2a3f90b8fd7e32926c3c
123197bb64bfddba19f21bf8944013482cff912a
3750 F20101114_AAAWAI li_y_Page_036thm.jpg
dd3f9a63932abfcfb395728c5510c28f
2eaab7c6c8aba5327ab5ebc1660b021f49b1f5b9
F20101114_AAAVTZ li_y_Page_112.txt
0b8aef288f214e60aabe7759da6087aa
a7a64589142195ade6a7822998bc5c8d991332f1
6870 F20101114_AAAVVC li_y_Page_089.QC.jpg
584e8da737acbbc3006e995271581a13
14465f1c72875816afd7e16ca6850ccc840edd65
6779 F20101114_AAAVUN li_y_Page_029thm.jpg
8ac5789c322fa36e402193d353b82ea7
d9c40715ccde9742d35e2627a7f373ce8ba73a1a
10073 F20101114_AAAWAJ li_y_Page_038.QC.jpg
6ee8a96fe1bd46202f799aa71d6b266c
fbeb94646996cb48e2cc705cf8343f70ba783528
3552 F20101114_AAAVVD li_y_Page_003.QC.jpg
af04174b5d5e5eae74427f20d81a25cd
129a3786064fa8f0fc7e77ed4efbb3eb9ea5927d
6405 F20101114_AAAVUO li_y_Page_062thm.jpg
b9faac070e3d9d4a2332977ac4eee06c
0b12b53c6e8d027442bdbc69e91b1bc1387474b1
4009 F20101114_AAAWAK li_y_Page_039thm.jpg
e0040f581f105ff0f3003899f795c59f
defd57ebeb0461dc6bd34639bb1dbcff8bab57e9
8096 F20101114_AAAVVE li_y_Page_033.QC.jpg
a346954f83bfb751a5cae6bf5c6734d2
5c6cc1869165d7056e0d02b7b32a644c16f16c20
8486 F20101114_AAAVUP li_y_Page_008.QC.jpg
db313db6194389a8e9499ddc20c476f8
87058e530f7ae250a882ebea260f26b95ae3655b
23295 F20101114_AAAWAL li_y_Page_041.QC.jpg
a669e99f3b90137c43e0912d56a13241
4a71010a6a8354ca8dae8c4b4e9d8fc093bf80f9
17790 F20101114_AAAVVF li_y_Page_071.QC.jpg
f226d462e0fbae407a1bb7d8ca303289
6320c4acaeaf0ee3ce573a634aa31b85791b31de
23539 F20101114_AAAVUQ li_y_Page_049.QC.jpg
d75a3d299a70e9c0b5e52551ff370b49
a2a2c59e8a695883f51baf1dbc608f782d8c288e
15299 F20101114_AAAWBA li_y_Page_054.QC.jpg
4089c1dc40ee428a93d513582ace2506
00d778cfd49dfdaa7b2b133db4bae492288d2ad4
6428 F20101114_AAAWAM li_y_Page_041thm.jpg
7566876fbc190951ff015799db6c728d
272a1efe6835be6508b17f6062a1e3c10d744606
12032 F20101114_AAAVVG li_y_Page_037.QC.jpg
553641c10851e2bbb87594c7e404e28b
718d4f16e59e479c516c5d85cd0f401b3b2bf6f9
25287 F20101114_AAAVUR li_y_Page_017.QC.jpg
0bd0f2fe0f14a76274aacea88df5577e
cbc853687490a4e5ee1719e928d3d5527b351bc2
4057 F20101114_AAAWBB li_y_Page_056thm.jpg
b1cfccf3efe69895e83094104f80271d
3038e7fd5ab0dba38b915c3aa15977a6375bd037
26503 F20101114_AAAWAN li_y_Page_043.QC.jpg
c1c73485b7cf362ebeddad95626fd988
4b0052bf0c65d6131d7fefd82daff909bb9c1763
24823 F20101114_AAAVUS li_y_Page_085.QC.jpg
d19847b1784bfe2d409a2fedb0bbbad6
028999fe4310649cea1b3e6be76d2d278e02eb2c
6659 F20101114_AAAWBC li_y_Page_058thm.jpg
7d2f82a790ccc7b057a69beff3089e12
54254d340dec20e6777dae7f3f333a2b6d2fccd8
6929 F20101114_AAAWAO li_y_Page_044thm.jpg
ae3fc25abfff1a35739e077ba52b1c64
ba42f2526c09be832e7d7d589e7e19c208d9906b
6868 F20101114_AAAVVH li_y_Page_057thm.jpg
8a2ca88678b5b5ade0768bf1bced7aa1
78758661f474824519fd8000a19fb5fe982ccc92
7124 F20101114_AAAVUT li_y_Page_105thm.jpg
1148c293df9733170d3ddecd0043a247
713f6a24af7c4624c5b674f246c1f8a6bb86799d
26625 F20101114_AAAWBD li_y_Page_059.QC.jpg
84b35b1c89434eec5620f3c6fd9e72a8
2ab892c47dc1bb2aed207b99a2a72cf82cd029bf
24826 F20101114_AAAWAP li_y_Page_045.QC.jpg
3a92cd78514e6b66b65ef88ff219cb42
50cef82a5f0127cff97b2bb82bc374b5e39b6d10
11037 F20101114_AAAVVI li_y_Page_055.QC.jpg
e28183806e91d0cc57a79d32a7072839
f34423b86c73493cc09511086a78bd8d086e0e17
10880 F20101114_AAAVUU li_y_Page_070.QC.jpg
9fa4f7cc0f23adbc86c1408dd9822f78
5fd466ad81c5ececd5693764512c8bbcd8f49057
7239 F20101114_AAAWBE li_y_Page_059thm.jpg
e700b09e239f362be93d2eaf06511e2f
b70bb7b759ffe4e21d0c4ed4fbe2817bdf7f2737
7200 F20101114_AAAWAQ li_y_Page_046thm.jpg
6bd4e6cbb965594b6541caa8b37ed70e
ccd09438d9b2882e105704e689a0eb736701fce8
22301 F20101114_AAAVVJ li_y_Page_047.QC.jpg
69855b4db6e270319984306572f9b512
077957e90beba88a5278659d80c754ccad81a9c1
25727 F20101114_AAAVUV li_y_Page_101.QC.jpg
cb5f153144d152443824c100da3d2950
0c909cf5b97d33abf68b7ff2235066794da0621f
7190 F20101114_AAAWBF li_y_Page_060thm.jpg
9fc3777823aa4b3d9a1c01dcae3e2c79
7774f123f3d8940b72498395b1191b14f9ea9eb5
8553 F20101114_AAAVVK li_y_Page_068.QC.jpg
a3aacca0d3131b7a8f863a951f3803b6
3104d6b8417fc99a9224e6b70fbbb5a22d605bed
3196 F20101114_AAAVUW li_y_Page_038thm.jpg
ecf138b00680e94cca0a54fd0e0d0907
89a1ae65c054bf81654eb8ee2bde1ad35b5bd262
23822 F20101114_AAAWBG li_y_Page_061.QC.jpg
403ed8e93c2146a5fde281ed6b6cc037
f2b4f39fcd12cc72889d87f9101cbbccd6d7bf2d
6420 F20101114_AAAWAR li_y_Page_047thm.jpg
cfbee2f746109a1c65f2be2ae2ff3b93
83b1450b00770f0df9cc32b0eced417f1efc6996
6814 F20101114_AAAVWA li_y_Page_106thm.jpg
10985d11d596df56c7c9133e1cdee109
63ac9bdbb69c1184e2ed57ee2d65670a8e49a210
2518 F20101114_AAAVVL li_y_Page_008thm.jpg
24b3cb84d6e0ec0766c34d8f6ba6ddfb
40425e7afe62743a48281f99bd9ab7c974effb58
20834 F20101114_AAAVUX li_y_Page_005.QC.jpg
6a27f806dd2c686219833ca598044d43
7d28daad575ecf60413b7f5a69e45311245de5a1
6719 F20101114_AAAWBH li_y_Page_061thm.jpg
624c565b6d94bda0856666e066b66e5d
a5de80bb86af74cf627d0db0dbeddd52452675e2
26331 F20101114_AAAWAS li_y_Page_048.QC.jpg
fe3b00878a6f6d2595ef2950babfec06
7ccce054443335397f64fe0ad24baca7c60beffb
8602 F20101114_AAAVWB li_y_Page_111.QC.jpg
52f3b3e1ca6c1f4c1da36dc89a55c1cc
9da7bc394a915efdbfc6aada25b1b9419b090e73
7149 F20101114_AAAVVM li_y_Page_040thm.jpg
5cc363ba6ec11ea1505fd34b20fdc73e
9055a834e78582d5143dc1caa9c41ca571fb5176
24785 F20101114_AAAVUY li_y_Page_006.QC.jpg
840f3b430df1bf6ff891014ee99c0a2e
b1ff9b0e08f77abca62a39e6ebc39717dad0be39
22837 F20101114_AAAWBI li_y_Page_062.QC.jpg
9239a9f053e58d5cd066e18a4c2cf8ee
83dac3b94fa869de645650775b70274a821a3d1d
6920 F20101114_AAAWAT li_y_Page_048thm.jpg
79dda984f037af453ed4b884fb9195ab
bcb0890828361d31aab3e42ff8277722715b44a7
1484 F20101114_AAAVWC li_y_Page_003thm.jpg
86e10fac90d63288c19b6884b93357ad
6ea0a8083aa39b4ceb53b7f45c9ca34238cb832a
25463 F20101114_AAAVVN li_y_Page_113.QC.jpg
5b724d9c9a4c62a1fa728313772d7674
78423ec05e7a90fe5602713b3e816c81a793dc10
7160 F20101114_AAAVUZ li_y_Page_075thm.jpg
250edbda36c0983dd2c7cf8a89a99108
d748c997f7bdb5beba16293d2c93b0faca4b452c
25479 F20101114_AAAWBJ li_y_Page_063.QC.jpg
5c71a83e9f92de0cb760c1c7e4a6be02
fdb56472783602c81803c57f4753e27df8a939c1
6450 F20101114_AAAWAU li_y_Page_049thm.jpg
b2260d747c7a7e5e7f22f205aa9fd150
7d29fba0250d4dcceffc090a1bc78dfd8b934e17
25590 F20101114_AAAVWD li_y_Page_080.QC.jpg
b1d32557363d688ae7766228fd6f59d7
d94e6ee64373f2b8c8d52ac37d8131f8c84aee9b
13425 F20101114_AAAVVO li_y_Page_093.QC.jpg
f8c2233aa6fc18a2feda35b454f91be6
985d54d4015b3c180353c1fe4987368a2172dc86
7043 F20101114_AAAWBK li_y_Page_063thm.jpg
147634ba115accb64593e3c3899a1d39
360f2a36cb3c97a02b5db9b73cf3d7eb851a3cb6
24052 F20101114_AAAWAV li_y_Page_050.QC.jpg
3b1282c8259d459aa5ab66c3c78dc1e3
1636bce6e0534fe746a6d6df2ce000c1e5f09c02
7158 F20101114_AAAVWE li_y_Page_081thm.jpg
6397c783d8072f8b353d44cb887693c0
50637ee78029dc96c32555d64bf9b9c2c2cc5515
7255 F20101114_AAAVVP li_y_Page_103thm.jpg
6e3039f76c810148427ab2836448121c
dd63d7686e0dcc6e4a5c7b61e5dc63d216a9b34f
7048 F20101114_AAAWBL li_y_Page_064thm.jpg
9507c570e3909c1294e3c2e7a0e4e69d
1f6a69afce907e3de98be0e1b3bc1dbdbda2aa9f
25994 F20101114_AAAWAW li_y_Page_051.QC.jpg
3a0de65851cb118a9acf3f9341845742
16405fd3cb84f76b1b71bcfa57e7af0b44f0dd5e
27131 F20101114_AAAVWF li_y_Page_102.QC.jpg
11151e9b94dfaee6ed9c8c8ae078b6cf
3de14aa5026b54aafc044eb349d20d56750b3cb3
6490 F20101114_AAAVVQ li_y_Page_077thm.jpg
1133cbada9adc5c573e4e6ff765d43c3
48dd9e9010a5fd82b21ba236146afb00ee2177f5
7192 F20101114_AAAWCA li_y_Page_079thm.jpg
b39464081ba8ec096f63abf823097256
54e5f2c27a4eae4354bc3c2ebc67d4949c15d5d9
6469 F20101114_AAAWBM li_y_Page_066thm.jpg
bd45cf4e6ef98eafb1438e1fcb0cec70
8082407bbb8b572cd7bb94f1ace30471182e7a80
5267 F20101114_AAAWAX li_y_Page_052.QC.jpg
3a0461ab218d64e7cf49257f27aa7f29
1be160812059bac83e327451bc44f41e414fe4e7
5727 F20101114_AAAVVR li_y_Page_124thm.jpg
27de77995ae42b88309193318a54196e
8cd641d1a564b0530bc8885a1b425ae1d81553ec
5179 F20101114_AAAVWG li_y_Page_054thm.jpg
63b6305f3d3a1cff632f599284f515ac
ddf5861009ac9f1c79aa2ff4b43e75ac23b0781f
22667 F20101114_AAAWCB li_y_Page_083.QC.jpg
354ac598dcedd3b80b38334f67ccb9d1
9ceb26670bc0d724e77a361e936f5f8dcc74af0d
22577 F20101114_AAAWBN li_y_Page_067.QC.jpg
d452e56bf5399beac8d563d2741d768a
55a28b70a0336d36c0623bb232e41e37332ff4a9
1797 F20101114_AAAWAY li_y_Page_052thm.jpg
ebc81be14e558d8639a110b1ea7218c7
9ab3655346fa3dd266b85e46ebda5f355e3977cf
26558 F20101114_AAAVVS li_y_Page_046.QC.jpg
0cdcf738b5b1010d3b16c64ab024e088
8da8d843752374f7cea4f39a4f40ee52e8c16329
6242 F20101114_AAAVWH li_y_Page_006thm.jpg
ae6786d4aaf6c6a14c4127a318468b87
5a0829a9037efad59b270d8a3b097bd9bf59e238
6732 F20101114_AAAWCC li_y_Page_084thm.jpg
af0d9a12e1012697f0ef2022feff89c8
74d28f9359e0bbf1ff82ac6a6110893c2ce3391c
6307 F20101114_AAAWBO li_y_Page_067thm.jpg
46582cfbea8a343de1da68e40a582916
f605baa4d188a8cbe4ae6061cd42dfab01f6391a
2945 F20101114_AAAWAZ li_y_Page_053thm.jpg
5930ac9a8feb96ce9dc70c5b65e37505
19bec2dd1cfcb00e14addc3751e0d2c0ca3d9747
5326 F20101114_AAAVVT li_y_Page_125thm.jpg
912f788494620402f42d6e0c2ce026a1
4d7555df462ad79856e956c1efbe6d1671ea276c
23932 F20101114_AAAWCD li_y_Page_086.QC.jpg
184f84e332e1000200f92e2e6963073a
2b6aa09d38460b28a829716d9e93f334bc61dc57
9513 F20101114_AAAWBP li_y_Page_069.QC.jpg
5518bca7e49383d8bb0621e0169d6307
fbb4b109c6618b63f07f4f8702eb40a65b10a169
6358 F20101114_AAAVVU li_y_Page_114thm.jpg
0f7fbc619ee39cbb5281d535224a260c
4e2f80466bca84e3479e9ad11b347e856383b31d
12878 F20101114_AAAVWI li_y_Page_010.QC.jpg
d899a9fe98f25fc2e5563b4ccd232bff
048ca733218b18cff3e9b0cef5343cb2fe2deb8c
6650 F20101114_AAAWCE li_y_Page_087thm.jpg
4a848206c0ce86255461662afb30067f
7361184f459fa0ad39a6ecbb385a5bc013c0e9a3
3186 F20101114_AAAWBQ li_y_Page_069thm.jpg
93c37dbf4d859606e05e8a5a8316e39d
64a6f12fd9df77a43a1112d13dec7246263be54d
4713 F20101114_AAAVVV li_y_Page_092thm.jpg
c8cefd377d544dbec25e716c943d33fd
6e8b4834b094d297eeb460c9d1cbbe6bdd017586
7277 F20101114_AAAVWJ li_y_Page_022thm.jpg
85eac74811db97cb4beb06ac28b35f29
e7953cefbe2c30e2ca09848440228e9301b213b6
27475 F20101114_AAAWCF li_y_Page_088.QC.jpg
1a5fd1bf5a7e1bf892391eb6c53e4efc
e7d7e551b44cc93d2b3ac2eb629a5165b40ccd61
5304 F20101114_AAAWBR li_y_Page_071thm.jpg
dd6230eb698c58f640d3caccd52790c5
5cc1d035ed171a01c04c8a35893694711e5a1c29
7171 F20101114_AAAVVW li_y_Page_065thm.jpg
4f3b44a69dd8f425d6222480cc9c0cd9
32e7e28f07228bb67ab71f610a20e33f804a8e6b
2734 F20101114_AAAVWK li_y_Page_068thm.jpg
82d26bf95757c222db6f1d2a56a46f1f
1a30c26e017f07a226b48a01fccc793a17fc301e
2332 F20101114_AAAWCG li_y_Page_089thm.jpg
941e496e3888e38ea97f462ca965878e
2f2326ca8c04f36d53ebbec354034fa8694da453
7175 F20101114_AAAVVX li_y_Page_051thm.jpg
b6568bb641e20c0cbbd050a5b06f2d03
0ac0dbcc3b86f6b0147a7808f827bb632160c41b
7205 F20101114_AAAVXA li_y_Page_043thm.jpg
072534b1efe115e8c8fbb1e8c1c9f758
b0114dd4bf28c7ad9629e73e49779b896928172b
8604 F20101114_AAAVWL li_y_Page_053.QC.jpg
123a6543469e1aa004aee94e3b8d6a5f
0954833323db09f0e981e86beb6873df5316c58f
9824 F20101114_AAAWCH li_y_Page_091.QC.jpg
ee334d8cb1dc5d231f11d3c33701d12f
15993585904bb8e0dcf276653bf0341373a3d139
7090 F20101114_AAAWBS li_y_Page_072thm.jpg
26e22d97bfa392af2d7e53a67642735e
f4985982b6331192768dbbf1f5dc5e2897ceffd6
6800 F20101114_AAAVVY li_y_Page_101thm.jpg
73afd390e3faea0f0fec55bde7a83577
e5472b9205883cc5e0b3c6f9f86341d9374c5001
27833 F20101114_AAAVXB li_y_Page_009.QC.jpg
91708bd9c368a4a6b6c9c92596c8fa8e
176e3cfc1fd595d8cfcedbfbe20ca5977de7ba1d
24037 F20101114_AAAVWM li_y_Page_084.QC.jpg
2a1e50dc9990b2984fa935ce4a2172e2
cd3150622af12ed949c000948820c659bddfc1ec
12537 F20101114_AAAWCI li_y_Page_094.QC.jpg
e6878894462af714359daf567ec467b4
7dd318675bf37aa1d5c0d1eb5aa9de64fa0a17b4
6621 F20101114_AAAWBT li_y_Page_073thm.jpg
9ec18a4487f60cb7e482fa5d9eb7da5a
7b66ff519ee7e4768a96eeec4f778e2e277c126c
26458 F20101114_AAAVVZ li_y_Page_079.QC.jpg
ba3697150bdb0698a504af5f3b49c3da
1b4a1b87d8f34891ff59e76ad854653a5501b096
7303 F20101114_AAAVXC li_y_Page_102thm.jpg
9a1ec94623648242110374038e9cef47
1105bfd85b84af4361aa66b3abf27aa640cda023
2991 F20101114_AAAVWN li_y_Page_002.QC.jpg
5b9b6168b6fce5a60a3d9f0e6f11fd67
358d6b64836e6be2d18cae160e5da24705f39c8c
F20101114_AAAWCJ li_y_Page_096thm.jpg
d567355a44f6e0461e17b0bc194a4207
7413d1b040f3a3c79f374a010c471255eac0d667
6713 F20101114_AAAWBU li_y_Page_074thm.jpg
f69a2cd7f3c6c898f52441816f3d2b7c
205da68a3b147405d9f240f7cb559a2c403295cd
3611 F20101114_AAAVXD li_y_Page_010thm.jpg
f5bc7699cc9171340da24b5c228e2fb3
707603981dff9e810554f5e161e39988db9e264a
6809 F20101114_AAAVWO li_y_Page_042thm.jpg
e503a3a2a6a2741396c90a58d0fb6595
7124923554ed2d92efcaec973d88c69a768136fc
4071 F20101114_AAAWCK li_y_Page_097thm.jpg
29bdb818c486b31a76e0fdedc23b2f5e
0ae24e7e8ac386363950491093488ae7300fc65d
26167 F20101114_AAAWBV li_y_Page_075.QC.jpg
959308aa6fcced2f703d4f2ed2e10b00
010c7215daf033a76a262c7e0161c56a87644b8a
27716 F20101114_AAAVXE li_y_Page_120.QC.jpg
5d7078c8be1ab83f7e248fe4fbcf8cd0
a8fee6e73f4b27f0ae32292ca619c77f7bf164ed
3023 F20101114_AAAVWP li_y_Page_111thm.jpg
086982727e135c61fac680ecd11342c1
887260bc33e0cb622f3de9f7a275b63ed109526f
3705 F20101114_AAAWCL li_y_Page_098thm.jpg
a169871f93ae1f2df9169e5b443c095b
9ee3fa36d4ebc625f9cb86bd06b27b4c0751141d
26631 F20101114_AAAWBW li_y_Page_076.QC.jpg
07d5341a8c2f9d022d30971eee19620f
7a740de0302e4765fce7adf2bfa9098368438801
3799 F20101114_AAAVXF li_y_Page_070thm.jpg
96135090ab698a6e73b5e5c478d43165
77b5a1fa0f4c4f7d1a1ee379455643b6e6702cd7
25657 F20101114_AAAVWQ li_y_Page_040.QC.jpg
13df9afd4f3d7215f59ed2f8b8ff2af8
893da5f98f7cd0fdde582997863c6fd8e702d17d
22894 F20101114_AAAWDA li_y_Page_114.QC.jpg
2797568a95fbda404009456df4eade57
e5de76437f9e63b7ca3cc38b3f423815798b9eb8
6928 F20101114_AAAWCM li_y_Page_099thm.jpg
46e9a53a8a07af591f8c1562056d82be
4de5f529f4ab79571c01a85c24fb1a076d43c6dc
7217 F20101114_AAAWBX li_y_Page_076thm.jpg
c1ef4d706c0ee5ba662390f6a872eac4
efbae8d4b6669651d0c9cf263144a6bff497e8a1
25314 F20101114_AAAVXG li_y_Page_057.QC.jpg
db48471bbf481fb435c58854b1913ca9
59ee99a9924bbe5450f71432a4b863a5faf025e0
24788 F20101114_AAAVWR li_y_Page_042.QC.jpg
b6ce9eca0505a0b1772e93fc461bf42b
659e1ef161137602a7f7f9bfd650b4a1b851468f
4569 F20101114_AAAWDB li_y_Page_116.QC.jpg
a2e33e14445fe644187e8de8ede3c3c1
3aa6530907f44988dd3ca5654b2d6e942ee2eb27
26371 F20101114_AAAWCN li_y_Page_100.QC.jpg
da6456ff6ffa2434541ff569cf147ffa
d2036a8e5758ab85aee9c5c3c8b4758fa0d20005
26904 F20101114_AAAWBY li_y_Page_078.QC.jpg
d6858febee1226b3b08fa348df3bcf52
29df520f0889863b92ced21ae5107ccf7e7ea4c7
4310 F20101114_AAAVXH li_y_Page_094thm.jpg
f0de5069d9f0748778efe8d7cebd8f7d
9a5c7069f6f5313fee9919389e7eb00ff29b670a
26573 F20101114_AAAVWS li_y_Page_065.QC.jpg
ed07e58b06c9e11a672fcc115e3ba511
377cf3e06988574ee8a3d74e24144530c5bc319f
26647 F20101114_AAAWDC li_y_Page_118.QC.jpg
23357a84d3c174795d5dcd9ab350a0b8
5d2048b394a2da7f613d9ef0b0c670c8c893037b
6824 F20101114_AAAWCO li_y_Page_100thm.jpg
314c6548eac6cdf6fee5647de4ca8923
bbeb4add877918758236899dd4a17a1a50b7edd7
7006 F20101114_AAAWBZ li_y_Page_078thm.jpg
a2a1e9a9598139605921aba4f5063a5d
e75fc33c6ce0bca394397d4ce634c09b9e2621c6
16967 F20101114_AAAVXI li_y_Page_090.QC.jpg
a8a14858085942133d5d9d3e5d294521
f32a543e4ea27f008df723da5c2d3c551374c585
23753 F20101114_AAAVWT li_y_Page_027.QC.jpg
11b2a69eba33cfcad7ff6734f87d2dd2
76eafc54dc35acece541bf42ef29cee77083b5b2
81290 F20101114_AAAVAA li_y_Page_065.jpg
e02a075cafabcbdf707933381c08fa19
41bbdee409a107417d34879f586c0e86928942a9
7289 F20101114_AAAWDD li_y_Page_118thm.jpg
ac0a4382203c3b23797005b9361cf7c0
3fe1d649f90c34b839c80b85db11d98aee63daae
26100 F20101114_AAAWCP li_y_Page_103.QC.jpg
9a222027406e44c7e45a9ece30b65867
99e98972c235660d1cee08b6513b4fd462469091
23910 F20101114_AAAVWU li_y_Page_020.QC.jpg
88a371aba76cd45e44e6c6556798a100
329a7b5374a346db77b52738047b6288af85f92c
70794 F20101114_AAAVAB li_y_Page_066.jpg
ef158fd332d06bd6c8af9c1d5957dcbd
1aa59eb0735a07d65cfa6facf70820e4ddad17fb
7413 F20101114_AAAWDE li_y_Page_119thm.jpg
5c3eae04e92ab54cbbc4614ba0dad16a
0bb8666e2b4876fafc039e67174fbdbc8dddfa27
24574 F20101114_AAAWCQ li_y_Page_104.QC.jpg
c08ed0f9076e8928a8ec864eeb38a153
be39d6218f04a55213cad415c46c27cf107cd197
25188 F20101114_AAAVXJ li_y_Page_087.QC.jpg
5d908f512dd5c6f26b3b57153274ef5f
c6aad6663e0be2081c198596b3e16f20fba45e83
3850 F20101114_AAAVWV li_y_Page_055thm.jpg
10cd5229f3957fd5a9a49079cfb0d330
c8a3306492e3c85fb54b349dbfcf02d3277d23b2
70379 F20101114_AAAVAC li_y_Page_067.jpg
fe4d71f10f15b87b8a80242f9bbe3fe1
50f0c153788b3f5e8ccf4a15edd41765d368aaed
7036 F20101114_AAAWDF li_y_Page_120thm.jpg
b91bb0ecea2e6d47f203433344e723c6
6bca467b7db8b8e3c8c964e0288978474a7206d5
6781 F20101114_AAAWCR li_y_Page_104thm.jpg
420dbcd735df78c6afe8d1873f5c682e
c8ed0c0c212c67fa05793e5e4c78d6a60ce22d6c
24921 F20101114_AAAVXK li_y_Page_115.QC.jpg
4d68da8e10a9c84c816f85ee5bfa2a09
aad5a8293f6d6bc461ded581a10a5a84ab228268
26973 F20101114_AAAVWW li_y_Page_060.QC.jpg
b467816bf0c87791d35e081f2667d210
cecbc5b9a8f2f64b7e0ab6257b3d672ec13cfe50
26963 F20101114_AAAVAD li_y_Page_068.jpg
09637ed554dce5f2d0eea745b73b57ca
e0440bf697c3e3e030b5e080a3a59764c253a1a8
27106 F20101114_AAAWDG li_y_Page_121.QC.jpg
4b6f92383f13673266656b28f7422e09
7e47fe91d3d49e4d736194d0463b042bb1e3d22d
25571 F20101114_AAAWCS li_y_Page_106.QC.jpg
f52c37bbc8b6642fdc86db41fcac8cd0
abf9f2fc4b8f1943b13e9ac25f9c7151d572349e
24426 F20101114_AAAVYA li_y_Page_073.QC.jpg
05e802243f2a2d32c951343a8bee0704
2fd225c50fa2ecba43d6c609211187199ab16216
7033 F20101114_AAAVXL li_y_Page_028thm.jpg
d7e0704ba827396072753ea16b5f778f
dac8c89e6111acd820daae32882f58024983c3f2
25506 F20101114_AAAVWX li_y_Page_058.QC.jpg
0a15b0b813e227256a287a6eb3f683af
db96707ecbe4edf832a55016467d05bfd436bceb
30373 F20101114_AAAVAE li_y_Page_069.jpg
b330f2a8e68e713353cd4626cf550df7
fd520486617bcc0c78e8d46cdd0997c94285598d
7065 F20101114_AAAWDH li_y_Page_121thm.jpg
54d8346395407b7b0cd1635254f2e744
3fa6f1ad779a051167fd9e58230ada6ff4d81174
28566 F20101114_AAAVYB li_y_Page_021.QC.jpg
de84e4811d89b3354ad1cecfcb5d083c
ec38d34fe0d1db28dff2d12ed018d9ab6e773d63
5091 F20101114_AAAVXM li_y_Page_004thm.jpg
60190f908cfe924e18b5eb9a2e556c2b
6f0edeb8d0296d3ec64a05ee4745e8dd53522fa6
25458 F20101114_AAAVWY li_y_Page_099.QC.jpg
49b026bdc0fa2e7222a1db582aeba9f3
c7173a0fa64279a7f56bae189db49720b9c77839
30352 F20101114_AAAVAF li_y_Page_070.jpg
690ac5b15b2311aa85c4eb45bae1580b
f9dee1516f4155c2030c08043293cde6f7e8b09d
7331 F20101114_AAAWDI li_y_Page_122thm.jpg
480647fa429f3a710813d1b516e102ec
5b492dc471987e230b333aef8e8bcd3f29337901
25513 F20101114_AAAWCT li_y_Page_107.QC.jpg
22beb084d613408cf3b75ef97645db8a
362414f982521662a25aab20bd94866bb890e989
4543 F20101114_AAAVYC li_y_Page_093thm.jpg
c7395d8e267cbb3ed0bbe0955c64df8b
cb3122244e7c9b092dfdbd2ce0e85cf4229c1bbb
7027 F20101114_AAAVXN li_y_Page_018thm.jpg
bf73ac3cf29ffd93da0d6f9532678e47
46b64785a6bc52274df09c36371ac10131f91def
28239 F20101114_AAAVWZ li_y_Page_119.QC.jpg
2982717ab31330fbf160ca1a6ae912d7
b9721910ed5be376d342857cfcad2a1210848e70
58227 F20101114_AAAVAG li_y_Page_071.jpg
b1abaeafa6868f581d3eac3b2b50e45f
d75d578349df2f7fd10484f477e684d9292d505f
28600 F20101114_AAAWDJ li_y_Page_123.QC.jpg
ba4347f1d86edafe029047f5724e8d58
098fe0be904cd803f34b1c45360e338027fe0e62
6711 F20101114_AAAWCU li_y_Page_107thm.jpg
2eb0e2c709777e985acc77df573c3611
0b50409c0b2ffc70e7e1d7388ada74b93d4042bf
23109 F20101114_AAAVYD li_y_Page_011.QC.jpg
c6a6ca9070779ffb8448a2b097249578
8617ae3a51ef91c364e75b96d0e30a935d7853f5
26846 F20101114_AAAVXO li_y_Page_081.QC.jpg
967a4251d8a573ce170cc5711e51eab7
f381903aabb5a273a6ec6ab80ec04a43e8ac3cdc
80077 F20101114_AAAVAH li_y_Page_072.jpg
0c4640118e834418275419f11bce5427
790ff20ea7bd149216df204fa8ef47671df72e6d
7475 F20101114_AAAWDK li_y_Page_123thm.jpg
5d95dad2c35955a36678aadb5b2dc681
86a9d3d7167939d426290551781d782a117f7488
18098 F20101114_AAAWCV li_y_Page_108.QC.jpg
25b57a20f9b837939fda5ba0fbca170b
c9b8f5bfbe97dd0b1a002ba483cd1fdd92416ffa
1658 F20101114_AAAVYE li_y_Page_116thm.jpg
fa6b2a0ce4a3123745b7d24447b108ab
8419440e94a07c719565bb67a84056a90a30db7e
3002 F20101114_AAAVXP li_y_Page_091thm.jpg
0193259ab1e4daa523a706d872d98657
de1de83eb2e7adc7e5fa8bf2dd353fb9bd0fa933
74583 F20101114_AAAVAI li_y_Page_073.jpg
4523f55e87da17a341a01077ac198952
30b02e6a77ebcb3be88b47eb7bfbaca1ebd31faf
22629 F20101114_AAAWDL li_y_Page_124.QC.jpg
c9994dfb804d2f48e1b7c6d99481d59c
3d81e40ac0d8a51359d3da03e8538f7ed5001bcf
5153 F20101114_AAAWCW li_y_Page_108thm.jpg
04daff8b919b7412a23e688b6fe48066
ac9bf91363e5e5310e750b14ec55ae2ccedf5da6
19111 F20101114_AAAVYF li_y_Page_125.QC.jpg
a20967462132943383e39a13cf6b0ec1
f78d2c75a36ab0c480605b813eb51ee789c48cd8
24566 F20101114_AAAVXQ li_y_Page_074.QC.jpg
c75d613e648c8976e914729e1eafd7d1
81f4326fb9a5be2cf24add4091261fde74300c29
75418 F20101114_AAAVAJ li_y_Page_074.jpg
eb840fe07d60cf201d280d65e229dac8
40cce806c8a473deb99610edf276c3580e3466dd
11100 F20101114_AAAWCX li_y_Page_109.QC.jpg
2f9e28155c7fe9828502c5883acb2529
24d404dcfdf54d2a5be4daa8c7397abfe9f349b5
4964 F20101114_AAAVYG li_y_Page_090thm.jpg
383c1dfbb0b0ad3c65b252db435311bd
198e9efbb0cd967bf7eb6d6dfb973d9c98eb1473
12680 F20101114_AAAVXR li_y_Page_097.QC.jpg
26b7c760b390de4e58506820f7deed10
388f94d457efddc9f5b9d559a550dc9985bcb1d6
81066 F20101114_AAAVAK li_y_Page_075.jpg
c1de4fd1c3d48eeb014dbde2e28f9543
06115a9120a6d3f679efc2db062ba767ded02e24
25529 F20101114_AAAWCY li_y_Page_112.QC.jpg
bec0f7d79cce2a4e8d1872396e6b0a79
324ebdd678efa15f98b43da1a1c6b72dd4583b0e
18230 F20101114_AAAVYH li_y_Page_004.QC.jpg
530dee5ceaf390bba9443ccb8ed620d9
c594cef58d4f0ac9ceec40477b8193e6e1d77aca
9271 F20101114_AAAVXS li_y_Page_095.QC.jpg
eafc8f7fc9f9e94c3b517add23cafe41
61709f5906a53180987e89a94663b1220c15ffe1
79721 F20101114_AAAVAL li_y_Page_076.jpg
749339586fd927f303a051933a88686a
7199cd11d9882777def898a1e57c5c74958c1140
6828 F20101114_AAAWCZ li_y_Page_112thm.jpg
77d2b1c48314eccaf9f9452c94f38740
81c600ff7e90e13c7492ce703da3e3cc8e09a597
22067 F20101114_AAAVYI li_y_Page_019.QC.jpg
6bd22a2d5265798ac83d7ed91d403a2a
4400ecc269b088013765c3c0266eca8465d02f05
22764 F20101114_AAAVXT li_y_Page_077.QC.jpg
f36f173ebd3b0c018c39fb3688829b31
36fcda32b896302adca2971ed1778d6546896bb4
46722 F20101114_AAAVBA li_y_Page_092.jpg
c1b4d2c7cd3fa5a174283e87c1b81edb
c85dbaeb88c6057046358a88d157f78f8499356f
70721 F20101114_AAAVAM li_y_Page_077.jpg
bd170cee3a61c650ffbc36cc197dc129
293438783d3fbb7f00c78ee132a692bd2f16c458
3170 F20101114_AAAVYJ li_y_Page_095thm.jpg
f8450091ce33f0d5d641c1c87e233474
b3c76331929e3ed20d6cb324df6b2a8700c03b3b
6694 F20101114_AAAVXU li_y_Page_086thm.jpg
46ea2135a42af8b636603d35642ed3e6
afaeaa885303e8ff5084ba109d6b3d5be4949650
39310 F20101114_AAAVBB li_y_Page_093.jpg
ca6f11617e33c6bbce3cb80658a35c46
f7c6af50da953812e93d59cc96e1621fce2d0daf
80888 F20101114_AAAVAN li_y_Page_078.jpg
5ba3bc704a5ef9717117786ff1589c7d
d7573dcbb36c75a0580ea27884a3223ad336e4f5
5398 F20101114_AAAVXV li_y_Page_024thm.jpg
d38319c766892e4ea7f5572807f08ba3
ad4eaba561b45235cb2a10451ee8f67530c4a331
36727 F20101114_AAAVBC li_y_Page_094.jpg
e535fe27b307a37492fd0ac650d5c827
d94e77e36b65740768657d1cec46d2a61820bcf9
78874 F20101114_AAAVAO li_y_Page_079.jpg
bd9270fdcc93346eeb3b613092365624
653823802cac37d99f3434e7295741af903aaa6c
9164 F20101114_AAAVYK li_y_Page_032.QC.jpg
e7921f825d02cf5745c7e956aefbfead
a2bd4993a8ba1f0369eef30cf2b437af813d0ff0
6543 F20101114_AAAVXW li_y_Page_117thm.jpg
f062a4870b9bba8acfa848c114284a78
40df640222ff8f49fb8b5fd4cd33c6f7d486e3d8
25253 F20101114_AAAVBD li_y_Page_095.jpg
78a3419ad3c9fd9bb486b656f1848878
ee7c1c5eae5e600f8e391f747a2a459c03d8d6fb
77468 F20101114_AAAVAP li_y_Page_080.jpg
d42b0e03d1cc47b8142ccab1929b48bb
19ee61ba37df1a21acd7cce938af3dab22ddc496
25958 F20101114_AAAVYL li_y_Page_105.QC.jpg
6399c1fe4ace3e0185561d4a3d4660b5
a480d30473b2ae7168969caaaaab3d8c8615957a
6875 F20101114_AAAVXX li_y_Page_113thm.jpg
6d06fb93165fe09f32d897dd0efec2a9
bbaea59bcef4c1571865dabf605604d95dec96f5
28414 F20101114_AAAVBE li_y_Page_096.jpg
da5beb728bbca1e9b92cecaf7ef288ee
8345d6929ac974ab221a730780b212e165c09731
81894 F20101114_AAAVAQ li_y_Page_081.jpg
2983ddf073c39cc10c7a4be7800b9927
a987b7ec92e5c96ed79cdc27614aef8b91809529
12016 F20101114_AAAVZA li_y_Page_012.QC.jpg
4f280551e33ae69f5982ae56fcb5fe09
39526f5a5728664235739bc181b2fdc1b5216156
25847 F20101114_AAAVYM li_y_Page_064.QC.jpg
1a4d45b73b37188a99f602cdedb8395a
9b9a40720e8ceaf008a19788383f0b8cf1e4f48d
26340 F20101114_AAAVXY li_y_Page_072.QC.jpg
0a4f991b0483f84d649d5723ce06329e
202334b05ebcfe7a7d8fae05755ecebfbe2cbada
37124 F20101114_AAAVBF li_y_Page_097.jpg
3f237fef2daa9708ff7eb9eff36efd43
6c0b77dec7f326a272faf97051b8f9a733e6e368
77598 F20101114_AAAVAR li_y_Page_082.jpg
be1998c83b581edaa7e6e10403bfa61b
070c988b80e2243693f23b1b7a4fd8fcfa1bb4d9
3669 F20101114_AAAVZB li_y_Page_012thm.jpg
13a406cb3caba65e544dad81c4763547
584d1cce1b6437a2ef420cd16dd7da0e4c97b1ec
9964 F20101114_AAAVYN li_y_Page_098.QC.jpg
c34d9719beb65f54ef6342cb3d75f5a5
21f72128504fafd2134378b949d8e088e7d6c6e0
6523 F20101114_AAAVXZ li_y_Page_050thm.jpg
62dfe916380b4366d5a4009aadc40386
b3ed9024a383ae1b2aa37176df530bab80b7cc0e
28720 F20101114_AAAVBG li_y_Page_098.jpg
d759e387550dd47b50559ffcb146dfb0
9a076abf3d5c1ba403888ceaf3122826585cb7ca
6237 F20101114_AAAUWA li_y_Page_083thm.jpg
aee967f711b47699ecbb9a77970ca7f5
60875612cc8be6708196678bfd8e251f58a25586
24714 F20101114_AAAVZC li_y_Page_013.QC.jpg
78dfaa2875ca5f022b13d559a72542fa
e893407d8398ba2aafd2b149e22f8a6b5384cfb4
12268 F20101114_AAAVYO li_y_Page_056.QC.jpg
c56ae8a533a882d17b5594b0e896acf7
f4d8a03f051332d3e7ef4aede460c6320a64ed92
78991 F20101114_AAAVBH li_y_Page_099.jpg
bd6e6a6a97bb1ac4e1de3249ab147702
0f66be3e2aaa9e565dca1932e91c8dc7901dfbff
55769 F20101114_AAAUWB li_y_Page_059.pro
3197c43c5b9193176aa877bfab28eec2
31f3dd164af3b7cc599eb07747c99ff51e875f28
69108 F20101114_AAAVAS li_y_Page_083.jpg
4dc38eb65670b8fe8e225a68d4044fd0
5598d5ca65ad01347da87b6c71e7c82d0f270247
6878 F20101114_AAAVZD li_y_Page_013thm.jpg
8b1e2553a89bcb1e2b0c315b87783804
fcde5661d7df7dcff2ace3db4e9dcf1ac8f15cd3
3613 F20101114_AAAVYP li_y_Page_109thm.jpg
3e45662f3e44fcf60c1b283a7cd97578
4f04cb6c27457efb61f68a879b42ac22d3d0d001
79659 F20101114_AAAVBI li_y_Page_100.jpg
b7b52722a6f7c3c23fdf92a324e5fece
94b23c28d437c975275df48b17e87ae9c31025aa
6528 F20101114_AAAUWC li_y_Page_035.pro
4dd1fa0ba292b63b7ff51b0c420f327e
6acecb5144dd85f9ac103e8d64adfe3a4d17fe9a
72625 F20101114_AAAVAT li_y_Page_084.jpg
25f20f4e0373680fef03071bb3f4fc24
8330743ed0301b36b5146848b1a570b91820bd27
25756 F20101114_AAAVZE li_y_Page_014.QC.jpg
880c5053b548e4d9c7e07f8863fade1a
c21843d92f967fe33dbe6635a7fa7a017d7d6460
6841 F20101114_AAAVYQ li_y_Page_085thm.jpg
a9f496f0ef4f4493fcd86be14de5b559
3fb4b8fbe790d28c099dfb93f638f7c1238e7de8
77852 F20101114_AAAVBJ li_y_Page_101.jpg
94df21fee4be0212d75415602e3c5d48
17ec93dc69d11bfc8ddf50faadb089f9a797cfd9
1931 F20101114_AAAUWD li_y_Page_114.txt
1b2499ad25930cc166d65a50eb8673bf
fc4844ea00ae6c8766c23045eaf36ac25caeeb08
73923 F20101114_AAAVAU li_y_Page_085.jpg
8b2306a3bd97e52c64007ed84f989c32
49e5e84b1999c37d6feb72236a2732c85bf991f4
7103 F20101114_AAAVZF li_y_Page_014thm.jpg
890cdc28457341adacf3599eefdcc472
081395ee62629e3153221257e28cfab03885b6c1
7216 F20101114_AAAVYR li_y_Page_088thm.jpg
866955e3259e887e081b83e5fde71e96
76d57e95179cfa5e9c49714a24ec2b32866e468f
81275 F20101114_AAAVBK li_y_Page_102.jpg
5b4b8d901e55f3eba3a5d190e6bdf5d0
e37816c51feac1b2ca250de6021bd10955167d23
21412 F20101114_AAAUWE li_y_Page_012.pro
a7bf6de6168071223115286a7ffa6994
90b1ec1f2821ec94ca465eaba356946190a107ad
72956 F20101114_AAAVAV li_y_Page_086.jpg
66e4909b56a37c15461a450494e4db4c
a248fbcfca29adf11acc401dd056aba0fc76fc37
25115 F20101114_AAAVZG li_y_Page_015.QC.jpg
685f958354d11b428b84631834a70cc1
8c31ad5946b9367e7bff7a7ca40c0481f074384e
55106 F20101114_AAAUVQ li_y_Page_029.pro
451247c16a4c80546435e3db9fec7000
fdd5b781aeebd5f27b1e10bc873c9690f318a529
183998 F20101114_AAAVYS UFE0021202_00001.xml FULL
e03ed9631299323ddacde7106f5fea6f
f4c39a91eee6168bf45c0157a08a7e2938bc7ab8
78937 F20101114_AAAVBL li_y_Page_103.jpg
a6f444f42d40f409661ee91af8138fcc
62bfa20c06c266565eb95d74c2e74f70fdc49ca3
22699 F20101114_AAAUWF li_y_Page_071.pro
ed9b1dd0c8329ec9fd3759e63f7f4a4a
4b5c43d4647d9068b680f0d1462ed20d8f8acd41
75994 F20101114_AAAVAW li_y_Page_087.jpg
fb51da903e18aa903e98bab60a58ee6a
02115ccd6124b6a9557cbd55c3200def19fa21c3
6808 F20101114_AAAVZH li_y_Page_015thm.jpg
3d18dfb37b4dbf25d2775a59ff9cd82f
79012d5790c07b7b0d9c613fccf1ae7e9c9183a5
24813 F20101114_AAAUVR li_y_Page_117.QC.jpg
664a1c38aaa361b9c3c3f82d8e542710
097eeebdda331f215b2a7a0b92ab7f387cc640e8
7304 F20101114_AAAVYT li_y_Page_001.QC.jpg
1ccd1a1dac1fd57e807a34ff953c5db0
6c9df341cf2a3cbd22bfa640051987e330807956
99523 F20101114_AAAVCA li_y_Page_119.jpg
c52462232e204dc75c8b119407f3f261
2a6203155b6ff1efec64417cfb289a2c96cc1e2e
73433 F20101114_AAAVBM li_y_Page_104.jpg
4f5c1c7fa8ddfa9381d164a19829fe2d
810282879b80b4aa0536088520a5c3d27095ce86
115561 F20101114_AAAUWG li_y_Page_112.jp2
deac0580b12bd7f6c392c2437441aef4
ae6fd6ec45cc69327a4db9ce5f561d64b5377224
81790 F20101114_AAAVAX li_y_Page_088.jpg
caaf3914ffc0131a9b9ba6f5fec21964
432e15525d2250c19eb761cd55e4e2b92ac5163b
25733 F20101114_AAAVZI li_y_Page_016.QC.jpg
0219cdaab785651f54b5b9fa4cc22c28
e8446f65e68f3c07a40165aff5380da60eed3012
9977 F20101114_AAAUVS li_y_Page_096.QC.jpg
c6e52178bfb6ab4799bef1810e8514e9
413066979894ee6d5b8264fc51492d8d18742e46
1292 F20101114_AAAVYU li_y_Page_002thm.jpg
eb05ee82d4e6bb9567202565845d32db
f0957ac70ea498c6d4457f9639386769647de5a0
102417 F20101114_AAAVCB li_y_Page_120.jpg
26d2c45afc8cf821cfccc4076fd5b061
5dacf7e51739b48c646c0f0cb78425e8f421edd3
81997 F20101114_AAAVBN li_y_Page_105.jpg
33df39758b475b123a381917ad1c3537
beadc91186d5f2e75d09e9cc05b4da22876fc2db
2634 F20101114_AAAUWH li_y_Page_118.txt
206d2f70a1004f2bdec77b7730a3b372
55f9d5097e1dec06e5d345aa8a1895761e2afa72
56845 F20101114_AAAVAY li_y_Page_090.jpg
24a8aa6d31fc60bdebce2e31ed6e23cd
0eaa131bc5ebb324702f2ffe37e4dbb2759c6921
7015 F20101114_AAAVZJ li_y_Page_016thm.jpg
ce5ec847b08a31c66c329176c299d485
cf6f08802d689492c67777088ca55b2b4f66fe60
5424 F20101114_AAAVYV li_y_Page_005thm.jpg
1cfd16494ef91f521b7c5b112b742103
52c04807b27bd4d6d112bd3f7cda38cc2bc57203
92542 F20101114_AAAVCC li_y_Page_121.jpg
8fbea09d70fc98ebf3386ffdffe802ed
03f1e9516ec7615353136da70a33ab9f73d6a55d
76000 F20101114_AAAVBO li_y_Page_106.jpg
7fbc6dfa2672c76002c4cdc16bb14104
02515953092513332052991216511fb49c4ef16d
1517 F20101114_AAAUWI li_y_Page_125.txt
9300d50f92110e5b2e62000d1db3cca9
5969e898e70fc01f3bdb4aa6e2ee02ce37edfda1
29556 F20101114_AAAVAZ li_y_Page_091.jpg
1a8f57d8edd82e61ec2cdb60c11c10ac
5e3896325e45577ae384e89c81f1e4cf68593a6f
7038 F20101114_AAAVZK li_y_Page_017thm.jpg
e2f3741cd0a13b8e27c94f43894c815b
abd245dc5f53955bb1ebe8c995c81db35b78e232
2185 F20101114_AAAUVT li_y_Page_078.txt
68753f695bf09e57fc0554f7e6344d2c
3ae9061cd6dbcd9acd40e8bc8f49f47fe8eee87f
7064 F20101114_AAAVYW li_y_Page_007.QC.jpg
62191b88ea3163afa65707b1e752be41
ccc964b01db96ebaad43c8cd957230a869ae97b6
90154 F20101114_AAAVCD li_y_Page_122.jpg
071eff68ac86b44fd54eda390946b00d
4cfd5e97b3de5b6714f5bd261c7cab2fa882273f
76875 F20101114_AAAVBP li_y_Page_107.jpg
a3ff2b9570c4f6e884b06982bb2c83ba
c1c46edf1d44b5dd20d601e31499b1e05fb79dab
52132 F20101114_AAAUVU li_y_Page_042.pro
9b767993608ffee50495fb0d2d464ef3
e569ed07d42cd442688a04a410539c9262e54b3f
2268 F20101114_AAAVYX li_y_Page_007thm.jpg
3386baff2d6cfacb7e2c20e70c4b097b
2e1622802ac86a04104f88038df1b7388106409c
100428 F20101114_AAAVCE li_y_Page_123.jpg
436718c98aaa65dbad87c5d4ce6496d4
4214cd3ffc68ba3958637301eac87f5e7ef455d4
31318 F20101114_AAAVBQ li_y_Page_109.jpg
2a9bbf890fbc06fe334b4c2ee0917fcb
7371d6c14fef82efe373af9d0c8397751f56ee03
5995 F20101114_AAAUWJ li_y_Page_025thm.jpg
b9caf2fa69ed1b917fc2a72b0b1c70fe
3ada9eaa8852ccd8db9ee17cbfa0e0f8e077ff43
26072 F20101114_AAAVZL li_y_Page_018.QC.jpg
960d3af14ccc238d3fb48f61decdfb1e
a4fee4b80b5295c2625ef06f9fe5ae2602c50cd5
14228 F20101114_AAAUVV li_y_Page_092.QC.jpg
c05b7a2fa11bc59852e36d4f20665edf
58d175474157f33c91868bdc275b867ce311c273
7178 F20101114_AAAVYY li_y_Page_009thm.jpg
87f356393a675162178f1229bb522497
8e30b3467400c02c178d7c3bb88e0bf776a08d24
82399 F20101114_AAAVCF li_y_Page_124.jpg
19e9dc80c98a4080d256029762a68e1b
07db412eb3fd7b61c9680e082d7da3f8bb3c046c
25481 F20101114_AAAVBR li_y_Page_110.jpg
775b6e2018560a247b65ef1f64390e90
5ed18b399fa67603f5ca2e8034400991cdf2cf4c
25292 F20101114_AAAUWK li_y_Page_044.QC.jpg
5d3b269daa34bd273ee80ea2cd3dbf96
81e72115b894e0f89e595b8aa4fddf95907288eb
6135 F20101114_AAAVZM li_y_Page_019thm.jpg
c5323639329ac28a6fdd601445112627
8d40dd0e1b1706082084b73d3cd27d1702d91386
1887 F20101114_AAAUVW li_y_Page_083.txt
f544c7ae95b2ac52a059c9f36375d5e4
d154d96364a5f13d9d5446717b34e6a3488db24b
6562 F20101114_AAAVYZ li_y_Page_011thm.jpg
70e4830c1d2d0868d6a4e44c205597af
19b2731a9ebb475349537b5ad0703248932e16ef
59542 F20101114_AAAVCG li_y_Page_125.jpg
207774bbfea8e720b764b897b4157a38
11e8a1d7bcdf3269660cc7d7285d5b74b75ab452
72404 F20101114_AAAUXA li_y_Page_061.jpg
441fe4742a1f50abd6340732855dac43
61237a8a8be22edee0a20db8b1463e726af2c2a0
23977 F20101114_AAAVBS li_y_Page_111.jpg
9aab731a1eed21b525ab2664bdb12f27
78703995a2673ddc025b3aa178ae64af5360d534
F20101114_AAAUWL li_y_Page_065.tif
0548ccdf870888d15946afb5394ae2b2
ae075e3510a2c6dfe1d06a6a611e69873f4914da
6590 F20101114_AAAVZN li_y_Page_020thm.jpg
2eeb2cb76f6a07e8f75780372e8842b2
f3f697c3e419ff7e3594a5af726c9ff30bee5d1b
23213 F20101114_AAAUVX li_y_Page_066.QC.jpg
f35b4af933faf4c72f5582a4c1b4a7e8
961d682b1e490f75078545eb5d9032b19d78c0d7
25052 F20101114_AAAVCH li_y_Page_001.jp2
8f7a6bc72ba8f81dfb523de36ddf8fb3
f775618ed5e0c72312af9fd90fd00fdbe7d4bdd9
81425 F20101114_AAAUXB li_y_Page_108.jp2
d99c2616e2dd61cb67a11234acb9ae20
c129fc0226d8d1031d58291c8a2abb003ea736b9
F20101114_AAAUWM li_y_Page_084.txt
6a1da6060479a1263765a45856622931
d8477fd5f0dc96cc58005ecbdb9a44260d3da672
7680 F20101114_AAAVZO li_y_Page_021thm.jpg
a063570da350c9e7f855965fb1183ef7
be42707d84d31eabea499c5d4cc43a46ef92a618
21091 F20101114_AAAUVY li_y_Page_089.jpg
edb7616f8b56194ba993242e4682b931
f8ac78a11a2312f2d1b573a9597d53f3ae92afa4
F20101114_AAAVCI li_y_Page_002.jp2
055e9d6bbfffad6ae0a6924b10da4716
c88a6650f7e590323d85fd0bcb22117adc3190ec
27939 F20101114_AAAUXC li_y_Page_038.jp2
7eb76aed1ffb680e30af45d10e689593
160de593b5555b893d89c719e0d5a728eddcbdfc
77558 F20101114_AAAVBT li_y_Page_112.jpg
a0228410752a8f7b97e106f826ef97c4
b02b1132a2163940a5e728fb82b98b568600b14e
46319 F20101114_AAAUWN li_y_Page_124.pro
6e0f15bce222ccc6a48c5f42ab20cdf1
f9c1280765e6c460df7990384eeec6e7bd1409bf
27047 F20101114_AAAVZP li_y_Page_022.QC.jpg
97986c8d8d3cfae8026fbf13ff5ba0bf
fc1ee7212c0c80f13a0cc697cb67ea1e23c77c3f
120629 F20101114_AAAUVZ li_y_Page_117.jp2
babf243031a56511e6d18b3619198115
806ec52eb960152fca9189c9b16ec8fab8ae627e
8483 F20101114_AAAVCJ li_y_Page_003.jp2
4a7dda7a083c3e829d8ef7b8b3ecedc8
2ad7d7eab1230a9babc9226c4b8144026ae03e59
49648 F20101114_AAAUXD li_y_Page_020.pro
d9713a41458f3f1c0d862b6b25c43f04
49962018705c52a1c2c5242a94ac6a6b7c6ecb82
76977 F20101114_AAAVBU li_y_Page_113.jpg
bbab5a68f497df2aa9c335cbbde8607e
c7ea2320d4cbb498dd30a2fde287bf7d50f99edf
6785 F20101114_AAAUWO li_y_Page_027thm.jpg
1bddd84e8b3a4865cafb95e4c540442a
99d3f58782812cd2fde59256f59b527c09efb827
24229 F20101114_AAAVZQ li_y_Page_023.QC.jpg
6390837b2aa60a439f7dc0fd7745ca8c
88df2fcb1d636e8110aea1bf1f6168959a0f7aaa
82186 F20101114_AAAVCK li_y_Page_004.jp2
a91a2210638ce140cf4014488d77307b
e424458c694735d05fb714d637211798d9cb2f15
12142 F20101114_AAAUXE li_y_Page_039.QC.jpg
554953a0ac25846cbfccbe8ec0179c25
4762ab4290946b4c64f021471d7dfe3a538e6760
69292 F20101114_AAAVBV li_y_Page_114.jpg
45d6869e1739e4fbbfe2f1e7480788b2
8d369d83187d8486ed9446bdb186279f81857dbe
F20101114_AAAUWP li_y_Page_059.tif
44874a1310699961f7e8a091612e7753
c4506510c3ecb1c0117897f8bb391f8abf556716
6859 F20101114_AAAVZR li_y_Page_023thm.jpg
e5ce7daae80a7feaf6424a90c9436625
5c082f0ff6a47292a928848a602741c1cc1cef4d
1051982 F20101114_AAAVCL li_y_Page_005.jp2
dbd566c456370407c2a9f497791ebdcc
f389f97c01a8a9e51761a20caa36409f741d820f
F20101114_AAAUXF li_y_Page_082thm.jpg
d930591b488d2b0b585cf21a166c9ba6
bcb64fb388e47b33460d9aa660b151cd09ef58cd
75869 F20101114_AAAVBW li_y_Page_115.jpg
1d82bb47c5937295c311faa9368f761a
508dae7650b3c4b49cdb67e4fe8b70399481cdb3
61389 F20101114_AAAUWQ li_y_Page_120.pro
032a37ea69a429755d9ac98c24898595
ee5b62f296eb4cec6f32768dfd42b6b718d46b9e
19701 F20101114_AAAVZS li_y_Page_024.QC.jpg
7667e320f6a1ff47f660f3289073a372
0305c60f2d8532e8f9e2168832e7b14a5a88cc65
1051980 F20101114_AAAVCM li_y_Page_006.jp2
226d9e1c2a8969bda6af701c777d829b
26b69c275e99e8b6f2c837829ca87d7132d1b66b
5120 F20101114_AAAUXG li_y_Page_031thm.jpg
aa51669c20a36b7dd888413070fd32a2
3d57f424bd3cf6a48a09ea68eb742bce1c070c9b
13430 F20101114_AAAVBX li_y_Page_116.jpg
f814924536f405b10e030272dc1fd54f
5fed037497bb213a541f89fe91666af873df417c
634 F20101114_AAAUWR li_y_Page_095.txt
589409202f84790ab52772c00c967d7a
5fb386de6ea181fa12121490c4d3c62fb210b06c
21143 F20101114_AAAVZT li_y_Page_025.QC.jpg
aed1bb2df1b46444b7390732982259d9
820dc28e227be48f33a264eb0acdd07ad845c5d5
122858 F20101114_AAAVDA li_y_Page_022.jp2
56c2a4f01dafbec17a11ab1f2f2ccc23
81738e8e7f31ffef7e384f6674c3dbc3efa4e0dd
540908 F20101114_AAAVCN li_y_Page_007.jp2
ca5b509aafcaa8d2a01e5b59ea913df8
671b85d73945285c919d8952567750dc38544f6b
35605 F20101114_AAAUXH li_y_Page_034.jpg
f87d85c8768532c2202a4237bb374c08
f07f2acfd712c39cf5f5d844a62b260ddd318e39
78833 F20101114_AAAVBY li_y_Page_117.jpg
7197277d3126c2b37373ac8fbd73266c
53d2d86d743e6afc05461ec1613be3f393283a09
6796 F20101114_AAAUWS li_y_Page_045thm.jpg
9d0bf5a6b7d35f31ccb7a62db81ff883
42e5929b1cf6181703851c27fbf7f3166de08d9e
21130 F20101114_AAAVZU li_y_Page_026.QC.jpg
ae6b5522296fc917342168ed2184707d
e426a46d333654b7eb55d9b3b11e26afda9de5d9
110644 F20101114_AAAVDB li_y_Page_023.jp2
e10cdb46de9bfdb335ef998f69e42616
8d54339e92a608a1fb13c777281ce80ceb86c4dd
574262 F20101114_AAAVCO li_y_Page_008.jp2
2a6fef6869f627235dc84395a2dc4d02
db1acc6d9f8ea09af0f152ce4ed5f70e19b19f2d
F20101114_AAAUXI li_y_Page_033.tif
dc6c876342b99ceef43a8793d6ee0c8b
2e9fdc3a1093a6029b55295cd4adc37152f51c84
87777 F20101114_AAAVBZ li_y_Page_118.jpg
deb1395431b657231efae8d3eac736d0
74db33132aefdcd7b3f34183a8a7092163433ad3
9121 F20101114_AAAUWT li_y_Page_110.QC.jpg
34a2bb881b58b768d10724b6a24d51a9
3b94f63be74e01c1c0ff91b798f28dca95d2e548
6228 F20101114_AAAVZV li_y_Page_026thm.jpg
0a83b76de741b811b3e0112e77241d62
815350172c3992d872d101c9940131954c8e7075
87149 F20101114_AAAVDC li_y_Page_024.jp2
29aad896a44dfe32eb209bfb0cd2674a
74aeae8b762f3ff8152875966c5f43a74b0e92e6
25425 F20101114_AAAVZW li_y_Page_028.QC.jpg
c036ae7e91f630d9437b61f70988adc4
043425d1e8001c2b89528188922f731b4c117975
1051973 F20101114_AAAVCP li_y_Page_009.jp2
e9c995c9ee2a737a743002d9264820c5
001172a5b9327d36da2ecfcc07572f1da2fe54a2
3838 F20101114_AAAUXJ li_y_Page_037thm.jpg
4edcd4a8109d3473b48125818cf2af9a
e206b31d6cb7666d022ea44534693da8f012e17e
55524 F20101114_AAAUWU li_y_Page_108.jpg
67f65b2878befaf340a9304acafdbc44
d03af70004ec5ee67e773db6386f0c66cc3d27e2
91227 F20101114_AAAVDD li_y_Page_025.jp2
a0d2c17c7fc17de9b6127b44e974be0b
813519302091076a128f43e1229c740f6e5533dc
25165 F20101114_AAAVZX li_y_Page_029.QC.jpg
8b3deb53a04848363193312e785ff61a
8e51f76b69fc4a868ffa0a1dd8a9edcb344923f6
1051975 F20101114_AAAVCQ li_y_Page_010.jp2
c53b9436a4acfb64b3ce67b11c1d99ba
5940cddd46318d08bf107670008055d3301583b1
113348 F20101114_AAAUWV li_y_Page_013.jp2
85338f13ad46958d0de03522adbc39de
04a4435e53a8db9f4f0e572e5f5a2bcd58bea3eb
94430 F20101114_AAAVDE li_y_Page_026.jp2
f03d300ca33f412d0e145b1bf09a8dbd
96c075a12e7b244d8a83743e530b2e72fa376457
26581 F20101114_AAAVZY li_y_Page_030.QC.jpg
dcb99d3b1ca985e2fcdf7b8f8dde6c49
43098efeffc4f1125eef1a053b892c77ce8c7907
107669 F20101114_AAAUXK li_y_Page_084.jp2
12589e3f68cc33df5b9c7138092a121c
4b55485bacc04aab8d698ebc329de560087fa0fa
3082 F20101114_AAAUWW li_y_Page_110thm.jpg
ebe4917cc3327d585d5f18301e110ad7
834b82a2173a38ccb321f1cad7ed56bf3f8add46
108703 F20101114_AAAVDF li_y_Page_027.jp2
3f504a9c18bae071892ad3625a294a56
5dcc5c54943dabddc2cbb382faaecef6ead518c9
106744 F20101114_AAAVCR li_y_Page_011.jp2
cd3fab5bf4c8296711bc4580b620c05d
453b2b2d383a85210eb908dd8f01bc294ae66494
7024 F20101114_AAAVZZ li_y_Page_030thm.jpg
b800e71e9d8a1d90122f3f4959af8bae
e568ce2dcef60c6089090b0ca3a93e6909ffc2b1
2023 F20101114_AAAUXL li_y_Page_045.txt
6e0eb33d828dcc7d9425a7103650d911
5de5ca469970b1a2482808eae485dd706fa7cc28
25617 F20101114_AAAUWX li_y_Page_082.QC.jpg
b3ade409c88d11d3136fe5fca6548bdf
f8845d0a43d1ec9d99909d4447f3ed2c0ea4f063
115514 F20101114_AAAVDG li_y_Page_028.jp2
a5eff3b1b8468c176ce4d4c6565cd03c
772699eb7753731c75c97dc9b8143a972b806b00
74282 F20101114_AAAUYA li_y_Page_011.jpg
ef3cd8122850ae6ea3972c998c046e68
7aa6ad502a8987bd53365e2e728898aa84dff5c8
50853 F20101114_AAAVCS li_y_Page_012.jp2
f38bafbbafc1f5b2f5e1529f62fbd249
af579698ff2b523cb2faffbdf463db56f0361223
F20101114_AAAUXM li_y_Page_096.tif
21452df7c599598a0edeb0e806f89000
6085164ab6de83fa6d86e827e8cf210a6e94ae19
6401 F20101114_AAAUWY li_y_Page_093.pro
004d2896d22b6b83bc568cb1a87e8b0e
d4aba7267bbfbc9e0f64ce1692c8f8d5eaa9d7a9
116390 F20101114_AAAVDH li_y_Page_029.jp2
7e23b29cdf64a79d2c383d29392b4f79
b466ea0f28e72048d2de4e8680370dbc5fe49a5e
36966 F20101114_AAAUYB li_y_Page_012.jpg
ca6de18adb4d2a326056aff9c0f7765f
048c163a3a7352a955cebc475f48800e24ee534e
117031 F20101114_AAAVCT li_y_Page_014.jp2
f6e1509c778ddee72f611a6b29aecfb2
9bae20e57c717269d386c84fbdb650ddf5edf8f4
142366 F20101114_AAAUXN UFE0021202_00001.mets
279918a966add42092e660d83e9e7a60
542cde20381357b73ccc02cebaa2a79189507242
116674 F20101114_AAAUWZ li_y_Page_017.jp2
98ca808d7684a3e2d5b6ce54a14cd5ee
58d4948a21c3b340bfe696e22ba08c9677f1adeb
122559 F20101114_AAAVDI li_y_Page_030.jp2
08cebc3df830cdcbffb6996f8be0738a
2dc95a00422e94c3c580b5a0cbe1c633408a1bc6
75785 F20101114_AAAUYC li_y_Page_013.jpg
5b6908982dc73b4211e3d67065f1d4ad
2e2178c7fdd9963ad2b831039d5c8d10fea9965d
83085 F20101114_AAAVDJ li_y_Page_031.jp2
fffd4964b464869db3561a25a8e2dc16
1c24efb8d0168951b78cbe2caf2a1d9c45df1b2d
78426 F20101114_AAAUYD li_y_Page_014.jpg
6063abba8e12b4c9893bfc7e07545ba1
c9502ffe79006098fb7d57cfd30ef9e385f5baa8
117085 F20101114_AAAVCU li_y_Page_015.jp2
5f4a895d451a23c0960d0497e78529b8
ef3ca8235a654182b35a6f41bd6a4dfd4b995d6c
25913 F20101114_AAAVDK li_y_Page_032.jp2
90fbd1b385eb741b5e8421de277cc1cf
991d3d2be4cd1dfca4f566a9df273ac74eaa6455
76583 F20101114_AAAUYE li_y_Page_015.jpg
3b3c065f85f66dcf68485062038bcc8a
385c39fddbaab68aa6b16fde01a4c96605ddfe0d
117496 F20101114_AAAVCV li_y_Page_016.jp2
df657a42c0ffd19669c6baa69c929e7c
cafa97e28aa191620d7e6647a62fb0058fff7909
23591 F20101114_AAAUXQ li_y_Page_001.jpg
fdcb43897913dea8084cff9a1d06dd38
216baad2b4bcb8b9ba478aebc3f8f1123997bc1d
244931 F20101114_AAAVDL li_y_Page_033.jp2
4df5cec197a701a17bfd555e3c37219c
b69886b50058c5f87a0eddf1d15e0fb169d46e33
77758 F20101114_AAAUYF li_y_Page_016.jpg
f629baf63c7bbdc44f6c5312fb233b31
837433045761e943d4d6cea8deca03c1625fb6f0
119179 F20101114_AAAVCW li_y_Page_018.jp2
08c969f3f8c21cb1524672d96239f5fa
b6749b60b439d5d74869393dedf4bde21155834a
9558 F20101114_AAAUXR li_y_Page_002.jpg
03bd66b6ef4340aff24978c41294c055
c762d50e84dfb15cc7bc2ef95b48cfc716977796
107896 F20101114_AAAVEA li_y_Page_049.jp2
5f1b06e6a986eeea03a11741f1b77255
aacedad31e448af98d10c451f635e6478873db21
37054 F20101114_AAAVDM li_y_Page_034.jp2
b86bbf4fae79ee72ed3f47f3258b4157
ad6adf15b7480838b13b57afa17708fe4a1fb6fb
77965 F20101114_AAAUYG li_y_Page_017.jpg
d89b9a12b8d0b080fcd7b1f2eb5cfe69
7e252aeaf46dac93fa1c8ff5c479633fbeecfdab
98968 F20101114_AAAVCX li_y_Page_019.jp2
c7aa2cd330d90d380daeae885bda3951
5285e6394e5a295b131c14820e2e6892cb7f2648
11953 F20101114_AAAUXS li_y_Page_003.jpg
5fa813229419a3860be0b3c6cf074ad3
9dd671a4619ad0f943ff1cbcf69ceb8c6d65a3a0
110995 F20101114_AAAVEB li_y_Page_050.jp2
72f427d4299ef243e4df504c54d45e32
d7ffc432dccd83c7ba2e9c4e754df9de6c76b20e
25693 F20101114_AAAVDN li_y_Page_035.jp2
3638d0172164077179f2a189895c4e13
29a6628141caea6fb8aa26bae82b2c7cffe865bb
78693 F20101114_AAAUYH li_y_Page_018.jpg
47ee8d6d28e6e0c870cd99efcddbe21d
b52f574b02241546cf03f161ff2dcf01eaefbc62
106552 F20101114_AAAVCY li_y_Page_020.jp2
65d1f30ed458bbb4f7beaca601e2be9d
c1cd0f89dca35b1f6eadb3cf096ea81d2f444b56
57333 F20101114_AAAUXT li_y_Page_004.jpg
72460afd5a56a10f741fbf8df713247c
c04ed57d5b18d0d7c9e750c17d3c5ee4011109ee
120067 F20101114_AAAVEC li_y_Page_051.jp2
0ab1df109391f753dd989577740be434
c8da69873ca805dc5fc09c956f588fc584022fd4
31326 F20101114_AAAVDO li_y_Page_036.jp2
76b465f7c2267904f078bd83c10fbce7
f66087981083a207e1839e3b5f2bdc3d5dde1003
66747 F20101114_AAAUYI li_y_Page_019.jpg
cafac3302b9b4b5ac9776af6e8f8e9de
8d459f43e767c8296d80073aefa6b90f6f128c2a
1051936 F20101114_AAAVCZ li_y_Page_021.jp2
b57ec1bf224b3adfbabdd529f8e669e2
dbe39ca88479666b37175835d599b3e0c182e3d1
78638 F20101114_AAAUXU li_y_Page_005.jpg
6a6c2067d97dc45f6fd6957f81eb67fb
74ad7ac5b9d19177212313646ff46f7f352f2679
13935 F20101114_AAAVED li_y_Page_052.jp2
6ba5c862c7a7963979f0365205c48e3f
ccf4f6390d88e3dd2625378801589e779984f399
35744 F20101114_AAAVDP li_y_Page_037.jp2
b84289a49f8fd2059991326b56b28a69
e4fc219fc48a159f52b5d9b88ae1142087abfd85
73020 F20101114_AAAUYJ li_y_Page_020.jpg
1fdd184a276ee819b07077d2e562329e
05c2719f46017c4b29bb56473a97d31a5c341bc6
100763 F20101114_AAAUXV li_y_Page_006.jpg
5f23d4dea6510529b4aea49df8e519b6
bb48ce3d98c454d1e78dd8fba65a867901af6f84
255958 F20101114_AAAVEE li_y_Page_053.jp2
bb97110ca9638b44c0028057702d96b2
5dc1c27ee9dc043fc438446e2bafc2dab3f407c8
35432 F20101114_AAAVDQ li_y_Page_039.jp2
fcaa61e6f549795b50d543620954d0ee
af976d52f2349c0f351b96a33b3bde289f879b4f
88828 F20101114_AAAUYK li_y_Page_021.jpg
78a7bc6ecbfa2ff622f88e65353d98e3
8272a856907bb8119bb024ac0047b1f96c0486a7
26416 F20101114_AAAUXW li_y_Page_007.jpg
8985375373f988f8ff80ec6e7ae76f34
e8cd939550c48888ebd4aa79e57bc1cd28a39119
46479 F20101114_AAAVEF li_y_Page_054.jp2
58881cf97d57936791d4eca808bbf4ee
cc907c08156630a1e0a911e5083ac60490675abd
118618 F20101114_AAAVDR li_y_Page_040.jp2
82f499fe5d25d72fa841f1cb73d9f662
e5f33e08bcc3a8e68b0fa479f834676c04d01beb
27460 F20101114_AAAUXX li_y_Page_008.jpg
67f5e756c099f4c1fbfba97ed625d9bc
8f44d190ddf3cb5608e06aa6df5d609d56e5445d
349583 F20101114_AAAVEG li_y_Page_055.jp2
ef5a315048c95e56dfc01f7ceab6cf11
6ee267bf35ba402749c7f076b6a285c9ea2ec452
26712 F20101114_AAAUZA li_y_Page_038.jpg
d4cb0ee33e4a310bc4e1cfe64d382b7f
204f43ba0ec62ca519b6a31094e9175e0d2fb252
107031 F20101114_AAAVDS li_y_Page_041.jp2
8dc6737f92b365a290f09ed9d551ef71
41f1b02ac46846cfa66e0795b8815e6740138c0a
81605 F20101114_AAAUYL li_y_Page_022.jpg
5b63c70ef7a5c1067cb9e750642a11cf
6c064ba4871e2db53c25e62b62e6d19195655621
99231 F20101114_AAAUXY li_y_Page_009.jpg
3bba362ca0d2a56567b1cd7cb88432d4
9a57010ec2b9f1014d71b71769bbbd794f488bbf
438993 F20101114_AAAVEH li_y_Page_056.jp2
1092c465435ce4300bb8a2b8263c703a
71f2f45eed649fa11e2fd16a83deae46f19d9af3
30372 F20101114_AAAUZB li_y_Page_039.jpg
a50fe422152b132d5fae9290665af79b
cad40047e41d75cfd1b68720cd82b5835bc82ba8
113020 F20101114_AAAVDT li_y_Page_042.jp2
52336cbb21192e57d384199e2ad3d0dd
a7b62dfaae0b3a7f4e8eaf88cc5d28fb875fbb2b
74770 F20101114_AAAUYM li_y_Page_023.jpg
431c324c1d3a4f30bb1709324a99b5cf
3f2ed6d58a1bcbb4131c9b1e553cd8ba1f2ce077
46228 F20101114_AAAUXZ li_y_Page_010.jpg
b30afd640b422bc7c6b2ac45a187e669
be6828e1ed098c2f19d47e3ae4fa48f6072f5b7e
113794 F20101114_AAAVEI li_y_Page_057.jp2
4e34e0c66f426f728083ad4bb98b6329
296958b4975fe4aff4922d2be1f8b6ea4399ce70
80172 F20101114_AAAUZC li_y_Page_040.jpg
f4ea2d0d11ce225079d6c91f44721d9d
44b35c198043e055f8923bd7ecd71db4fd73866c
121336 F20101114_AAAVDU li_y_Page_043.jp2
c872e7d7fb9456351e672ef318f4a94d
654091f9654a75c7cccb978c968f5ff864b4e07d
58775 F20101114_AAAUYN li_y_Page_024.jpg
5e479727ca3e531493f1888872426e68
595f5707cbbef6934a174f7b61c1691b20abb9df
114157 F20101114_AAAVEJ li_y_Page_058.jp2
603df0e76c5915282785e8e183bff5af
3e8f7ef482dbd784d74496175a71c7843d89b4f5
70723 F20101114_AAAUZD li_y_Page_041.jpg
c6eeae40aeccacacde954bdc64822df7
624e2f21af05866b0bca05d06e5cf91a2606c3a0
61271 F20101114_AAAUYO li_y_Page_025.jpg
42b93864e2bdc385075589b284afa042
51a14ac7f35777367a76a693c26f29f32c7f5ef2
121379 F20101114_AAAVEK li_y_Page_059.jp2
af49fc1af59dba5ea41ddb8f6f1b0c20
0d565e7fc238873b6a282f6b453cbe94e1a53aea
75684 F20101114_AAAUZE li_y_Page_042.jpg
8a7d84a0e3b32ff08869a0e8abaaeed6
786ebf1e3c8a29817114246daa3f05bdea80a17e
115045 F20101114_AAAVDV li_y_Page_044.jp2
2e5cb831d61daedf0673c29d9d12a675
ba9499745b2962dc0a436c92cd6a396135055da6
62747 F20101114_AAAUYP li_y_Page_026.jpg
afa970eff97fdfc44756593ef8aee7b7
bdb6740b31c82fd87a77d92afa9c942f7f29e0d1
123291 F20101114_AAAVEL li_y_Page_060.jp2
ecbb36fc3aee90e7534dc291123b4ccc
60fedd5e502165702191ec0649eeb52b6bb1cbb1
79981 F20101114_AAAUZF li_y_Page_043.jpg
fb2443dc1b417aa1a7cd69fa6f2082b4
053541802ac898f03585dbc851bc98828fd86d3a
112070 F20101114_AAAVDW li_y_Page_045.jp2
aaf7438a0f5c7dfb9b601b12f31d66c8
32218d256e33e0bf6645a0763af754d037ebcb84
73196 F20101114_AAAUYQ li_y_Page_027.jpg
b50ac35ffc63fe37139f6297bec6df6c
3e589c55c5a2c5470d7eac4c78a0fdc1e0e966fb
107916 F20101114_AAAVEM li_y_Page_061.jp2
e7c19d7f979ec61afc47add29d5405b4
1f8dbcd5dad28f106d5e36c4aafdca10bd8d1f9f
76183 F20101114_AAAUZG li_y_Page_044.jpg
2bffa5ddd842c9ae0dc4823d10b47e47
ded9ab35ca0d92b3f50cc9566f8bb04282874053
117993 F20101114_AAAVDX li_y_Page_046.jp2
cd57310a5acc0238146b3bb0d7152661
3f9113cc4e0c2cd07d6331eb662d496e01b887f1
76833 F20101114_AAAUYR li_y_Page_028.jpg
965922fa0081a3ad0459f60669558612
1c7be18dc638c1cb2f1c8320e6656d28a10e6cc1
119306 F20101114_AAAVFA li_y_Page_075.jp2
ab585a917ab60a0bd1ac782e70e53ac0
002defbd1f16a0fa44d622c3dc5d25915db09833
103729 F20101114_AAAVEN li_y_Page_062.jp2
afe1831b871e83d4223826d8807c946c
4e1892c8f37a411618199210f7bb4407e4cf196a
75539 F20101114_AAAUZH li_y_Page_045.jpg
00e2cb222a292116921db01e5f791866
3974499a750811ed87a3990159d5f0282d3ce478
102267 F20101114_AAAVDY li_y_Page_047.jp2
c7fc6838271b07b0e12d48d36eeabf82
229cb6e53467d362933bf6a00b0c179bda86279b
77133 F20101114_AAAUYS li_y_Page_029.jpg
17cbc4487061ff4e11624d43c525e304
06eb8818289fcf02cbe2ba7a896bae8903dc794e
121043 F20101114_AAAVFB li_y_Page_076.jp2
ffaad999ae961f425aeb369840923263
e6aa69be2db4bc02c3e621dc83fdab0ec4258bd6
119666 F20101114_AAAVEO li_y_Page_063.jp2
71d617ea490a752b59beada8a692da5f
c0c5171cbcc28761f2eb91f73839852699167a56
79735 F20101114_AAAUZI li_y_Page_046.jpg
0af5b887c5d693ba81d19f21ccf2b7c3
3a8e9ef7ecc75f48a5d6dc4c0d24312526c05996
121089 F20101114_AAAVDZ li_y_Page_048.jp2
5f5baa2dcc029686515e6e13420cf601
035aac3d61c416ddb801e7685ca51bc5d5a2a466
79885 F20101114_AAAUYT li_y_Page_030.jpg
f024dcff05440516b01759d4e774a9ac
0ced2fdd5f4864ceeb492808ffb8d3dbc9dbb979
107649 F20101114_AAAVFC li_y_Page_077.jp2
2b0a55997dedb729fbfa6da721c048a1
071790e862e45185b4bc7eceff31f7bcd191cd78
118454 F20101114_AAAVEP li_y_Page_064.jp2
763c1b708e2e85179aaca3fcad25409a
cd1b8aab1369a8434f9b92267d8a2773d742ffca
67802 F20101114_AAAUZJ li_y_Page_047.jpg
1483431904a8ab4929c10446a1a796b7
6e5e054c55f21325aaeba876ca7762f555d4f062
56166 F20101114_AAAUYU li_y_Page_031.jpg
1dfa8a4effed86ed1cae2670793a8c23
61c6627b0d8054b9d64f7457cdfa805432deda76
121203 F20101114_AAAVFD li_y_Page_078.jp2
7085327af5f8d1740162d7f172101045
02d719f778302c7f4aa29dbe497c4653ffa66efc
122312 F20101114_AAAVEQ li_y_Page_065.jp2
954a039fa1aacd7df8cf81cf931840f9
56182086af0f1566caddc695f6e06b094385019d
79275 F20101114_AAAUZK li_y_Page_048.jpg
449a590cebf346206a5e996fe1db67bf
3bf2d8c63b8c342ee21c8a1c440f5468fd1d8aec
27315 F20101114_AAAUYV li_y_Page_032.jpg
94b096cf7f8568ae6df50ed521950247
6ad035fec9886c2ba87ebec11ee700b1b9cf492b
119685 F20101114_AAAVFE li_y_Page_079.jp2
889546b60fc85a4b25f1673a2b00ed0e
cf561f0d578126a7026aad7278cf6ca7fe95e646
105376 F20101114_AAAVER li_y_Page_066.jp2
9d72e0b12c0b38875f77e7772a132f04
38c0b677d9e8178af3a40a6fb9479c5dc0a0a6f2
72138 F20101114_AAAUZL li_y_Page_049.jpg
dadb89b23439ba371b536e3a2046ec21
49c274dbf32b4e21e16dbab29b66d1b43212e05a
25921 F20101114_AAAUYW li_y_Page_033.jpg
b4774015655225cf84f065618e3b8ddd
b7f328d351d4468d07116969a4b77aae7ca8295d
117201 F20101114_AAAVFF li_y_Page_080.jp2
87b9716a4ea851b948aa5f3450f1be2d
cefd2c93fee371666af1ff5b04804eb62a511948
104231 F20101114_AAAVES li_y_Page_067.jp2
4afcf802d483834f93d33d1a69aefb85
ac08858c9a12f0b811af78f3617bd6890c4d1acf
24917 F20101114_AAAUYX li_y_Page_035.jpg
07b5b78ce14050ddcab4f1cdad8ac2ef
bc1ca61a22fcf1528f87212fbb23774ba09ee2de
122781 F20101114_AAAVFG li_y_Page_081.jp2
0daca2e482ab8d13ca1e3fab21ca32ca
4792882f76d322521bfd26e0a0996d26a6e35f01
31600 F20101114_AAAVET li_y_Page_068.jp2
41c150c5ec8bcf4d143b73e1ad169331
4f804b14fb3a1dcf74aa2069d67c0a1e76407ac3
73300 F20101114_AAAUZM li_y_Page_050.jpg
230c24f9c65546eff5752896342095af
d11f63381e50c83d2d9de1226a211beff2eaf821
30946 F20101114_AAAUYY li_y_Page_036.jpg
fd36905c430ed7ea6908447bdbc6ce31
eefb7028c5e7f9ac80e58e7f9dac88ce35205024
119181 F20101114_AAAVFH li_y_Page_082.jp2
73e970b211a6848f624fe97a3a474528
848ce4c62fa64fe764ba4ac7d8a0c09b35719a1e
277830 F20101114_AAAVEU li_y_Page_069.jp2
d3096da623e244613ea1404bffc1948e
34b1451ee71fb6c55e264330161b9b7b6dd1c922
80029 F20101114_AAAUZN li_y_Page_051.jpg
04eff8afbbc5878b89f45520bebbb994
b6771addbe90e4a0f2646652cc9311f491e8bfd8
34490 F20101114_AAAUYZ li_y_Page_037.jpg
bb7ada1d15bc71d430260a69258996f2
d6bb8e293d7b7220869c455fd21457c502ce5461
103202 F20101114_AAAVFI li_y_Page_083.jp2
59f346c414ae0540863cd9e6a320a69f
65d05f6866b739625c4147a1ae2fbf53f66f80cf
30428 F20101114_AAAVEV li_y_Page_070.jp2
e5626c87e2222a3f60a7d67aeda97986
f4343291b2ed209378490031ff81de865d8c6bcc
15264 F20101114_AAAUZO li_y_Page_052.jpg
6d0fce2e3b2c07373cc906e868690d6b
02946c5ad7c82fe3a1ec65116fdc88cac334fc3d
110589 F20101114_AAAVFJ li_y_Page_085.jp2
0f89c47a0a5edd2a089a8b1544f7c398
8daa8dbec563625fe2de8df3b5b856ed10afc5c2
27925 F20101114_AAAUZP li_y_Page_053.jpg
4cd8adf7a436d96cb5457104a6e134a3
7040a5205499a199949c281e73e9b28c15587b8c
111248 F20101114_AAAVFK li_y_Page_086.jp2
3e4b4bb47f6ec71687faf246f1410400
bd63aa2da4c0d65f0d44c7a5eaaa354d87d2959c
612089 F20101114_AAAVEW li_y_Page_071.jp2
6049f4b7a3846dab52551fd1f931473b
1a633dbb14c1c5b3b077fe2a6ffd8c5d764e4b05
44412 F20101114_AAAUZQ li_y_Page_054.jpg
997c8f7baf4c6f0b16b24d96023440ef
a88183a9386e7855fb12e4ce8232e09b335a1a66
115041 F20101114_AAAVFL li_y_Page_087.jp2
2416842d61c85ee2e47488d28aef7395
7cf1083e30e20f82b636f917d0dfa409972add73
120015 F20101114_AAAVEX li_y_Page_072.jp2
98aace295ad85b60ec9ae1ba3a093bf6
8a85f493cfb2c1f0711e89091cd712cb4338d9ec
34316 F20101114_AAAUZR li_y_Page_055.jpg
b15dbb44037c989ec0b6ea7ea237649a
b346cdea63c3a87eb9ea6141496267aa933cd085
121288 F20101114_AAAVGA li_y_Page_102.jp2
cd737e09c38a78fc9f8a9584e21fc496
c3b64373e38d390ac757024b8650715bfc2784b0
125175 F20101114_AAAVFM li_y_Page_088.jp2
826b419873a32309c7c8bb4e3a49ee28
d763dcd3334244627daeae358e6a33dbd37eb3b9
111137 F20101114_AAAVEY li_y_Page_073.jp2
70fe36f7c4cdd79c957ae4ad79fee3df
ed78c31d388800aa695e1f80daa9926607c88cf9
38128 F20101114_AAAUZS li_y_Page_056.jpg
6f6ec7652656147d919e40b4afa55d17
a42e0a199f7d8da8247ea3dc771b7c4dec1d0d66
119038 F20101114_AAAVGB li_y_Page_103.jp2
e84399eac4597123cedd71252a91a8a0
247a1dfbeaa31020817545007557741af233833e
24317 F20101114_AAAVFN li_y_Page_089.jp2
21d1a680b5462f7103e9c796e8bb19c6
bb0874b4707d63d2cd3e236d392c7811977b2652
114223 F20101114_AAAVEZ li_y_Page_074.jp2
0bd3a869c99f58caf83b56104b384f8e
1463af3531496a9256f5abcfc8edcd68cbd2d9d7
77186 F20101114_AAAUZT li_y_Page_057.jpg
64f0ce40b67ceb3e40541896c968c68e
4ec2ca93a54dce3113249fd2aea3e54946bb9025
113121 F20101114_AAAVGC li_y_Page_104.jp2
13ed6fc6f560df970f2b394d06a5bfb7
43df8e4822502d4e15726c1b26792a43397675e2
79869 F20101114_AAAVFO li_y_Page_090.jp2
8b308d98c92cc87ce1ff32549599a18f
31c61c7332a42d97ab00e2d19d76d5294f6847cf
76256 F20101114_AAAUZU li_y_Page_058.jpg
0d2551726e7219f1080c802e6f020fa5
9a4e65818ecb1a577ed461e272c05e8d0060af22
1051953 F20101114_AAAVGD li_y_Page_105.jp2
2da7418205908638656452344351e3cd
8be3aa22b70fdd7e4be230b50925ac155d14d830
291017 F20101114_AAAVFP li_y_Page_091.jp2
85788a81e5e36bd8ef754a6517366a69
46237de666267a17209a04ff70d3f7cb9c5910e0
80549 F20101114_AAAUZV li_y_Page_059.jpg
7a108161f9b80f264378b4443fc11aa3
c68d0e49c2096c231a4ac52fab110ad2aed0dcd0
113092 F20101114_AAAVGE li_y_Page_106.jp2
8aa7bacd34d52bc5fd6b5b41c3a69cd5
1e42c018886b15cbbb5465265f96698d3004cf14
870123 F20101114_AAAVFQ li_y_Page_092.jp2
1c1525ba26f8da2e15d48f0cafe101e9
de3b7027afe136499688d13819646e720156cb44
81813 F20101114_AAAUZW li_y_Page_060.jpg
f6b4e76f05e49bd465afc28f8cfcc606
37f73a24df6e3aea984faefa72501edbd697faeb
116972 F20101114_AAAVGF li_y_Page_107.jp2
0eeff6656900ebe5f682d22dd1643313
3fe0d8b417d1603e9edeecf13b2a350e492706b6
387119 F20101114_AAAVFR li_y_Page_093.jp2
daa0ae19c94eb023d3ab79d7a8bfb38a
cfda24ce6c69ebabb9279dcfaffeb8d495d2969d
69335 F20101114_AAAUZX li_y_Page_062.jpg
26a4e689f3609e93ab1b09e5bcc5e1cc
737fb1bd2985b314e1ad7d045cc6b2a7531fdddd
30002 F20101114_AAAVGG li_y_Page_109.jp2
30e3debd137924b0cef4516ff3a68250
7163f46f40fecfa9c9e2fd4c65cf2e1f41753df7
42252 F20101114_AAAVFS li_y_Page_094.jp2
98bb4282e87a528fcb7eb1540bf6a0fc
1a833184aa3e58aace704f208da0ae23abb0be11
78533 F20101114_AAAUZY li_y_Page_063.jpg
9c9353f75e7a7761ad3b3ae6f5c1b8d7
3735eba21476f4e475adfdd6cf5913998a5c8c3e
27187 F20101114_AAAVGH li_y_Page_110.jp2
43a5bf59d20c2e1960c741ee54a8ac6f
f78425b83aaf1a54d189416f745fad298a10ed56
26777 F20101114_AAAVFT li_y_Page_095.jp2
d6347b33ec59b4215b15a9b0eaa8fc28
f82dbf82b1b53b76471b414f41d7624b93752605
78064 F20101114_AAAUZZ li_y_Page_064.jpg
ccc724eacdd864dcebe847162058355d
c6079a554ddd79010cb67834048a2903cab6f4e8
24575 F20101114_AAAVGI li_y_Page_111.jp2
4ae20949c9422c44d6e88211bf878e44
cd83e6197b30f93ad6d2fe69144a2291f43c5e7e
28089 F20101114_AAAVFU li_y_Page_096.jp2
5def99bbc6e2831c316799b05520707a
d574ad16629d78c26e78c542edf6dd1b35e0c80b
115495 F20101114_AAAVGJ li_y_Page_113.jp2
3fae90f30e4a7db063d47ef0fadef67f
ac91f1c2f8f9e6927d0597c971342087fc29dca7
422855 F20101114_AAAVFV li_y_Page_097.jp2
c8302f1e611602eac4a3a9e8b37a4550
0b73ecf547d5d82ad61cf4987ed02bca97d54adc
104505 F20101114_AAAVGK li_y_Page_114.jp2
acdea0e3d55eb314d0b8c157d7946653
88672d332d30c65e6061cdaeaab5f84e9d277ab3
256599 F20101114_AAAVFW li_y_Page_098.jp2
ebbfb8ebb5a0fbd26a2d6059580c47ec
a021f3007d64b2ae36a709d9ce7480451ba9b7ea







REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING
NANOSTRUCTURED SILICA/TITANIA/VANADIA COMPOSITES




















By

YING LI
















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

2007































O 2007 Ying Li



































To my wife and my parents for their constant love, understanding, and support.









ACKNOWLEDGMENTS

I sincerely thank Dr. Chang-Yu Wu (my supervisory committee chair) for his invaluable

guidance, inspiration, and encouragement. I am truly grateful for his continuous and enthusiastic

support of my graduate research and my career development. My supervisory committee

members (Dr. Jean Andino, Dr. Kevin Powers, Dr. Jean-Claude Bonzongo, and Dr. Wolfgang

Sigmund) have generously given their time and expertise to better my work. I thank them very

much for their guidance and their good-natured support.

I would like to thank Patrick Murphy, who has been assisting me for one and a half years

in my research proj ect. He has given me a big hand in building the reaction system and

conducting the experiments. I thank Sameer Matta for his assistance in doing the experiments

and making the pellets in the past two semesters. My thanks also go to Jie Gao, who helped me

analyze the amount of mercury captured on the catalysts. I greatly appreciate the help from Yu-

Mei Hsu for miscellaneous lab items. I must acknowledge as well many other fellow students in

our air resources group who kindly assisted my research.

I am grateful to the National Science Foundation for the financial support and VICI

Metronics, Inc. for providing the mercury permeation tube. Finally, I am thankful to the

scholarship and encouragement from the Air & Waste Management Association.











TABLE OF CONTENTS


page

ACKNOWLEDGMENT S .............. ...............4.....

LI ST OF T ABLE S .........._.... ...............8..._.........

LIST OF FIGURES .............. ...............9.....

AB S TRAC T ........._. ............ ..............._ 1 1...

CHAPTER

1 INTRODUCTION ................. ...............13......... .....


Mercury and Its Health Effects ........_................. ........._._ .......1
Mercury Emissions and Regulations .............. ...............13....
Mercury Speciation and Control Technologies .............. ...............14....
Mercury Removal by SiO2-TiO2: Unknowns and Challenges .................... ...............1
Mercury Measurement Interference .............. ...............17....
Research Objectives............... ...............1

2 KINETIC STUDY FOR PHOTOCATALYTIC OXIDATION OF ELEMENTAL
MERCURY ON A SILICA-TITANIA NANOCOMPOSITE .............. ....................2


Back ground ........._._ ...... .___ ...............20....
M materials and M ethods .............. ..... .. ..............2
Synthesis of SiO2-TiO2 Nanocomposite ................. ...............21................
Apparatus and Procedure ................. ...............22........... ....
M odel Description .............. ...............24....
Results and Discussion .............. ...............26....
Effect of Hgo Concentration ................ ...............26................
Effect of W ater Vapor ................. ...............28.......... ....
Sum m ary ................. ...............3.. 1..............

3 ROLE OF MOISTURE IN ADSORPTION, PHOTOCATALYTIC OXIDATION,
AND REEMISSION OF ELEMENTAL MERCURY ON A SILICA-TITANIA
NANACOMPOSITE ............ ..... .__ ...............40...


Back ground ............ ..... ._ ...............40....
Experimental ..........................__ ......_ .............4
Synthesis of SiO2-TiO2 Nanocomposite ................. ...............41................
Experimental Setup .............. ...............42....












Results and Discussion ................... ...............43..
Role of Moisture in Hgo Capture ................. ...............43...............
Role of Moisture in Hgo Reemission ................. ...............45...............
Mechanisms of Hgo Capture and Reemission ................. ...............46...............
Sum m ary ................. ...............5.. 1..............

4 REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING A SILICA-
TITANIA NANOCOMPOSITE ........._._ ...... .... ...............57...


Back ground ........._._ ...... .... ...............57...
Experimental M ethods ..........._ ....... .... .. ...............58..
Synthesis of the SiO2-TiO2 Nanocomposite ................. ...............5.. 8......... ...
Experimental Setup and Procedure .............. ...............59....
Results and Discussion .............. ...............61....
Baseline Test ............... .... ... ................6
Effects of Individual Flue Gas Components .............. ...............63....
Hg Removal in Simulated Flue Gases ................. ...............67...............
Summary ................. ...............67.................

5 DEVELOPMENT OF SILICA/VANADIA/TITANIA COMPOSITES FOR
REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS .............. ....................72


Back ground .................. ...............72.......... .....
M materials and M ethods .............. ...............74....
Catalyst Preparation.................. .............7
Catalyst Characterization Techniques .............. ...............74....
Catalyst Activity Measurement .............. ...............75....
Results and Discussion ................. .. ...............76..
Characterization of the Catalysts ................. ...............76................
Mercury Removal Using Pellet Catalysts ................. ...............77........... ...
Mercury Removal Using Powder Catalysts .............. ...............79....
Mercury Removal Mechanisms ................. ...............82................
Role of 02 ................. ...............82........... ....
Role of HC I .............. ...............83....
Role of NO2 ................. ...............85........... ....
Role of N O .............. ...............86....
Role of SO2 ................. ...............87........... ....
Role of H20 ................. ...............87........... ....
Summary ................. ...............88.................

6 UV-AB SORPTION-BASED MEASUREMENTS OF OZONE AND MERCURY: AN
INVESTIGATION ON THEIR MUTUAL INTERFERENCE .............. ....................9


Back ground ................. ...............99..............
M ethods ................. ........ ... ............10
Descriptions of Hg and Ozone Instruments ....._._.__ ..... ..__... ......_._.........10
Experimental Setup and Procedures ....._._.__ ..... ..__... .....___ ...........0












Results and Discussion ............... ........ .. ...........10
Interference of Ozone on Hg Measurement .............. ...............103....
Interference of Hg on Ozone Measurement .........__ ....... __ ......__............106
Summary ............ ..... .._ ...............108...


7 CONCLUSIONS AND RECOMMENDATIONS ............_...... .__ .........._.......112


LIST OF REFERENCES ............_...... .__ ...............117...


BIOGRAPHICAL SKETCH ............_...... .__ ...............125...










LIST OF TABLES


Table page

4-1 Experimental conditions for investigation of the flue gas effects ................. ................ .68

5-1 Experimental parameters for activity measurement of the catalysts in pellet form...........90

5-2 Experimental parameters for activity measurement of the catalysts in powder form........90

5-3 BET surface areas of the catalysts ................ ............... ......... ........ ...._..90

5-4 Amounts of Hg captured and oxidized on the catalysts in a 6-hr test............. ..._............90










LIST OF FIGURES


Fiare page

2-1 Experimental system for kinetic studies. ............. ...............32.....

2-2 Description of Hgo photocatalytic oxidation on SiO2-TiO2 HRHOCOmposites. .................. .33

2-3 Photocatalytic oxidation of Hgo at different inlet Hgo concentrations without water
vapor. ............. ...............34.....

2-4 Inverse of Hgo photocatalytic oxidation rate versus the inverse of inlet Hgo
concentration (without water vapor). .............. ...............35....

2-5 Rate of Hgo photocatalytic oxidation versus inlet Hgo concentration without water
vapor (solid circles: experimental data; solid line: L-H model). ............. ....................36

2-6 Photocatalytic oxidation of Hgo at a constant inlet Hgo concentration of 0.66 CI-mol
m-3 with variation in water vapor concentration. .............. ...............37....

2-7 Inverse of Hgo photocatalytic oxidation rate versus water vapor concentration at a
constant inlet Hgo concentration of 0.66 CI-mol m-3............. ...............38...

2-8 Rate of Hgo photocatalytic oxidation versus inlet Hgo concentration at different water
vapor concentrations (markers: experimental data; lines: L-H model). ............................39

3-1 Experimental system for studies on the role of water vapor. ..........__.... ..._.._..........53

3-2 Dimensionless Hgo concentration at the reactor outlet (A, [H120] = 0 ppmy; B, [H120]
= 13000 ppmy; C, [H120] = 23000 ppmy) .............. ...............54....

3-3 Hgo reemission from SiO2-TiO2 HRHOCOmposite after 3-h pretreatment (The inset
shows the dimensionless Hg concentration during the pretreatment). ............. ................55

3-4 Mechanisms of Hg capture and reemission on the surface of SiO2-TiO2
nanocomposite. ............. ...............56.....

4-1 Photocatalytic reaction system under flue gas conditions. ............. .....................6

4-2 Hg speciation at the outlet of the reactor in the baseline test. ................ .....................70

4-3 Effects of flue gas components on Hg capture and oxidation under various conditions
of a) H120, b) HC1, c) SO2, d) NO, e) NO2, and f) simulated flue gases ............................71

5-1 Experimental system for the fixed-bed study using powder catalysts ............... .... ...........91

5-2 XRD patterns of(a) SV2, (b) SV5, (c) SV8, (d) SV10, (e) ST12, and (f) ST12V5..........92

5-3 Catalytic removal of Hg using the pellet catalysts under various conditions ................... .93











5-4 Outlet Hg concentration as a function of time using 500 mg powder of (a) SV2, (b)
SV5, (c) SV8, (d) SV10, and (e) ST12V5 ................ .......... ....... ...........9

5-5 Outlet Hg concentration as a function of time using 250 mg powder of SV5 (20-hr
test)............... ...............96.

5-6 The role of flue gas components on Hg removal using 250 mg SV5 powder under
dry conditions............... ...............9

5-7 The role of water vapor on Hg removal using 250 mg SV5 powder .............. .... ........._..98

6-1 Experimental setup. A) Ozone interference on Hg measurement. B) Hg interference
on ozone measurement. ........._.._.. ...._... ...............109...

6-2 Measurement interference of ozone on the RA-915+ Hg analyzer as a function of
ozone concentration (The error bars represent one standard deviation). ........._.._............110

6-3 Measurement interference of Hg on the ozone analyzer as a function of Hg
concentration. ........._.._._ ........... ..111._._._ ......









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

REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING
NANO STRUCTURED SILICA/TITANIA/VANADIA COMPO SITES

By

Ying Li

August 2007

Chair: Chang-Yu Wu
Major: Environmental Engineering Sciences

As a highly toxic pollutant, mercury (Hg) tends to bioaccumulate in the food chain and

exerts adverse effects on human health. The U.S. EPA issued the Clean Air Mercury Rule in

2005 to permanently cap and reduce Hg emissions from coal-fired power plants. A low-cost

methodology using a SiO2-TiO2 HRHOCOmposite as a photocatalyst has been recently developed

to effectively remove elemental Hg (Hgo) under room conditions. In this research, a bench-scale

fixed-bed reactor system has been established and Hgo removal on the SiO2-TiO2 HRHOCOmposite

was examined under both room and flue gas conditions. A kinetic study showed that Hg

oxidation on the SiO2-TiO2 HRHOCOmposite under UV irradiation followed the Langmuir-

Hinshelwood rate expression. The flue gas components were found to have significant effects on

Hgo removal using the SiO2-TiO2 HRHOCOmposite. HCI and SO2 promoted Hgo oxidation, while

water vapor and NO significantly inhibited Hg removal. The mechanisms of these promotional

and inhibitory effects were thoroughly explored in this research.

The active phase of the selective catalytic reduction (SCR) catalyst, V20s, was added to the

SiO2 or SiO2-TiO2 COmposites in an effort to improve the catalytic activity for Hg removal. No

UV light activation is needed for the V20s doped catalysts, which is a great advantage over the

SiO2-TiO2 COmposites. The Hgo removal efficiency increased as the V20s loading increased from









2 to 8% but decreased as it further increased to 10%. The optimal V20s loading was found to be

somewhere between 5 and 8%. The SiO2-TiO2-V205 exhibited a greater ability of oxidizing Hg

compared to the SiO2-V205. It was suggested that the Hg oxidation on the V20s doped catalysts

follows an Eley-Rideal mechanism where HC1, NO, and NO2 are first adsorbed on the catalyst

surface and then react with gas-phase Hgo

This research also reported that atomic absorption spectrometry based continuous mercury

monitors are subj ect to interference by ozone due to its strong absorption bands near the Hg

absorption line. On the other hand, Hg interferes with ozone measurement which is based UV

adsorption. These mutual interference can consequently affect the risk assessment of human

exposure to both Hg and ozone.









CHAPTER 1
INTRODUCTION

Mercury and Its Health Effects

The 1990 Clean Air Act Amendments (CAAA) listed 189 hazardous air pollutants (HAPs).

Among them, mercury (Hg) has attracted significant attention due to its increased levels in the

environment and well-documented food chain transport and bioaccumulation (Brown et al.,

1999). Human exposure by direct inhalation of Hg in the air is not a predominant public health

concern because the Hg concentration in the air is typically very low. However, Hg in ambient

air can eventually be re-deposited on land surfaces or directly into rivers, lakes, and oceans, and

then biologically enter the food chain. In aquatic systems, Hg is often converted by bacteria to

methylmercury (CH3Hg ), which is a neurotoxin and can be magnified through the aquatic food

chain hundreds of thousands of times (Ravichandran, 2004). Hg and its compounds act as

dangerous and insidious poisons and can be adsorbed through the gastrointestinal tract and also

through the skin and lungs (Bidstrup, 1964). High-concentration of Hg can cause impairment of

pulmonary and kidney function, chest pain and dyspnousea (Berglund and Bertin, 1969). An

extreme example of the health effects of Hg is the high-dosage exposure from the consumption

of methylmercury-contaminated fish by the residents living near Minamata Bay in Japan in the

1950s that resulted in fatalities and severe neurological damage (Mishima, 1992).

Mercury Emissions and Regulations

According to the M~ercury Study Report to Congress prepared by the U.S. Environmental

Protection Agency (USEPA) (USEPA, 1997a), the maj or anthropogenic Hg emission sources are

coal-fired boilers (33%), municipal waste combustors (19%), industrial and commercial boilers

(18%), and medical waste incinerators (10%). Hg emissions from manufacturing sources are

generally lower compared to combustion sources with the exception of chlor-alkali plants using









the mercury cell process and portland cement manufacturing plants (USEPA, 1997a). Ever since

the 1990 Clean Air Act, the U. S. EPA has issued a series of rules to regulate Hg emissions from

solid waste combustors/incinerators (USEPA, 1997c, 2001b), and mercury cell chlor-alkali

plants (USEPA, 2003). Since coal-fired utility boilers are currently the largest single-known

source of anthropogenic Hg emissions (one-third of the 150 tons of Hg emitted annually) in the

United States, the U.S. EPA issued the Clean Air Mercury Rule (CAMR) in 2005 to permanently

cap and reduce Hg emissions from coal-fired power plants (USEPA, 2005a). CAMR will be

implemented in two phases, with the first phase cap of 38 tons in 2010 followed by a Einal cap of

15 tons in 2018. The Einal cap requires an approximately 70% reduction from the 1999 emission

levels.

Mercury Speciation and Control Technologies

There are three basic forms of Hg in the coal-derived flue gas: (1) elemental Hg (Hgo), (2)

oxidized Hg (Hg2+), predominantly HgCl2 due to the large excess of chlorine species in the flue

gas, and (3) particle-bound Hg (Hg,). During combustion, Hg is released from coal as Hgo, and

as the flue gas cools, some of the Hgo can be oxidized or bound on the fly ash. The Hg speciation

in the flue gas is determined by various factors including coal properties, boiler operating

conditions, flue gas composition, and the time-temperature profile (Romero et al., 2006). Hg2+

and Hg, are relatively easy to remove from the flue gas using typical air pollution control devices

(APCDs). Hg,, bound on fly ash particles, is collected in electrostatic precipitators (ESPs) and/or

baghouses. Hg2+ is Soluble in water and is readily captured by wet flue gas desulfurization (FGD)

equipment. Hgo is volatile and insoluble in water, and thus, it is difficult to be captured using

these conventional control technologies. Unfortunately, Hg speciation studies showed that Hgo is

the dominant species in flue gas when burning low rank (subbituminous or lignite) coals.









Therefore, the need exists for a low cost Hg oxidation/capturing process that can be applied for

the flue gas treatment.

Many methodologies have been proposed for Hg emission control from flue gas. Among

them, the technology of sorbent inj section, particularly activated carbon inj section (ACI), has been

investigated most intensively (Pavlish et al., 2003). Both Hgo and Hg2+ can be captured by the

sorbent and collected in ESPs and/or baghouses. This technology has been successfully

implemented in the municipal waste incinerator industry, where 90% Hg removal can be

achieved. However, the application of ACI in coal-fired utility boilers is far more challenging

due to the shorter gas residence time, the lower equilibrium adsorption capacity and mass-

transfer rate, and the compromise of fly ash properties by the inj ected sorbent. The high cost of

ACI also limits its application (Pavlish et al., 2003).

Recently, selective catalyst reduction (SCR) catalysts are found to be capable of oxidizing

Hgo in addition to its ability of removing nitrogen oxides (NOx) from the flue gas (Benson et al.,

2005; Lee, Srivastava et al., 2004; Lee et al., 2006; Niksa and Fujiwara, 2005b; Senior, 2006).

The extent of Hgo oxidation through SCR processes varies under different operating conditions

burning different types of coal. However, the SCR process is treated as a "black box" and the

mechanisms of Hg oxidation over SCR catalysts are yet to be understood.

A novel methodology using titanium dioxide (TiO2) based nanostructured sorbents has

been demonstrated to be very effective for capture of Hgo under ultraviolet (UV) irradiation (Lee

et al., 2001; Pitoniak et al., 2003; Wu et al., 1998). Wu et al. (1998) and Lee et al. (2001)

reported a high level of Hgo capture in simulated combustor exhaust using in-situ generated TiO2

particles, while Pitoniak et al. (2003) used a highly porous silica (SiO2) gel doped with TiO2

nanoparticles and achieved synergistic adsorption and photocatalytic oxidation of Hgo in a fixed-









bed reactor. The high surface area and open structure of the SiO2-TiO2 HRHOCOmposite allow

effective irradiation by UV light and thus minimize the mass-transfer resistance for Hgo

(Pitoniak et al., 2003; Wu et al., 1998). Using this material, Hg removal efficiency remained over

90% even after a 477-hr treatment (Pitoniak et al., 2005). Potential applications of the SiO2-TiO2

nanocomposite for Hg removal lie in two main areas. First, like ACI, a powdered form of the

nanocomposite can be inj ected into combustion exhaust upstream of a particle control device

(e.g. ESP). Second, a pellet form of the nanocomposite can be used in packed-bed columns to

treat Hg emissions from flue gas. In this case, the device is preferably installed between an ESP

and a wet scrubber.

Mercury Removal by SiO2-TiO2: Unknowns and Challenges

While the SiO2-TiO2 HRHOCOmposites have demonstrated prominent effectiveness for Hg

removal, the past studies were mainly conducted under room conditions and there is little

knowledge so far about the performance of this novel material under real or simulated flue gas

conditions. As is known, typical coal-derived flue gas consists of a high concentration of water

vapor (normally 5~15 % v/v) and various minor gas components such as HC1, SO2, and NOx.

Among the factors that affect the efficiency of Hgo capture by the TiO2 photocatalyst, moisture

content in the Hgo-laden gas was reported to be one of the most important (Pitoniak et al., 2003;

Rodriguez et al., 2004). However, the understanding of the water vapor effects on Hg removal in

literature is limited and some of the findings are controversial. It has also been reported that the

minor acid gases are important to the heterogeneous adsorption and/or oxidation of Hg on

activated carbons or fly ash under flue gas conditions (Carey et al., 1998; Norton et al., 2003). In

addition, the typical flue gas temperature at the cold-end of the boiler convective pass is in the

range of 120 ~150 oC, much higher than the room temperature. Thus, it is expected that the









nature of Hg capture on the SiO2-TiO2 HRHOCOmposite would be different under flue gas

conditions from that reported under room conditions in the past studies.

Furthermore, the kinetics for catalytic oxidation of Hg is highly uncertain (Presto and

Granite, 2006). The lack of understanding presents a severe limitation in predicting the extent of

Hg oxidation in larger scale applications. In literature, kinetic modeling studies on Hg capture

have mainly focused on activated carbon adsorption (Chen et al., 1996; Flora et al., 1998;

Meserole et al., 2000; Rostam-Abadi et al., 1997), while modeling on photocatalytic oxidation

using TiO2 has mainly focused on degradation of volatile organic compounds (VOCs) (Kim and

Hong, 2002; Raillard et al., 2004; Shang et al., 2002; Son et al., 2004). The kinetics for Hg

oxidation using the SiO2-TiO2 HRHOCOmposite is yet to be investigated.

While the performance of the SiO2-TiO2 HRHOCOmposite under flue gas conditions is

unknown at this point, it is reasonable to carry out a parallel study that focuses on the

development of a modified or even new catalyst which would be more effective on Hg removal

in flue gas. Considering the fact that industrial SCR catalysts, with an active phase of V205

supported on TiO2, are capable of oxidizing Hgo in addition to its ability of removing NOx

(Benson et al., 2005; Lee, Srivastava et al., 2004; Lee et al., 2006; Niksa and Fujiwara, 2005b;

Senior, 2006), the addition of V205 to the existing SiO2-TiO2 HRHOCOmposite would be expected

to enhance the catalytic activity. Meanwhile, since the Hg oxidation across SCR catalysts is

usually treated as a "black box" in pilot- or full-scale studies, a better understanding of the

fundamental nature of the catalytic reactions is of great importance to the advancement of the

catalysts.

Mercury Measurement Interference

Both Hg emission regulations and development of Hg control technologies require that

reliable methods be used for accurate Hg measurement. Currently, the EPA accepted methods for









Hg measurement in the United States are manual procedures based on wet-chemistry such as

EPA Methods 29 and 101A (for total mercury) and the Ontario Hydro Method (for speciated

mercury) (Laudal et al., 2004). However, continuous mercury monitors (CMMs) have distinct

advantages over these manual methods in that CMMs are able to provide a real-time or near-real-

time response for Hg measurements and to perform long-term emission measurement. On the

other hand, a significant disadvantage of CMMs lies in their measurement interference.

Atomic absorption spectrometry (AAS) is one of the maj or techniques applied to current

CMMs. In the case of AAS, the concentration of Hgo in a gas sample is determined by measuring

the light that is absorbed by Hg atoms at their characteristic wavelengths (usually at the

resonance line of 254 nm). Thus, interference can occur when other components of the sample

gas possess strong absorption bands near this wavelength (254 nm). Since the 254 nm Hg

emission line also falls into the absorption spectra of ozone, which is capable of absorbing UV

light below 290 nm, the presence of ozone in the sampling environment may impact the Hg

measurement by AAS based CMMs. Granite and Pennline (2002) studied photochemical

oxidation of Hg and speculated that photosensitized formation of ozone may interfere with Hg

measurement by absorbing UV radiation. However, no quantitative data were reported in

literature on the magnitude of ozone interference.

Research Objectives

To reveal the unknowns and to embrace the challenges mentioned above, five obj ectives

are proposed in this doctoral research. The first obj ective is to study the kinetics of the Hgo

photocatalytic oxidation on the SiO2-TiO2 HRHOCOmposite. The competitive adsorption of water

vapor in Hgo photocatalytic oxidation will be established in a kinetic expression as well. This

modeling study is of importance in predicting Hgo removal efficiency and is useful for designing

an effective reactor, under photocatalytically oxidizing conditions.









The second obj ective is to perform a mechanistic study probing the role of moisture on

Hgo capture (adsorption and/or photocatalytic oxidation) using a SiO2-TiO2 HRHOCOmposite. To

provide an overall evaluation of the performance of the SiO2-TiO2 HRHOCOmposite, possible

reemission of captured Hg species will also be examined. The corresponding mechanisms of Hgo

removal and reemission in the presence of water vapor will be investigated as well.

The third obj ective is to install a fixed-bed photocatalytic reactor and to investigate the

performance of the SiO2-TiO2 HRHOCOmposite under simulated flue gas conditions. The effects

of the flue gas components on the removal of Hgo by the SiO2-TiO2 HRHOCOmposite as well as

the surface reaction mechanisms will be explored. An improved understanding of the role of the

flue gas components can help evaluate the potential of applying this novel material for effective

Hg control in coal-fired power plants.

The fourth obj ective is to develop a method to dope V205 on the SiO2 or SiO2-TiO2

composites in an effort to improve the catalytic activity for Hg removal. The SiO2-V205 and

SiO2-TiO2-V205 composites will be synthesized and characterized. The catalytic abilities of

those composites on Hg removal will be tested in a fix-bed reactor. The reaction mechanisms on

the catalytic removal of Hg over the new catalysts will be investigated.

The last but not the least important obj ective of this research is to quantitatively investigate

the mutual interference of ozone and Hg on their measurements. This study may be of particular

importance to the ambient and indoor measurements of ozone and Hg because these two air

pollutants coexist in the environment.









CHAPTER 2
KINETIC STUDY FOR PHOTOCATALYTIC OXIDATION OF ELEMENTAL MERCURY
ON A SILICA-TITANIA NANOCOMPOSITE*

Background

A solid understanding of the kinetics of photocatalytic oxidation of Hgo is of great

importance to make an effective design of the photocatalytic reactor and to predict the reaction

rate in larger scale applications. Lee et al. (2004) studied Hgo oxidation by TiO2 nanoparticles

with UV irradiation in a differential bed reactor (DBR) and an aerosol flow reactor (AFR), and

correlated the overall reaction rate with the initial Hgo concentration and UV intensity. However,

the kinetic parameters on water vapor dependence were not available in that study, while water

vapor is an important component in the flue gas and plays a critical role in the chemistry of

mercury in coal-fired boilers (Edwards et al., 2001; Niksa et al., 2001). Rodriguez et al. (2004)

developed a mechanistic model to predict Hgo capture with in situ-generated TiO2 nanoparticles

by solving the equilibrium equations for electron-hole pair generation/consumption They also

compared their mechanistic model with the Langmuir-Hinshelwood (L-H) model used by Obee

(1996) for characterizing photocatalytic oxidation of certain organic compounds. At low water

vapor concentrations, the Hg capture rate predicted by the mechanistic model (Rodriguez et al.,

2004) was proportional to the square root of the water vapor concentration, whereas the L-H

model (Obee, 1996) indicated first-order dependence. At high water vapor concentrations, both

models predicted a constant Hg capture rate that was independent of the water vapor

concentration.





* Reprinted with permission from Li, Y., Wu, C.-Y., 2007. Kinetic Study for Photocatalytic Oxidation of Elemental
Mercury on a SiO2-TiO2 N8HOcomposite. Environ. Eng. Sci. 24, 3-12, published by Mary Ann Liebert, Inc., New
Rochelle, NY









Some other modeling studies have been done on Hg capture using activated carbon.

Rostam-Abadi et al. (1997) applied an empirical equation to the mass balance for Hgo sorption

on carbon particles in a duct flow reactor and derived the minimum C/Hg ratio required to reduce

Hgo at a certain inlet Hgo concentration. Chen et al. (1996) derived an equation to model mercury

capture when it is limited by both mass transfer and capacity by assuming that adsorption at the

surface obeys Henry's law. A conceptually similar approach was used by Flora et al. (1998)

based on the Langmuir isotherm and by Meserole et al. (2000) based on the Freundlich equation.

Several other studies (Kim and Hong, 2002; Raillard et al., 2004; Shang et al., 2002; Son et al.,

2004) have been conducted on photocatalytic oxidation of various volatile organic compounds

(VOCs) by TiO2, and the experimental data matched well with the L-H kinetic model. This

intriguing L-H nature of a wide range of VOCs warrants the investigation on the correlation

between the kinetics of Hgo photocatalytic oxidation by TiO2 and the L-H rate expression,

whereas no relevant research has been done so far. In addition, the L-H model takes advantages

over the other models previously described in incorporating the effect of competitive adsorption

of water vapor. Therefore, the research goal of this chapter was to study the kinetics of the Hgo

photocatalytic oxidation on a SiO2-TiO2 HRHOCOmposite by using the L-H model to analyze the

kinetic data. The role of water vapor in Hgo photocatalytic oxidation was established as well.

This kinetic modeling study is of importance in predicting Hgo removal efficiency and is useful

for designing an effective reactor, under photocatalyzed oxidizing conditions.

Materials and Methods

Synthesis of SiO2-TiO2 NRHOcomposite

The SiO2-TiO2 HRHOCOmposite was synthesized following a sol-gel method (Pitoniak et al.,

2003) using deionized water, ethanol, tetraethyl orthosilicate (TEOS) with HNO3 and HF added

as catalysts to increase the hydrolysis and condensation rates. First, the chemicals were added to









a polymethylpentene container, and then TiO2 nanoparticles (Deggusa, P25) were added to the

batch with a magnetic stir plate providing sufficient mixing. After that, the solution suspended

with TiO2 nanoparticles was pipetted into polystyrene 96-well assay plates before the gelation

occurred. The pellets were later aged at room temperature for 2 days and then at 65 OC for

another 2 days. After aging, the pellets were removed from the plates, rinsed with deionized

water to remove any residual acid or ethanol. Next, the pellets were placed in a programmable

oven and heated at 103 OC for 18 h to remove any residues of liquid solution within the silica

network and then at 180 oC for 6 h to harden the gel. Finally the temperature was slowly

decreased back to room temperature over a 90 min period. The final size of an individual

cylindrical pellet was approximately 5 mm in length and 3 mm in diameter. The loading of TiO2

in the nanocomposite was 12 wt%, which corresponded to the optimum performance of Hgo

removal using the SiO2-TiO2 HRHOCOmposite (Pitoniak et al., 2003). The average BET (Brunauer,

Emmett, and Teller equation) surface area of the nanocomposite was measured to be 280 m2 -1

using a Quantachrome NOVA 1200 Gas Sorption Analyzer (Boynton Beach, FL).

Apparatus and Procedure

Figure 2-1 shows the schematic diagram of the experimental system. An incoming cylinder

air was divided into three streams, the flowrates of which were controlled by mass flow

controllers (MFC, Model. FMA 5400/5500, Omega Engineering, Inc., Stamford, CT). The total

flowrate remained constant at 2 L/min. One of the air streams was allowed to pass through a

water bubbler for a humid flow or to bypass it for a dry flow. The second stream served as

dilution to adjust the humidity level. The third stream passed through the surface of a liquid Hgo

reservoir and introduced the saturated Hgo vapor into the system. The Hgo reservoir was placed

in an ice-water bath to maintain a constant Hgo vapor pressure. Downstream of all the gases was

the fixed-bed photocatalytic reactor, the lower part of which is a cylindrical tube of fused quartz









4.5 cm in diameter and 20 cm in length. The reactor was mounted with a fused quartz center with

a diameter of 2 cm, which was used to house a UV lamp. The UV light has a peak wavelength of

365 nm with an intensity of 4 mW/cm2 measured by a UVX radiometer (with a UVX-36 sensor

probe). At the bottom of the reactor is a glass frit used to hold the SiO2-TiO2 pellets within the

bed. A thermocouple (TC, Type K, Omega Engineering, Inc.) was used to monitor the

temperature on the surface of the pellets. The Hgo concentration at the reactor outlet was

measured by a RA-915+ Hg analyzer (OhioLumex Co., Cleveland, OH), which is based on

Zeeman Atomic Absorption Spectrometry using High Frequency Modulated light polarization

(ZAAS-HFM) (Sholupov et al., 2004). The inlet Hgo concentration was obtained when the Hgo

laden air bypassed the reactor. Finally, the air stream passed through a carbon trap before it was

exhausted into the fume hood.

Two sets of experiments were performed in this study. In the first set, no water vapor was

introduced into the air stream but with variations in the inlet Hgo concentration (0.19 to 1.28 CI-

mol m-3 or 38 to 256 Clg m-3). In the second set, the inlet Hgo concentration remained constant

but with changes in water vapor concentration (0 to 0.95 mol m-3). In each experiment, the Hgo

laden air was allowed to pass through the reactor for one hour to ensure that the Hgo adsorption

on the SiO2-TiO2 HRHOCOmposite reached equilibrium, which was monitored by the online Hg

analyzer. Then, the photocatalytic reaction was started by turning on the UV lamp and the Hgo

concentrations were recorded for a certain period of time until no more reduction in Hgo

concentration was observed. All the experiments were conducted under room conditions. In each

test, 2.5 grams of fresh SiO2-TiO2 pellets were used, which corresponded to an average of 4 mm

bed thickness (approximately one single layer of pellets).









Model Description

Photocatalytic oxidation of Hgo occurs when the SiO2-TiO2 HRHOCOmposite is under UV

irradiation as shown in Figure 2-2. The hole-electron pairs generated on the TiO2 particle

surfaces lead to the formation of highly reactive hydroxyl (OH) radicals, which are responsible

for Hgo oxidation to form HgO (Pitoniak et al., 2003; Wu et al., 1998). The mechanism can be

described as the following reactions:

TiO2 + hv e + h' (2-1)

H20 ++ H' + OH (2-2)

h' + OH OH (2-3)

h' + H20 -OH +H' (2-4)

OH +Hgo HgO (2-5)

Among the factors that affect the efficiency of Hgo capture by the SiO2-TiO2

nanocomposite, water vapor content in the Hgo laden air was reported to be one of the most

important (Pitoniak et al., 2003). On one hand, surface moisture on TiO2 nanoparticles is

necessary for generating OH radicals (Reactions 2-4) which are responsible for photocatalytic

Hgo oxidation. On the other hand, at high water vapor concentrations, competitive adsorption

may reduce the number of sites available for Hgo (Pitoniak et al., 2003; Rodriguez et al., 2004).

Similar to the studies by other researchers (Canela et al., 1998; Obee, 1996; Obee and Hay,

1997), the rate of photocatalytic oxidation of Hgo is defined as

(C "' C ""'t )x Q
r = (2-6)


where Cy"'n is Hgo concentration at the inlet of the reactor, CHow is Hgo concentration at the

outlet of the reactor at steady state, Q is the volumetric flow rate of the Hgo laden air (2 L min'









or 0. 12 m3 h- ), and A, is the effective surface area of the pellets that is exposed to UV light. It

should be noted that only a thickness of 0. 1 mm from the surface of the pellets and only the areas

facing the UV light can effectively contribute to Hgo oxidation (Pitoniak et al., 2005). Thus, Ae

can be calculated as

A, = SA x mx f x f, (2-7)

where SA is the specific surface area of the pellets (280 m2 -1)~, m iS the mass of pellets used (2.5

g), fV is the volume fraction of the 0. 1 mm thickness layer that UV light can penetrate (estimated

to be 0. 15), and fp is the packing factor that accounts for the fraction of the surface areas exposed

to UV light (estimated to be 0.5).

To correlate the experimental data of photocatalytic oxidation rate of Hgo, the L-H rate

equation was used. If the concentration of water vapor is constant, the L-H expression can be

simplified as


r =k K~Cg(2-8)
1+ KgHg

where r is the reaction rate (p-mol m-2 h- ), k is the L-H rate constant (p-mol m-2 h- ), Kii is the

Langmuir adsorption constant of Hgo (m3 CI-mOl-1), and Cag is the Hgo concentration (CI-mol m-3.

Cag is normally assigned to be the bulk or inlet concentration, Ca"'n (Obee, 1996; Obee and Hay,

1997). The inverse of Equation 2-8 gives

1 11 1
-= ~- -+- (2-9)
rkK C~ k

If the assumed L-H expression is valid for Hg photocatalytic oxidation, a plot of r-l vs.

Ca-~ should be linear. Subsequently, the values of k and K~g can be derived from the

combination of the intercept and the slope of the linear line. From these values, the

photocatalytic Hgo oxidation rate can be predicted by the L-H model.









Similar to the modeling studies conducted by other researchers on photocatalytic oxidation

of organic pollutants (Obee and Hay, 1997; Shang et al., 2002), when water vapor is present, the

inhibitory effect of water vapor on Hgo photocatalytic oxidation can be assumed according to the

following L-H form


r =k K~gHg (2-10)
1+ KgC~g K,C,

where Kw is the Langmuir adsorption constant of water and Cw is the water vapor concentration.

The inverse of Equation 2-10 gives

1 K 1 1
C, +- 1+(-)
r kGKgCg wk KgC, 2g1

The value ofKH, can be obtained from previous analysis when water vapor is not present.

When C4~ remains at a constant level and only C,, varies, a plot of r-' versus C,. should be linear

if it follows the L-H model expression. Then the values of k and Kw can be derived from the plot.

Results and Discussion

Effect of Hgo Concentration

Figure 2-3 shows the outlet Hgo concentration as a function of UV illumination time at six

different inlet levels ranging from 0. 19 to 1.28 CI-mol m-3 (3 8 to 256 Clg m-3) when water vapor

was not present. The outlet Hgo concentration dropped quickly when UV was first turned on for

a few minutes and then gradually leveled off. From 20 to 30 min, no significant change in outlet

Hgo concentration was observed and the pellet surface temperature remained almost constant

(42.7 & 0.3 OC). Therefore, 30 min was taken as the time the system reached steady state.

Experiments were repeated three times at each inlet Hgo concentration level. The average Hgo

removal efficiency ranged from 90 to 95% but was not an apparent function of the inlet Hgo

concentration.









At each inlet Hgo concentration level, the photocatalytic oxidation rate r can be calculated

from Equation 2-6 and the average value can be obtained. A plot of r- vs. Cing- is shown in

Figure 2-4 and the observed linear relationship indicates that the kinetics of Hgo photocatalytic

oxidation fits the L-H model very well. From Equation 2-9, values of the L-H rate constant k and

the Langmuir adsorption constant Kng were calculated to be k = 0.024 CI-mol m-2 h-l and K~g=

0.094 m3 p-mOl-1. Substituting the values of k and Kii, back into Equation 2-8, the photocatalytic

oxidation rates at different inlet Hgo concentrations can be predicted by the L-H model.

In the kinetic study of Hgo photocatalytic oxidation on TiO2 particles by Lee et al. (2004),

the reaction orders with respect to initial Hgo concentration (which ranged from 1-10 Clg m-3 Of

0.005-0.05 CI-mol m-3) were reported to be 1.4 for the differential bed reactor (DBR) and 1.1 for

the aerosol flow reactor (AFR). They also suggested that the higher value obtained for the DBR

might be due to inherent experimental errors. In this work, the fix-bed reactor design is similar to

the DBR used by Lee et al. (2004). With the inlet Hgo concentration ranging from 0. 19 to 1.28 CI-

mol m-3 in this study, the value ofKIngTrr is far less than 1. Thus, Equation 2-8 can be simplified



r = kK,,,C,, (2-12)

Equation 2-12 shows that the reaction order with respect to the initial Hgo concentration is

1, which is representative of a practical sorbent process (Lee, Biswas et al., 2004). Lee et al.

(2004) also correlated the overall reaction order with respect to the UV intensity and reported an

order of 0.3 5 for the DBR and 0.39 for the AFR. In this study, the effect of UV intensity was not

investigated.

Useful prediction results can be obtained from the L-H model as shown in Figure 2-5,

which is characterized by a steep rise of Hgo photocatalytic oxidation rate at inlet concentrations









approximately less than 20 CI-mol m-3 and subsequent mild increase at higher concentrations.

Due to the limitations on the capability of the Hg generation unit and the measurement range of

the Hg analyzer, experimental data greater than 20 CI-mol m-3 WeTO HOt available in this study.

Further research is needed on validating the L-H feature of Hgo photocatalytic oxidation in the

high concentration range. On the other hand, it should be noted that typical Hg concentrations in

coal-fired power plant flue gases are less than 0.05 CI-mol m-3(10 Clg m-3) (Pavlish et al., 2003),

which locates this process at the very lower end of the steep-rise range. This further demonstrates

the great potential of the SiO2-TiO2 HRHOCOmposite for Hgo removal from emission sources even

with much higher Hgo concentrations.

Effect of Water Vapor

Water vapor experiments were conducted at a constant inlet Hgo concentration of 0.66 CI-

mol m-3 with variations in the water vapor concentration, as shown in Figure 2-6. As the water

vapor concentration increased from 0 to 0.95 mol m-3, the steady-state Hgo removal efficiency (at

30 min) also decreased from 93% to 24%. This demonstrates a significant inhibitory effect of

water vapor on photocatalytic Hgo oxidation. Experiments were repeated three times at each

water vapor concentration level. The average values of r-l versus C, at a constant inlet Hgo

concentration are plotted in Figure 2-7. The linear relationship between them shows a good

match of the experimental data with the L-H model expression in humid air (Equation 2-11). The

intercept and the slope of the linear plot give the L-H rate constant k = 0.03 1 CI-mol m-2 h-l and

the Langmuir adsorption constant of water K,= 4.39 m3 mol-1. The previously obtained Kng

(0.094 m3 CI-mOl-1 or 9.4 x 104 m3 mol-1) is four orders of magnitude larger than K,, which

indicates that the adsorption ability of the SiO2-TiO2 HRHOCOmposite is much greater for Hgo than

for water vapor. However, water vapor plays a very important role in Hgo removal because in the









flue gas Hgo concentration is at such trace levels (seven to eight orders of magnitude smaller)

compared to that of water vapor.

Now that all the kinetic parameters have been estimated, the L-H model can be used to

predict the rate of Hgo photocatalytic oxidation at any level of inlet Hgo concentration and water

vapor concentration. Figure 2-8 compares the experimental data of the Hgo photocatalytic

oxidation rate with L-H model predictions in humid air. For the six experimental conditions

shown in Figure 2-8, the deviations of the experimental data from the L-H model predictions are

less than 15%, which are within an allowable range of experimental error. This result once again

verifies the L-H nature of Hgo photocatalytic oxidation by the SiO2-TiO2 HRHOCOmposite, and

suggests that it is appropriate to apply the L-H model to predict the photocatalytic reaction rate.

Using the L-H model, the rate of Hgo oxidation by SiO2-TiO2 under coal combustion flue gas

conditions can be predicted. At an inlet Hgo concentration of 0.05 CI-mol m-3 (10 Clg m-3) and a

water vapor concentration of 10 vol%, the reaction rate is calculated to be 7.7x10-6 CI-mOl m-2 h-l

and the Hgo removal efficiency is around 7% in the current system (Q = 0.12 m3 h-l and Ae

52.5 m2 or 2.5 g of pellets used). However, a 95% removal efficiency can be achieved by

increasing Ae by 14 fold (using 35 g of pellets or 56 mm bed height) which is practically

applicable in a bench-scale reactor like the one used in this work. In addition, increasing the UV

power level can be another option to reduce the required amount of catalysts as the reaction rate

is proportional to the UV intensity (Lee, Biswas et al., 2004).

It is generally believed (Kim and Hong, 2002; Pitoniak et al., 2003; Raillard et al., 2004;

Rodriguez et al., 2004; Shang et al., 2002) that the inhibitory effect of water vapor on

photocatalytic reactions at relatively high water vapor concentrations is due to the competition

between water vapor and the pollutants at the TiO2 surface, i.e., a high concentration of water









vapor blocks the adsorption sites from pollutants. Unlike the mechanistic model developed by

Rodriguez et al. (2004) where water vapor promoted Hgo capture by in situ-generated TiO2

particles at low water vapor concentrations (<2000 ppmy or 0.0815 mol m-3 at 250C), the Hgo

capture rate by the SiO2-TiO2 HRHOCOmposite in this study reached maximum in dry air and

decreased as the water vapor concentration increased. An explanation for the highest

photocatalytic oxidation rate without water vapor may be related to the silanol (Si-OH) groups

on the surface of the SiO2-TiO2 HRHOCOmposite. The sol-gel reactions are performed in

water/alcohol systems that cannot avoid the reverse reactions during the sol-gel process, i.e.

hydrolysis and alcoholysis for silanol formation (Yang and Chen, 2005). Yang and Chen (2005)

reported that a SiO2 nanolayer around TiO2 nanocrystals can enhance the efficiency of

photocatalysis because the transfer of electrons to the silica sites and the hole scavenging by the

hydroxides at the TiO2-SiO2 interface prevent the electrons and holes from recombination. In the

SiO2-TiO2 HRHOCOmposite produced in this work, the hydroxyl groups from silanols may act as

traps for the holes generated by TiO2 under UV irradiation and thus, an adequate number of

hydroxyl radicals may be produced resulting in photocatalytic oxidation of Hgo even in the

absence of water vapor. Similar Eindings were reported by Kim and Hong (2002) that

photodegradation of methanol by TiO2 reached the highest rate at considerably low water

concentrations, which was explained due to the production of hydroxyl radicals from hydroxyl

groups of methanol itself. In this manner, hydroxyl radicals generated from water molecules

might be insignificant, and addition of water vapor may only prohibit Hgo photocatalytic

oxidation by blocking the Hgo adsorption sites on the surface of the SiO2-TiO2 HRHOCOmposite.

In the system of Rodriguez et al. (2004), Hgo photocatalytic oxidation rate increased with water

vapor at low water vapor concentrations, which may be because water vapor was the only source









for hydroxyl radical production. Comparisons between this study and that by Rodriguez et al.

(2004) suggest that hydrophilic adsorbents (such as SiO2-TiO2 HRHOCOmposite) may have better

performance in Hgo removal at dry or very low humidity environment, and on the other hand,

hydrophobic materials (such as TiO2 nanoparticles) may yield a larger Hgo removal rate as the

humidity increases. However, the performance of both types of materials will be inhibited at very

high water vapor concentrations.

Summary

The kinetics of Hgo photocatalytic oxidation on a SiO2-TiO2 HRHOCOmposite under UV

irradiation was studied through experiments in a fixed-bed reactor. A Langmuir-Hinshelwood

model was used to analyze the kinetic data. Good agreement between the experimental data and

the L-H model was demonstrated, indicating the validity of using the L-H model to describe the

kinetics of Hgo photocatalytic oxidation. Model predictions demonstrate a great potential of the

SiO2-TiO2 HRHOCOmposite for Hgo removal even at very high Hgo concentrations. The rate of

photocatalytic Hgo oxidation increased when the inlet Hgo concentration increased and it reached

a maximum value in the absence of water vapor. The addition of water vapor was found to

inhibit Hgo photocatalytic oxidation, which may be explained by the competitive adsorption of

water vapor with Hgo on the TiO2 SUTrfCO.











Photocatalytic
Reactor



UV ampComputer


3-Way
Valve


Cylinder Air


Carbon
Trap


MCIce-Wl~ater
Bath


Figure 2-1. Experimental system for kinetic studies.










OHgo


H20
0


Hgo IuZ

OH -OH


HgO HgO
SiO2-TiO2 nanOcomposite

Figure 2-2. Description of Hgo photocatalytic oxidation on SiO2-TiO2 HRHOCOmposites.










10
03-0- Cgin = 1.28 C1-mol m"
E c Cn'= .86 pmol m-
E~~ --cgin= 0.66 p-mol m-3
1W-9- Cgin = 0.53 p-mol m"
-0 "gn= 0.31 C1-mol m"
~t-0- Cgin = 0.19 tr-mol m"





0.01
o



0 5 10 15 20 25 30

IIlumination Time (min)

Figure 2-3. Photocatalytic oxidation of Hgo at different inlet Hgo concentrations without water
vapor.








3000

2500 r = 453.5C + 42.8
R2 = 0.9986
o2000

1500

E1000

500


0 1 2 3 4 5 6

1/CHg (m3 p-mOl-1
Figure 2-4. Inverse of Hgo photocatalytic oxidation rate versus the inverse of inlet Hgo
concentration (without water vapor).










0.024


0.020


r** 0.0030
0.016
hi I /0.0025

-0.012 h
O E
E o 0.0015

I E/
0.0005

0.004 .l 0.0000
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

CHg (p-mol m-3
0.000
0 20 40 60 80 100

CHg (p--mol m- )


Figure 2-5. Rate of Hgo photocatalytic oxidation versus inlet Hgo concentration without water
vapor (solid circles: experimental data; solid line: L-H model).










1.0
I-0- C, = 0.95 mol m-3
E -0--[ C, = 0.74 mol m-3
o 0.8 ^- C, = 0.54 mol m-3
E _9- Cw= 0.42 mol m-3
-0- C, = 0.31 mol m-3
c o0.6 -- ~c,= omolm-3



a 0.4



o


O
0.0
0 5 10 15 20 25 30

IIlumination Time (min)

Figure 2-6. Photocatalytic oxidation of Hgo at a constant inlet Hgo concentration of 0.66 CI-mol
m-3 with variation in water vapor concentration.








3000

250 r-' = 2261C, + 546.6
m R2 = 0.9794
o 2000

1500

S1000

500


0.0 0.2 0.4 0.6 0.8 1.0

C, (mol m )

Figure 2-7. Inverse of Hgo photocatalytic oxidation rate versus water vapor concentration at a
constant inlet Hgo concentration of 0.66 CI-mol m-3










0.0020
C, = 0.31 mol m-3
-C, = 0.54 mol m-3
C, = 0.95 mol m-3
0.0015 A, = 0.31 mol m-3
I 5a C, = 0.54 mol m-3
hi I C,= 0.95 mol m-3

-0.0010 -



0.0005 r



0.0000
0.0 0.2 0.4 0.6 0.8 1.0 1.2

C~g (Ip-mol m- )

Figure 2-8. Rate of Hgo photocatalytic oxidation versus inlet Hgo concentration at different water
vapor concentrations (markers: experimental data; lines: L-H model).









CHAPTER 3
ROLE OF MOISTURE IN ADSORPTION, PHOTOCATALYTIC OXIDATION, AND
REEMISSION OF ELEMENTAL MERCURY ON A SILICA-TITANIA NANACOMPOSITE*

Background

Among the factors that affect the efficiency of Hgo capture by the TiO2 photocatalyst,

moisture content in the Hgo-laden gas was reported to be one of the most important (Pitoniak et

al., 2003; Rodriguez et al., 2004). Using in-situ generated TiO2 nanoparticles in a flow reactor,

Rodriguez et al. (2004) developed a mechanistic model and reported that Hgo capture was

promoted by low water vapor concentrations (700 to 1800 ppmy) but remained constant at higher

water vapor concentrations. They also speculated that very high water vapor concentrations can

inhibit Hgo oxidation by occupying available adsorption sites. Using a SiO2-TiO2 HRHOCOmposite

in a Eixed-bed flow reactor, Pitoniak et al. (2003) reported that when the relative humidity

increased from 15% to 90% at room temperature, the rate of Hgo adsorption decreased but that

the rate of photocatalytic oxidation remained constant.

While there is limited understanding of the effect of water vapor on Hgo capture on TiO2

surfaces, many research studies have investigated the effect of water vapor on photodegradation

of organic pollutants in air streams using TiO2 nanoparticles or thin films. Obee and Hay (1997)

reported that moisture in the range of 0 to 25000 ppmy inhibited photooxidation of ethylene by a

TiO2-COated glass plate. Shang et al. (2002) found that water vapor at concentrations of 3.7-22.4

g m-3 inhibited photocatalytic oxidation of heptane in a quartz reactor coated with TiO2 particles.

It was also reported that water vapor strongly inhibits the oxidation of trichloroethylene (TCE)

and acetone (Kim and Hong, 2002) but enhances the oxidation of toluene (Augugliaro et al.,



* Reprinted with permission from Li, Y., Wu, C.-Y., 2006. Role of Moisture in Adsorption, Photocatalytic
Oxidation, and Reemission of Elemental Mercury on a SiO2-TiO2 NaHOcomposite. Environ. Sci. Technol. 40, 6444-
6448. Copyright 2006 American Chemical Society.









1999; Kim and Hong, 2002). The results obtained using batch reactors coated with TiO2 thin

films were consistent with those using flow reactors packed with TiO2 particles. These findings

indicate that water vapor can either promote or inhibit photocatalyitc oxidation of different

organic pollutants. This uncertainty about the effect of water vapor necessitates further

investigation of the effect of moisture on Hgo oxidation by TiO2-based photocatalysts.

The goal of this chapter was to probe the role of moisture on Hgo capture (adsorption

and/or photocatalytic oxidation) using a SiO2-TiO2 HRHOCOmposite. This study did not aim to

explore the maximum Hg removal efficiency under certain emission conditions. Thus, the

experimental conditions used in this study (65 OC gas temperature and up to 23,000 ppmy H120)

were designed to explore the range of emission conditions encountered in various combustion

and manufacturing processes, and not to be representative of any specific process. In this study,

possible reemission of captured Hg species was also examined, to provide an overall evaluation

of the performance of the SiO2-TiO2 HRHOCOmposite. The corresponding mechanisms of Hgo

removal and reemission were investigated as well.

Experimental

Synthesis of SiO2-TiO2 NRHOcomposite

The SiO2-TiO2 HRHOCOmposite was made by a sol-gel method using deionized water,

ethanol, and tetraethyl orthosilicate (TEOS). Nitric acid (HNO3) and hydrogen fluoride (HF)

were used as catalysts to increase the hydrolysis and condensation rates. A detailed synthetic

procedure has been reported by Pitoniak et al. (2003). The nanocomposite was prepared in the

form of cylindrical pellets approximately 5 mm in length and 3 mm in diameter. The weight

fraction of TiO2 in the prepared Sio2-TiO2 HRHOCOmposite was approximately 13%.









Experimental Setup

A schematic diagram of the experimental system is shown in Figure 3-1. The incoming air

flow was divided into three streams, the flowrates of which were controlled by mass flow

controllers (FMA 5400/5 500, Omega). The total flowrate remained constant at 2 L/min. One of

the air streams was allowed to pass through a water bubbler (to provide a humid flow) or to

bypass it for dry conditions. The second stream served as dilution to adjust the humidity level.

The third stream passed through the surface of a liquid Hgo reservoir and introduced Hgo-vapor-

laden air into the system. The Hgo reservoir was placed in an ice-water bath to maintain a

constant Hgo vapor pressure. After the three streams converged, a humidity sensor (HX94C,

Omega) was used to measure the relative humidity, from which the partial pressure (volume

fraction) of water can be calculated.

Downstream was the fixed-bed photocatalytic reactor, the lower part of which was a

cylindrical tube of fused quartz 4.5 cm in diameter and 20 cm in length. The gas stream passed

through the reactor from top to bottom. The reactor was mounted with a fused quartz center 2 cm

in diameter, which was used to house a UV lamp. The UV light delivered 4 mW/cm2 intensity

measured by a UVX radiometer (with a UVX-36 sensor probe) at a peak wavelength of 365 nm.

Aluminum foil was wrapped around the cylindrical tube to reflect UV energy. A heating mantle

(regulated by a temperature controller) was used to heat the reactor to each selected temperature,

which was monitored by a thermocouple (type K, Omega). At the bottom of the reactor a glass

frit was used to hold the SiO2-TiO2 pellets within the bed. In this study, 2.5 g of fresh pellets

was used in each test, which gave an average bed thickness of 4 mm (approximately one layer of

pellets). Before each test, the pellets were heated at 130 OC for 3 h to remove any moisture that

may have adsorbed from the storage environment.









A RA-915+ Hg analyzer (OhioLumex) was used to measure Hgo concentration at the outlet

of the reactor. The Hg analyzer is based on Zeeman Atomic Absorption Spectrometry, which is

selective only for Hgo and capable of providing a real-time response every 1 s. The calibration of

the Hg analyzer was conducted by the manufacturer using a Dynacal" permeation device, which

is certified traceable to NIST (National Institute of Standards and Technology) standards. In this

study, the high-concentration mode of the analyzer was used (with a detection limit of 0.5 Clg m-3

and an upper measurable concentration of 200 Clg m-3). A condenser was installed upstream of

the Hg analyzer to remove excess moisture in the gas stream to minimize possible interference

from water vapor. Although a Hg speciation converting unit could be used to analyze both Hg(II)

and Hgo in the gas phase, it was not installed in this study. This is because HgO is the only

product for the reaction between Hgo and hydroxyl radicals (Pal and Ariya, 2004; Sommar et al.,

2001) and its extremely low saturation vapor pressure, 9.2x10-12 Pa at 25 OC (Schroeder and

Munthe, 1998), causes HgO to deposit on the catalyst in the reactor. Baseline Hgo concentration

was obtained when the Hgo-laden air bypassed the reactor. The Hgo removal efficiency was

obtained by comparing the outlet Hgo concentration with the baseline level. Finally, the air

stream was passed through a carbon trap before it was exhausted into the fume hood.

Results and Discussion

Role of Moisture in Hgo Capture

To investigate the effect of moisture on Hgo capture by the SiO2-TiO2 HRHOCOmposite,

experiments were conducted at different water vapor concentrations (0, 13000, and 23000 ppmy).

Blank runs without the nanocomposite were performed at concentrations of around 65 Clg m-3

Hgo and 0 and 23000 ppmy water vapor. Less than 0.5% reduction in Hgo concentration was

observed when the Hgo-laden air passed through the reactor with or without UV light. Tests were

also conducted to examine any possible interference of water vapor on measurements by the Hg









analyzer. While exactly the same experimental parameters were maintained, variations in the Hg

readings were less than 1% when switching from a dry to a humid (23000 ppmy H20) condition,

indicating negligible interference from water vapor in this range.

In each experiment, the baseline Hgo concentration (Hg0BL, ~65 Cpg m-3) WaS measured

first. Then the Hgo-laden air was passed through the reactor for 10 min and concentrations

defining the initial adsorption (IA) of Hgo were recorded. Next, the UV light was turned on for

10 min--allowing photocatalytic oxidation of Hgo to take place--and then turned off for 5 min.

Another two such UV on/off cycles were repeated, and concentrations defining the final removal

(FR) efficiency were recorded. After that, Hg0BL WaS checked again, and the Hgo-laden air was

passed through the reactor without UV for 15 min. Finally concentrations to compute the

adsorption at the end of the test (EA) were recorded.

The temperature at the pellet surface was maintained near 65 OC throughout the

experiment. Measured 5 mm above the pellets, the temperature of the gas passing through the

pellets was approximately 1.5 OC lower than that of the pellets. To maintain a relatively constant

temperature, the heating mantle around the reactor was turned off when the UV lamp was

switched on in each cycle and back on when UV was switched off. The fluctuations in

temperature of both pellets and gas were measured to be +2 oC. A preliminary test showed that

the change in the final removal efficiency was less than 5% when the average temperature of the

pellets increased from 65 to 70 OC. This indicated that the fluctuations within this range of

temperature had negligible effects on the reaction rate.

Figure 3-2 shows the dimensionless Hgo concentration [Hgo/Hg0BL] aIt the outlet of the

reactor in dry and humid conditions. Under dry conditions (Figure 3-2A), Hgo removal efficiency

increased with successive cycles and FR reached 95% during the fourth irradiation. Adsorption









was initially insignificant (IA = 5%), but it was enhanced at the end of the test (EA = 22%). In

contrast, when the water vapor concentration was increased to 13000 ppmy (Figure 3-2B), Hgo

removal increased only slightly with successive cycles and FR ended at only 51%. When the

water vapor concentration was further increased to 23000 ppmy (Figure 3-2C), FR decreased to

28%.

In addition, Hgo adsorption was found to be insignificant throughout the test under these

two humid conditions. A separate test was conducted extending the adsorption time to 2 h

(without UV), and no adsorption was observed. This verified that the recorded values of IA and

EA were measured at adsorption equilibrium. Experiments under the three humidity conditions

were repeated and similar results were recorded. The results indicated that increased humidity

can significantly suppress both Hgo adsorption and photocatalytic oxidation on the

nanocomposite.

Role of Moisture in Hgo Reemission

Potential reemission of Hg species from the SiO2-TiO2 HRHOCOmposite after their capture

is an important factor in evaluating the overall performance of the nanocomposite. The

experiment to examine Hg reemission began by exposing 2.5 g of fresh pellets for 3 h to UV

light in a stream containing approximately 300 Clg/m3 Hgo vapor in dry air at room temperature.

During the 3-h pretreatment, Hgo capture efficiency by the nanocomposite reached about 90%

after 30 min and remained relatively steady between 90 and 95% for the rest of the time (as

shown in the insert in Figure 3-3). The Hg species retained on the nanocomposite are predicted

to be a mixture of HgO (due to photocatalytic oxidation) and Hgo (due to adsorption enhanced by

HgO) (Pitoniak et al., 2005).

After the pretreatment, Hgo release from the nanocomposite at room temperature was

examined by feeding Hgo-free air into the reactor (see Figure 3-3). During the first 5 min, dry air









(condition I) was allowed to pass through the pretreated pellets and only traces of reemitted Hgo

were found. However, immediately after the air stream was switched from dry to humid ([H20]

= 23000 ppmy, condition II), a significant release of Hgo was observed, peaking briefly at 68

Clg/m3, about 23% of the Hgo feeding level during pretreatment. Even though the Hgo reemission

level decreased over time, it remained relatively high (43 Clg/m3) after 20 min. Hgo reemitted

during this 20-min period was calculated to be approximately 2. 1% of the total Hg species (Hgo

and HgO) retained on the pellets during the 3-h pretreatment.

At 25 min, UV light was turned on while the air stream remained humid (condition III),

and the Hgo concentration quickly dropped to approximately 10 Clg/m3. This suggests that a large

portion of the Hgo reemitted was photooxidized during irradiation. At 3 5 min, the UV light was

turned off (return to condition II), and the rate of Hgo reemission did not recover to the previous

high level, but further decreased to approximately 6.5 Clg/m3. Turning the UV on again (return to

condition III) at 45 min caused the Hgo concentration to return to an extrapolation of the line

observed during the first period at condition III. When the condition was switched at 55 min to

dry air with UV on (condition IV), the Hgo concentration decayed to approximately zero in 5

min. The results of the reemission test were repeatable using another batch of pellets undergoing

the same pretreatment procedure.

One may draw three conclusions about Hgo reemission under the conditions tested:

(1) Hgo reemission does not occur in dry air, with or without UV light. (2) Introducing water

vapor causes significant Hgo reemission, which slowly decreases over time. (3) Exposure to UV

light in humid air can either inhibit or promote Hgo reemission.

Mechanisms of Hgo Capture and Reemission

To explain these intriguing findings, a comprehensive model is developed in this work

elucidating the fate of Hg species on the surface of this SiO2-TiO2 HRHOCOmposite. In this model,









as illustrated in Figure 3-4, Hgo capture is accomplished by photocatalytic oxidation and

adsorption, while Hgo reemission results from desorption and photocatalytic reduction.

Photocatalytic oxidation of Hgo occurs during UV irradiation of the SiO2-TiO2

nanocomposite, as shown in Figure 3-4a. Hgo is oxidized by OH radicals generated on the TiO2

surface, and HgO has been reported to be the final oxidation product in literature (Lee et al.,

2001; Pal and Ariya, 2004; Pitoniak et al., 2003; Pitoniak et al., 2005; Rodriguez et al., 2004;

Sommar et al., 2001). Pal and Ariya (2004) experimentally identified HgO as the only product

for gas-phase reaction of Hgo with OH radicals, while HgOH was suggested to be an

intermediate reaction product (Pal and Ariya, 2004; Rodriguez et al., 2004; Sommar et al., 2001)

that has a very short lifetime (Goodsite et al., 2004). The fast removal of Hgo in this study (see

Figure 3-2) was consistent with rates reported by researchers (Pitoniak et al., 2003; Pitoniak et

al., 2005) using similar experimental systems. It was faster than OH-Hg reactions reported under

simulated atmospheric conditions (Calvert and Lindberg, 2005; Pal and Ariya, 2004; Sommar et

al., 2001), most likely due to the much higher concentration of OH radicals produced on the

high-surface-area, open-structured Sio2-TiO2 HRHOCOmposite. The overall mechanism of

photocatalytic oxidation can be described by the following reactions:

TiO2 + hv e + h' (3-1)

H20 ++ H' + OH (3 -2)

h' + OH OH (3-3)

h' + H20 -OH +H' (3 -4)
OH +Hgo oves" HgO(35


It should be noted that the specifications of the cylinder air indicated a water vapor

concentration of less than 24 ppmy. This low concentration of water vapor could not be detected









by the humidity sensor, and thus, the air was considered "dry." However, this water vapor

concentration is 3 orders of magnitude higher than that of Hgo used in this study. Further, the

SiO2-TiO2 pellets have a great capacity to adsorb water vapor. Tests demonstrated that 2.5 g of

SiO2-TiO2 pellets can adsorb an average of 0. 16 g of water vapor in 30 min when exposed to an

air flow containing 23000 ppmy water vapor. Therefore, even though the pellets were pretreated

at 130 oC for 3 h before each experiment to remove any interior moisture, they adsorbed enough

moisture (compared to the trace amount of Hgo) from the cylinder air to support the generation of

OH radicals and subsequent photocatalytic oxidation.

Physisorption of Hgo is minor if the SiO2-TiO2 HRHOCOmposite is not exposed to UV light,

but it can be enhanced by HgO that is deposited on the composite's surface after photocatalytic

oxidation takes place (Figure 3-4b). This is likely due to the high affinity between Hgo and HgO,

which was characterized by a decrease in the contact angle of Hg on the HgO-enriched sorbent

surface (Pitoniak et al., 2005). However, this enhanced adsorption ability was not observed in

humid conditions due to the inhibitory effect of water vapor.

It is generally believed that water vapor inhibits photocatalytic reactions by blocking the

available adsorption sites on the surface of TiO2 catalysts (Kim and Hong, 2002; Pitoniak et al.,

2003; Shang et al., 2002). The results of this study suggest that desorption of bound Hgo by a

high concentration of water vapor (Figure 3 -4c) also contributes to the reduced Hgo capture rate.

The reason is likely related to the photoinduced superhydrophilicity of TiO2 SUTrfCOS (Fujishima

et al., 2000; Wang et al., 1997). During the process of photocatalysis, the electrons tend to reduce

the Ti(IV) cations to the Ti(III) state, and the holes oxidize the Ol- anions. In the process,

oxygen atoms are ej ected, creating oxygen vacancies. Water molecules can then occupy these

oxygen vacancies and create adsorbed OH groups, which tend to make the surface hydrophilic.









Hence, in our experiments, when humid air passed through the reactor, the superhydrophilic

surface of TiO2 may attract excessive amount of water and result in ej section of adsorbed Hgo

(which is superhydrophobic) from the surface (Figure 3-4c). This mechanism is similar to the

self-cleaning (stain proofing) quality of TiO2-COated glass (Fujishima et al., 2000; Guan, 2005),

from which organic stains are washed away by rainfall (or water) on the superhydrophilic

surface.

A sharp decrease in Hgo concentration was detected when switching from the initial

condition II to III (Figure 3-3). A very possible reason is that the desorbed Hgo was reoxidized

upon UV irradiation and re-captured onto the pellets. However, the increase in Hgo concentration

when switching from the second condition II to III indicated that UV irradiation contributed to

the reemission of Hgo from the pellets. This can be explained by the photocatalytic reduction of

HgO to Hgo by the free electrons generated on TiO2 surface under UV light (Figure 3-4d). The

mechanism is expressed in Reaction 3-6, which has a reduction potential of 0.098V vs. normal

hydrogen electrode (NHE) (Meites, 1963):

HgO +H20 +2e 4 Hg +20H- (3-6)

To validate the occurrence of Reaction 3 -6 in this study, its redox potential was compared

with those in other photocatalytic reactions on TiO2 reported in the literature. Fujishima et al.

(2000) reported reduction of 02 to H202 on a TiO2 photocatalyst. Zhang et al. (2004) used TiO2-

modified sewage sludge carbon for photocatalytic removal and recovery of Hg2+ in the form of

Hgo from water. The redox potentials of 02/H202 and Hg2+/ Hg are 0.28V (Fujishima et al.,

2000) and 0.85V (Meites, 1963), respectively. Thus, it is reasonable to infer that Reaction 3-6,

which has a much lower redox potential (0.098V), can occur in our system. The necessity of H20









in Reaction 3-6 is also consistent with the finding that no Hgo reemission occurred in dry air with

UV irradiation.

Separate experiments were conducted in pure Ol and N2, respectively, and the results

further supported the hypothesis that photocatalytic reduction caused the reemission of Hgo

Under humid conditions and UV irradiation, a higher Hgo reemission level was observed in N2

than in Ol. Because the redox potential of 02/02 (-0.28 V) is lower than that of HgO/Hg, Ol is a.

stronger electron trap than HgO. The absence of 02 inCTreSed the chance for HgO to trap

electrons, and thus more HgO was reduced to Hg in pure N2.

The intriguing finding that UV irradiation can result in either inhibition or promotion of

Hgo reemission (as shown in Figure 3-3) can then be explained by the competition between

photocatalytic oxidation of reemitted Hgo to form HgO (Reaction 3-5) and photocatalytic

reduction of HgO to form Hgo (Reaction 3-6), accompanied by the physical desorption of Hgo

caused by water vapor at the same time. At the moment when the condition was first changed

from II to III, the concentration of desorbed Hgo was very high, and thus photocatalytic oxidation

prevailed over reduction, which resulted in a sharp decrease in Hgo concentration. Over time, the

rate of photocatalytic oxidation decreased as the Hgo desorption rate decreased. When the

condition was switched from II to III for the second time, the rate of photocatalytic oxidation

dropped to a lower level than that of photocatalytic reduction, so the level of Hgo reemission

increased.

The above discussion boils down to a conclusion that Hgo capture on the SiO2-TiO2

nanocomposite in a humid environment under UV irradiation is controlled by four mechanisms:

adsorption; photocatalytic oxidation; desorption; photocatalytic reduction. Water vapor

concentration is a significant parameter affecting the Hgo capture efficiency. The inhibitory









effect of water vapor is due to its competitive occupancy of the available adsorption sites,

displacement of adsorbed Hgo, and participation in the photocatalytic reduction of HgO to Hgo

Hgo reemission seems to be inevitable since humidity exists in most environmental

conditions. However, the mechanisms discussed above imply that appropriate application of UV

irradiation can be utilized to mitigate this Hgo reemission. The difference between the first and

second condition II in Figure 3-3 shows that UV treatment caused a significant drop of the Hgo

reemission level. However, further exposure to UV light caused an increase in the Hgo

reemission due to the dominant photocatalytic reduction later on. Therefore, determining the

optimal time of UV treatment and avoiding further exposure to UV sources (including sunlight)

are critical to achieving minimal Hgo reemission.

Summary

A novel silica-titania (SiO2-TiO2) HRHOCOmposite has been developed to effectively

capture elemental mercury (Hgo) under UV irradiation. Moisture has been reported to have an

important impact on this nanocomposite' s performance. In this work, the role of moisture on Hgo

removal and reemission as well as the corresponding mechanisms was investigated. Hgo removal

experiments were carried out in a Eixed-bed reactor at 65 OC using air as the carrier gas. Without

UV irradiation, Hgo adsorption was found to be insignificant, but it could be enhanced by the

photocatalytic oxidation product, mercuric oxide (HgO), possibly due to the high affinity

between HgO and Hgo. Under dry conditions 95% of Hgo can be removed; however, increased

humidity levels remarkably suppress both Hgo adsorption and photocatalytic oxidation.

Introducing water vapor can also result in significant reemission of captured Hgo from the

nanocomposite, which may be ascribed to the repellant effect of water vapor adsorbed on the

superhydrophilic TiO2 surface. Exposure to UV light was found either to prohibit Hgo reemission

when photocatalytic oxidation of reemitted Hgo prevailed or to promote Hgo reemission when









photocatalytic reduction of HgO to Hgo dominated later on. The results indicate that

minimization of Hgo reemission can be achieved by appropriate application of UV irradiation.













UV Lamp


3-Way
Valve


Condenser Cmue
Humidity I Heating \,

Cylinder Air Water Sesrvatl
Bubbler SiO2-TiO2
MFC Pellets
Cold Water
I 1 \~7/3-Way Hg Analyz'


MIFC ro
2-Way Valve

-8Hg C rbn

MIFC Ice-Water Bath Vn



Figure 3-1. Experimental system for studies on the role of water vapor.



















0.8


.2 0.6


0.4


0.2


0 10 20 30 40 50 60 70 80 90
Time (min)

1.2


(BBL IA Hg0BL = 63pg/1m3 BL EA



0.8


.I 0.6

0.4R
I~I

UV off UV off | UV off
0.2
UVon UVon UVon UVon


0 10 20 30 40 50 60 70 80 90
Time (min)


1.2


(CBL IA Hg0BL =62 pg ml BL/ EA



0.8 i k L FR



II I.

0.4
UV off UV off UV off|

0.2
UV on UV on UV on UV on


0 10 20 30 40 50 60 70 80 90
Time (min)

Figure 3-2. Dimensionless Hgo concentration at the reactor outlet (A, [H20] = 0 ppmy;
B, [H20] = 13000 ppmy; C, [H20] = 23000 ppmy)













II: Dry air w/o UV; II: Humid air w/o UV; III: Humid air w/ UV; IV: Dry air w/UVI


0 5 10 15 20 25


30 35 40 45 50 55 60 65


Time (min)


Figure 3-3. Hgo reemission from SiO2-TiO2 HRHOCOmposite after 3-h pretreatment (The inset
shows the dimensionless Hg concentration during the pretreatment).










(a) Photocatalytic oxidation


(b) Adsorption


SHgO /IHgo
HgO

SiO2-Tio2 nanOcomposite


HgO HgO
SiO2-Tio2 nanOcomposite


(c) Desorption

H20


(d) Photocatalytic reduction


HgO u
HgO H20

SiO2-Tio2 nanOcomposite


SHgO HgO

SiO2-Tio2 nanOcomposite


Tio2 _) Hgo HgO H20 O OH e

Figure 3-4. Mechanisms of Hg capture and reemission on the surface of SiO2-TiO2
nanocomposite.









CHAPTER 4
REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING A SILICA-TITANIA
NANOCOMPO SITE"

Background

The SiO2-TiO2 HRHOCOmposite has exhibited very high efficiency of Hgo removal (up to

99%) under room conditions with low relative humidity (Pitoniak et al., 2003). However, as

reported in Chapters 2 and 3, at room temperature but higher water vapor concentrations (up to

23,000 ppmy), Hgo capturing on the nanocomposite was hindered due to the competitive

adsorption of water vapor on the active sites, and the extent of prohibition in Hgo removal was

proportional to the water vapor concentration (Li and Wu, 2006, 2007). It should be noted that in

coal-fired boiler flue gas, the concentration of water vapor typically accounts for several percent

in volume, much higher than that in the room conditions. Thus, it is expect that water vapor may

have a greater inhibitory effect on Hgo removal in flue gas. One the other hand, the catalyst

developed in this work was designed for application in the cold-end of the boiler convective pass

(e.g. between the electrostatic precipitator and the wet scrubber), where the typical flue gas

temperature (120 to 1500C) is higher than room temperature. In this aspect, the competitive

adsorption of water vapor on the catalyst would be smaller at higher temperatures. These two

counteracting effects warrant further investigation on the performance of the SiO2-TiO2

nanocomposite for Hgo removal in flue gas.

Typical coal-derived flue gas consists of various minor gas components such as HC1, SO2,

and NOx, concentrations of which vary when burning different types of coal. It has been

reported that these minor gases are important to the heterogeneous adsorption and/or oxidation of

Hgo on activated carbons or fly ash in flue gas conditions (Carey et al., 1998; Norton et al.,


* Reprinted with permission from Li, Y., Murphy, P.D., Wu, C.Y. Removal of Elemental Mercury from Flue Gas
Using A Silica-Titania Nanocomposite. Submitted to Fuel Process. Technol.









2003). Carey et al. (1998) reported that the adsorption capacity for both Hgo and HgCl2 by Darco

FGD carbon dramatically increased as HCI concentration increased from 0 to 50 ppm but

decreased as the SO2 COncentration increased from 0 to 500 ppm. Norton et al. (2003) reported

that in the presence of fly ash, NO2, HC1, and SO2 resulted in greater levels of Hg oxidation

while NO inhibited Hg oxidation. It is expected that the nature of Hg capture on the SiO2-TiO2

nanocomposite would be different in flue gas conditions from that reported in room conditions in

our previous studies.

In this work, a photocatalytic reactor packed with SiO2-TiO2 HRHOCOmposite was installed.

The goal of this research was to identify the effects of the flue gas components on the removal of

Hgo by SiO2-TiO2 HRHOCOmposite and to explore possible surface interaction mechanisms. An

improved understanding of the role of the flue gas components can help evaluate the potential of

applying this novel material as an effective Hg control strategy for coal-fired power plants.

Experimental Methods

Synthesis of the SiO2-TiO2 NRHOcomposite

The SiO2-TiO2 HRHOCOmposite was made by a sol-gel method using deionized water,

ethanol and tetraethyl orthosilicate (TEOS). Nitric acid (HNO3) and hydrogen fluoride (HF) were

used as catalysts to increase the hydrolysis and condensation rates. A detailed procedure was

described in our previous study (Li and Wu, 2007). The nanocomposite was prepared in the form

of cylindrical pellets approximately 5 mm in length and 3 mm in diameter. The weight fraction

of TiO2 in the prepared Sio2-TiO2 pellets was approximately 12%, which corresponded to the

optimum performance of Hgo removal in room conditions (Pitoniak et al., 2003). The average

BET (Brunauer, Emmett, and Teller equation) surface area of the nanocomposite was measured

to be 280 m2 -1l USing a Quantachrome NOVA 1200 Gas Sorption Analyzer (Boynton Beach,

F L).









Experimental Setup and Procedure

A schematic diagram of the experimental setup is shown in Figure 4-1. The simulated flue

gas consisted of three maj or gases: 02, CO2, and N2. The N2 flOW WaS divided into three

branches. One of the N2 streams converged with the Ol and CO2 to form the main gas flow,

which was allowed to pass through a heated water bubbler to introduce water vapor into the

system. The second stream ofN2 SeTVed to dilute the main flow so as to adjust the humidity of

the total gas stream. The third stream of N2 paSSed through a Dynacal" Hgo permeation tube

(VICI Metronics) and introduced the saturated Hgo vapor into the system. The permeation tube

was placed in a U-shape glass tube which was immersed in a constant-temperature (90 + 0.2 OC)

water bath to ensure a constant Hgo permeation rate. Hgo concentration in the system was

controlled in the range of 75~80 Clg m-3. Minor gases including HC1, SO2, NO, and NO2 WeTO

introduced into the main flow individually or in combination. A mass flow controller (MFC) was

used to control each of the gas flow with a total gas flow rate controlled to be 2.0 L/min. The gas

concentrations were designated to be within the range of typical flue gas composition (Senior et

al., 2000): 4% 02, 12% CO2, 4~16% H20, 10~50 ppm HC1, 400~1200 ppm SO2, 50~300 ppm

NO, 10~30 ppm NO2, and balanced with N2. The experimental conditions for investigation of the

flue gas effects are listed in Table 4-1.

Downstream of all the gas flows was the packed-bed photocatalytic reactor placed

horizontally. The SiO2-TiO2 pellets were packed in a U-shape quartz tube with an inner diameter

of 13 cm. A heating cord was wrapped around the U-tube so that the flue gas temperature can be

controlled at around 135 OC, which was monitored by a Teflon thermocouple (Type K, Omega).

A UV lamp was placed in a separate quartz tube centered in the reactor and 5 cm above the

centerline of the U-tube. The UV light had a peak wavelength of 365 nm with an intensity of 4

mW/cm2 measured by a UVX radiometer (with a UVX-36 sensor probe). A stream of cooling air









was continuously purged through the UV lamp to lower the lamp temperature to around 60 oC.

The entire reactor was placed inside an aluminum cylinder so that the UV light could be reflected

back to the pellets and a maximum utilization of the UV energy could be achieved.

A wet-chemistry conversion system (Laudal et al., 2004; McLarnon et al., 2005) and a RA-

915+ Hg analyzer (OhioLumex) were used to measure gas-phase Hg speciation (Hgo and Hg(II))

downstream the reactor. The Hg analyzer is based on Zeeman Atomic Absorption Spectrometry

(ZAAS), which is selective only for Hgo. In the conversion system, the sampling gas was divided

into two streams, one for measuring Hgo and the other for total Hg (HgT). The solution used for

Hgo measurement consisted of 10% potassium chloride (KC1), which captures Hg(II) and allows

only Hgo to pass through. HgT measurement was accomplished using an acidic 10% stannous

chloride (SnCl2) SOlution, which reduces Hg(II) to Hgo, thus producing HgT. The concentration

of Hg(II) can then be calculated by the difference between HgT and Hgo. The two streams

converged to a 10% sodium hydroxide (NaOH) before entering the Hg analyzer. The NaOH

solution captured acid gases, such as HCI and SO2, to prevent corrosion of the detecting cell of

the Hg analyzer. In addition, as part of the conversion process, a NaOH solution was used to

remove SO2 before the sampling gas entered the SnCl2 SOlution, as SO2 can interfere with the

reduction of Hg(II) by SnCl2 (Laudal et al., 2004). A condenser was installed upstream of the Hg

analyzer to remove excess moisture in the gas stream. This aimed to avoid condensation of water

vapor inside the Hg detection cell and thus to minimize possible interference from water vapor.

The Hg analyzer was capable of providing a real-time response every 1 s. The calibration of the

Hg analyzer was conducted by the manufacturer using a Dynacal" permeation device. In this

study, the high-concentration mode of the Hg analyzer was used (with a detection limit of 0.5 Clg

m-3 and an upper measurable concentration of 200 Clg m-3). Finally, the gas stream was passed









through a carbon trap before it was exhausted into the fume hood. The entire system was Teflon

lined. To avoid condensation of the water vapor along the pathway, all the lines before the

condenser were heated by heating tapes to above 900C.

Results and Discussion

Baseline Test

Tests were first conducted to examine any possible interference caused by the flue gas

components on the measurement of the Hg analyzer. Balanced with pure N2, 8% H120, 50 ppm

HC1, 1200 ppm SO2, 300 ppm NO, and 30 ppm NO2 were individually introduced to the system

without the presence of Hg. In all cases, no significant Hg readings were observed (i.e. the

interference was less than the detection limit, 0.5 Clg m-3) with or without the UV irradiation.

This indicated negligible interference by the flue gas components in the concentration ranges

studied in this paper. In addition, tests were performed by introducing 80 Clg m-3 Hgo to an empty

reactor (i.e. no catalyst) with or without UV irradiation. Less than 0.5% reduction in Hgo

concentration was observed, which indicated that the loss of Hgo on the reactor wall was

negligible.

To examine the effect of individual flue gas components, a baseline test without the minor

gases (Set 1) was first conducted. While a larger amount of SiO2-TiO2 pellets could be used to

achieve a Hg removal efficiency greater than 90%, only 8 g of pellets (~1 g TiO2) WeTO USed to

better manifest possible enhancement by the minor gases in subsequent tests. As shown in Figure

4-2, the inlet Hg concentration was measured within the first 10 min when the gas stream

bypassed the reactor (Period A). Next, the gas stream was passed through the reactor without UV

light for another 10 min (Period B) and then the UV light was turned on to activate the

photocatalytic reaction (Period C). The concentrations of HgT and Hgo at the outlet of the reactor









were recorded in alternation and were averaged every 2 min for Periods A and B and every 10

min for Period C.

In Period A, the concentration of HgT WaS equal to that of Hgo, confirming that the Hg

source in this study was only Hgo. In Period B when the gas passed through the reactor, the outlet

Hg concentration first dropped by approximately 10%, probably due to the physical adsorption

by the porous SiO2-TiO2 pellets. However, it quickly recovered to the same level of the inlet

concentration indicating the adsorption was saturated. A significant decrease in Hg concentration

was detected only when the UV light was turned on in Period C. The concentration of HgT

dropped to 59% of the inlet level during the first 10 min of UV irradiation and slowly decreased

to 53% in the next 80 min. At 100 min, the rate of the decrease of HgT WaS approaching zero,

and thus it was assumed that the performance of the catalyst reached a relatively stable level at

this point. At 100 min, the outlet concentration of Hgo decreased to 34% of the inlet Hgo level

while Hg(II) slowly increased to 19%. The amount of Hg captured on the pellets can be

expressed as

Hgcap = HgTin -- HgTout (4-6)

where Hgcap represents captured Hg, and HgTin, and HgTout represent HgT at the inlet and outlet of

the reactor, respectively. Since the inlet Hg source is 100% Hgo and negligible Hgo capture was

observed without UV, it is reasonable to assume that the captured Hg species under UV

irradiation was only Hg(II) due to the photocatalytic oxidation. Hence, the total amount of Hgo

oxidized can be expressed as

Hgoxi = Hgoin Hgoout (4-7)

where Hgox represents oxidized Hg, and Hgoin and Hgoout represent Hgo at the inlet and outlet of

the reactor, respectively. It should be noted that HgTin WaS equal to Hgoin in this study. In the









baseline test, the percentage of oxidized Hg was calculated to be 66%, while 47% was captured

by the pellets and 19% penetrated through the reactor. In other word, the inlet Hgo partial

pressure was around 9.5 ppb and approximately 6.3 ppb (66%) of Hgo was oxidized to form HgO

over the SiO2-TiO2 HRHOCOmposite. The saturated vapor pressure of HgO at 135 OC, calculated

according to Borderieux et al. (2004), is 6.0 ppb. The measured partial pressure of HgO (6.3 ppb)

is very close to the saturated vapor pressure, indicating that most of the HgO formed would stay

in the gas-phase at 13 5 OC. Then a fraction of the gaseous HgO (47%) was captured by the

nanocomposite, and the rest (19%) penetrated through the reactor.

Effects of Individual Flue Gas Components

The effects of individual flue gas components were examined and the results were

compared with the baseline. At least two runs were performed at each of the experimental

condition listed in Table 1. The average values of Hg capture and oxidation efficiencies are

shown in Figure 4-3, where the error bars represent the envelope of minimum and maximum

values.

An inhibitory effect of water vapor on Hg removal was observed as shown in Figure 4-3a.

Experiments were first conducted in a relatively dry condition (<0.1%) by bypassing the water

bubbler and then in humid conditions with an increasing water vapor concentration. In the dry

condition, the efficiencies of both Hg capture and oxidation reached over 99%. As the water

vapor concentration increased from 4% to 16% (baseline was 8%), the Hg capture efficiency

decreased from 73% to 18%, and the Hg oxidation efficiency decreased from 88% to 32%. The

inhibitory effect of water vapor is very likely due to its competitive adsorption with Hgo on the

active sites, and as shown in Figure 4-3 a, the extent of inhibition on Hg removal is proportional

to the concentration of water vapor. This trend agrees with the results obtained at room

temperature and low water vapor concentrations (< 2.3 %) (Li and Wu, 2006), which indicates









that even at a higher temperature (13 5 OC in this work) the competitive adsorption of water vapor

is still significant. It has been reported that Hgo is not adsorbed (or is only weakly adsorbed) on

the surface of sorbents including unburned carbon and selective catalytic reduction (SCR)

catalysts (Niksa and Fujiwara, 2005a, 2005b). The weakly bonded Hgo can even be desorbed

from the surface of SiO2-TiO2 COmposite by water vapor at high concentrations (Li and Wu,

2006). The desorption process would also lead to a lower efficiency of Hgo removal. Since the

concentration of water vapor in this study was seven to eight orders of magnitude higher than

that of Hgo, the inhibitory effect of water vapor could be very significant even at higher

temperatures.

It should also be noted that penetration of oxidized Hg from the reactor (i.e. the difference

between Hg oxidized and captured) occurred in humid conditions (4 16% H20) but not in the

dry condition. This can be explained by the competitive adsorption of water vapor with the gas-

phase oxidized Hg. Since Hgo is not adsorbed (or is only weakly adsorbed) on the sorbent

surface (Niksa and Fujiwara, 2005a, 2005b), it is very possible that a portion of the oxidized Hg,

which is the product of the reaction between Hgo and OH radicals, existed in the gas phase in the

vicinities of the reaction sites. In the dry condition, the oxidized Hg in the gas phase was

adsorbed and thus there was no penetration. However, in humid conditions, water vapor

competes with the oxidized Hg and consequently not all oxidized Hg in the gas phase can be

adsorbed. The superhydrophilic surface of TiO2 after exposure to UV irradiation can further

enhance the adsorption of water vapor (Li and Wu, 2006) but reduce the capture of oxidized Hg.

As a result, penetration of oxidized Hg was usually observed in humid conditions in this work.

The effect of HCI on Hg removal was found to be promotional (Figure 4-3b). In the range

of 10 to 50 ppm HC1, Hg capture efficiency increased to approximately 75% and Hg oxidation









efficiency increased to over 90%. However, the extent of promotion was not apparently related

to the HCI concentration in the range studied. The promotional effect of HCI is consistent with

the literature that HCI promotes heterogeneous Hg oxidation (Presto and Granite, 2006). It has

been reported that in the presence of an appropriate catalyst (e.g. metal oxides), a Deacon

process (Pan et al., 1994) could convert HCI in flue gas to Cl2 at high temperatures (300-400 oC),

thereby enhancing Hgo oxidation (or chlorination). Niksa et al. proposed that Hg oxidation

occurs via an Eley-Rideal mechanism, where adsorbed HCI reacts with gas-phase (or weakly

adsorbed) Hgo (Niksa and Fujiwara, 2005a, 2005b). The mechanism was also consistent with the

observation of enhanced Hgo sorption to halogen-promoted sorbents and fly ashes in literature

(Granite et al., 2000; Maroto-Valer et al., 2005). In this work, the Deacon process was less likely

to occur because of the relatively low flue gas temperature (13 5 OC). Instead, it is more likely to

follow the Eley-Rideal mechanism. In the Eley-Rideal mechanism, HCI may first be adsorbed on

the surface of SiO2-TiO2 HRHOCOmposite, and then react with gas-phase Hgo. This, together with

the Hgo oxidation by OH radicals, can result in a higher Hg removal efficiency. Further

investigation is needed to confirm the reaction mechanism.

As shown in Figure 4-3c, SO2 WaS found to have a promotional effect on Hg capture and

oxidation and the promotion was proportional to the concentration of SO2 in the range of 0~1200

ppm. The Hg capture and oxidation efficiencies reached 73% and 91% respectively at 1200 ppm

SO2. The effect of SO2 on heterogeneous Hg oxidation was not conclusive in literature. It has

been reported that SO2 COmpetes with HCI for sites on activated carbon and fly ash sorbents and

thus inhibits mercury oxidation and adsorption in flue gas (Laudal et al., 2000; Laumb et al.,

2004). Carey et al. reported that the adsorption capability of a Darco FGD carbon for both Hgo

and HgCl2 decreased as the SO2 COncentration increased from 0 to 500 ppm but neither capacity









changed significantly above 500 ppm SO2 (Carey et al., 1998). However, in some cases, SO2

appears to enhance Hgo oxidation (Eswaran and Stenger, 2005; Norton et al., 2003). Eswaran and

Stenger reported a promotional effect of SO2 on Hgo oxidation over a selective catalytic

reduction (SCR) catalyst (Eswaran and Stenger, 2005). The mechanism was proposed as:

SO2 +o 4 024SO3 (4-8)

Hg + SO3 + ,0 o, HgSO4 (4-9)

Similar mechanism can be used to explain the promotional effect of SO2 in this work, where SO3

was formed through the oxidation of SO2 by OH radicals (Bai et al., 2006), which were

generated on the SiO2-TiO2 pellets under UV irradiation.

The effect of NO on Hg removal was found to be inhibitory at a significant but relatively

constant level in the concentration range of 50~300 ppm NO (Figure 4-3d). The Hg capture and

oxidation efficiencies both decreased to around 10% in the presence ofNO. NO has been

reported as an inhibitor for heterogeneous Hgo oxidation on fly ash (Norton et al., 2003), but the

mechanism was not clear. In this study, it is very likely that the scavenging of OH radicals by

NO hindered the photocatalytic oxidation of Hgo. The inhibition occurred via (Niksa et al., 2001)

OH + NO + M 4 HONO + M (4-10)

In the presence of 10~30 ppm NO2, the efficiencies of Hg capture and oxidation were

slightly lower than those in the baseline (Figure 4-3e). However, the effect of NO2 in this range

can be considered as insignificant compared to other flue gas components. It has been reported in

literature that NO2 can enhance heterogeneous oxidation of Hgo in the presence of fly ash

(Norton et al., 2003) or iron oxides (Borderieux et al., 2004; Galbreath et al., 2005), though this

effect is often considered of minor importance compared to chlorination.









Hg Removal in Simulated Flue Gases

The performance of the SiO2-TiO2 HRHOCOmposite was finally tested in two simulated flue

gases, the compositions of which were in line with those reported in literature (Pavlish et al.,

2003). Flue gas 1 (FGl, Set 7) represents burning of high rank (bituminous) coals that contain

higher chlorine and sulfur contents. Flue gas 2 (FG2, Set 8) represents burning of low rank

(subbituminous and lignite) coals, which contain less chlorine and sulfur but more moisture. As

shown in Figure 4-3f, Hg removal in Flue Gas 1 was close to that in baseline, indicating that the

prohibitory effect of 300 ppm NO counteracted the promotional effects of 30 ppm HCI and 1200

ppm SO2. Hg removal in Flue Gas 2 was less than in Flue Gas 1, very likely due to the higher

concentration of H20 and lower concentrations of HCI and SO2. Hence, high rank coals are

preferable to low rank coals for the application of the SiO2-TiO2 HRHOCOmposite. Minimizing the

adverse effect of NO so as to improve the overall performance of the catalyst would be an

important task for future research.

Summary

A novel SiO2-TiO2 HRHOCOmposite has been synthesized to removal Hgo from simulated

coal-fired power plant flue gas. The flue gas components were found to have significant effects

on Hg removal efficiency in a fixed bed study. HCI and SO2 promoted Hg oxidation and capture,

while H20 and NO inhibited Hg removal and the effect of NO2 WaS not significant. Experiments

of Hg removal in simulated flue gases showed that high rank coals are preferable to low rank

coals because of the lower moisture and higher HCI and SO2 COncentrations in the flue gas. It is

essential, however, to minimize the adverse effect of NO to improve the catalytic performance of

the SiO2-TiO2 HRHOCOmposite.












Set 1 (baseline) 8
Set 2 0, 4, 12, 16
Set 3 8 10, 30, 50
Set 4 8 -400, 800, 1200
Set 5 8 --50, 100, 300
Set 6 8 --- 10, 20
Set 7 (FGl) 8 30 1200 300
Set 8 (FG2) 12 10 400 300
Note: All the conditions contained 4% 02, 12% CO2, 75~80 Clg m Hgo (inlet), and balanced
with N2; the temperature was controlled at approximately 135 oC.


Table 4-1. Experimental conditions for investigation of the flue gas effects
H20 (%) HCI (ppm) SO2 (ppm) NO (ppm) NO2 (ppm)


,, 30
10
10













Aluminum SiO2-TiO2
Wall U-tube Pellets UVLm


Figure 4-1. Photocatalytic reaction system under flue gas conditions.









120

A B, C A: Bypass Reactor -*- ~Hg(T)I
S100 B: Pass Reactor, wlo UV -n- Hg(0)
C: Pass Reactor, wl UV A- Hg(II)I

Sso
o
60


O 40






0 10 20 30 40 50 60 70 80 90 100 110
Time (min)


Figure 4-2. Hg speciation at the outlet of the reactor in the baseline test












,> 100 b) 100
90 OHg captured 90 OHg Captured
80 H~~~Hg oxidized 80 HgOiie
S70- 70
S60 60

S40- 1 40
S30- 30

10 10
H20<0.1% H20=4% H20=8% H20=12% H20=16% Baseline HCI=10ppm HCI=30ppm HCI=50ppm



C) 100 d) 100
90 01 IHg Captured 90 II o OHg Captured
80 W~ xdzd80 -1Hog Oxidized
.N 70 .N 70
60 60

5 0 5 0
40 40


Baseline SO2=400ppm SO2=800ppm SO2=1200ppm Baseline NO=50ppm NO=100ppm NO=300ppm


e)100 100
90 OHg Captured 90 OHg Captured
80 H~ xdzd80 -1Hog Oxidized
.N70 .N70
S60 60


il la
Baeln NO2 0pp NO=0p NO=3pp Baeln r=u~s =Iue Ias2




ofa)ein H20 b H1,pp c)O2, d)p NO, e) ND2, and f) simulated flue gases2









CHAPTER 5
DEVELOPMENT OF SILICA/VANADIA/TITANIA COMPOSITES FOR REMOVAL OF
ELEMENTAL MERCURY FROM FLUE GAS

Background

As indicated in Chapter 4, the performance of the SiO2-TiO2 HRHOCOmposite is

significantly affected by the flue gas composition. NO, particularly, has a dramatically inhibitory

effect on Hg oxidation. To minimize the adverse effect of NO, there are two feasible ways. One

way is to remove NO from the flue gas before passing the gas through the photocatalytic bed.

Actually, in industrial practices, NO removal from the flue gas is usually achieved by selective

catalytic reduction (SCR) of NO with ammonia (NH3) (Parvulescu et al., 1998). However, the

results in Chapter 4 show that even at a relatively low concentration (i.e. 50 ppm), NO still

greatly inhibits Hg oxidation and capture by the SiO2-TiO2 HRHOCOmposite. This undesired

impact implies that unless NO can be completely or nearly completely removed from the flue

gas, the adverse effect of NO is inevitable and a larger amount of the catalyst must be used to

compensate the effect. Apparently, this is not a cost-effective method.

The other way to enhance the catalytic performance in Hg oxidation is to modify the

composition of the catalyst or even develop a new catalyst that is more effective under the flue

gas conditions. Recently, it has been reported that the SCR catalyst is capable of oxidizing Hgo in

addition to its ability of removing NO (Benson et al., 2005; Lee, Srivastava et al., 2004; Lee et

al., 2006; Niksa and Fujiwara, 2005b; Senior, 2006). The extent of Hgo oxidation through SCR

processes varies under different operating conditions burning different types of coal. It was

reported that Cl species in the flue gas promote the Hgo oxidation across SCR (Lee et al., 2006;

Senior, 2006), although the exact mechanism is still poorly understood. In contrast, the

mechanisms of the reduction of NO with NH3 by various SCR catalysts have been extensively

investigated (Parvulescu et al., 1998; Weckhuysen and Keller, 2003). Bosch and Janssen (1988)










reported a broad survey of metal oxide catalysts active for the reduction of NOx with NH3, and

they indicated vanadium oxide (V20s) to be the most active and selective catalyst. The active

sites on industrial SCR catalysts are V20s species supported on TiO2 (Busca et al., 1998). The

nature of the support, on the other hand, is also an important factor for the catalytic activity

(Weckhuysen and Keller, 2003). TiO2 aS a Support normally has drawbacks such as low surface

area and low resistance to sintering, and thus, a common practice is to use a silica support coated

with TiO2 (Kobayashi et al., 2005; Martra et al., 2000; Tesser et al., 2004). In addition, Shikada

et al. (1981) reported that the order of activity of NO reduction for supported V20s is SiO2-TiO2

> y-Al203 > SiO2. The SiO2-TiO2 COmposite used in this research for Hg removal has a high

surface area and is in line with the composition of the high-activity support reported by Shikada

et al. (1981). Hence, it is speculated that an addition of V20s to the SiO2-TiO2 COmposite could

further enhance the capability of oxidizing Hg.

In this chapter, V20s was doped onto the SiO2-TiO2 matrix to form a SiO2-TiO2-V205

composite. The catalytic ability of the new material in Hg removal was compared to that of the

SiO2-TiO2 COmposite. In addition, in order to investigate whether V20s is the active species for

Hg oxidation, SiO2-V20s composite was also synthesized and tested for its catalytic

performance. Since SiO2 is generally considered as an inert support, the hypothesis is that the

activity of the SiO2-V20s composite would represent the nature of the doped V20s species.

Therefore, the obj ectives of this chapter are: (1) to synthesize and characterize the SiO2-V20s and

SiO2-TiO2-V205 composites; (2) to test the catalytic abilities of those composites in Hg removal

in a fix-bed reactor; and (3) to investigate the reaction mechanisms of the catalytic removal of

Hg.









Materials and Methods


Catalyst Preparation

The details of the procedure of synthesizing the SiO2-TiO2 COmposite are described in

Chapter 2. When synthesizing the SiO2-TiO2-V205 composite, the same procedure was used but

with one more step of adding V205. Vanadium triisopropoxide oxide (VTPO) (Alfa Aesar) was

used as the precursor of V205. A known amount of VTPO was first dissolved in well stirred

ethanol to form an orange-brown solution. It was then added dropwise to the prepared silica sol

under vigorous stirring. TiO2 nanoparticles (P25 Degussa) were finally added to the mixture

before it started to gel. When synthesizing the SiO2-V20s composite, the step of adding TiO2

nanoparticles was skipped. In this study, the synthesized Sio2-V20s composite had a weight

fraction of V20s ranging from 2% to 10%. The synthesized Sio2-TiO2-V205 composite had 12

wt.% of TiO2 and varied contents of V205. The composites were originally made in the pellet

form (3 mm in diameter and 5 mm in length). Powder form of the composites was also obtained

by grinding the pellets and sieving through meshes. Both the pellets and powders of the

composites were tested in a fix-bed study. The powders used have a mesh size of 40x 100 (425

x 150 Clm). The names of the catalysts were abbreviated by way of STxVy, where S represents

SiO2, T represents TiO2, V represents V20s, and x and y represent the weight percentages of the

TiO2 and V20s, respectively.

Catalyst Characterization Techniques

The BET (Brunauer, Emmett, and Teller equation) surface areas of the catalysts were

measured using a Quantachrome NOVA 1200 Gas Sorption Analyzer (Boynton Beach, FL). The

powder samples were outgassed at 180 OC for 3 hours before the analysis. X-ray diffraction

(XRD) patterns of the powders were recorded with a Philips APD 3720 diffractometer using Cu-

Ku radiation (h = 0. 1542 nm) in the range of 15 to 400 (26) with a step size of 0.020









Catalyst Activity Measurement

As reported in Chapter 4, the SiO2-TiO2 COmposite needs activation by UV light for

catalytic oxidation of Hg. Thus, the necessity of UV light activation for V20s based composites

was first investigated. Pellet form of the catalysts was then used because the space between the

pellets allows penetration of UV light so that a maximum exposure of the catalyst to the light can

be achieved. In contrast, for a catalyst bed densely packed with powders, it is difficult for the UV

light to reach the central part of the bed. However, if tests indicate that the UV light is

unnecessary, powders are preferable to pellets because of better contact of the gas with the inner

pore surfaces of the catalyst.

Table 5-1 lists the experimental parameters for measuring the photocatalytic activities

using pellet catalysts. The three catalysts (SiO2-TiO2, SiO2-V205, and Sio2-TiO2-V205) were

tested following the same procedure as described in Chapter 4 using the reactor system

previously illustrated in Figure 4-1. Experiments were conducted under two flue gas conditions,

FG1 and FG2, as listed in Table 4-1 in Chapter 4. FG1 has relatively higher concentrations of

HCI and SO2 and lower concentration of water vapor, which represents flue gas burning high

rank coals. In contrast, FG2 represents flue gas burning low rank coals.

Next, the powder catalysts were tested with a modified reactor as shown in Figure 5-1. The

experimental parameters are listed in Table 5-2. No UV irradiation was supplied in this modified

system. The U-tube quartz reactor (13 mm ID) was immersed in an oil bath heated by a hotplate

to a constant temperature at 135 OC (+ 0.5 OC). The catalyst powders were packed in between

glass wools in the reactor. The flue gas (FG3) in this series of experiments contained 4% 02,

12% CO2, 8% H20, 10 ppm HC1, 400 ppm SO2, 300 ppm NO, 10 ppm NO2, and balanced with

N2. The inlet Hg concentration was maintained in a relatively constant range of 15.0~16.5 ppb.

The total flow rate was controlled at 1.5 L/min. Finally, to explore the reaction mechanisms and









the roles of the flue gas components in the catalytic reactions, the catalytic activity of a fixed

amount of catalyst (SV5) was examined with introduction of individual flue gas components.

Results and Discussion

Characterization of the Catalysts

The BET specific surface areas of the catalysts are listed in Table 5-3. All the catalysts

exhibit high surface areas (> 250 m2/g). Without any doping, the pure silica gel had the highest

surface area, 341.8 m2/g. The inclusion of 12% TiO2 to the silica gel (ST12) slightly reduced the

surface area to 3 19.4 m2/g. The doping of V205 (2 10%) to the silica gel moderately reduced

the surface area, but all the SiO2-V205 catalysts had the similar level of surface areas and did not

exhibit any clear dependence on the V205 loading. The surface areas of the SiO2-TiO2-V205

catalysts were close to those of the SiO2-V205 catalysts.

The XRD patterns of the catalysts are shown in Figure 5-2. No visible crystal phase of

V205 (peak at 26 = 26. 10) was detected for SV2 and SV5, which indicated that the vanadium

contents were highly dispersed on these catalysts (Kobayashi et al., 2006). A very small peak of

crystalline V205 was detected for SV8, and SV10 showed a relatively broader and more

prominent peak of crystalline V205. The molecular structures of the vanadium oxides at different

surface loadings have been widely reported in literature (Parvulescu et al., 1998; Rodella et al.,

2001; Weckhuysen and Keller, 2003). At low surface vanadia concentrations mainly monomeric

vanadyl (V4+) species are formed containing one terminal V=0 bond and three bridging V-O-

support bonds. As the vanadia loading increases, the monomeric species react to form polymeric

vanadates (V +) which consist of a terminal V=0 bond with one V-O-support and two bridging

V-O-V bonds. The presence of these monomeric and polymeric vanadium oxide species has

been identified by Raman and/or Infrared (IR) spectroscopy. As the vanadia loading further

increases, a fraction of the vanadia aggregates to form amorphous and crystalline V205 clusters.









The XRD results in this study indicated that crystalline V205 begins to grow as the vanadia

loading increases to somewhere between 5 and 8%.

The XRD pattern of ST12 showed a strong anatase phase (peaks at 26 = 25.30 and 38.00)

and a weak rutile phase (peak at 26 = 27.60) of TiO2. This agrees with the composition of the

TiO2 nanoparticles (P25 Degussa AG) which is an 80/20 mixture of anatase/rutile phases. For the

catalyst of ST12V5, no crystal phase of V205 was detected because of the relatively low vanadia

loading. It has been reported that TiO2-anatase is metastable and tends to convert to the

thermodynamically stable form rutile and that V205 favors this anantase-to-rutile phase

transformation (Busca et al., 1998). This may explain the finding that the peak of anatase TiO2 in

ST12V5 was lower than in ST12. However, no increase in the rutile phase in ST12V5 was

observed.

Mercury Removal Using Pellet Catalysts

Mercury removal tests were first conducted using pellet catalysts to investigate their

photocatalytic activities. The results are summarized in Figure 5-3. It was found that the catalytic

abilities of SV2 and ST12V2 on Hg removal were almost the same with or without UV

irradiation. Thus, only the results without UV are demonstrated in Figure 5-3 for SV2 and

ST12V2. In contrast, ST12 had to be activated under UV irradiation and without UV there was

negligible removal of Hg. Apparently, the addition of V205 is advantageous. It simplifies the

system by eliminating the UV devices and reduces the cost by saving the energy of UV

irradiation.

Figure 5-3 also compares the behavior of the catalysts under different flue gas conditions

and mass loadings. As described in Chapter 4, the oxidized Hg is calculated as the difference

between inlet and outlet Hgo concentrations, and the captured Hg is calculated as the difference









between the inlet and outlet HgT COncentrations. In flue gas 1 (FGl), 8 g of ST12V2

demonstrated a very high efficiency of Hg capture/oxidation (~ 92%), compared to only about

50% capture/oxidation of cg by 8g of ST12. This increase in Hg removal is obviously caused by

the addition of the 2% V205. When reducing the mass loading of the ST12V2 from 8 g to 4 g, the

efficiency of Hg oxidation decreased a little to 73% but that of Hg capture dramatically

decreased to 27%. The reduction in Hg capturing ability might result from the decrease in total

available surface area due to lower mass loading. It is interesting to find that compared to 4 g of

SV2, 4 g of ST12V2 had a higher Hg oxidation efficiency but a lower Hg capturing efficiency.

The enhanced oxidizing ability of ST12V2 is in line with the literature where the SiO2-TiO2

supported V205 has a higher activity than SiO2 supported V205 for reduction of NO (Shikada et

al., 1981). It has also been reported that the V-O-support bonds are the most critical structures

for the catalytic oxidation of methanol to formaldehyde (Weckhuysen and Keller, 2003). Thus, it

is possible that the V-O-Ti bonds have superior activities than the V-O-Si bonds, which

resulted in a higher oxidation of Hg on the SiO2-TiO2-V205 composite than on the SiO2-V205

composite. The reason for the penetration of oxidized Hg from ST12V2 may be related to the

selectivity of the reactions, i.e. ST12V2 may have a higher capacity of converting Hgo to volatile

oxidized Hg species such as Hg(NO3)2. This agrees with the suggestions by other researchers

that the V-O-support bonds are determinant for activity and selectivity of different reactions

(Tesser et al., 2004; Weckhuysen and Keller, 2003). More discussion relevant to the formation of

Hg(NO3)2 is provided later in this chapter.

In flue gas 2 (FG2), the behavior of the catalysts followed a similar pattern as in FGl. For

4 g of ST12V2 and 4g of SV2, no obvious change in the Hg removal efficiency was observed in

FG2 compared to in FGl. For the case of 8 g ST12V2, the oxidized Hg was at the same level as









in FG1 but the captured Hg was much lower than that in FGl. The difference may be due to the

higher water vapor concentration in FG2 which competes with oxidized Hg for the adsorption

sites. It is also possible that a larger fraction of certain volatile Hg compounds (such as

Hg(NO3)2) were produced in FG2. However, identification of the oxidized Hg species in the

current system is technically difficult due to their trace amounts.

Mercury Removal Using Powder Catalysts

The previous experiments using pellet catalysts showed no necessity of UV light activation

for the SiO2-V20s and Sio2-TiO2-V205 catalysts. Hence, no UV light was used for the study of

powder catalysts. Tests were first performed to verify that the glass wools (used as the support of

powders) and pure silica powders were inert to Hg removal. Then, with a feed of approximately

16 ppb Hg, each flue gas component (4% 02, 8% H20, 50 ppm HC1, 1200 ppm SO2, 300 ppm

NO, or 30 ppm NO2) balanced with N2 WaS individually introduced to the system. In all cases,

the change of the Hg concentration was observed to be within 15%. Since the Hg permeation

tube has an error of 2%, the result indicates that homogeneous oxidation of Hg in the gas-phase

was negligible.

The first set of experiments was carried out using 500 mg of SiO2-V20s catalysts

(corresponding to a bed height of 17 mm in average) but with different V20s loadings (2 10

wt.%). The results are shown in Figure 5-4. In the beginning, the flue gas bypassed the reactor to

obtain the inlet Hg concentration. It then passed through the reactor and the extent of the

catalytic oxidation was recorded. After a 6-hr period, the flue gas bypassed the reactor again and

the inlet Hg concentration was checked. It was observed that the Hg removal efficiency

increased as the V20s loading increased from 2 to 8%. For SV2 (Figure 5-4a) and SV5 (Figure 5-

4b), the outlet concentration of HgT and Hgo were at the same level, indicating that all the

oxidized Hg has been captured. For SV2, the Hg removal efficiency was initially around 22%









and slowly decreased to 12% in 6 hours. For SV5, the Hg removal efficiency was initially around

65% and slowly decreased to 45% in 6 hours. For SV8, as shown in Figure 5-4c, the outlet HgT

concentration was initially dropped to a very low level (corresponding to 93% capture), and then

it increased and was relatively stabilized at around 32% of the inlet Hg concentration (i.e. 68%

capture). The outlet Hgo concentration was found to be always lower than the HgT COncentration,

indicating that a portion of the oxidized Hg penetrated the reactor. The Hg oxidation efficiency at

the end of the 6-hr test was around 77%, higher than that of SV2 and SV5. For SV10, as shown

in Figure 5-4d, a different pattern of Hg removal was observed. The outlet HgT COncentration

slowly dropped to 40% of the inlet level in the 6-hr period, unlike the previous cases where the

outlet HgT COncentrations dropped to their minimum levels immediately after the flue gas passed

through the reactor. The outlet Hgo concentration was relatively stable at 30% of the inlet level

during the last 3 hours. Both the Hg capturing and oxidation efficiencies of SV10 were lower

than those of SV8.

Next, 500 mg of ST12V5 was tested for its catalytic activity and the result is shown in

Figure 5-4e. The outlet Hgo concentration remained almost constantly low at 15% of the inlet

level (i.e. 85% oxidation). The outlet HgT COncentration initially dropped to 11% of the inlet

level but quickly increased and stabilized around 65% (i.e. 35% capture). Similar to the findings

observed in the study of pellet catalysts (Figure 5-3), ST12V5 powder has a higher oxidation but

lower capturing efficiency compared to SV5.

The amounts of Hg captured/oxidized on the various catalysts in the 6-hr test are

summarized in Table 5-4. To better compare the effectiveness of the catalysts, the capabilities of

the catalysts were normalized to unit mass of catalyst and unit mass of V205 (the active phase),

respectively. Table 5-4 showed that the amount of Hg captured per gram of the SiO2-V205









catalyst increased as the V20s loading increased from 2 to 8% but decreased as the V20s loading

further increased to 10%. When normalized to per gram of V20s, the capacity of SV5 was the

highest; either lower (2%) or higher (8% and 10%) V20s loadings reduced the capacity. These

results suggested that the optimal V20s loading for an optimal catalytic activity is somewhere

between 5 and 8%. This is coincident with the XRD result that the maximum loading of V20s

without the formation of crystalline V20s is somewhere between 5 and 8%. It should be noted

that ST12V5 exhibited an even greater ability of oxidizing Hg than SV5, although its Hg

capturing ability was lower than SV5. Applications of ST12V5 can be beneficial to coal-fired

power plants equipped with wet-scrubbers where oxidized Hg can be easily captured. Therefore,

ST12V5 surpasses SV5 in terms of the total potential for Hg removal.

As shown in Figure 5-4, the Hg removal efficiencies slowly decreased in the 6-hr test for

SV2, SV5, and SV8. To investigate whether a 100% breakthrough would happen, a smaller

amount (250 mg) of SV5 was subj ect to a test with a longer period of time, as shown in Figure 5-

5. The outlet HgT (or Hgo) concentration initially dropped to 65% of the inlet level when the flue

gas passed the bed. Then the Hg concentration slowly increased to 80% in a 9-hr period.

Thereafter, the Hg removal efficiency remained relatively stable at approximately 20% for

another 3 hr (in Day 1). With the same batch of catalyst under the same flue gas conditions, the

experiment was continued in Day 2 after a pause of 40 hours. Only the HgT COncentration was

recorded in Day 2 since Hgo was found to be at the same level of HgT in Day 1. This time when

the flue gas passed through the bed, the HgT COncentration first dropped to 68%, very close to the

lowest level of HgT COncentration (65%) in Day 1. Then the HgT COncentration increased to 81%

in 2 hr and remained at this level for another 4 hr. This 20-hr test suggested that a steady-state

catalytic activity was reached (approximately 20% Hg removal) using either fresh (Day 1) or









used (Day 2) catalyst. This transition to a steady-state activity is in agreement with the reported

kinetics of oxidative dehydrogenation of propane over V205/TiO2 in literature (Grabowski et al.,

2002; Sloczynski, 1996). Those researchers observed that the rate of reduction is limited by the

adsorption of reductant at the surface of the vanadium phase and that reoxidation of the reduced

catalysts proceeds via two stages: (i) a quick surface reaction, and (ii) a slow process limited by

bulk diffusion through the growing coat of the oxidation product. They also indicated that the

surface reoxidation rate is considerably higher than that of the reduction. According to these

theories and Eindings, the rate of Hg oxidation in this work should quickly reach a maximum

point at the beginning and then gradually decrease due to the reduced number of available

vanadium sites and/or the increasing difficulty of 02 diffusing through the coat of oxidized Hg.

However, this decrease in the catalytic activity proceeded so slowly that the catalyst behaved like

reaching a steady-state activity for a relatively long time (Grabowski et al., 2002). This agrees

with the result that the captured Hg on the 250 mg SV5 during the 20-hr test was calculated to

account for only 0.2 mol % of the total vanadium sites. Future study is needed to investigate the

long-term performance of the catalyst.

Mercury Removal Mechanisms

To explore the Hg removal mechanisms on the SiO2-V205 catalyst, it is important to

understand the role of the flue gas components in the catalytic reactions. Hence, experiments

were conducted by mixing Hg with individual flue gas components (e.g. HC1, SO2, and NOx)

and/or in combination with 02. The role of water vapor was also examined. 250 mg of fresh SV5

was used in each test since 5% V205 was previously found to be close to the optimal loading.

Role of 02

The role of 02 on Hg removal was first investigated because Ol is an important oxidant in

flue gas. A background test was conducted using high purity N2 (>99.995%, Airgas) as the









carrier gas and around 5% of Hg removal was detected, as shown in Figure 5-6. Considering that

the carrier gas contained a maximum of 50 ppm impurity very likely consisting of Ol, the 5% Hg

removal might be contributed by the trace amount of Ol. When 4% 02 WaS introduced, the Hg

removal increased to about 15%. Hg removal further increased to 26% as 02 inCTreSed to 20%.

Granite et al. (2000) studied various metal oxides for catalytic Hg removal from flue gas and

they proposed that lattice oxygen of the metal oxides can serve as the oxidant of Hg, forming

mercuric oxide (HgO). It has also been reported in literature that lattice oxygen is the most

abundant reactive intermediates that are responsible for oxidative dehydrogenation of alkanes

over V205 based catalysts (Argyle et al., 2004; Grabowski et al., 2002). Gas-phase Ol, on the

other hand, reoxidizes the reduced metal oxides, replenishing the lattice oxygen (Grabowski et

al., 2002; Granite et al., 2000). In this work, a similar redox cycle is proposed for the catalytic

oxidation of Hg on SiO2-V205 in the presence of 02:

V205 + Hg a V204 + HgO (5-1)

V204 +202 4 V205 (5-2)

The overall reaction can be written as:

Hg + 4 02 4 HgO (5-3)

02 playS an important role in the redox mechanism, which is supported by finding that Hg

oxidation on the catalyst is proportional to the concentration of 02.

Role of HCI

HCI was found to enhance Hg oxidation over the SiO2-V205 catalyst (Figure 5-6). 10 ppm

HCI (balanced with N2) resulted in 15% Hg removal, and as HCI increased to 50 ppm, the Hg

removal increased to 25%. The combination of 50 ppm HCI with 20% 02 further improved the

Hg removal to 39%. It has been reported in literature that Hgo is not adsorbed (or is only weakly









adsorbed) on the surface of unburned carbon and SCR catalysts (Niksa and Fujiwara, 2005a,

2005b). The researchers also proposed that Hg oxidation occurs via an Eley-Rideal mechanism,

where adsorbed HCI reacts with gas-phase (or weakly adsorbed) Hg; however, the specific

reaction pathway was not given. In this work, it was also found that Hg is not (or only weakly)

adsorbed on the SiO2-V20s catalyst, which is verified by the result that there was only 5%

removal of Hg in the environment of pure N2. The Eley-Rideal mechanism can be used as well to

explain the reaction on the SiO2-V20s catalyst in the presence of HC1. The pathway of HCI

adsorption can be inferred from literature about HCI adsorption on other types of metal oxide

surfaces. Parfitt et al. (1971) conducted an infra-red study of HCI adsorption on rutile surface

and found an increase in surface hydroxyl (OH) groups due to the introduction of HC1. The OH

groups further react with excess HCI to form Cl and water. Tseng et al. (2003) reported the

adsorption of HCI on the CuO surface forming a hydroxychloride (Cu(OH)C1) intermediate.

Similar to the mechanisms reported in those studies, it is proposed that the formation of adsorbed

Cl on the V20s surface occurs via:

V O- V+HCls~V OH V- C (5-4)

V OH + HCI ++ V Cl + H,O (5-5)

Actually, the V-OH structures are one type of the active sites readily present on the surface of

the vanadia based catalysts (Busca et al., 1998; Kantcheva et al., 1994; Parvulescu et al., 1998).

Thus, the reaction with HCI can directly start from Reaction 5-5. The adsorbed Cl species then

react with gas-phase Hg through the Eley-Rideal mechanism to generate an intermediate HgCl

species, which then further reacts with HCI or Cl to form a more stable mercuric chloride, HgCl2.

The overall reaction can be written as:

O, + 2HC + Hg~ go HgC + H,O(56









It should be noted that chlorination of Hg may take place without the presence of Ol, aS Shown in

Figure 5-6. In this case, V20s is consumed to form V204. The addition of 20% 02 to 50 ppm HCI

enhanced the total oxidation of Hg, very likely due to the oxidation of V204 to V20s, i.e. the

regeneration of the catalyst.

Role of NO2

The effect of NOx was also found to be promotional in Hg oxidation over the SiO2-V205

catalyst (Figure 5-6). In the presence of 10 ppm NO2, 57% of Hg was oxidized while 50% was

captured. It should be noted that the cylinder gas of NO2 WaS balanced with both N2 and 02 with

02 COncentration three times as much as NO2. Thus, 30 ppm 02 WaS present in the gas together

with 10 ppm NO2. When 20% 02 WaS added, the Hg oxidation remained at a similar level.

Increasing the concentration of NO2 to 30 ppm (with 20% 02) inCTreSed the Hg oxidation to 68%

and Hg capture to 57%. The results indicated that NO2 greatly promoted Hg oxidation with or

without Ol. It has been reported in literature that NO2 significantly improves heterogeneous

oxidation of Hg on fly ash (Norton et al., 2003) and on activated carbon based sorbents (Miller et

al., 2000). Mercuric nitrate (Hg(NO3)2) WaS suggested to be the reaction product initiated by NO2

(Laumb et al., 2004; Miller et al., 2000). Other researchers have reported that adsorption of NO2

on TiO2 supported V20s catalysts was the first step in the process of selective catalytic reduction

of NOx (Kantcheva et al., 1994; Parvulescu et al., 1998). Kantcheva et al. (1994) indicated two

pathways for the NO2 adsorption on V20s involving V=0 and V-OH sites:

V5' + N' 24 V N'0 (5-7)

2V5' WH+3NO2 42V' -NO3 +H2 + NO (5-8)

In this work, similar mechanism can be applied to explain the oxidation of Hg in the

presence ofNO2. NO2 is fifSt adsorbed on V20s via Reactions 5-7 and 5-8 and then transformed









to adsorbed nitrate species, which react with gaseous Hg to form Hg(NO3)2 via the Eley-Rideal

mechanism. The overall reaction can be written as:

02 + 2 NO2 + Hg a Hg(NO3 )2 (5 -9)

Considering the low melting point of Hg(NO3)2, 79 OC (Weast et al., 1983), it is likely that the

formed Hg(NO3)2 is VOlatile at the reactor temperature (135 OC) and thus part of it may be

released from the reactor in the gas-phase, as shown in Figure 5-6. This formation of volatile

Hg(NO3)2 initiated by NO2 is in agreement with the findings by other researchers. Using a

carbon-based sorbent to remove Hgo, Miller et al. (2000) observed nearly 100% breakthrough of

a volatile oxidized Hg species in a gas mixture of SO2 and NO2. In a foll0w-up study conducted

by Olson et al. (2002) using both MnO2 and carbon sorbents, they identified this volatile Hg

species to be Hg(NO3)2 by trapping the effluent in cold acetonitrile followed by analysis using

gas chromatography mass spectroscopy (GC-MS). Considering the much higher

melting/decomposing point of HgCl2 (277 OC) and HgO (500 oC) (Weast et al., 1983), the

penetration of these two less volatile Hg species is less likely or occurs to a very small extent, as

shown in Figure 5-6.

Role of NO

As also shown in Figure 5-6, in the presence of 300 ppm NO, there was 29% removal of

Hg and an addition of 20% 02 further increased the removal to 48%. The oxidation and capture

of Hg increased with the increase of NO concentration. Decreasing the NO concentration to 100

ppm moderately decreased the Hg removal efficiency. The Hg removal may be ascribed to the

adsorbed NO species on the V20s surface. It is generally agreed that NO adsorbs as nitrosyl and

dinitrosyl surface species on reduced vanadia surfaces, whereas it does not adsorb over fully

oxidized surfaces (Busca et al., 1998). The adsorbed NO can be oxidized on the surface, giving









rise to species like NO NO2, and NO3-, or it can be reduced by reduced catalyst centers (Busca

et al., 1998). It is likely that these adsorbed species derived from NO are responsible for the

observed Hg oxidation, but the exact reaction pathways and products are unknown. It has been

reported that gas-phase NO can first be oxidized by 02 and then is adsorbed as NO2 on the SCR

catalyst (Parvulescu et al., 1998). This route is less likely to occur on the SiO2-V20s catalyst in

this work, because there was no penetration of the oxidized Hg through the reactor in the

presence of NO, unlike that observed in the presence of NO2.

Role of SO2

Figure 5-6 shows that the effect of SO2 on Hg removal was insignificant in the

concentration range of 400~1200 ppm. When combining 20% 02 with 400 ppm SO2, the Hg

oxidation/capture efficiency was very close to or slightly lower than that without SO2. It has been

reported that SO2 COmpetes with HCI for sites on activated carbon and fly ash sorbents and thus

inhibits Hg oxidation and adsorption in flue gas (Laudal et al., 2000; Laumb et al., 2004). Other

studies, however, reported that SO2 can promote heterogeneous oxidation of Hg over fly ash

sorbents (Norton et al., 2003) or SCR catalysts (Eswaran and Stenger, 2005). Hence, further

investigations involving the combination of SO2 with other flue gas components (e.g. HCI and

NOx) would warrant a more comprehensive understanding of the role of SO2 for Hg removal on

the SiO2-V20s catalyst.

Role of H20

It has been well discussed in Chapters 2 to 4 that a high concentration of water vapor can

inhibit Hg oxidation or capture on SiO2-TiO2 in TOOm or flue gas conditions. In this work, H20

was also found to have a dramatic inhibitory effect on Hg removal over SiO2-V20s (Figure 5-7).

Under flue gas conditions (FG3) using 250 mg SV5, when the gas was switched from humid (8%

H20) to dry, the Hg removal efficiency increased from 20% to 66%. A H20 concentration as low









as 0.6% (6000 ppm) also exhibited inhibitory effects. In a test with 10 ppm HCI and 20% 02,

switching the gas from dry to a 0.6% H120 caused a decrease in Hg capture from 30% to 19%.

Similarly, in another test with 10 ppm NO2 and 20% 02, the introduction of 0.6% H120 caused a

decrease in Hg capture from 43% to 24%. The competitive adsorption of water vapor on the

active sites may have prohibited the adsorption of reactive species such as HCI and NOx. The

Endings agree with the literature where Laumb et al. (2004) reported that the absence of water

vapor increased the adsorption of Cl compounds on activated carbon in flue gas condition.

Summary

In this chapter, the active phase of SCR catalysts, V20s, was doped on the SiO2 and Sio2-

TiO2 supports. Compared to SiO2-TiO2 COmposites, improvements in Hg removal from flue gas

were observed in Eixed-bed studies using both pellet and powder forms of the SiO2-V20s and

SiO2-TiO2-V205 catalysts. No UV light activation is needed for the vanadia based catalysts,

which simplifies the system and reduces the operating cost. For SiO2-V20s catalysts, the Hg

removal efficiency increased as the V20s loading increased from 2 to 8% but decreased as the

V20s loading further increased to 10%. The results suggested that the optimal V20s loading for

an optimal catalytic activity is somewhere between 5 and 8%. The SV5 catalyst reached a

steady-state activity during a 20-hr test and no deactivation of the catalyst was observed. The

SiO2-TiO2-V205 catalysts have an even greater ability of oxidizing Hg compared to SiO2-V205,

which can be advantageous to power plants equipped with wet-scrubbers where oxidized Hg can

be easily captured. The different supports of V20s may account for the difference in their

catalytic activity. The mechanisms of Hg oxidation on the SiO2-V20s catalyst have also been

investigated. It was found that the Hg oxidation may follow an Eley-Rideal mechanism where

HC1, NO, and NO2 are first adsorbed on the catalyst and then react with gas-phase Hg. While

HC1, NO, and NO2 promote Hg oxidation, SO2 has an insignificant effect and water vapor









dramatically inhibits Hg oxidation. Penetration of oxidized Hg was observed in the presence of

NO2, which is probably due to the formation of volatile Hg(NO3)2. It was verified in this work

that V205 is the active center for the Hg oxidation, but further studies are needed to better

understand the enhancement of Hg oxidation by the SiO2-TiO2-V205 catalysts.









Table 5-1. Experimental parameters for activity measurement of the catalysts in pellet form
Catalyst Composition Mass (g) Carrier gas UV light
ST12 12% TiO2 + 88% SiO2 8.0 FGl, FG2 On/Off
ST12V2 12% TiO2 + 2% V20s + 86% SiO2 4.0, 8.0 FGl, FG2 On/Off
SV2 2% V20s + 98% SiO2 4.0 FGl, FG2 On/Off



Table 5-2. Experimental parameters for activity measurement of the catalysts in powder form
Catalyst Composition Mass (mg) Carrier gas
SV2 2% V20s + 98% SiO2 500 FG3
SV5 5% V20s + 95% SiO2 500 FG3; individual gases
SV8 8% V20s + 92% SiO2 500 FG3
SV10 10% V20s + 90% SiO2 500 FG3
ST12V5 12% TiO2 + 5% V20s + 83% SiO2 500 FG3



Table 5-3. BET surface areas of the catalysts
Sample BET specific surface area (m /g)
Silica gel 341.8
SV2 263.4
SV5 283.2
SV8 273.8
SV10 262.9
ST12 319.4
ST12V2 258.0
ST12V5 262.5




Table 5-4. Amounts of Hg captured and oxidized on the catalysts in a 6-hr test
Catalyst Hg Captured Hg Captured Hg Oxidized Hg Oxidized
(Clg/g-catalyst) (Clg/g-V20s) (Clg/g-catalyst) (Clg/g-V20s)
SV2 (500 mg) 22.5 1130 22.5 1130
SV5 (500 mg) 75.1 1500 75.1 1500
SV8 (500 mg) 83.7 1050 106.3 1320
SV10 (500 mg) 63.1 630 86.6 870
ST12V5 (500mg) 52.2 1040 117.4 2350












Catalyst


MFC HCI SO2 Vent
Water Bath


Figure 5-1. Experimental system for the fixed-bed study using powder catalysts












x Anatase TiO2
+ Rutile TiO2
o Crystalline V20s


(e)


(d)


(b)




15 20 25 30 35 40
2 9(degree)


Figure 5-2. XRD patterns of (a) SV2, (b) SV5, (c) SV8, (d) SV10, (e) ST12, and (f) ST12V5.










100
90
80
70
60
50
40
30
20
10
0


usin I III


wlo UV


wi uv


wio uv


wio uv


wi uv


Figure 5-3. Catalytic removal of Hg using the pellet catalysts under various conditions


WlO UV
5 Hg Captured
SHg Oxidized
wlo UV












1.2
(8 a Pass bed Bypass
1. +6

o .~,o+000+++~ooopso
50.8 **+0+0* 0.*044000+




S0.4
HgI(T)

i 0.2 o Hg(0)


0.0
0 50 100 150 200 250 300 350 400
Time (min)

1.2

(b) 6 Pass bed I H( JIBypass
S1.0 ooI
I-Y o Hg(0)


E 0.8



S0.6




0.0
0 50 100 150 200 250 300 350 400
Time (min)

1.2
-Pass bed Bypass
( 1.0 + Hg(T) oi
m~ oHg(0)



S0.6


~0.4
rn ** ** ***

~i0.2 ooo o 0 oo ooo
o o .o 0
0.0
0 50 100 150 200 250 300 350 400
Time (min)


Figure 5-4. Outlet Hg concentration as a function of time using 500 mg powder of (a) SV2, (b)
SV5, (c) SV8, (d) SV10, and (e) ST12V5.





Pass bed

*/ +Hg(T)
o Hg(0)




o 4
** **
ot oo
OOO OO


Bypass


*r, ***


0 50 100 150 200 250 300 350 400
Time (min)


1.2




o
Z 0.8



0.4

E 0.2


S0.0


Pass bed

so


Bypass

o4


* *
**


, *


oooo


0 0o


ooo ooo


0 50


100 150 200 250 300 350 400
Time (min)


Figure 5-4. Continued.


1.2


(d) "i.0


Z 0.8

2 .6
E 0.

S0.4


a,0.2
O
0.0


* Hg(T)

o Hg(0)





1.2









0.6 -


0.4


0.2


0 0


Bypass i Pass bed


Pass bed


0~0~0~


*~*~*


I
I
Day 1
I
I
1


I
I


Day 2


0 3 6 9 12 15 18 21
Time (hour)


Figure 5-5. Outlet Hg concentration as a function of time using 250 mg powder of SV5 (20-hr
test).


I Hg(T)

So Hg(0)









100%

|5 Hg~ Capturedl
-aI 80% I Hg Oxidized
.s NO2 NO SO2
2 02 HCI
O 60%


3 40%


I 20%


0%




Fig re -6.Th r ole of \ fue gas co p net on Hg re ova usin 2 0~ mg ~ SV o d r n e r
c~:onditons. \ eO;9\ (`.LO ;9~





100%


80%


60%


40%


20%


O% -C~
cS\
\O
13
(~C3


cs\ o\
~
\0;


cs\ o\
\O ~'
o~ ~
\0'


Figure 5-7. The role of water vapor on Hg removal using 250 mg SV5 powder .


H Hg Captured
H Hg Oxidized









CHAPTER 6
UV-AB SORPTION-BASED MEASUREMENTS OF OZONE AND MERCURY: AN
INVESTIGATION ON THEIR MUTUAL INTERFERENCE"

Background

The regulations on Hg emissions and the development of Hg control technologies require

that reliable methods be used for accurate Hg measurement. Currently, the EPA accepted

methods for Hg measurement in the United States are manual procedures based on wet-

chemistry such as EPA Methods 29 and 101A (for total mercury) and the Ontario Hydro Method

(for speciated mercury) (Laudal et al., 2004). However, continuous mercury monitors (CMMs)

have distinct advantages over these manual methods in that CMMs are able to provide a real-

time or near-real-time response for Hg measurements and to perform long-term emission

measurement to truly characterize a process' temporal emissions. On the other hand, a significant

disadvantage of CMMs lies in their measurement interference, which may vary depending on

the principle of the Hg detection technique.

Atomic absorption spectrometry (AAS) is one of the maj or techniques applied to current

CMMs. In the case of AAS, the concentration of elemental Hg in a gas sample is determined by

measuring the light that is absorbed by Hg atoms at their characteristic wavelengths (usually at

the resonance line of 254 nm). Thus, interference can occur when other components of the

sample gas possess strong absorption bands near this wavelength (254 nm). The EPA' s

Environmental Technology Verification (ETV) program (USEPA, 2001a) identified sulfur

dioxide (SO2), nitrogen oxides (NOx), hydrogen chloride (HC1) and chlorine (Cl2) as interference

gases by assessing the CMMs' responses to each gas as well as to a mixture of the gases.

Although pretreatment or conditioning systems can be used to remove or negate the effects of


* Reprinted with permission from Li, Y., Lee, S.-R., Wu, C.-Y., 2006. UV-Absorption-Based Measurements of
Ozone and Mercury: An Investigation of Their Mutual Interferences. Aerosol Air Qual. Res. 6, 418-429.









these interfering gases prior to the sample delivery to the detectors (Laudal et al., 2004), they

may increase the complexity and cost of the instrumentation and impair the real-time feature of

the CMMs. Consequently, many types of AAS based CMMs do not have pretreatment systems.

Since the 254 nm Hg emission line also falls into the absorption spectra of ozone, which is

capable of absorbing UV light below 290 nm, the presence of ozone in the sampling environment

may impact the Hg measurement by AAS based CMMs. Granite and Pennline (2002) studied

photochemical oxidation of Hg and speculated that photosensitized formation of ozone may

interfere with Hg measurement by absorbing UV radiation. However, no quantitative data were

reported on the magnitude of ozone interference. Therefore, the first obj ective of this study was

to quantitatively investigate the interference of ozone on Hg measurement. This study may be of

particular importance to ambient and indoor Hg measurement because ozone and Hg coexist in

these conditions.

Monitoring ground level of ozone, another significant air pollutant, is also required by

US EPA. Control of ozone is expensive, with costs estimated in the billions of dollars (USEPA,

2005b). Hence, deployment of accurate ozone measurement is of great importance to

demonstrate compliance with the National Ambient Air Quality Standard (NAAQS) for ozone.

Many methods have been developed for ozone measurement where UV absorption and gas-phase

chemiluminescence are the major techniques used nowadays. The method of UV absorption is

based on the principle that upon exposure to UV light ozone will absorb some of the light and the

intensity difference is directly proportional to the concentration of ozone. Frequently the UV

light source is a 254 nm emission line from a Hg discharge lamp. Known interference on this

type of ozone detection method include gaseous hydrocarbons with strong absorption at 254 nm,

such as aromatic hydrocarbons (i.e., benzene and substituted benzene rings) (NARSTO, 1999).









Since 254 nm is exactly one of the Hg absorption lines, it is speculated that even a small amount

of Hg in the sample gas may absorb a considerable amount of UV light. The U. S EPA (1999b)

reported that at a baseline ozone concentration of approximately 75 ppb, the addition of 0.04 ppb

(300 ng m-3 at room temperature) Hg caused an increase in measured ozone concentration by

12.8% at low humidity (RH = 20 ~ 30%) and 6.4% at high humidity (RH = 70 ~80%) using a

UV photometric ozone monitor. The interference of Hg using another two types of ozone

monitors were above 30% at either low or high humidity. However, the interference data were

reported only at one Hg level (0.04 ppb). More data at other levels of Hg are needed to determine

the relationship between Hg concentration and its corresponding interference. Therefore, the

second obj ective of this study was to quantitatively investigate Hg interference on ozone

measurement. This is of importance in accurate ozone measurement in ambient and indoor

conditions.

Methods

Descriptions of Hg and Ozone Instruments

The CMM used in this study is a RA-915+ Hg analyzer (OhioLumex Co.), which is

capable of recording Hg concentrations every second. It employs Zeeman AAS using High

Frequency Modulated light polarization (ZAAS-HFM) (Sholupov et al., 2004), which combines

the approach of AAS with a simultaneous background correction provided by the Zeeman

splitting of the Hg resonance line (254 nm). In the RA-915+ Hg analyzer, the emissions from a

Hg discharge lamp are subj ected to a strong magnetic field, which causes the three-fold splitting

of the Hg resonance line (x, o+ and o-, respectively). Two of these components (o+ and o-) have

identical intensity when Hg is absent in the analytical cell. When Hg is present in the cell, the

difference between the intensities of the two components is proportional to the Hg concentration.

The calibration was conducted by the manufacturer using Dynacal" permeation device (VICI









Metronics Inc.) which is certified traceable to NIST (National Institute of Standards and

Technology) standards. Two optical cells are available for different ranges of Hg concentration.

A single-path cell is available for measuring higher Hg concentrations from 0.5 to 200 Clg m-3. A

multi-path cell with an effective length of 9.6 m is used to enhance the sensitivity of analysis,

and thus the detection limit can reach as low as 2 ng m-3. The RA-915+ Hg analyzer may not be

suitable for ambient Hg measurement since the ambient Hg concentration has been reported to be

approximately 1.5 to 1.9 ng m-3 in the northern hemisphere (Ebinghaus et al., 2002; Weiss-

Penzias et al., 2003), which is lower than the detection limit of this instrument. However, the

RA-915+ Hg analyzer has been used for measurement in stationary sources or Hg contaminated

sites (including indoor areas) (Kinsey et al., 2004; Pogarev et al., 2002; Sholupov et al., 2004).

A M146 dynamic gas calibration system (Thermo Electron Instrument) served as the

ozone generating source using its internal ozonator. The precision of the ozone concentration that

can be generated is 1 ppb. A M49 UV photometric ozone analyzer (Thermo Electron Instrument)

was used to measure the ozone concentration in the sample gas. The UV light source in the

ozone analyzer is a 254 nm emission line from a Hg discharge lamp. The full scale of the ozone

analyzer was set to be from 0 to 500 ppb. Its precision is 2 ppb while the noise is within 1 ppb.

This type of ozone monitor is equipped with a standard (manganese-dioxide) scrubber, which

was reported to suffer the lowest interference from Hg compared to other two types of ozone

monitors (USEPA, 1999b). A M49-PS UV photometric ozone calibrator (Thermo Electron

Instrument) was applied to calibrate both the ozone generator and analyzer.

Experimental Setup and Procedures

A schematic diagram of the experimental system is shown in Figure 6-1. All the

experiments were carried out at room temperature (25 A loC). To test the ozone interference on

Hg measurement (Figure 6-1A), zero air was produced using a zero air supplier (M111, Thermo










Electron Instrument) and then passed through the ozone generator to provide designated ozone

concentrations. The total air flow rate at the outlet of the ozone generator was controlled to be 5

L/min. The ozone laden air was then divided into two streams and connected to the RA-915+ Hg

analyzer and the ozone analyzer respectively. The reading from the Hg analyzer would indicate

potential interference caused by ozone. The interference from a designated range of ozone

concentrations were measured. Each time before changing the ozone concentration, the entire

system was purged by ozone-free air until the readings from both Hg and ozone analyzers were

zero. This was to minimize the experimental error from residual ozone in the system. Finally, the

sample gas was cleaned through an activated carbon trap before exhausted to the vent hood.

Figure 6-1B shows the experimental setup for testing Hg interference on ozone

measurement. The incoming zero air with a flow rate of 5 L/min was split into two streams. One

stream was passed through the surface of a liquid Hg reservoir, which was placed in an ice-water

bath to maintain a constant Hg vapor pressure. By doing this, saturated Hg vapor was introduced

into the system. The other stream served as dilution air and was used to adjust the Hg

concentration. Both air streams were controlled by mass flow controllers (MFC, Model. FMA

5400/5500, Omega Engineering, Inc.). The RA-915+ Hg analyzer was used to measure the Hg

concentration in the sample gas while the ozone analyzer was used to detect the potential

interference caused by Hg. Similarly, the interference from a designated range of Hg

concentrations were measured, and each time before changing the Hg concentration, the entire

system was cleaned until the readings from both Hg and ozone analyzers were zero.

Results and Discussion

Interference of Ozone on Hg Measurement

The NAAQS for ozone is 80 ppb for an 8-hour average and 120 ppb for a 1-hour average

(USEPA, 1997b). Thus, the ozone concentration generated in this work ranged from 0 to 120









ppb. The corresponding interference on the RA-915+ Hg analyzer is shown in Figure 6-2. The

blank test showed that no interference was detected when no ozone was fed into the gas stream.

As the ozone concentration increased, the reading on the Hg analyzer was almost linearly

elevated. At each ozone concentration level, tests were repeated for three times. For ozone

concentrations at 80 and 120 ppb, the interference on Hg measurement reached approximately

46 and 63 ng m-3, respectively. The relationship between ozone concentration and its

corresponding interference can be approximated as:

Case = 0.5559 x Co, (6-1)

where Case is the equivalent Hg concentration, i.e., the interference on Hg analyzer (in unit of


ng m-3) and Co, is the ozone concentration (in unit of ppb). 1 ng m-3 of Hg is equivalent to 1.22 x

10-4 ppb at room temperature. For a convenient understanding of the significance of the

interference magnitudes, units used in Equation 6-1 and Equation 6-2 are in line with the EPA

standards (USEPA, 1997b, 1999a).

This observed ozone interference may have an important effect on Hg measurement. For

example, Ferrara (1999) used the RA-915+ Hg analyzer to measure the Hg distribution over the

area of Idria, where one of the largest European Hg mines was located. The Hg concentration

was reported to range from 50 to 170 ng m-3 in the central part of Idria near the Hg mines and

dumps. Suppose that 40 ppb ozone existed in that local atmosphere, an overestimate of Hg

concentration at about 22 ng m-3 WOuld have been involved according to the results from this

work. Indoor Hg measurement can also be affected by ozone interference. Although the indoor

ozone levels are typically less than those outdoors (Weschler, 2000), they can be much greater

when strong ozone generating sources are present such as photocopiers, electrostatic filters, and

ozone generators (Godish, 2001). In addition, the use of ozone for the removal of indoor air









contaminants has been widely promoted in the United States (Li et al., 2002). In those areas with

elevated ozone concentrations, the interference of ozone on Hg measurement may result in a

significant overestimate of Hg concentrations when using RA-915+ or other similar Hg

analyzers. It has been indicated by Singhvi et al. (2001) that water vapor can have positive

interference on UV Hg analyzers. The magnitude of this interference was not reported, and it

could vary for different UV Hg analyzers. If this positive interference of water vapor were

applied to the RA-915+ Hg analyzer used in this study, the total interference on Hg measurement

(caused by both ozone and water vapor) could be even larger.

The interference of ozone on Hg measurement can also impact the risk assessment of

human exposure to Hg. The Reference Concentration (RfC) for elemental Hg specified by the

EPA is 300 ng m-3 based on central nervous system (CNS) effects in humans (USEPA,

1999a). Hg concentrations above this level may result in a further investigation of hazardous

exposure. According to the findings in this work, an ozone concentration in the range of 0-120

ppb can exert a positive bias in Hg measurement up to 21% (63 ng m-3) of the EPA RfC (300 ng

m-3). The interference may be especially critical for conditions where the measured Hg

concentration is slightly above the RfC level, because after subtracting the bias caused by ozone

interference, the actual Hg concentration may not exceed the RfC any more. Therefore,

eliminating ozone from sample gas is essential to obtain accurate Hg concentration and thus is

important for risk assessment of human exposure to Hg. Helmig (1997) reviewed ozone removal

techniques in sampling of atmospheric volatile organic trace gases. These techniques may be

applicable to ozone removal in Hg sampling as long as they do not tamper with Hg

concentrations.









Interference of Hg on Ozone Measurement

A designated range of Hg concentrations were fed into the gas stream to investigate the

possible Hg interference on ozone measurement. Due to the limitation of the Hg vapor

generating unit used in this work (the dilution ratio was relatively small compared with the

saturated Hg vapor concentration), the minimum Hg concentration introduced was about 2300 ng

m-3. In addition, Hg levels were controlled so that the interference on ozone readings were

within the measurement range of the ozone analyzer (0-500 ppb). A blank test showed no

interference when no Hg was present. As the Hg concentration in the sample gas increases, the

corresponding reading on the ozone analyzer also increases (as shown in Figure 6-3). An

approximately linear relationship can be obtained as expressed in the following equation:

C10 = 0. 1165 x CH (6-2)

where Co3=, iS the equivalent ozone concentration, i.e., the interference on ozone analyzer (in

unit of ppb) and CI, is the Hg concentration (in unit of ng m-3.

Equation 6-2 implies that Hg can exert a significant interference on ozone measurement,

which is very likely due to the UV absorption by Hg when passing through the ozone analyzer

(USEPA, 1999b). Although the Hg level in the ambient and indoor environment is typically

lower than the minimum Hg concentration (2300 ng m-3) tested in this work, the results obtained

can be used as a reference to predict the practical conditions. It should be noted that

extrapolation beyond the tested concentration range might involve errors that could impact the

accuracy of the prediction. However, given the high value of R2 (0.9957) of the regression

analysis which has incorporated the origin point, it is suggested that the error associated with the

extrapolation to the range of0O to 2300 ng m-3 may not be significant. Carpi and Chen (2001)

reported that the highest Hg concentration measured at 12 indoor sites in New York City was









523 ng m-3. This Hg level would result in 61 ppb interference on ozone measurement using a

UV-absorption-based ozone analyzer, provided that Equation 6-2 is valid at lower Hg

concentration levels. The interference, 61 ppb, added to the normal ambient ozone concentration

(0-50 ppb) (Lim and Turpin, 2002), may have many chances to exceed the NAQQS for ozone

(80 ppb). In the case when Hg concentration is equal to the EPA RfC (300 ng m-3), an

interference of 35 ppb would be involved, which is comparable to the average ambient ozone

concentration. Since indoor ozone concentration is typically lower than that outdoors (Weschler,

2000) and indoor Hg concentrations generally higher than that outdoors due to various indoor Hg

contamination sources (such as accidental spills of Hg from natural gas meters, Hg

thermometers, fluorescent light bulbs, etc.) (Carpi and Chen, 2001), the impact of Hg

interference on indoor ozone measurement may be much greater. These results indicate that it is

essential to eliminate the Hg interference to obtain correct ozone concentration at Hg

contaminated places.

Now that the linear relationship between Hg concentration and corresponding

interference on ozone measurement was successfully established, it is necessary to compare the

results with those by the EPA (1999b) where the same type of ozone monitor was used. At a Hg

concentration of 0.04 ppb (328 ng m-3 at 250C), Equation 6-2 predicts an ozone interference of

38 ppb, whereas the EPA (1999b) reported approximately 10 ppb at low humidity (RH = 20 to

30%) and 5 ppb at high humidity (RH = 70 to 80%). Since in this work the experiments were

conducted using dry zero air, comparison of the above results suggests that higher humidity may

diminish the interference caused by Hg. A possible reason is the deposition of water vapor in the

optical cell attenuates the incident UV light, causing a negative interference by water vapor

itself. The finding also suggests that the reported interference may not be so significant in









ambient and indoor air as in the dry air used in this work. Further research is needed to verify

this hypothesis.

Summary

Mutual interference of UV-absorption-based measurements of ozone and Hg were

investigated in this study. It was found that ozone in the range of 0 to 120 ppb can exert an

interference of up to 63 ng m-3 on an AAS based Hg analyzer. A linear relationship was

established between the ozone concentration and corresponding interference on Hg

measurement. On the other hand, it was found that Hg can also result in significant interference

on an ozone analyzer based on UV absorption. Results showed that Hg at a concentration of 300

ng m-3 can potentially cause a bias in ozone measurement of approximately 35 ppb. These

mutual interference may consequently affect the risk assessment of human exposure to both Hg

and ozone. It should be noted that the results were obtained from a relatively small set of

equipment (only one Hg and one ozone analyzer). There are many different types of Hg or ozone

analyzers based on principle of UV absorption. It is possible that certain types of analyzers are

not compromised by the interference found in this work. Thus, the findings in this work should

be considered as analyzer-specific. Further investigation is needed to determine whether the

trends observed in this work can be extended to other types of analyzers.











Hg Analyzer


Ozone Generator


Carbon Trap


Zero Air


Vent


MIFC


Hg Analyzer


2-Way
Valve


Carbon Trap


Zero Air


Ice-Water
Bath


Figure 6-1. Experimental setup. A) Ozone interference on Hg measurement. B) Hg interference
on ozone measurement.










80

70
Cg,e, = 0.5559xCos
so 60 R 2 = 0.9835



c 40



I 20

10


0 20 40 60 80 100 120 140
Ozone Concentration (ppb)



Figure 6-2. Measurement interference of ozone on the RA-915+ Hg analyzer as a function of
ozone concentration (The error bars represent one standard deviation).











600


500 C 0,e = 0.1165x C H
..... ~R 2 = 0 9 9 57

8 400

300


a 200
O
100



0 500 1000 1500 2000 2500 3000 3500 4000 4500

Hg Concentration (ng m-3


Figure 6-3. Measurement interference of Hg on the ozone analyzer as a function of Hg
concentration.









CHAPTER 7
CONCLUSIONS AND RECOMMENDATIONS

This doctoral research has focused on the fixed-bed studies of Hgo removal using

nanostructured Sio2/TiO2/V205 composites. Three types of composites were tested for their

catalytic activities in both pellet and powder forms: (1) TiO2 nanoparticles doped on high surface

area SiO2 gel (SiO2-TiO2), (2) V20s doped on Sio2 gel (SiO2-V20s), and (3) V20s doped on the

SiO2-TiO2 support (SiO2-TiO2-V205). Experiments were conducted under both room conditions

and flue gas conditions. The kinetics of the Hg oxidation was studied, and the interactions

between Hg and the flue gas components on the catalyst surfaces were investigated. The catalytic

activities of the three catalysts were also compared. The following conclusions have been

obtained from this research.

* Conclusion 1: A Langmuir-Hinshelwood (L-H) model can be used to express the kinetics of

photocatalytic oxidation of Hgo on the SiO2-TiO2 HRHOCOmposite under room conditions.

Good agreement between the experimental data and the L-H model was demonstrated. The

model predicted a great potential of the SiO2-TiO2 HRHOCOmposite for Hgo removal even at

very high Hgo concentrations. The rate of photocatalytic Hgo oxidation increased when the

inlet Hgo concentration increased and it reached a maximum value in the absence of water

vapor. The addition of water vapor was found to inhibit Hgo photocatalytic oxidation, which

may be explained by the competitive adsorption of water vapor with Hgo on the TiO2 SUTrfCO.

* Conclusion 2: The mechanisms of Hgo removal and reemission from used catalysts were

investigated in a fixed-bed reactor at 65 OC using air as the carrier gas. Without UV

irradiation, Hgo adsorption was found to be insignificant, but it could be enhanced by the

photocatalytic oxidation product, HgO, possibly due to the high affinity between HgO and

Hgo. Under dry conditions 95% of Hgo can be removed; however, increased humidity levels









remarkably suppress both Hgo adsorption and photocatalytic oxidation. Introducing water

vapor can also result in significant reemission of captured Hgo from the nanocomposite,

which may be ascribed to the repellant effect of water vapor adsorbed on the

superhydrophilic TiO2 surface. Exposure to UV light was found either to prohibit Hgo

reemission when photocatalytic oxidation of reemitted Hgo prevailed or to promote Hgo

reemission when photocatalytic reduction of HgO to Hgo dominated later on. It is concluded

that Hgo capture on the SiO2-TiO2 HRHOCOmposite in a humid environment under UV

irradiation is controlled by four mechanisms: adsorption, photocatalytic oxidation,

desorption, and photocatalytic reduction. The observed inhibitory effect of water vapor is

contributed by its competitive occupancy of the available adsorption sites, displacement of

adsorbed Hgo, and participation in the photocatalytic reduction of HgO to Hgo

* Conclusion 3: Hgo removal using the SiO2-TiO2 HRHOCOmposite has been investigated in

simulated flue gas of coal-fired power plants. The flue gas components were found to have

significant effects on Hg removal efficiency. H20 inhibited Hg oxidation and capture and the

inhibitory effect was proportional to the H20 concentration. HCI enhanced Hgo oxidation

probably following an Eley-Redeal mechanism where adsorbed Cl species reacts with gas-

phase Hgo. SO2 promoted Hg oxidation, possibly forming mercury sulfate species. NO

significantly inhibited Hg removal by scavenging OH radicals that are necessary for Hgo

oxidation. The effect of NO2 WaS found to be insignificant. Experiments in simulated flue

gases also showed that high rank coals are preferable to low rank coals because of the lower

moisture and higher HCI and SO2 COncentrations in the flue gas.

* Conclusion 4: A method has been developed to dope V20s on the SiO2 and Sio2-TiO2

supports. Improvements in Hg removal from flue gas were observed in fixed-bed studies









using both pellet and powder forms of the SiO2-V20s and Sio2-TiO2-V205 catalysts. No UV

light activation is needed for the V20s doped catalysts. For SiO2-V20s catalysts, the Hg

removal efficiency increased as the V20s loading increased from 2 to 8% but decreased as

the V20s loading further increased to 10%. The results suggested that the optimal V20s

loading for a maximum catalytic activity is somewhere between 5 and 8%. The SiO2-TiO2-

V20s catalysts have an even greater ability of oxidizing Hg compared to SiO2-V20s, which

can be advantageous to power plants equipped with wet-scrubbers. where oxidized Hg can be

easily captured. It was found that the Hg oxidation on the V20s doped catalysts may follow

an Eley-Rideal mechanism where HC1, NO, and NO2 are first adsorbed on the catalyst and

then react with gas-phase Hg. On the contrary, water vapor dramatically inhibits Hg

oxidation while SO2 has an insignificant effect.

*Conclusion 5: Mutual interference of UV-absorption-based measurements of ozone and Hg

were also investigated in this research. It was found that ozone in the range of 0 to 120 ppb

can exert an interference of up to 63 ng m-3 on an AAS based Hg analyzer and the

interference is linearly related to the ozone concentration. On the other hand, it was found

that Hg can have significant interference on ozone analyzers that are based on UV

absorption. Hg at a concentration of 300 ng m-3 can potentially cause a bias in ozone

measurement of approximately 35 ppb, an average ozone concentration in the air under

normal conditions. These mutual interference may consequently affect the risk assessment

of human exposure to both Hg and ozone.

Based on the findings of this doctoral research, future research is recommended on the

following topics:









1. It has been verified in this research that V205 is the active center for the Hg oxidation, but the

SiO2-TiO2-V205 composites exhibit much higher oxidation abilities than the SiO2-V205

composites. This warrants further studies to understand the enhancement of Hg oxidation on

the superior TiO2 support and to explore the optimal V205/TiO2 ratio. Finding the optimal

V205/TiO2 loading is of great importance to the advancement of a most cost-effective

catalyst. Surface analysis techniques such as X-ray photoelectron spectroscopy (XPS) are

recommended to be used to examine the surface structure of the catalysts before and after the

Hg oxidation experiments. XPS may provide information on the oxidation states of the

vanadium and the dispersion of the vanadia on the surface by providing the V/Ti ratio on the

surface. This information would be helpful for a better understanding of the catalyst

characteristics as well as the reaction mechanisms.

2. A long term test on the performance of the SiO2-TiO2-V205 composites on Hg removal is

recommended. Although this research demonstrated a stabilized activity of the catalyst in a

20-hr test, a longer testing time, probably a few hundred hours, would be considered

sufficient to verify its long-term performance. However, some modifications of the current

system may be needed for the long-term test.

3. Exploring the potential of the SiO2-TiO2-V205 composites as a multipollutant control

strategy, i.e. a low temperature SCR catalyst for control of both Hg and NOx. It is promising

because the catalyst contains the same active phases as the commercial SCR catalysts.

However, the effect of the inj ected ammonia (NH3) on Hg removal is yet to be investigated.

4. A pilot-scale study using the SiO2-TiO2-V205 composites is recommended as a scale-up of

the current bench-scale system. Kinetic studies may be necessary to determine the amount (or

total surface area) of the catalyst that is needed for treating a higher flow rate of flue gas. A









good configuration of the packed-bed reactor may also be critical to achieve a high efficiency

of Hg removal in the pilot-scale study.










LIST OF REFERENCES


Argyle, M. D., Chen, K. D., Resini, C., Krebs, C., Bell, A. T., Iglesia, E., 2004. Extent of
reduction of vanadium oxides during catalytic oxidation of alkanes measured by in-situ
UV-visible spectroscopy. J. Phy. Chem. B. 108, 2345-2353.

Augugliaro, V., Coluccia, S., Loddo, V., Marchese, L., Martra, G., Palmisano, L., Schiavello,
M., 1999. Photocatalytic oxidation of gaseous toluene on anatase TiO2 catalyst:
mechanistic aspects and FT-IR investigation. Appl. Catal. B-Environ. 20, 15-27.

Bai, M., Zhang, Z., Bai, M., Yi, C., Bai, X., 2006. Removal of SO2 from Gas Streams by
Oxidation using Plasma-Generated Hydroxyl Radicals. Plasma Chem. Plasma Process.
26, 177-186.

Benson, S. A., Laumb, J. D., Crocker, C. R., Pavlish, J. H., 2005. SCR catalyst performance in
flue gases derived from subbituminous and lignite coals. Fuel Process. Technol. 86, 577-
613.

Berglund, F., Bertin, M., 1969. Chemical Fallout. Tomas Publisher, Springfield.

Bidstrup, P. C., 1964. Toxicity of Mercury and its Compounds. Elsevier, Amsterdam.

Borderieux, S., Wu, C.-Y., Bonzongo, J.-C., Powers, K., 2004. Control of Elemental Mercury
Vapor in Combustion Systems Using Fe203 Nanoparticles. Aerosol Air Qual. Res. 4, 74-
90.

Bosch, H., Janssen, F., 1988. Formation and control of nitrogen oxides. Catal. Today. 2, 369-
379.

Brown, T. D., Smith, D. N., Hargis, R. A., O'Dowd, W. J., 1999. Mercury measurement and its
control: What we know, have learned, and need to further investigate. J. Air. Waste
Manage. 49, 628-640.

Busca, G., Lietti, L., Ramis, G., Berti, F., 1998. Chemical and mechanistic aspects of the
selective catalytic reduction of NOx by ammonia over oxide catalysts: A review. Appl.
Catal. B-Environ. 18, 1-36.

Calvert, J. G., Lindberg, S. E., 2005. Mechanisms of mercury removal by 03 and OH in the
atmosphere. Atmospheric Environment. 39, 3355-3367.

Canela, M. C., Alberici, R. M., Jardim, W. F., 1998. Gas-phase destruction of H2S using
TiO2/UV-VIS. J. Photochem. Photobio. A-Chem. 112, 73-80.

Carey, T. R., Hargrove, C. W., Richardson, C. F., Chang, R., 1998. Factors affecting mercury
control in utility flue gas using activated carbon. J. Air Waste Manage. Assoc. 48, 1166-
1174.










Carpi, A., Chen, Y. F., 2001. Gaseous elemental mercury as an indoor air pollutant. Environ. Sci.
Technol. 35, 4170-4173.

Chen, S., Rostam-Abadi, M., Chang, R., 1996. Mercury removal from combustion flue gas by
activated carbon injection: mass transfer effects. Prepr. Pap.-Am. Chem. Soc., Div. Fuel
Chem. 41, 442-446.

Ebinghaus, R., Kock, H. H., Coggins, A. M., Spain, T. G., Jennings, S. G., Temme, C., 2002.
Long-term measurements of atmospheric mercury at Mace Head, Irish west coast,
between 1995 and 2001. Atmos. Environ. 36, 5267-5276.

Edwards, J. R., Srivastava, R. K., Kilgroe, J. D., 2001. A study of gas-phase mercury speciation
using detailed chemical kinetics. J. Air Waste Manage. Assoc. 51, 869-877.

Eswaran, S., Stenger, H. G., 2005. Understanding Mercury Conversion in Selective Catalytic
Reduction (SCR) Catalysts. Energy Fuels. 19, 2328-2334.

Ferrara, R., 1999. Mercury Contaminated Sites. Springer, Berlin.

Flora, J. R. V., Vidic, R. D., Liu, W., Thurnau, R. C., 1998. Modeling powdered activated carbon
inj section for the uptake of elemental mercury vapors. J. Air Waste Manage. Assoc. 48,
1051-1059.

Fujishima, A., Rao, T. N., Tryk, D. A., 2000. Titanium dioxide photocatalysis. J. Photochem.
Photobio. C: Photochem. Reviews. 1, 1-21.

Galbreath, K. C., Zygarlicke, C. J., Tibbetts, J. E., Schulz, R. L., Dunham, G. E., 2005. Effects of
NOx, a-Fe203, y-Fe203, and HCI on mercury transformations in a 7-kW coal combustion
system. Fuel Process. Technol. 86, 429-448.

Godish, T., 2001. Indoor Environmental Quality. Lewis Publishers, Boca Raton, Florida.

Goodsite, M. E., Plane, J. M. C., Skov, H., 2004. A theoretical study of the oxidation of Hgo to
HgBr2 in the troposphere. Environ. Sci. Technol. 38, 1772-1776.

Grabowski, R., Pietrzyk, S., Sloczynski, J., Genser, F., Weislo, K., Grzybowska-Swierkosz, B.,
2002. Kinetics of the propane oxidative dehydrogenation on vanadia/titania catalysts
from steady-state and transient experiments. Appl. Catalysis A-Gen. 232, 277-288.

Granite, E. J., Pennline, H. W., 2002. Photochemical removal of mercury from flue gas. Ind.
Eng. Chem. Res. 41, 5470-5476.

Granite, E. J., Pennline, H. W., Hargis, R. A., 2000. Novel sorbents for mercury removal from
flue gas. Ind. Eng. Chem. Res. 39, 1020-1029.

Guan, K. H., 2005. Relationship between photocatalytic activity, hydrophilicity and self-cleaning
effect of TiO2/SiO2 f11ms. Surface & Coatings Technology. 191, 155-160.










Helmig, D., 1997. Ozone removal techniques in the sampling of atmospheric volatile organic
trace gases. Atmos. Environ. 31, 3635-3651.

Kantcheva, M., Bushev, V., Klissurski, D., 1994. Study Of The NO2-NH3 Interaction On A
Titania (Anatase) Supported Vanadia Catalyst. J. Catalysis. 145, 96-106.

Kim, S. B., Hong, S. C., 2002. Kinetic study for photocatalytic degradation of volatile organic
compounds in air using thin film TiO2 photocatalyst. Appl. Catal. B-Environ. 35, 305-
315.

Kinsey, J. S., Anscombe, F. R., Lindberg, S. E., Southworth, G. R., 2004. Characterization of the
fugitive mercury emissions at a chlor-alkali plant: overall study design. Atmos. Environ.
38, 633-641.

Kobayashi, M., Kuma, R., Masaki, S., Sugishima, N., 2005. TiO2-SiO2 and V205/TiO2-SiO2
catalyst: Physico-chemical characteristics and catalytic behavior in selective catalytic
reduction of NO by NH3. Appl. Catal. B-Environ. 60, 173-179.

Kobayashi, M., Kuma, R., Morita, A., 2006. Low temperature selective catalytic reduction of
NO by NH3 over V20s supported on TiO2-SiO2-MoO3. Catal. Letters. 112, 37-44.

Laudal, D. L., Brown, T. D., Nott, B. R., 2000. Effects of flue gas constituents on mercury
speciation. Fuel Process. Technol. 65, 157-165.

Laudal, D. L., Thompson, J. S., Pavlish, J. H., Brickett, L. A., Chu, P., 2004. Use of continuous
mercury monitors at coal-fired utilities. Fuel Process. Technol. 85, 501-511.

Laumb, J. D., Benson, S. A., Olson, E. A., 2004. X-ray photoelectron spectroscopy analysis of
mercury sorbent surface chemistry. Fuel Process. Technol. 85, 577-585.

Lee, C. W., Srivastava, R. K., Ghorishi, S. B., Hastings, T. W., Stevens, F. M., 2004.
Investigation of selective catalytic reduction impact on mercury speciation under
simulated NOx emission control conditions. J. Air Waste Manage. Assoc. 54, 1560-1566.

Lee, C. W., Srivastava, R. K., Ghorishi, S. B., Karwowski, J., Hastings, T. W., Hirschi, J. C.,
2006. Pilot-scale study of the effect of selective catalytic reduction catalyst on mercury
speciation in Illinois and powder river basin coal combustion flue gases. J. Air Waste
Manage. Assoc. 56, 643-649.

Lee, T. G., Biswas, P., Hedrick, E., 2001. Comparison of Hgo capture efficiencies of three in situ
generated sorbents. AICHE J. 47, 954-961.

Lee, T. G., 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.

Li, T. H., Turpin, B. J., Shields, H. C., Weschler, C. J., 2002. Indoor hydrogen peroxide derived
from ozone/d-limonene reactions. Environ. Sci. Technol. 36, 3295-3302.










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.

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.

Lim, H. J., Turpin, B. J., 2002. Origins of primary and secondary organic aerosol in Atlanta:
Results' of time-resolved measurements during the Atlanta supersite experiment. Environ.
Sci. Technol. 36, 4489-4496.

Maroto-Valer, M. M., Zhang, Y. Z., Granite, E. J., Tang, Z., Pennline, H. W., 2005. Effect of
porous structure and surface functionality on the mercury capacity of a fly ash carbon and
its activated sample. Fuel. 84, 105-108.

Martra, G., Arena, F., Coluccia, S., Frusteri, F., Parmaliana, A., 2000. Factors controlling the
selectivity of V20s supported catalysts in the oxidative dehydrogenation of propane.
Catal. Today. 63, 197-207.

McLarnon, C. R., Granite, E. J., Pennline, H. W., 2005. The PCO process for photochemical
removal of mercury from flue gas. Fuel Process. Technol. 87, 85-89.

Meites, L., Ed. 1963. Handbook of Analytical Chemistry. New York, McGraw Hill.

Meserole, F. B., Richardson, C. F., Miller, S. D., Searcy, K., Chang, R. 2000, Estimated the costs
of electric utility mercury control using sorbent inj section. Proceedings of the Air and
Waste Management Association 93rd Annual Meeting, Salt Lake City, UT.

Miller, S. J., Dunham, G. E., Olson, E. S., Brown, T. D., 2000. Flue gas effects on a carbon-
based mercury sorbent. Fuel Process. Technol. 65, 343-363.

Mishima, A., 1992. Bitter Sea: The Human Cost of Minamata Disease. Kosei Pub. Co., Tokyo,
Japan.

North American Research Strategy for Tropospheric Ozone (NARSTO), 1999. Measurement
Methods Compendium Ozone Method. Oak Ridge National Laboratory, Oak Ridge,
TN. http://narsto.ornl_ gov/Compendium/methods/o3 .shtml. Date last accessed: June
2007.

Niksa, S., Fujiwara, N., 2005a. Predicting extents of mercury oxidation in coal-derived flue
gases. J. Air Waste Manage. Assoc. 55, 930-939.

Niksa, S., Fujiwara, N., 2005b. A predictive mechanism for mercury oxidation on selective
catalytic reduction catalysts under coal-derived flue gas. J. Air Waste Manage. Assoc. 55,
1866-1875.










Niksa, S., Helble, J. J., Fujiwara, N., 2001. Kinetic modeling of homogeneous mercury
oxidation: The importance of NO and H20 in predicting oxidation in coal-derived
systems. Environ. Sci. Technol. 35, 3701-3706.

Norton, G. A., Yang, H. Q., Brown, R. C., Laudal, D. L., Dunham, G. E., Erjavec, J., 2003.
Heterogeneous oxidation of mercury in simulated post combustion conditions. Fuel. 82,
107-116.

Obee, T. N., 1996. Photooxidation of sub-parts-per-million toluene and formaldehyde levels an
titania using a glass-plate reactor. Environ. Sci. Technol. 30, 3578-3584.

Obee, T. N., Hay, S. O., 1997. Effects of moisture and temperature on the photooxidation of
ethylene on Titania. Environ. Sci. Technol. 31, 2034-2038.

Olson, E. S., Sharma, R. K., Pavlish, J. H., 2002. On the analysis of mercuric nitrate in flue gas
by GC-MS. Anal. Bioanal. Chem. 374, 1045-1049.

Pal, B., Ariya, P. A., 2004. Gas-phase HO-Initiated reactions of elemental mercury: Kinetics,
product studies, and atmospheric implications. Environ. Sci. Technol. 38, 5555-5566.

Pan, H. Y., Minet, R. G., Benson, S. W., Tsotsis, T. T., 1994. Process for Converting Hydrogen-
Chloride to Chlorine. Ind. Eng. Chem. Res. 33, 2996-3003.

Parfitt, G. D., Ramsboth.J, Rocheste.Ch, 1971. Infra-Red Study of Hydrogen Chloride
Adsorption on Rutile Surfaces. Trans. Faraday Society. 67, 3100-&.

Parvulescu, V. I., Grange, P., Delmon, B., 1998. Catalytic removal of NO. Catal. Today. 46, 233-
316.

Pavlish, J. H., Sondreal, E. A., Mann, M. D., Olson, E. S., Galbreath, K. C., Laudal, D. L.,
Benson, S. A., 2003. Status review of mercury control options for coal-fired power
plants. Fuel Process. Technol. 82, 89-165.

Pitoniak, E., Wu, C. Y., Londeree, D., Mazyck, D., Bonzongo, J. C., Powers, K., Sigmund, W.,
2003. Nanostructured silica-gel doped with TiO2 for mercury vapor control. J. Nanopart.
Res. 5, 281-292.

Pitoniak, E., Wu, C. Y., Mazyck, D. W., Powers, K. W., Sigmund, W., 2005. Adsorption
enhancement mechanisms of silica-titania nanocomposites for elemental mercury vapor
removal. Environ. Sci. Technol. 39, 1269-1274.

Pogarev, S. E., Ryzhov, V., Mashyanov, N., Sholupov, S., Zharskaya, V., 2002. Direct
measurement of the mercury content of exhaled air: a new approach for determination of
the mercury dose received. Anal. Bioanal. Chem. 374, 1039-1044.

Presto, A. A., Granite, E. J., 2006. Survey of Catalysts for Oxidation of Mercury in Flue Gas.
Environ. Sci. Technol. 40, 5601-5609.










Raillard, C., Hequet, V., Le Cloirec, P., Legrand, J., 2004. Kinetic study of ketones
photocatalytic oxidation in gas phase using TiO2-COntaining paper: effect of water vapor.
J. Photochem. Photobio. A-Chem. 163, 425-431.

Ravichandran, M., 2004. Interactions between mercury and dissolved organic matter a review.
Chemosphere. 55, 319-331.

Rodella, C. B., Nascente, P. A. P., Franco, R. W. A., Magon, C. J., Mastelaro, V. R., Florentino,
A. O., 2001. Surface characterisation of V205/TiO2 catalytic system. Phy. Stat. Solidi A-
Appl. Res. 187, 161-169.

Rodriguez, S., Almquist, C., Lee, T. G., 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.

Romero, C. E., Li, Y., Bilirgen, H., Sarunac, N., Levy, E. K., 2006. Modiaication of boiler
operating conditions for mercury emissions reductions in coal-fired utility boilers. Fuel.
85, 204-212.

Rostam-Abadi, M., Chen, S. G., Hsi, H.-C., Rood, M., Chang, R., Carey, T. R., Hargrove, B.,
Richardson, C., Rosenhoover, W., Meserole, F. 1997, Novel vapor phase mercury
sorbents. Proceedings of the EPRI-DOE-EPA Combined Utility Air Pollutant Control,
Washington, DC.

Schroeder, W. H., Munthe, J., 1998. Atmospheric mercury An overview. Atmospheric
Environment. 32, 809-822.

Senior, C. L., 2006. Oxidation of mercury across selective catalytic reduction catalysts in coal-
fired power plants. J. Air Waste Manage. Assoc. 56, 23-31.

Senior, C. L., Sarofim, A. F., Zeng, T. F., Helble, J. J., Mamani-Paco, R., 2000. Gas-phase
transformations of mercury in coal-fired power plants. Fuel Process. Technol. 63, 197-
213.

Shang, J., Du, Y. G., Xu, Z. L., 2002. Photocatalytic oxidation of heptane in the gas-phase over
TiO2. Chemosphere. 46, 93-99.

Shikada, T., Fujimoto, K., Kunugi, T., Tominaga, H., Kaneko, S., Kubo, Y., 1981. Reduction of
Nitric-Oxide with Ammonia on Vanadium-Oxide Catalysts Supported on
Homogeneously Precipitated Silica-Titania. Ind. Eng. Chem. Prod. Res. Dev. 20, 91-95.

Sholupov, S., Pogarev, S., Ryzhov, V., Mashyanov, N., Stroganov, A., 2004. Zeeman atomic
absorption spectrometer RA-915+for direct determination of mercury in air and complex
matrix samples. Fuel Process. Technol. 85, 473-485.

Singhvi, R., Turpin, R., Kalnicky, D. J., Patel, J., 2001. Comparison of field and laboratory
methods for monitoring metallic mercury vapor in indoor air. Journal of Hazardous
Materials. 83, 1-10.










Sloczynski, J., 1996. Kinetics and mechanism of reduction and reoxidation of the alkali metal
promoted vanadia-titania catalysts. Appl. Catalysis A-Gen. 146, 401-423.

Sommar, J., Gardfeldt, K., Stromberg, D., Feng, X. B., 2001. A kinetic study of the gas-phase
reaction between the hydroxyl radical and atomic mercury. Atmospheric Environment.
35, 3049-3054.

Son, H. S., Lee, S. J., Cho, I. H., Zoh, K. D., 2004. Kinetics and mechanism of TNT degradation
in TiO2 photocatalysis. Chemosphere. 57, 309-317.

Tesser, R., Maradei, V., Di Serio, M., Santacesaria, E., 2004. Kinetics of the oxidative
dehydrogenation of ethanol to acetaldehyde on V205/TiO2-SiO2 catalysts prepared by
grafting. Ind. Eng. Chem. Res. 43, 1623-1633.

Tseng, H.-H., Wey, M.-Y., Liang, Y.-S., Chen, K.-H., 2003. Catalytic removal of SO2, NO and
HCI from incineration flue gas over activated carbon-supported metal oxides. Carbon. 41,
1079.

US EPA, 1997a. Mercury Study Report to Congress. EPA-452/R-97-003.

US EPA, 1997b. National ambient air quality standards for ozone, Einal rule. Federal Register 62
(138).

US EPA, 1997c. Standards of Performance for New Stationary Sources and Emission Guidelines
for Existing Sources: Hospital/Medical/Infectious Waste Incinerators; Final Rule. Fed.
Regist., 62, 48347-448391.

US EPA, 1999a. Integrated Risk Information System (IRIS) on Elemental Mercury. National
Center for Environmental Assessment, Office of Research and Development,
Washington, DC.

US EPA, 1999b. Laboratory Study to Explore Potential Interferences to Air Quality Monitors.
EPA-454/C-00-002.

US EPA, 2000. Regulatory Finding on the Emissions of Hazardous Air Pollutants from Electric
Utility Steam Generating Units. Federal Register 65 (245), 79825-79831.

US EPA, 2001a. Environmental Technology Verifieation Program, Verification Reports and
Statements for each tested Mercury CEM. Research Triangle Park, NC.
http://www.epa.gov/etv/verifieations/vcenterl-.html. Date last accessed: June 2007.

US EPA, 2001Ib. Standards of Performance for Large Municipal Waste Combustors; Final Rule
and Proposed Rule. Fed. Regist., 66, 57823-57828.

US EPA, 2003. National Emission Standards for Hazardous Air Pollutants: Mercury Emissions
from Mercury Cell Chlor-Alkali Plants; Final Rule. Fed. Regist., 68, 70903-70946.










US EPA, 2005a. Clean Air Mercury Rule; Standards of Performance for New and Existing
Stationary Sources: Electric Utility Steam Generating Units. 40 CFR Parts 60, 63, 72, and
75.

US EPA, 2005b. Regulatory Impact Analyses for the Particulate Matter and Ozone NAAQS and
Proposed Regional Haze Rule. Research Triangle Park, NC.
http ://www.epa.gov/ttn/oarpg/naaqsfin/ria.html. Date last accessed: June 2007.

Wang, R., Hashimoto, K., Fujishima, A., Chikuni, M., Kojima, E., Kitamura, A., Shimohigoshi,
M., Watanabe, T., 1997. Light-induced amphiphilic surfaces. Nature. 388, 431-432.

Weast, R. C., Astle, M. J., Beyer, W. H., Eds. 1983. CRC Handbook of Chemistry and Physics.
Boca Raton, FL, CRC Press.

Weckhuysen, B. M., Keller, D. E., 2003. Chemistry, spectroscopy and the role of supported
vanadium oxides in heterogeneous catalysis. Catal. Today. 78, 25-46.

Weiss-Penzias, P., Jaffe, D. A., McClintick, A., Prestbo, E. M., Landis, M. S., 2003. Gaseous
elemental mercury in the marine boundary layer: Evidence for rapid removal in
anthropogenic pollution. Environ. Sci. Technol. 37, 3755-3763.

Weschler, C. J., 2000. Ozone in indoor environments: Concentration and chemistry. Indoor Air.
10, 269-288.

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.

Yang, C. S., Chen, C. J., 2005. Synthesis and characterization of silica-capped titania nanorods:
An enhanced photocatalyst. Applied Catalysis a-General. 294, 40-48.

Zhang, F.-S., Nriagu, J. O., Itoh, H., 2004. Photocatalytic removal and recovery of mercury from
water using TiO2-mOdified sewage sludge carbon. J. Photochem. Photobio. A: Chemistry.
167, 223-228.









BIOGRAPHICAL SKETCH

Ying Li was born in 1977 in Luzhou, Sichuan Province, China. He earned his B.S. degree

in thermal engineering in June 1999 at Zhejiang University, China. He continued his master

studies in thermal engineering at Zhejiang University and received his M. S. degree in March

2002. Ying Li came to the United States in August 2002. He studied in mechanical engineering

at Lehigh University, Pennsylvania, and worked as a research assistant at the Energy Research

Center. He earned his second M. S. degree at Lehigh University in 2004. He joined the research

group of Dr. Chang-Yu Wu at the University of Florida in August 2004 and started pursuing his

Ph.D. degree in the Department of Environmental Engineering Sciences. His doctoral research

focused on mercury removal from power plant flue gas using novel catalysts.

Ying Li was the president of the student chapter of Air & Waste Management Association

(A&WMA) at the University of Florida in 2006-2007. He was awarded the Milton Feldstein

Memorial Scholarship from A&WMA in 2006. He was awarded the Axel Hendrickson

Scholarship and the Clair Fancy Scholarship from Florida Section A&WMA in 2005 and 2006,

respectively. He won 1st place in the student paper poster competition (doctoral level) in the 2006

A&WMA Annual Conference. He won 1st place in the student poster competition (graduate

level) in the 2006 Florida A&WMA Annual Conference.





PAGE 1

1 REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING NANOSTRUCTURED SILICA/TITANIA/VANADIA COMPOSITES By YING LI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIR EMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Ying Li

PAGE 3

3 To my wife and my parents for their constant love, understanding, and support.

PAGE 4

4 ACKNOWLEDGMENTS I sincerely thank Dr. Chang Yu Wu (my supervisory co mmittee chair) for his invaluable guidance, inspiration, and encouragement. I am truly grateful for his continuous and enthusiastic support of my graduate research and my career development. My supervisory committee members (Dr. Jean Andino, Dr. Kevin Powe rs, Dr. Jean Claude Bonzongo, and Dr. Wolfgang Sigmund) have generously given their time and expertise to better my work. I thank them very much for their guidance and their good natured support. I would like to thank Patrick Murphy, who has been assistin g me for one and a half years in my research project. He has given me a big hand in building the reaction system and conducting the experiments. I thank Sameer Matta for his assistance in doing the experiments and making the pellets in the past two semeste rs. My thanks also go to Jie Gao, who helped me analyze the amount of mercury captured on the catalysts. I greatly appreciate the help from Yu Mei Hsu for miscellaneous lab items. I must acknowledge as well many other fellow students in our air resources g roup who kindly assisted my research. I am grateful to the National Science Foundation for the financial support and VICI Metronics, Inc. for providing the mercury permeation tube. Finally, I am thankful to the scholarship and encouragement from the Air & Waste Management Association.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LI ST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 13 Mercury an d Its Health Effects ................................ ................................ ............................... 13 Mercury Emissions and Regulations ................................ ................................ ...................... 13 Mercury Speciation and Control Technologies ................................ ................................ ...... 14 Mercury Removal by SiO 2 TiO 2 : Unknowns and Challenges ................................ ............... 16 Mercury Measurement Interference ................................ ................................ ....................... 17 Research Objectives ................................ ................................ ................................ ................ 18 2 KINETIC STUDY FOR PHOTOCATALYTIC OXIDATION OF ELEMENTAL MERCURY ON A SILICA TITANIA NANOCOMPOSITE ................................ ............... 20 Background ................................ ................................ ................................ ............................. 20 Materials and Methods ................................ ................................ ................................ ........... 21 Synthesis of SiO 2 TiO 2 Nanocomposite ................................ ................................ .......... 21 Apparatus and Procedure ................................ ................................ ................................ 22 Model Description ................................ ................................ ................................ .................. 24 Results and Discussion ................................ ................................ ................................ ........... 26 Effect of Hg 0 Concentration ................................ ................................ ............................ 26 Effect of Water Vapor ................................ ................................ ................................ ..... 28 Summary ................................ ................................ ................................ ................................ 31 3 ROLE OF MOISTURE IN ADSORPTION, PHOTOCATALYTIC OXIDATION, AND REEMISSION OF ELEMENTAL MERCURY ON A SILICA TITANIA NANACOMPOSITE ................................ ................................ ................................ .............. 40 Background ................................ ................................ ................................ ............................. 40 Experimental ................................ ................................ ................................ ........................... 41 Synthesis of SiO 2 TiO 2 Nanocomposite ................................ ................................ ......... 41 Experimental Setup ................................ ................................ ................................ ......... 42

PAGE 6

6 Results and Discussion ................................ ................................ ................................ ........... 43 Role of Moisture in Hg 0 Capture ................................ ................................ ..................... 43 Role of Moisture in Hg 0 Reemission ................................ ................................ ............... 45 Mechanisms of Hg 0 Capture and Reemission ................................ ................................ 46 Summary ................................ ................................ ................................ ................................ 51 4 REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING A SILICA TITANIA NANOCOMPOSITE ................................ ................................ ............................. 57 Background ................................ ................................ ................................ ............................. 57 Experimental Methods ................................ ................................ ................................ ............ 58 Synthesis of the SiO 2 TiO 2 Nanocomposite ................................ ................................ ... 58 Experimental Setup and Procedure ................................ ................................ ................. 59 Results and Discussion ................................ ................................ ................................ ........... 61 Baseline Test ................................ ................................ ................................ ................... 61 Effects of Individual Flue Gas Components ................................ ................................ ... 63 Hg Removal in Simulated Flue Gases ................................ ................................ ............. 67 Summary ................................ ................................ ................................ ................................ 67 5 DEVELOPMENT OF SILICA/VANADIA/TITANIA COMPOSITES FOR REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS ................................ ........ 72 Background ................................ ................................ ................................ ............................. 72 Materials and Methods ................................ ................................ ................................ ........... 74 Catalyst Preparation ................................ ................................ ................................ ......... 74 Catalyst Characterization Techniques ................................ ................................ ............. 74 C atalyst Activity Measurement ................................ ................................ ....................... 75 Results and Discussion ................................ ................................ ................................ ........... 76 Characterization of the Catalysts ................................ ................................ ..................... 76 Mercury Removal Using Pellet Catalysts ................................ ................................ ........ 77 Mercury Removal Using Powder Catalysts ................................ ................................ .... 79 Mercury Remov al Mechanisms ................................ ................................ ....................... 82 Role of O 2 ................................ ................................ ................................ ................. 82 Role of HCl ................................ ................................ ................................ .............. 83 Role of NO 2 ................................ ................................ ................................ .............. 85 Role of NO ................................ ................................ ................................ ............... 86 Role of SO 2 ................................ ................................ ................................ ............... 87 Role of H 2 O ................................ ................................ ................................ .............. 87 Summary ................................ ................................ ................................ ................................ 88 6 UV ABSORPTION BASED MEASUREMENTS OF OZONE AND MERCURY: AN INVESTIGATION ON THEIR MUTUAL INTERFERENCES ................................ ........... 99 Background ................................ ................................ ................................ ............................. 99 Methods ................................ ................................ ................................ ................................ 101 Descriptions of Hg and Ozone Instruments ................................ ................................ ... 101 Experimental Setup and Procedures ................................ ................................ .............. 102

PAGE 7

7 Results and Discussion ................................ ................................ ................................ ......... 103 Interference of Ozone on Hg Measurement ................................ ................................ .. 103 Interference of Hg on Ozone Measurement ................................ ................................ .. 106 Summary ................................ ................................ ................................ ............................... 108 7 CONCLUSIONS AND RECOMMENDATIONS ................................ ............................... 112 LIST OF REFERENCES ................................ ................................ ................................ ............. 117 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 125

PAGE 8

8 LIST OF TABLES Table page 4 1 Experimental conditions for investigation of the flue gas effects ................................ ...... 68 5 1 Experimental parameters for activity measurement of the catalysts in pellet form ........... 90 5 2 Experimental parameters for activity measurement of the catalysts in powder form ........ 90 5 3 BET surface areas of the catalysts ................................ ................................ ..................... 90 5 4 Amounts of Hg captured and oxidized on the catalysts in a 6 hr test ................................ 90

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Experimental system for kinetic studies. ................................ ................................ ........... 32 2 2 Description of Hg 0 photocatalytic oxidation o n SiO 2 TiO 2 nanocomposites. ................... 33 2 3 Photocatalytic oxidation of Hg 0 at different inlet Hg 0 concentrations without water vapor. ................................ ................................ ................................ ................................ 34 2 4 Inverse of Hg 0 photocatalytic oxidation rate versus the inverse of inlet Hg 0 concentration (without water vapor). ................................ ................................ ................. 35 2 5 Rate of Hg 0 photocatalytic oxidation versus inlet Hg 0 concentratio n without water vapor (solid circles: experimental data; solid line: L H model). ................................ ....... 36 2 6 Photocatalytic oxidation of Hg 0 at a constant inlet Hg 0 concentration of 0.66 mol m 3 with variation i n water vapor concentration. ................................ ............................... 37 2 7 Inverse of Hg 0 photocatalytic oxidation rate versus water vapor concentration at a constant inlet Hg 0 concentration of 0.66 mol m 3 ................................ .......................... 38 2 8 Rate of Hg 0 photocatalytic oxidation versus inlet Hg 0 concentration at different water vapor concentrations (markers: experimental data; lines: L H model). ............................ 39 3 1 Experimental system for studies on the role of water vapor. ................................ ............. 53 3 2 Dimensionless Hg 0 concentration at the reactor outlet (A, [H 2 O] = 0 ppm v ; B, [H 2 O] = 13000 ppm v ; C, [H 2 O] = 23000 ppm v ) ................................ ................................ ........... 54 3 3 Hg 0 reemission from SiO 2 TiO 2 nanocomposite after 3 h pretreatment (The inset shows the dimensionless Hg concentration during the pretreatment). .............................. 55 3 4 Mechanisms of Hg capture and reemission on the surface of SiO 2 TiO 2 nanocomposite. ................................ ................................ ................................ .................. 56 4 1 Photocatalytic reaction system under flue gas conditio ns. ................................ ................ 69 4 2 Hg speciation at the outlet of the reactor in the baseline test ................................ ............. 70 4 3 Effects of flue gas components on Hg capture a nd oxidation under various conditions of a) H 2 O, b) HCl, c) SO 2 d) NO, e) NO 2 and f) simulated flue gases ............................ 71 5 1 Experimental system for the fixed bed study using powder catalysts ............................... 91 5 2 XRD patterns of (a) SV2, (b) SV5, (c) SV8, (d) SV10, (e) ST12, and (f) ST12V5. ......... 92 5 3 Catalytic removal of Hg using the pellet cata lysts under various conditions .................... 93

PAGE 10

10 5 4 Outlet Hg concentration as a function of time using 500 mg powder of (a) SV2, (b) SV5, (c) SV8, (d) SV10, and (e) ST12V5. ................................ ................................ ........ 94 5 5 Outlet Hg concentration as a function of time using 250 mg powder of SV5 (20 hr test). ................................ ................................ ................................ ................................ .... 96 5 6 The role of flue gas components on Hg removal using 250 mg SV5 powder under dry conditions. ................................ ................................ ................................ .................... 97 5 7 The role of water vapor on Hg removal using 250 mg SV5 powder ............................... 98 6 1 Experimental setu p. A) Ozone interference on Hg measurement. B) Hg interference on ozone measurement. ................................ ................................ ................................ .... 109 6 2 Measurement interference of ozone on the RA 915+ Hg analyzer as a function of ozone concentration ( The error bars represent one standard deviation). ......................... 110 6 3 Measurement interference of Hg on the ozone analyzer as a function of Hg concentration. ................................ ................................ ................................ ................... 111

PAGE 11

11 Abstra ct 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 REMOVAL OF ELEMENTAL MERCURY FROM FLUE GAS USING NANOSTRUCTURED SILICA/TITANI A/VANADIA COMPOSITES By Ying Li August 2007 Chair: Chang Yu Wu Major: Environmental Engineering Sciences As a highly toxic pollutant, mercury (Hg) tends to bioaccumulate in the food chain and exerts adverse effects on human health. The U.S. EPA issued th e Clean Air Mercury Rule in 2005 to permanently cap and reduce Hg emissions f rom coal fired power plants. A low cost methodology using a SiO 2 TiO 2 nanocomposite as a photocatalyst has been recently developed to effectively remove elemental Hg (Hg 0 ) under r oom conditions. In this research, a bench scale fixed bed reactor system has been established and Hg 0 removal on the SiO 2 TiO 2 nanocomposite was examined under both room and flue gas conditions A kinetic study showed that Hg oxidation on the SiO 2 TiO 2 nan ocomposite under UV irradiation followed the Langmuir Hinshelwood rate expression. The flue gas components were found to have significant effects on Hg 0 removal using the SiO 2 TiO 2 nanocomposite HCl and SO 2 promoted Hg 0 oxidation, while water vapor and NO significantly inhibited Hg removal. The mechanisms of these promotional and inhibitory effects were thoroughly explored in this research. The active phase of the selective catalytic reduction (SCR) catalyst, V 2 O 5 was added to the SiO 2 or SiO 2 TiO 2 compo sites in an effort to improve the catalytic activity for Hg removal. No UV light activation is needed for the V 2 O 5 doped catalysts, which is a great advantage over the SiO 2 TiO 2 composites. The Hg 0 removal efficiency increased as the V 2 O 5 loading increased from

PAGE 12

12 2 to 8% but decreased as it further increased to 10%. The optimal V 2 O 5 loading was found to be somewhere between 5 and 8%. The SiO 2 TiO 2 V 2 O 5 exhibited a greater ability of oxidizing Hg compared to the SiO 2 V 2 O 5 It was suggested that the Hg oxidatio n on the V 2 O 5 doped catalysts follows an Eley Rideal mechanism where HCl, NO, and NO 2 are first adsorbed on the catalyst surface and then react with gas phase Hg 0 This research also reported that a tom ic absorption spectrometry based co ntinuous mercury mo nitors are subject to interferences by ozone due to its strong absorption bands near the Hg absorption line. On the other hand, Hg interfere s with ozone measurement which is based UV adsorption. These mutual interferences can consequently affect the risk a ssessment of human exposure to both Hg and ozone.

PAGE 13

13 CHAPTER 1 INTRODUCTION Mercury and Its Health Effects The 1990 Clean Air Act Amendments (CAAA) listed 189 hazardous air pollutants (HAPs) A mong them, mercury (Hg) has attracted significant attention due to its increased levels in the environment and well documented food chain transport and bio accumulation (Brown et al., 1999) Human exposure by direct inhalation of Hg in the air is not a predominant public health concern because the Hg concentration in the a ir is typically very low. However, Hg in ambient air can eventually be re deposited on land surfaces or directly into rivers, lakes, and oceans, and then biologically enter the food chain. In aquatic systems, Hg is often converted by bacteria to methylmerc ury ( CH 3 Hg + ) which is a neurotoxin and can be magnified through the aquatic food chain hundreds of thousands of times (Ravichandran, 2004) Hg and its compounds act as dangerous and insidious poisons and ca n be adsorbed through the gastrointestinal tract and a lso through the skin and lungs (Bidstrup, 1964) Hi gh conce ntration of Hg can cause impairment of pulmonary and kidney function, chest pain and dyspnousea (Berglund and Bertin, 1969) An extreme e xample of the health effects of Hg is the high dosage exposure from the consumption of methylmercury contaminated fish by the residents living near Minamata Bay in Japan in the 1950s that resu lted in fatalities and severe neurological damage (Mishima, 1992) Mercury Emissions and Regulations According to the Mercury Study Report to Congress prepared by the U.S. Environmental Protection Agency (USEPA) (USEPA, 1997a) the major anthropogenic Hg emission sources are coal fired boilers (33%), municipal waste combustors (19%), industrial and c ommercial boilers (18%), and medical waste incinerators (10%). Hg emissions from manufacturing sources are generally lower compared to combustion sources with the exception of chlor alkali plants using

PAGE 14

14 the mercury cell process and portland cement manufactu ring plants (USEPA, 1997a) Ever since the 1990 Clea n Air Act, the U.S. EPA has issued a series of rules to regulate Hg emissions from solid waste combustors/incinerators (USEPA, 1997c, 2001b) and mercury cell chlor alkali plants (USEPA, 2003) Since c oal fired utility boilers are currently the largest single known source of anthropogenic Hg emissions ( one th ird of the 150 tons of Hg emitted annually ) in the United States the U.S. EPA issued the Clean Air Mercury Rule (CAMR) in 2005 to permanently cap and reduce Hg emissions from coal fired power plants (USEPA, 2005a) CAMR will be i mplemented in two phases, with the first phase cap of 38 tons in 2010 followed by a final cap of 15 tons in 2018. The final cap requires an approximately 70% reduction from the 1999 emission levels Mercury Speciation and Control Technologies There are thr ee basic forms of Hg in the coal derived flue gas: (1) elemental Hg (Hg 0 ), (2) oxidized Hg (Hg 2+ ), predominantly HgCl 2 due to the large excess of chlorine species in the flue gas, and (3) particle bound Hg (Hg p ). During combustion, Hg is released from coal as Hg 0 and as the flue gas cools, some of the Hg 0 can be oxidized or bound on the fly ash. The Hg speciation in the flue gas is determined by various factors including coal properties, boiler operating conditions, flue gas composition, and the time tempe rature profile (Romero et al., 2006 ) Hg 2+ and Hg p are relatively easy to remove from the flue gas using typical air pollution control devices (APCDs). Hg p bound on fly ash particles, is collected in electrostatic precipitators (ESPs) and/or baghouses. Hg 2+ is soluble in water and is read ily captured by wet flue gas desulfurization (FGD) equipment. Hg 0 is volatile and insoluble in water, and thus, it is difficult to be captured using these conventional control technologies. Unfortunately, Hg speciation studies showed that Hg 0 is the domina nt species in flue gas when burning low rank (subbituminous or lignite) coals.

PAGE 15

15 Therefore, the need exists for a low cost Hg oxidation/capturing process that can be applied for the flue gas treatment. Many methodologies have been proposed for Hg emission co ntrol from flue gas. Among them, the technology of sorbent injection, particularly activated carbon injection (ACI), has been investigated most intensively (Pavlish et al., 2003) Both Hg 0 and Hg 2+ can be captured by the sorbent and collected in ESPs and/or baghouses. This technology has been successfully implemented in the municipal waste incinerator industry, where 90% Hg removal can be achieved. However, the application of ACI in coal fired utility boilers is far more challenging due to the shorter gas residence time, the lower equilibrium adsorption capacity and mass transfer rate, and the compromise of fly ash properties by the injected sorbent. The high cost of ACI also limits its application (Pavlish et al., 2003) Recently, selective catalyst reduction (SCR) catalysts are found to be capable of oxidizing Hg 0 in addition to its ability of removing nitrogen oxides (NOx) from the flue gas (Benson et al., 2005; Lee, Srivastava et al., 2004; Lee et al., 2006; Niksa and Fujiwara, 2005b; Senior, 2006 ) The extent of Hg 0 oxidation through SCR processes varies under different operating conditions mechanisms of Hg oxidation over SCR catalysts are yet to be unde rstood. A novel methodology using titanium dioxide (TiO 2 ) based nanostructured sorbents has been demonstrated to be very effec tive for capture of Hg 0 under ultraviolet (UV) irradiation (Lee et al., 2001; Pitoniak et al., 2003; Wu et al., 1998) Wu et al. (1998) and Lee et al. (2001) reported a high level of Hg 0 capture in simulated combustor exhaust using in situ generated TiO 2 particles, while Pitoniak et al. (2003) used a highly porous silica (SiO 2 ) gel doped with TiO 2 nanoparticles and achieved synergistic adsorption and photocatalytic oxidation of Hg 0 in a fixed

PAGE 16

16 bed reactor. The high surface area and open structure of the SiO 2 TiO 2 nanocomposite allow effective irradiation by UV light and thus minimize the mass transfer resistance for Hg 0 (Pitoniak et al., 2003; Wu et al., 1998) Using this material, Hg removal efficiency remained over 90% even after a 477 hr treatment (Pitoniak et al., 2005) Potential applications of the SiO 2 TiO 2 nanocomposite for Hg removal lie in two main areas. First, like ACI, a powdered form of the nanocomposite can be in jected into combustion exhaust upstream of a particle control device (e.g. ESP). Second, a pellet form of the nanocomposite can be used in packed bed columns to treat Hg emissions from flue gas. In this case, the device is preferably installed between an E SP and a wet scrubber Mercury Removal by SiO 2 TiO 2 : Unknowns and Challenges While the SiO 2 TiO 2 nanocomposites have demonstrated prominent effectiveness for Hg removal, the past studies were mainly conducted under room conditions and there is little know ledge so far about the performance of this novel material under real or simulated flue gas conditions. As is known, typical coal derived flue gas consists of a high concentration of water vapor (normally 5~15 % v/v) and various minor gas components such as HCl, SO 2 and NOx. Among the factors that affect the efficiency of Hg 0 capture by the TiO 2 photocatalyst, moisture content in the Hg 0 laden gas was reported to be one of the most important (Pitoniak et al., 2003; Rodriguez et al., 2004) Howeve r, the understanding of the water vapor effects on Hg removal in literature is limited and some of the findings are controversial. It has also been reported that the minor acid gases are important to the heterogeneous adsorption and/or oxidation of Hg 0 on activated carbons or fly ash under flue gas conditions (Carey et al., 1998; Norton et al., 2003) In addition, the typical flue gas temperature at th e cold end of the boiler convective pass is in the range of 120 ~150 C, much higher than the room temperature. Thus, it is expected that the

PAGE 17

17 nature of Hg capture on the SiO 2 TiO 2 nanocomposite would be different under flue gas conditions from that reporte d under room conditions in the past studies. Furthermore, the kinetics for catalytic oxidation of Hg is highly uncertain (Presto and Granite, 2006) The lack of understanding presents a severe limitation in predicting the extent of Hg oxidation in larger scale applications. In literature, kinetic modeling studies on Hg capture have ma inly focused on activated carbon adsorption (Chen et al., 1996; Flora et al., 1998; Meserole et al., 2000; Rostam Abadi et al., 1997) while modeling on photocatalytic oxidation using TiO 2 has mainly focused on degradation of volatile organic compounds (VOCs) (Kim and Hong, 2002; Raillard et al., 2004; Shang et al ., 2002; Son et al., 2004) The kinetics for Hg oxidation using the SiO 2 TiO 2 nanocomposite is yet to be investigated. While the performance of the SiO 2 TiO 2 nanocomposite under flue gas conditions is unknown at this point, it is reasonable to carry out a parallel study that focuses on the development of a modified or even new catalyst which would be more effective on Hg removal in flue gas. Considering the fact that industrial SCR catalysts, with an active phase of V 2 O 5 supported on TiO 2 are capable of o xidizing Hg 0 in addition to its ability of removing NOx (Benson et al., 2005; Lee, Srivastava et al., 2004; Lee et al., 2006; Niksa and Fujiwara, 2005b; Senior, 2006) the addition of V 2 O 5 to the existing SiO 2 TiO 2 nanocomposite would be expected to enhance the catalytic activity. Meanwhile, since the Hg oxidation acro ss SCR catalysts is or full scale studies, a better understanding of the fundamental nature of the catalytic reactions is of great importance to the advancement of the catalysts. Mercury Measurement Interference B oth Hg emission regulations and development of Hg control technologies require that reliable methods be used for accurate Hg measurement. Currently, the EPA accepted methods for

PAGE 18

18 Hg measurement in the United States are manual procedures based on wet chemist ry such as EPA Methods 29 and 101A (for total mercury) and the Ontario Hydro M ethod (for speciated mercury) (Laudal et al., 2004) However, continuous mercury monitors (CMM s) have distinct advantages over these manual methods in that CMMs are able to provide a real time or near real time response for Hg measurements and to perform long term emission measurement. On the other hand, a significant disadvantage of CMMs lies in t heir measurement interferences. A t omic absorption spectrometry ( AAS) is one of the major techniques applied to current CMMs. In the case of AAS, the co ncentration of Hg 0 in a gas sample is determined by measuring the light that is absorbed by Hg atoms at their characteristic wave lengths (usually at the resonance line of 254 nm). Thus, interferences can occ ur when other components of the sample gas possess strong absorption bands near this wavelength (254 nm) Since the 254 nm Hg emission line also falls in to the absorption spectra of ozone, which is capable of absorbing UV light below 290 nm, the presence of ozone in the sampling environment may impact the Hg measurement by AAS based CMMs. Granite and Pennline (2002) studied photochemical oxidation of Hg and speculated that photosensitized formation of ozone m ay interfere with Hg measurement by absorbing UV radiation. However, no quantitative data were reported in literature on the magnitude of ozone interference. Research Objectives To reveal the unknowns and to embrace the challenges mentioned above, five ob jectives are proposed in this doctoral research. The first objective is to study the kinetics of the Hg 0 photocatalyti c oxidation on the SiO 2 TiO 2 nanocomposite The competitive adsorption of water vapor in Hg 0 photocatalytic oxidation will be established in a kinetic expression as well This modeling study is of importance in predicting Hg 0 removal efficiency and is useful for designing an effecti ve reactor, under photocatalytically oxidizing conditions.

PAGE 19

19 The second objective is to perform a mechanistic st udy probing the role of moisture on Hg 0 capture (adsorption and/or photocatalytic oxidation) using a SiO 2 TiO 2 nanocomposite. To provide an overall evaluation of the performance of the SiO 2 TiO 2 nanocomposite, possible reemission of captured Hg species wil l also be examined. The corresponding mechanisms of Hg 0 removal and reemission in the presence of water vapor will be investigated as well. The third objective is to install a fixed bed photocatalytic reactor and to investigate the performance of the SiO 2 TiO 2 nanocomposite under simulated flue gas conditions. The effects of the flue gas components on the removal of Hg 0 by the SiO 2 TiO 2 nanocomposite as well as the surface reaction mechanisms will be explored. An improved understanding of the role of the f lue gas components can help evaluate the potential of applying this novel material for effective Hg control in coal fired power plants. The fourth objective is to develop a method to dope V 2 O 5 on the SiO 2 or SiO 2 TiO 2 composites in an effort to improve the catalytic activity for Hg removal. The SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 composites will be synthesized and characterized. The catalytic abilities of those composites on Hg removal will be tested in a fix bed reactor. The reaction mechanisms on the catalytic r emoval of Hg over the new catalysts will be investigated. The last but not the least important objective of this research is to quantitatively investigate the mutual interferences of ozone and Hg on their measurements. This study may be of particular impor tance to the ambient and indoor measurements of ozone and Hg because these two air pollutants coexist in the environment.

PAGE 20

20 CHAPTER 2 KINETIC STUDY FOR PH OTOCATALYTIC OXIDATI ON OF ELEMENTAL MERC URY ON A SILICA TITANIA NANOCOMPOSIT E Background A soli d understanding of the kinetics of photocatalytic oxidation of Hg 0 is of great importance t o make an effective desig n of the photocatalytic reactor and to predict the reaction rate in larger scale applications Lee et al. (2004) studied Hg 0 oxidation by TiO 2 nanoparticles with UV irradiation in a differen tial bed reactor (DBR) and an aerosol flow reactor (AFR), and correlated the overall reaction rate with the initial Hg 0 concentration and UV intensity. However, the kinetic parameters on water vapor dependence were not available in that study, while water vapor is an important component in the flue gas and plays a critical role in the chemistry of mercury in coal fired boilers (Edwards et al., 2001; Niksa et al., 2001) Rodrguez et al. (2004) developed a mechanistic model to predict Hg 0 capture with in situ generated TiO 2 nanoparticles by solving the equilibrium equations for electron hole pair generation/consumption They also compared their mechanistic model with the Langmuir Hinshelwood (L H) model used by Obee (1996) for characterizing photocatalytic oxidation of certain organic compounds. At low water vapor concentrations, the Hg capture rate predicted by the mechanistic model (Rodriguez et al., 2004) was proportional to the square root of the water vapor concentration, whereas the L H model (Obee, 1996) indicated first order dependence. At high water vapor concentrations, both models predicted a constant Hg captur e rate that was independent of the water vapor concentration. Reprinted with permission from Li, Y., Wu, C. Y., 2007. Kinetic Study for Photocatalytic Oxidation of Elemental Merc ury on a SiO 2 TiO 2 Nanocomposite. Environ. Eng. Sci. 24, 3 12, published by Mary Ann Liebert, Inc., New Rochelle, NY

PAGE 21

21 Some other modeling studies have been done on Hg capture using activated carbon. Rostam Abadi et al. (1997) ap plied an empirical equation to the mass balance for Hg 0 sorption on carbon particles in a duct flow reactor and derived the minimum C/Hg ratio required to reduce Hg 0 at a certain inlet Hg 0 concentration. Chen et al. (1996) derived an equation to model mercury capture when it i s limited by both mass transfer and capacity by assuming that adsorption at the (1998) based on the Langmuir isotherm and by Meserole et al. (2000) based on the Freundlich equation. Several other studies (Kim and Hong, 2002; Raillard et al., 2004; Shang et al., 2002; Son et al., 2004) have been conducted on photocatalytic oxidation of various volatile organic compounds (VOCs) by TiO 2 and the experimental data matched well with the L H kinetic model. This intriguing L H nature of a wide range of VOCs warrants the investigation on the correlation between the kinetics of Hg 0 photocatalyti c oxidation by TiO 2 and the L H rate expression, whereas no relevant research has been done so far. In addition, the L H model takes advantages over the other models previously described in incorporating the effect of competitive adsorption of water vapor. Therefore, the research goal of this chapter was to study the kinetics of the Hg 0 photocatalytic oxidation on a SiO 2 TiO 2 nanocomposite by using the L H model to analyze the kinetic data. The role of water vapor in Hg 0 photocatalytic oxidation was establi shed as well. This kinetic modeling study is of importance in predicting Hg 0 removal efficiency and is useful for designing an effective reactor, under photocatalyzed oxidizing conditions. Materials and Methods Synthesis of SiO 2 TiO 2 N anocomposite The SiO 2 TiO 2 nano composite was synthesized following a sol gel method (Pitoniak et al., 2003) using deionized water, ethanol, tetraethyl orthosilicate ( TEOS ) with HNO 3 and HF added as catalysts to increase the hydrolys is and condensation rates. First, t he chemicals were added to

PAGE 22

22 a polymethylpentene container, and then TiO 2 nanoparticles (Deggusa, P25) we re added to the batch with a magnetic stir plate providing sufficient mixing. After that, t he solution suspended with TiO 2 nanoparticles wa s pipetted into polystyrene 96 well assay plates before the gelation occurred The pellets wer e later aged at room temperature for 2 days and then at 65 C for another 2 d ays. After aging, the pellets we re removed from the plates, rin sed with deionized water to remove any residual acid or ethanol. Next, t he pellets were placed in a programmable oven an d heated a t 103 C for 18 h to remove any residues of liquid solution within the silica network and then at 180 C for 6 h to harden the gel. Finally the temperature was slowly decreased back to room temperature over a 90 min period The final size of an individual cylindrical pellet w as approximately 5 mm in length and 3 mm in diameter. The loading of TiO 2 in the nanocomposite was 1 2 wt%, which corresponded to the optimum performance of Hg 0 removal using the SiO 2 TiO 2 nano composite (Pitoniak et al., 2003) The average BET (Brunauer, Emmett, and Teller equation) surface area of the nanocomposite was measured to be 280 m 2 g 1 using a Quantachrome NOVA 1200 Gas Sorption Analyzer (Boynton Beach, FL ). A pparatus and P rocedure Figure 2 1 shows the schematic diagram of the experimental system. An incoming cylinder air was divided in to three streams, the flowrates of which were controlled by mass flow controllers ( MFC, Model. F MA 5400/5500, Omega Engineering, Inc. Stamford, CT ) The total flowrate remained constant at 2 L/min. One of the air streams was allowed to pass through a water bubbler for a humid flow or to bypass it for a dry flow. The second stream served as dilution to adjust the humidity level The third stream passed through the surface of a liquid Hg 0 reservoir and introduced the saturated Hg 0 vapor into the system. The Hg 0 reservoir was plac ed in an ice water bath to maintain a constant Hg 0 vapor pressure. Downstream of all the gases wa s the fixed bed photocatalytic reactor, t he lower part of which is a cylindrical tube of fused quartz

PAGE 23

23 4.5 cm in diameter and 20 cm in length The reactor was m ounted with a fused quartz center with a diameter of 2 cm, which was used to house a UV lamp The UV light has a peak wavelength of 365 nm with an intensity of 4 mW/cm 2 measured by a UVX radiometer (with a UVX 36 sensor probe). At the bottom of the reactor is a glass frit used to hold the SiO 2 TiO 2 pellets within the bed A thermocouple (TC, Type K, Omega Engineering, Inc.) was used to monitor the temperature on the surface of the pellets. The Hg 0 concentration at the reactor outlet was measured by a RA 915 + Hg analyzer (OhioLumex Co., Cleveland, OH), which is based on Zeeman Atomic Absorption Spectrometry using High Frequency Modulated light polariz ation (ZAAS HFM) (Sholupov et al., 2004) The inlet Hg 0 concentrati on was obtained when the Hg 0 laden air bypassed the reactor. Finally, the air stream passed through a carbon trap before it was exhausted into the fume hood. Two sets of experiments were performed in this study. In the first set, no water vapor was introdu ced into the air stream but with variations in the inlet Hg 0 concentration ( 0.19 to 1.28 mol m 3 or 38 to 256 g m 3 ) In the second set, the inlet Hg 0 concentration remained constant but with changes in water vapor concentration (0 to 0.95 mol m 3 ) In each experiment, the Hg 0 laden air was allowed to pass through the reactor for one hour to ensure that the Hg 0 adsorption on the SiO 2 TiO 2 nanocomposite reached equilibrium, which was monitored by the online Hg analyzer. Then, the photocatalytic reaction was started by turning on the UV lamp and the Hg 0 concentrations were recorded for a certain period of time until no more reduction in Hg 0 concentration was observed. All the experiments were conducted under room conditions. In each test, 2.5 grams of fres h SiO 2 TiO 2 pellets were used, which corresponded to an average of 4 mm bed thickness (approximately one single layer of pellets)

PAGE 24

24 Model Description Photocatalytic oxidation of Hg 0 occurs when the SiO 2 TiO 2 nanocomposite is under UV irradiation as shown in Figure 2 2. The hole electron pairs generated on the TiO 2 particle surfaces lead to the formation of highly reactive hydroxyl (OH) radicals, which are responsible for Hg 0 oxidation to form HgO (Pitoniak et al., 2003; Wu et al., 1998) The mechanism can be described as the following reactions: (2 1) (2 2) (2 3) (2 4) (2 5) Among the factors that affect the efficiency of Hg 0 capture by the SiO 2 TiO 2 nanocomposite, w ater vapor content in the Hg 0 laden air was reported to be one of the most important (Pitoniak et al., 2003) On one hand, surface moisture on TiO 2 nanoparticles is necessary for generating OH radicals (Reaction s 2 4) which are responsible for photocatalytic Hg 0 oxidation. On the other hand, at high water vapor concentrations, competitive adsorption may reduce the number of sites available for Hg 0 (Pitoniak et al., 2003; Rodriguez et al., 2004) Simila r to the studies by other researchers (Canela et al., 1998; Obee, 1996; Obee and Hay, 1997) the rate of photocatalytic oxidation of Hg 0 is defined as (2 6) where C Hg in is Hg 0 concentration at the inlet of the reactor C Hg out is Hg 0 concentration at the outlet of the reactor at steady state, Q is the volumetric flow rate of the Hg 0 laden air (2 L min 1

PAGE 25

25 or 0.12 m 3 h 1 ), and A e is the effective surface area of the pellets that is exposed to UV light. It should be noted t hat only a thickness of 0.1 mm from the surface of the pellets and only the areas facing the UV light can effectively contribute to Hg 0 oxidation (Pitoniak et al., 2005) Thus, A e can be calculated as (2 7) where SA is the specific surface area of the pellets (280 m 2 g 1 ), m is the mass of pellets used (2.5 g), f V is the volume fraction of the 0.1 mm thickness layer that UV light can penetrate (estimated to be 0.15), and f P is the packing factor that accounts for the fraction of the surface areas exposed to UV light (estimated to be 0.5). To correlate the experi mental data of photocatalytic oxidation rate of Hg 0 the L H rate equation was used. If the concentration of water vapor is constant, the L H expression can be simplified as (2 8) where r is the reaction rate ( mol m 2 h 1 ), k is the L H rate constant ( mol m 2 h 1 ), K Hg is the Langmuir adsorption constant of Hg 0 (m 3 mol 1 ), and C Hg is the Hg 0 concentration ( mol m 3 ). C Hg is normally assigned to be the bulk or inlet concentration, C Hg in (Obee, 1996; Obee and Hay, 1997) The inverse of Equation 2 8 gives (2 9) If the assumed L H expression is valid for Hg photocatalytic o xidation, a plot of r 1 vs. C Hg 1 should be linear. Subsequently, the values of k and K Hg can be derived from the combination of the intercept and the slope of the linear line. From these values, the photocatalytic Hg 0 oxidation rate can be predicted by th e L H model.

PAGE 26

26 Similar to the modeling studies conducted by other researchers on photocatalytic oxidation of organic pollutants (Obee and Hay, 1997; Shang et al., 2002) when water vapor is present, the inhibitory effect of water vapor on Hg 0 photocatalytic oxidation can be assumed according to the following L H form (2 10) where K w is the Langmuir adsorption constant of water and C w is the water vapor concentration. The inverse of Equation 2 10 gives (2 11) The value of K Hg can be obtained from previous analysis when water vapor i s not present. When C Hg remains at a constant level and only C w varies, a plot of r 1 versus C w should be linear if it follows the L H model expression. Then the values of k and K w can be derived from the plot. Results and D iscussion Effect of Hg 0 Concentr ation Figure 2 3 shows the outlet Hg 0 concentration as a function of UV illumination time at six different inlet levels ranging from 0.19 to 1.28 mol m 3 (38 to 256 g m 3 ) when water vapor was not present. The outlet Hg 0 concentration dropped quickly w hen UV was first turned on for a few minutes and then gradually leveled off. From 20 to 30 min, no significant change in outlet Hg 0 concentration was observed and the pellet surface temperature remained almost constant (42.7 0.3 C). Therefore, 30 min wa s taken as the time the system reached steady state. Experiments were repeated three times at each inlet Hg 0 concentration level. The average Hg 0 removal efficiency ranged from 90 to 95% but was not an apparent function of the inlet Hg 0 concentration.

PAGE 27

27 At e ach inlet Hg 0 concentration level, the photocatalytic oxidation rate r can be calculated from Equation 2 6 and the average value can be obtained. A plot of r 1 vs. C Hg 1 is shown in Figure 2 4 and the observed linear relationship indicates that the kinetic s of Hg 0 photocatalytic oxidation fits the L H model very well. From Equation 2 9, values of the L H rate constant k and the Langmuir adsorption constant K Hg were calculated to be k = 0.024 mol m 2 h 1 and K Hg = 0.094 m 3 mol 1 Substituting the values o f k and K Hg back into Equation 2 8, the photocatalytic oxidation rates at different inlet Hg 0 concentrations can be predicted by the L H model. In the kinetic study of Hg 0 photocatalytic oxidation on TiO 2 particles by Lee et al. (2004) the reaction orders with respect to initial Hg 0 concentration (wh ich ranged from 1 10 g m 3 or 0.005 0.05 mol m 3 ) were reported to be 1.4 for the differential bed reactor (DBR) and 1.1 for the aerosol flow reactor (AFR). They also suggested that the higher value obtained for the DBR might be due to inherent experime ntal errors. In this work, the fix bed reactor design is similar to the DBR used by Lee et al. (2004) With the inlet Hg 0 concentration ranging from 0.19 to 1.28 mol m 3 in this study, the value of K Hg C Hg is far less than 1. Thus, Equation 2 8 can be simplified as (2 1 2) Equation 2 12 shows that the reaction order with respect to the initial Hg 0 concentration is 1, which is representative of a practical sorbent process (Lee, Biswas et al., 2004) Lee et al. (2004) also correlated the overall reaction order with respect to the UV intensity and reported an order of 0.35 for the DBR and 0.39 for the AFR. In this study, the effect o f UV intensity was not investigated. Useful prediction results can be obtained from the L H model as shown in Figure 2 5, which is characterized by a steep rise of Hg 0 photocatalytic oxidation rate at inlet concentrations

PAGE 28

28 approximately less than 20 mol m 3 and subsequent mild increase at higher concentrations. Due to the limitations on the capability of the Hg generation unit and the measurement range of the Hg analyzer, experimental data greater than 20 mol m 3 were not available in this study. Further research is needed on validating the L H feature of Hg 0 photocatalytic oxidation in the high concentration range. On the other hand, it should be noted that typical Hg concentrations in coal fired power plant flue gases are less than 0.05 mol m 3 (10 g m 3 ) (Pavlish et al., 2003) which locates this process at the very lower end of the steep rise range. This further demonstrates the great potential of the SiO 2 TiO 2 nanocomposite for Hg 0 removal from emission sources e ven with much higher Hg 0 concentrations. Effect of Water V apor Water vapor experiments were conducted at a constant inlet Hg 0 concentration of 0.66 mol m 3 with variations in the water vapor concentration, as shown in Figure 2 6. As the water vapor con centration increased from 0 to 0.95 mol m 3 the steady state Hg 0 removal efficiency (at 30 min) also decreased from 93% to 24%. This demonstrates a significant inhibitory effect of water vapor on photocatalytic Hg 0 oxidation. Experiments were repeated thr ee times at each water vapor concentration level. The average values of r 1 versus C w at a constant inlet Hg 0 concentration are plotted in Figure 2 7. The linear relationship between them shows a good match of the experimental data with the L H model expre ssion in humid air (Equation 2 11). The intercept and the slope of the linear plot give the L H rate constant k = 0.031 mol m 2 h 1 and the Langmuir adsorption constant of water K w = 4.39 m 3 mol 1 The previously obtained K Hg (0.094 m 3 mol 1 or 9.410 4 m 3 mol 1 ) is four orders of magnitude larger than K w which indicates that the adsorption ability of the SiO 2 TiO 2 nanocomposite is much greater for Hg 0 than for water vapor. However, water vapor plays a very important role in Hg 0 removal because in the

PAGE 29

29 fl ue gas Hg 0 concentration is at such trace levels (seven to eight orders of magnitude smaller) compared to that of water vapor. Now that all the kinetic parameters have been estimated, the L H model can be used to predict the rate of Hg 0 photocatalytic oxid ation at any level of inlet Hg 0 concentration and water vapor concentration. Figure 2 8 compares the experimental data of the Hg 0 photocatalytic oxidation rate with L H model predictions in humid air. For the six experimental conditions shown in Figure 2 8 the deviations of the experimental data from the L H model predictions are less than 15%, which are within an allowable range of experimental error. This result once again verifies the L H nature of Hg 0 photocatalytic oxidation by the SiO 2 TiO 2 nanocompo site, and suggests that it is appropriate to apply the L H model to predict the photocatalytic reaction rate. Using the L H model, the rate of Hg 0 oxidation by SiO 2 TiO 2 under coal combustion flue gas conditions can be predicted. At an inlet Hg 0 concentrat ion of 0.05 mol m 3 (10 g m 3 ) and a water vapor concentration of 10 vol%, the reaction rate is calculated to be 7.710 6 mol m 2 h 1 and the Hg 0 removal efficiency is around 7% in the current system ( Q = 0.12 m 3 h 1 and A e = 52.5 m 2 or 2.5 g of pelle ts used). However, a 95% removal efficiency can be achieved by increasing A e by 14 fold (using 35 g of pellets or 56 mm bed height) which is practically applicable in a bench scale reactor like the one used in this work. In addition, increasing the UV powe r level can be another option to reduce the required amount of catalysts as the reaction rate is proportional to the UV intensity (Lee, Biswas et al., 2004) It is generally believed (Kim and Hong, 2002; Pitoniak et al., 2003; Raillard et al., 2004; Rodriguez et al., 2004; Shang et al., 2002) t hat the inhibitory effect of water vapor on photocatalytic reactions at relatively high water vapor concentrations is due to the competition between water vapor and the pollutants at the TiO 2 surface, i.e., a high concentration of water

PAGE 30

30 vapor blocks the ad sorption sites from pollutants. Unlike the mechanistic model developed by Rodrguez et al. (2004) where water vapor promoted Hg 0 capture by in situ generated TiO 2 particles at low water vapor concentrations (<2000 ppm v or 0.0815 mol m 3 at 25C), the Hg 0 capture rate by the SiO 2 TiO 2 nanocomposite in this study reached maximum in dry air and decreased as the water vapor concentration increased. An explanation for the highest photocatalytic oxidation rate without water vapor may be related to the silanol (Si OH) groups on the surface of the SiO 2 TiO 2 nanocomposite. The sol gel reactions are performed in water/alcohol systems that cannot avoid the reverse reactions during the sol gel process, i.e. hydrolysis and alcoholysis for silanol forma tion (Yang and Chen, 2005) Yang and Chen (2005) reported that a SiO 2 nanolaye r around TiO 2 nanocrystals can enhance the efficiency of photocatalysis because the transfer of electrons to the silica sites and the hole scavenging by the hydroxides at the TiO 2 SiO 2 interface prevent the electrons and holes from recombination. In the Si O 2 TiO 2 nanocomposite produced in this work, the hydroxyl groups from silanols may act as traps for the holes generated by TiO 2 under UV irradiation and thus, an adequate number of hydroxyl radicals may be produced resulting in photocatalytic oxidation of Hg 0 even in the absence of water vapor. Similar findings were reported by Kim and Hong (2002) that photodegradation of methanol by TiO 2 reached the highest rate at considerably low water concentrations, which was explained due to the production of hydroxyl radicals from hydroxyl groups of methanol itself. In this manner, hydroxyl radicals generated from water molecules might be insignificant, and addition of water vapor may only prohibit Hg 0 photocatalytic oxidati on by blocking the Hg 0 adsorption sites on the surface of the SiO 2 TiO 2 nanocomposite. In the system of Rodrguez et al. (2004) Hg 0 photocatalytic oxidation rate increased with water vapor at low water vapor concentrations, which may be because water vapor was the only source

PAGE 31

31 for hydroxyl radical production. Comparisons between this study and that by Rodrguez et al. (2004) suggest that hydrophilic adsorbents (such as SiO 2 TiO 2 nanocomposite) may have better performanc e in Hg 0 removal at dry or very low humidity environment, and on the other hand, hydrophobic materials (such as TiO 2 nanoparticles) may yield a larger Hg 0 removal rate as the humidity increases. However, the performance of both types of materials will be i nhibited at very high water vapor concentrations. Summary The kinetics of Hg 0 photocatalytic oxidation on a SiO 2 TiO 2 nanocomposite under UV irradiation was studied through experiments in a fixed bed reactor. A Langmuir Hinshelwood model was used to analyz e the kinetic data. Good agreement between the experimental data and the L H model was demonstrated, indicating the validity of using the L H model to describe the kinetics of Hg 0 photocatalytic oxidation. Model predictions demonstrate a great potential of the SiO 2 TiO 2 nanocomposite for Hg 0 removal even at very high Hg 0 concentrations. The rate of photocatalytic Hg 0 oxidation increased when the inlet Hg 0 concentration increased and it reached a maximum value in the absence of water vapor. The addition of w ater vapor was found to inhibit Hg 0 photocatalytic oxidation, which may be explained by the competitive adsorption of water vapor with Hg 0 on the TiO 2 surface.

PAGE 32

32 Figure 2 1. Experimental system for kinetic studies.

PAGE 33

33 Figure 2 2. Description of Hg 0 photocatalytic oxidation on SiO 2 TiO 2 nanocomposites.

PAGE 34

34 Figure 2 3. Photocatalytic oxidation of Hg 0 at different inlet Hg 0 concentrations without water vapor.

PAGE 35

35 Figure 2 4. I nverse of Hg 0 photocatalytic oxidation rate versus the inverse of inlet Hg 0 concentration (without water vapor).

PAGE 36

36 Figure 2 5. Rate of Hg 0 photocatalytic oxidation versus inlet Hg 0 concentration without water vapor (solid circles: experimental data; solid line: L H model).

PAGE 37

37 Figure 2 6. Photocatalytic oxidation of Hg 0 at a constant inlet Hg 0 concentration of 0.66 mol m 3 with variation in water vapor concentration.

PAGE 38

38 Figure 2 7. I nverse of Hg 0 photocatalytic oxidation rate versus water vapor concentrati on at a constant inlet Hg 0 concentration of 0.66 mol m 3

PAGE 39

39 Figure 2 8. Rate of Hg 0 photocatalytic oxidation versus inlet Hg 0 concentration at different water vapor concentrations (markers: experimental data; lines: L H model).

PAGE 40

40 CHAPTER 3 ROLE OF MOISTURE IN ADSORPTION, PHOTOCAT ALYTIC OXIDATION, AN D REEMISSIO N OF ELEMENTAL MERCU RY ON A SILICA TITANIA NANACOMPOSIT E Background Among the factors that affect the efficiency of Hg 0 capture by the TiO 2 photocatalyst, moisture content in the Hg 0 laden gas was reported to be one of the most important (Pitoni ak et al., 2003; Rodriguez et al., 2004) Using in situ generated TiO 2 nanoparticles in a flow reactor, Rodr guez et al. (2004) developed a mechanistic model and reported that Hg 0 capture was promoted by low water vapor concentrations ( 700 to 1800 ppm v ) but remained constant at higher water vapor concentrations. They also speculated that very high water vapor concentrations can inhibit Hg 0 oxidation by occupying available adsorption sites. Using a SiO 2 TiO 2 nanocomposite in a fixed bed f low reactor, Pitoniak et al. (2003) reported that when the relative humidity increased from 15% to 90% at room temperature, the rate of Hg 0 adsorption decreased but that the rate of photocatalyti c oxidation remained constant. While there is limited understanding o f the effect of water vapor on Hg 0 capture on TiO 2 surfaces many research studies have investigated the effect of water vapor on photodegradation of organic pollutants in air streams us ing TiO 2 nanoparticles or thin films. Obee and Hay (1997) reported that moisture in the range of 0 to 25000 ppm v inhibited photooxidation of ethylene by a TiO 2 coated glass plate. Shang et al. (2002) found that water vapor at concentrations of 3.7 22.4 g m 3 inhibited photocatalytic oxidation of heptane in a quartz reactor coated with TiO 2 particles. It was also reported that water vapor strongly inhibits t he oxidation of t richloroethylene ( TCE) and acetone (Kim and Hong, 2002) but enhances the oxidation of toluene (Augugliaro et al., Reprinted with permission from Li, Y., Wu, C. Y., 2006. Role of Moisture in Adsorption, Photocatalytic Oxidation, and Reemission of Eleme ntal Mercury on a SiO 2 TiO 2 Nanocomposite. Environ. Sci. Technol. 40, 6444 6448. Copyright 2006 American Chemical Society.

PAGE 41

41 1999; Kim and Hong, 2002) The results obtained using batch reactors coated with TiO 2 thin films were consistent with those using flow reactors packed with TiO 2 particles. These findings indicate that water vapor can either promote or inhibit photocatalyitc oxidation of different organic pollutants. Th is uncertainty about the effect of water vapor necessitates further investigation of the effect of moisture on Hg 0 oxidation by TiO 2 based photocatalysts. The goal of this chapter was to probe the role of moisture on Hg 0 capture (adsorption and/or photoca talytic oxidation) using a SiO 2 TiO 2 nanocomposite. This study did not aim to explore the maximum Hg removal efficiency under certain emission conditions. Thus, the experimental conditions used in this study (65 C gas temperature and up to 23,000 ppm v H 2 O ) were designed to explore the range of emission conditions encountered in various combustion and manufacturing processes, and not to be representative of any specific process. In this study, possible reemission of captured Hg species was also examined, t o provide an overall evaluation of the performance of the SiO 2 TiO 2 nanocomposite. The corresponding mechanisms of Hg 0 removal and reemission were investigated as well. Experimental Synthesis of SiO 2 TiO 2 N anocomposite The SiO 2 TiO 2 nanocomposite was made by a sol gel method using deionized water, ethanol and tetraethyl orthosilicate (TEOS) N itric acid (HNO 3 ) and hydrogen fluoride (HF) were used as catalysts to increase the hydrolysis and condensation rates A detailed synthetic procedure has been report ed by Pitoniak et al. (2003) The nanocomposite was prepared in the form of cylindrical pellets approximately 5 mm in length and 3 mm in diameter. The weight fraction of TiO 2 in the prepared SiO 2 TiO 2 nanocomposite was approximately 13%.

PAGE 42

42 Experimental S etup A schematic diagram of the experimental system is shown in Figure 3 1 The incoming air flow was divided in to three streams, the flowrates of which were controlled by mass flow controllers ( FMA 5400/5500, Omega) The total flowrate remained constant at 2 L/min. One of the air streams was allowed to pass through a water bubbler (to provide a humid flow ) or to bypass it for dry conditions The second stream served as dilution to adjust the humidity level The third stream passed through the surface of a liquid Hg 0 reservoir and introduced Hg 0 vapor laden air into the system. The Hg 0 reservoir was placed in an ice water bath to maintain a constant Hg 0 vapor pressure. After the three streams converged a humidity sensor (HX94C, Omega) was used to measure the relative humidity from which the partial pressure (volume fraction) of water can be calculated Downstream wa s the fixed bed photocatalytic reactor, t he lower part of which wa s a cylindrical tube of fused quartz 4.5 cm i n diameter and 20 cm in length. The gas stream passed through the reactor from top to bottom. The reactor was mounted with a fused quartz center 2 cm in diameter, which was used to house a UV lamp The UV light delivered 4 mW/cm 2 i ntensity measured by a UVX radiometer (with a UVX 36 sensor probe) at a peak wavelength of 365 nm. Aluminum foil was wrapped around the cylindrical tube to reflect UV energy. A heating mantle (regulated by a temperature controller) was used to heat the rea ctor to each selected temperature, which was monitored by a thermocouple (type K, Omega). At the bottom of the reactor a glass frit was used to hold the SiO 2 TiO 2 pellets within the bed. In this study, 2.5 g of fresh pellets was used in each test which g ave an average bed thickness of 4 mm (approximately one layer of pellets) Before each test, the pellets were heated at 130 C for 3 h to remove any moisture that may have adsorbed from the storage environment.

PAGE 43

43 A RA 915+ Hg analyzer (OhioLumex) was used to measure Hg 0 concentration at the outlet of the reactor. The Hg analyzer is based on Zeeman Atomic Absorption Spectrometry which is selective only for Hg 0 and capable of providing a real time response every 1 s The calibration of the Hg analyzer was cond ucted by the manufacturer using a Dynacal permeation device, which is certified traceable to NIST (National Institute of Standards and Technology) standards. In this study, the high concentration mode of the analyzer was used (with a detection limit of 0. 5 g m 3 and an upper measurable concentration of 200 g m 3 ). A condenser was installed upstream of the Hg analyzer to remove excess moisture in the gas stream to minimize possible interference from water vapor. Although a Hg speciation converting unit co uld be used to analyze both Hg(II) and Hg 0 in the gas phase, it was not installed in this study. This is because HgO is the only product for the reaction between Hg 0 and hydroxyl radicals (Pal and Ariya, 2004; Sommar et al., 2001) and its extremely low saturation vapor pressure, 9.210 12 Pa at 25 C (Schroeder and Munthe, 1998) causes HgO to deposit on the catalyst in the reactor. Baseline Hg 0 concentration was obtained when the Hg 0 laden air bypassed the reactor. The Hg 0 removal efficiency was obtained by comparing the o utlet Hg 0 concentration with the baseline level. Finally, the air stream was passed through a carbon trap before it was exhausted into the fume hood. Results and Discussion Role of Moisture i n Hg 0 Capture To investigate the effect of moisture on Hg 0 captu re by the SiO 2 TiO 2 nanocomposite, experiments were conducted at different water vapor concentrations (0, 13000, and 23000 ppm v ). Blank runs without the nanocomposite were performed at concentrations of around 65 g m 3 Hg 0 and 0 and 23000 ppm v water vapor Less than 0.5% reduction in Hg 0 concentration was observed when the Hg 0 laden air passed through the reactor with or without UV light. Tests were also conducted to examine any possible interference of water vapor on measurements by the Hg

PAGE 44

44 analyzer. While exactly the same experimental parameters were maintained, variations in the Hg readings were less than 1% when switching from a dry to a humid (23000 ppm v H 2 O) condition, indicating negligible interference from water vapor in this range. In each experimen t, the baseline Hg 0 concentration (Hg 0 BL ~65 g m 3 ) was measured first. Then the Hg 0 laden air was passed through the reactor for 10 min and concentrations defining the initial adsorption (IA) of Hg 0 were recorded. Next, the UV light was turned on for 10 min allowing photocatalytic oxidation of Hg 0 to take place and then turned off for 5 min. Another two such UV on/off cycles were repeated, and concentrations defining the final removal (FR) efficiency were recorded. After that, Hg 0 BL was checked again, an d the Hg 0 laden air was passed through the reactor without UV for 15 min. Finally concentrations to compute the adsorption at the end of the test (EA) were recorded. The temperature at the pellet surface was maintained near 65 C throughout the experiment Measured 5 mm above the pellets, the temperature of the gas passing through the pellets was approximately 1.5 C lower than that of the pellets. To maintain a relatively constant temperature, the heating mantle around the reactor was turned off when the UV lamp was switched on in each cycle and back on when UV was switched off. The fluctuations in temperature of both pellets and gas were measured to be 2 C. A preliminary test showed that the change in the final removal efficiency was less than 5% when t he average temperature of the pellets increased from 65 to 70 C. This indicated that the fluctuations within this range of temperature had negligible effects on the reaction rate. Figure 3 2 shows the dimensionless Hg 0 concentration [Hg 0 /Hg 0 BL ] at the out let of the reactor in dry and humid conditions. Under dry conditions (Figure 3 2A), Hg 0 removal efficiency increased with successive cycles and FR reached 95% during the fourth irradiation. Adsorption

PAGE 45

45 was initially insignificant (IA = 5%), but it was enhan ced at the end of the test (EA = 22%). In contrast, when the water vapor concentration was increased to 13000 ppm v (Figure 3 2B), Hg 0 removal increased only slightly with successive cycles and FR ended at only 51%. When the water vapor concentration was f urther increased to 23000 ppm v (Figure 3 2C), FR decreased to 28%. In addition, Hg 0 adsorption was found to be insignificant throughout the test under these two humid conditions. A separate test was conducted extending the adsorption time to 2 h (without UV), and no adsorption was observed. This verified that the recorded values of IA and EA were measured at adsorption equilibrium. Experiments under the three humidity conditions were repeated and similar results were recorded. The results indicated that in creased humidity can significantly suppress both Hg 0 adsorption and photocatalytic oxidation on the nanocomposite. Role of Moisture i n Hg 0 R eemission Potential reemission of Hg species from the SiO 2 TiO 2 nanocomposite after their capture is an important fa ctor in evaluating the overall performance of the nanocomposite. The experiment to examine Hg reemission began by exposing 2.5 g of fresh pellets for 3 h to UV light in a stream containing approximately 300 g/m 3 Hg 0 vapor in dry air at room temperature. D uring the 3 h pretreatment, Hg 0 capture efficiency by the nanocomposite reached about 90% after 30 min and remained relatively steady between 90 and 95% for the rest of the time (as shown in the insert in Figure 3 3). The Hg species retained on the nanocom posite are predicted to be a mixture of HgO (due to photocatalytic oxidation) and Hg 0 (due to adsorption enhanced by HgO) (Pitoniak et al., 2005) After the pretreatment, Hg 0 release from the nanocomposite at room temperature was examined by feeding Hg 0 free air into the reactor (see Figure 3 3). During the first 5 min, dry air

PAGE 46

46 (condition I) was allowed to pass through the pretreated pellets and only traces of reemitted Hg 0 were found. However, immediately after the air stream was switched from dry to humid ([H 2 O] = 23000 ppm v condition II), a significant release of Hg 0 was observed, peaking briefly at 68 g/m 3 about 23% of the Hg 0 feeding level during pretreatment. Even though the Hg 0 reemission level decreased over time, it remained relatively high (43 g/m 3 ) after 20 min. Hg 0 reemitted during this 20 min period was calculated to be approxi mately 2.1% of the total Hg species (Hg 0 and HgO) retained on the pellets during the 3 h pretreatment. At 25 min, UV light was turned on while the air stream remained humid (condition III), and the Hg 0 concentration quickly dropped to approximately 10 g/ m 3 This suggests that a large portion of the Hg 0 reemitted was photooxidized during irradiation. At 35 min, the UV light was turned off (return to condition II), and the rate of Hg 0 reemission did not recover to the previous high level, but further decrea sed to approximately 6.5 g/m 3 Turning the UV on again (return to condition III) at 45 min caused the Hg 0 concentration to return to an extrapolation of the line observed during the first period at condition III. When the condition was switched at 55 min to dry air with UV on (condition IV), the Hg 0 concentration decayed to approximately zero in 5 min. The results of the reemission test were repeatable using another batch of pellets undergoing the same pretreatment procedure. One may draw three conclusions about Hg 0 reemission under the conditions tested: (1) Hg 0 reemission does not occur in dry air, with or without UV light. (2) Introducing water vapor causes significant Hg 0 reemission, which slowly decreases over time. (3) Exposure to UV light in humid ai r can either inhibit or promote Hg 0 reemission. Mechanisms of Hg 0 Capture and R eemission To explain these intriguing findings, a comprehensive model is developed in this work elucidating the fate of Hg species on the surface of this SiO 2 TiO 2 nanocomposit e. In this model,

PAGE 47

47 as illustrated in Figure 3 4, Hg 0 capture is accomplished by photocatalytic oxidation and adsorption, while Hg 0 reemission results from desorption and photocatalytic reduction. Photocatalytic oxidation of Hg 0 occurs during UV irradiation of the SiO 2 TiO 2 nanocomposite, as shown in Figure 3 4a. Hg 0 is oxidized by OH radicals generated on the TiO 2 surface, and HgO has been reported to be the final oxidation product in literature (Lee et al., 2001; Pal and Ariya, 2004; Pitoniak et al., 2003; Pitoniak et al., 2005; Rodriguez et al., 2004; Sommar et al., 2001) Pal and Ariya (2004) experimentally identified HgO as the only product for gas phase reaction of Hg 0 with OH radicals wh ile HgOH was suggested to be an intermediate reaction product (Pal and Ariya, 2004; Rodriguez et al., 2004; Sommar et al., 2001) that has a very short lifetime (Goodsite et al., 2004) The fast removal of Hg 0 in this study (see Figure 3 2) was consistent with rates reported by researchers (Pitoniak et al., 2003; Pitoniak et al., 2005) using similar experimental systems. It was faster than OH Hg reactions reported under simulated atmospheric conditions (Calvert and Lindberg, 2005; Pal and Ariya, 2004; Sommar et al., 2001) most likely due to the much higher concentration of OH radicals produced on the high s urface area, open structured SiO 2 TiO 2 nanocomposite. The overall mechanism of photocatalytic oxidation can be described by the following reactions: (3 1) (3 2) (3 3) (3 4) (3 5) It should be noted that the specifications of the cylinder air indicated a water vapor concentration of less than 24 ppm v This low concentration of water vapor could not be det ected

PAGE 48

48 concentration is 3 orders of magnitude higher than that of Hg 0 used in this study. Further, the SiO 2 TiO 2 pellets have a great capacity to adsorb water vapor. T ests demonstrated that 2.5 g of SiO 2 TiO 2 pellets can adsorb an average of 0.16 g of water vapor in 30 min when exposed to an air flow containing 23000 ppm v water vapor. Therefore, even though the pellets were pretreated at 130 C for 3 h before each exper iment to remove any interior moisture, they adsorbed enough moisture (compared to the trace amount of Hg 0 ) from the cylinder air to support the generation of OH radicals and subsequent photocatalytic oxidation. Physisorption of Hg 0 is minor if the SiO 2 TiO 2 nanocomposite is not exposed to UV light, but it can be enhanced by HgO that is oxidation takes place (Figure 3 4b). This is likely due to the high affinity between Hg 0 and HgO, which was characte rized by a decrease in the contact angle of Hg on the HgO enriched sorbent surface (Pitoni ak et al., 2005) However, this enhanced adsorption ability was not observed in humid conditions due to the inhibitory effect of water vapor. It is generally believed that water vapor inhibits photocatalytic reactions by blocking the available adsorption sites on the surface of TiO 2 catalysts (Kim and Hong, 2002; Pitoniak et al., 2003; Shang et al., 2002) The results of this study suggest that desorption of bound Hg 0 by a high concentration of water vapor (Figure 3 4c) also contributes to the reduced Hg 0 capture rate. The reason is likely related to the photoinduced s uperhydrophilicity of TiO 2 surfaces (Fujishima et al., 2000; Wang et al., 1997) During the process of photocatalysis, t he electrons tend to reduce the Ti(IV) cations to the Ti(III) state, and the holes oxidize the O anions. In the process, oxygen atoms are ejected, creating oxygen vacancies Water molecules can then occupy these oxygen vacancies and create adsorbed OH groups, which tend to make the surface hydrophilic.

PAGE 49

49 Hence, in our experiments, when humid air passed through the reactor, the superhydrophilic surface of TiO 2 may attract excessive amount of water and result in ejection of adsorbed Hg 0 (which is superhydrophobic) from the surface (Figure 3 4c). This mechanism is similar to the self cleaning (stain proofing) quality of TiO 2 coated glass (Fujishima et al., 2000; Guan, 2005) from which organic stains are washed away by rainfall (or water) on the superhy drophilic surface. A sharp decrease in Hg 0 concentration was detected when switching from the initial condition II to III (Figure 3 3). A very possible reason is that the desorbed Hg 0 was reoxidized upon UV irradiation and re captured onto the pellets. Ho wever, the increase in Hg 0 concentration when switching from the second condition II to III indicated that UV irradiation contributed to the reemission of Hg 0 from the pellets. This can be explained by the photocatalytic reduction of HgO to Hg 0 by the free electrons generated on Ti O 2 surface under UV light (Figure 3 4d). The mechanism is expressed in Reaction 3 6, which has a reduction potential of 0.098V vs. normal hydrogen electrode (NHE) (Meites, 1963) : (3 6) To validate the occurrence of Reaction 3 6 in this study, its redox potential was compared with those in other photocatalytic reactions on TiO 2 reported in the literature. Fujishima et al. (2000) reported reduction of O 2 to H 2 O 2 on a TiO 2 photocatalyst. Zhang et al. (2004) used TiO 2 modified sewage sludge carbon for photocatalytic removal and recovery of Hg 2+ in the form of Hg 0 from water. The redox potentials of O 2 / H 2 O 2 and Hg 2+ / Hg are 0.28V (Fujishima et al., 2000) and 0.85V (Meites, 1963) respectively. Thus, it is reasonable to infer that Reaction 3 6, which has a much lower redox potential (0.098V), can occur in our system. The necessity of H 2 O

PAGE 50

50 in Reaction 3 6 i s also consistent with the finding that no Hg 0 reemission occurred in dry air with UV irradiation. Separate experiments were conducted in pure O 2 and N 2 respectively, and the results further supported the hypothesis that photocatalytic reduction caused t he reemission of Hg 0 Under humid conditions and UV irradiation, a higher Hg 0 reemission level was observed in N 2 than in O 2 Because the redox potential of O 2 / O 2 ( 0.28 V) is lower than that of HgO/Hg, O 2 is a stronger electron trap than HgO. The absence of O 2 increased the chance for HgO to trap electrons, and thus more HgO was reduced to Hg in pure N 2 The intriguing finding that UV irradiation can result in either inhibition or promotion of Hg 0 reemission (as shown in Figure 3 3) can then be explaine d by the competition between photocatalytic oxidation of reemitted Hg 0 to form HgO (Reaction 3 5) and photocatalytic reduction of HgO to form Hg 0 (Reaction 3 6), accompanied by the physical desorption of Hg 0 caused by water vapor at the same time. At the m oment when the condition was first changed from II to III, the concentration of desorbed Hg 0 was very high, and thus photocatalytic oxidation prevailed over reduction, which resulted in a sharp decrease in Hg 0 concentration. Over time, the rate of photocat alytic oxidation decreased as the Hg 0 desorption rate decreased. When the condition was switched from II to III for the second time, the rate of photocatalytic oxidation dropped to a lower level than that of photocatalytic reduction, so the level of Hg 0 re emission increased. The above discussion boils down to a conclusion that Hg 0 capture on the SiO 2 TiO 2 nanocomposite in a humid environment under UV irradiation is controlled by four mechanisms: adsorption; photocatalytic oxidation; desorption; photocatalyt ic reduction. Water vapor concentration is a significant parameter affecting the Hg 0 capture efficiency. The inhibitory

PAGE 51

51 effect of water vapor is due to its competitive occupancy of the available adsorption sites, displacement of adsorbed Hg 0 and participa tion in the photocatalytic reduction of HgO to Hg 0 Hg 0 reemission seems to be inevitable since humidity exists in most environmental conditions. However, the mechanisms discussed above imply that appropriate application of UV irradiation can be utilized t o mitigate this Hg 0 reemission. The difference between the first and second condition II in Figure 3 3 shows that UV treatment caused a significant drop of the Hg 0 reemission level. However, further exposure to UV light caused an increase in the Hg 0 reemis sion due to the dominant photocatalytic reduction later on. Therefore, determining the optimal time of UV treatment and avoiding further exposure to UV sources (including sunlight) are critical to achieving minimal Hg 0 reemission. Summary A novel silica ti tania (SiO 2 TiO 2 ) nanocomposite has been developed to effectively capture elemental mercury (Hg 0 ) under UV irradiation. Moisture has been reported to have an 0 re moval and reemission as well as the corresponding mechanisms was investigated. Hg 0 removal experiments were carried out in a fixed bed reactor at 65 C using air as the carrier gas. Without UV irradiation, Hg 0 adsorption was found to be insignificant, but it could be enhanced by the photocatalytic oxidation product, mercuric oxide (HgO), possibly due to the high affinity between HgO and Hg 0 Under dry conditions 95% of Hg 0 can be removed; however, increased humidity levels remarkably suppress both Hg 0 adsor ption and photocatalytic oxidation. Introducing water vapor can also result in significant reemission of captured Hg 0 from the nanocomposite, which may be ascribed to the repellant effect of water vapor adsorbed on the superhydrophilic TiO 2 surface. Exposu re to UV light was found either to prohibit Hg 0 reemission when photocatalytic oxidation of reemitted Hg 0 prevailed or to promote Hg 0 reemission when

PAGE 52

52 photocatalytic reduction of Hg O to Hg 0 dominated later on. The results indicate that minimization of Hg 0 r eemission can be achieved by appropriate application of UV irradiation.

PAGE 53

53 Figure 3 1 E xperimental system for studies on the role of water vapor.

PAGE 54

54 Figure 3 2. Dimensionless Hg 0 concentration at the reactor outlet (A, [H 2 O] = 0 ppm v ; B, [H 2 O] = 13000 ppm v ; C, [H 2 O] = 23000 ppm v ) (A) (C) (B)

PAGE 55

55 Figure 3 3 Hg 0 reemission from SiO 2 TiO 2 nanocomposite after 3 h pretreatment (The inset shows the dimensionless Hg concentration during the pretreatment). Time (min) Hg 0 Concentration (g m 3 )

PAGE 56

56 Figure 3 4 Mechanisms of Hg capture and reemission on the surface of SiO 2 TiO 2 nanocomposite.

PAGE 57

57 CHAPTER 4 REMOVAL OF ELEMENTAL MERCURY FROM FLUE GA S USING A SILICA TITANIA NANOCOMPOSITE Background The SiO 2 TiO 2 nanocomposite has exhibited very high efficiency of Hg 0 removal (up to 99%) under room conditions with low relative humidity (Pitoniak et al., 2003) However, as r eported in Chapters 2 and 3, at room temperature but higher water vapor concentrations (up to 23,000 ppm v ), Hg 0 capturing on the nanocomposite was hindered due to the competitive adsorption of water vapor on the active sites, and the extent of prohibition in Hg 0 removal was proportional to the water vapor concentration (Li and Wu, 2006, 2007) It should be noted that in coal fired boiler flue gas, the concentration of water vapor typically accounts for several percent in volume, much higher than that in the room conditions. Thus, it is exp ect that water vapor may have a greater inhibitory effect on Hg 0 removal in flue gas. One the other hand, the catalyst developed in this work was designed for application in the cold end of the boiler convective pass (e.g. between the electrostatic precipi tator and the wet scrubber), where the typical flue gas temperature ( 120 to 150 C ) is higher than room temperature. In this aspect, the competitive adsorption of water vapor on the catalyst would be smaller at higher temperatures. These two counteracting e ffects warrant further investigation on the performance of the SiO 2 TiO 2 nanocomposite for Hg 0 removal in flue gas. Typical coal derived flue gas consists of various minor gas components such as HCl, SO 2 and NOx, concentrations of which vary when burning different types of coal. It has been reported that these minor gases are important to the heterogeneous adsorption and/or oxidation of Hg 0 on activated carbons or fly ash in flue gas conditions (Carey et al., 1998; Norton et al., Reprinted with permission from Li, Y., Murphy, P.D., Wu, C.Y. Removal of Elemental Mercury from Flue Gas Using A Silica Titania Nan ocomposite. Submitted to Fuel Process. Technol.

PAGE 58

58 2003) Carey et al. (1998) reported that the adsorption capacity for both Hg 0 and HgCl 2 by Darco FGD carbon dramatically increased as HCl concentration increased from 0 to 50 ppm but decreased as the SO 2 concentration increased from 0 to 500 ppm. Norton et al. (2003) reported that in the presence of fly ash, NO 2 HCl, and SO 2 resulted in greater levels of Hg oxidation while NO inhibited Hg oxidation. It is expected that the nature of Hg capture on the SiO 2 TiO 2 nanocomposite would be different in flue gas conditions from that reported in room conditions in our previous studies. In this work, a photocatalytic reactor packed with SiO 2 TiO 2 nanocomposite was installed. The goal of this research was to identify the effects of the flue gas components on the rem oval of Hg 0 by SiO 2 TiO 2 nanocomposite and to explore possible surface interaction mechanisms. An improved understanding of the role of the flue gas components can help evaluate the potential of applying this novel material as an effective Hg control str ategy for coal fired power plants. E xperimental Methods Synthesis of the SiO 2 TiO 2 N anocomposite The SiO 2 TiO 2 nanocomposite was made by a sol gel method using deionized water, ethanol and tetraethyl orthosilicate (TEOS) N itric acid (HNO 3 ) and hydrogen fl uoride (HF) were used as catalysts to increase the hydrolysis and condensation rates A detailed procedure was described in our previous study (Li and Wu, 2007) The nanocomposite was prepared in the form of cylindrical pellets approxi mately 5 mm in length and 3 mm in diameter. The weight fraction of TiO 2 in the prepared SiO 2 TiO 2 pellets was approximately 12%, which corresponded to the optimum performance of Hg 0 removal in room conditions (Pitoniak et al., 2003) The average BET (Brunauer, Emmett, and Teller equation) surface area of the nanocomposite was measured to be 280 m 2 g 1 using a Quantachrome NOVA 12 00 Gas Sorption Analyzer (Boynton Beach, FL).

PAGE 59

59 Experimental Setup and Procedure A schematic diagram of the experimental setup is shown in Figure 4 1. The simulated flue gas consisted of three major gases: O 2 CO 2 and N 2 The N 2 flow was divided into three branches. One of the N 2 streams converged with the O 2 and CO 2 to form the main gas flow, which was allowed to pass through a heated water bubbler to introduce water vapor into the system. The second stream of N 2 served to dilute the main flow so as to adju st the humidity of the total gas stream. The third stream of N 2 passed through a Dynacal Hg 0 permeation tube (VICI Metronics) and introduced the saturated Hg 0 vapor into the system. The permeation tube was placed in a U shape glass tube which was immersed in a constant temperature (90 0.2 C ) water bath to ensure a constant Hg 0 permeation rate. Hg 0 concentration in the system was controlled in the range of 75~80 g m 3 Minor gases including HCl, SO 2 NO, and NO 2 were introduced into the main flow indivi dually or in combination. A mass flow controller (MFC) was used to control each of the gas flow with a total gas flow rate controlled to be 2.0 L/min. The gas concentrations were designated to be within the range of typical flue gas composition (Senior et al., 2000) : 4% O 2 12% CO 2 4~16% H 2 O, 10~50 ppm H Cl, 400~1200 ppm SO 2 50~300 ppm NO, 10~30 ppm NO 2 and balanced with N 2 The experimental conditions for investigation of the flue gas effects are listed in Table 4 1. Downstream of all the gas flows was the packed bed photocatalytic reactor placed horiz ontally. The SiO 2 TiO 2 pellets were packed in a U shape quartz tube with an inner diameter of 13 cm. A heating cord was wrapped around the U tube so that the flue gas temperature can be controlled at around 135 C which was monitored by a Teflon thermocou ple (Type K, Omega). A UV lamp was placed in a separate quartz tube centered in the reactor and 5 cm above the centerline of the U tube. The UV light had a peak wavelength of 365 nm with an intensity of 4 mW/cm 2 measured by a UVX radiometer (with a UVX 36 sensor probe). A stream of cooling air

PAGE 60

60 was continuously purged through the UV lamp to lower the lamp temperature to around 60 C The entire reactor was placed inside an aluminum cylinder so that the UV light could be reflected back to the pellets and a m aximum utilization of the UV energy could be achieved. A wet chemistry conversion system (Laudal et al., 2004; McLarnon et al., 2005) and a RA 915+ Hg analyzer (OhioLumex) were used to measure gas phase Hg speciation (Hg 0 and Hg(II)) downstream the reactor. The Hg analyzer is based on Zeeman Atomic Absorption Spectrometry (ZAAS), which is selective only for Hg 0 In the conversion system, the sampling gas was divided into two streams one for measuring Hg 0 and the other for total Hg (Hg T ). The solution used for H g 0 measurement consisted of 10% potassium chloride (KCl), which captures Hg (II) and allows only Hg 0 to pass through Hg T measurement was accomplished using an acidic 10% stannous chloride (SnCl 2 ) solution which reduces Hg (II) to Hg 0 thus producing Hg T T he concentration of Hg(II) can then be calculated by the difference between Hg T and Hg 0 The two streams converged to a 10 % sodium hydroxide (NaOH) before entering the Hg analyzer T he NaOH solution capture d acid gases such as HCl and SO 2 to prevent corr osion of the detecting cell of the Hg analyzer. In addition, as part of the conversion process, a NaOH solution was used to remove SO 2 before the sampling gas entered the SnCl 2 solution, as SO 2 can interfere with the reduction of Hg(II) by SnCl 2 (Laudal et al., 2004) A condenser was installed upstream of the Hg analyzer to remove excess moisture in the gas stream. This aimed to avoid condensation of water vapor inside the Hg detection cell and thus to minimize possible interference from water vapor. The Hg analyzer was capable of providing a real time response every 1 s The calibration of the Hg analyzer was conducted by the manufacturer using a Dynacal permeation device In this study, the high concentration mode of the Hg analyzer was used (with a detection limit of 0.5 g m 3 and an upper measurable concentration of 200 g m 3 ). Finally, the gas stream was passed

PAGE 61

61 through a carbon trap before it was exhausted into the f ume hood. The entire system was Teflon lined. To avoid condensation of the water vapor along the pathway, all the lines before the condenser were heated by heating tapes to above 90 C R esults and Discussion Baseline Test Tests were first conducted to ex amine any possible interference caused by the flue gas components on the measurement of the Hg analyzer. Balanced with pure N 2 8% H 2 O, 50 ppm HCl, 1200 ppm SO 2 300 ppm NO, and 30 ppm NO 2 were individually introduced to the system without the presence of Hg. In all cases, no significant Hg readings were observed (i.e. the interference was less than the detection limit, 0.5 g m 3 ) with or without the UV irradiation. This indicated negligible interference by the flue gas components in the concentration rang es studied in this paper. In addition, tests were performed by introducing 80 g m 3 Hg 0 to an empty reactor (i.e. no catalyst) with or without UV irradiation. Less than 0.5% reduction in Hg 0 concentration was observed, which indicated that the loss of Hg 0 on the reactor wall was negligible. To examine the effect of individual flue gas components, a baseline test without the minor gases (Set 1) was first conducted. While a larger amount of SiO 2 TiO 2 pellets could be used to achieve a Hg removal efficiency greater than 90%, only 8 g of pellets (~1 g TiO 2 ) were used to better manifest possible enhancement by the minor gases in subsequent tests. As shown in Figure 4 2, the inlet Hg concentration was measured within the first 10 min when the gas stream bypassed the reactor (Period A). Next, the gas stream was passed through the reactor without UV light for another 10 min (Period B) and then the UV light was turned on to activate the photocatalytic reaction (Period C). The concentrations of Hg T and Hg 0 at the out let of the reactor

PAGE 62

62 were recorded in alternation and were averaged every 2 min for Periods A and B and every 10 min for Period C. In Period A, the concentration of Hg T was equal to that of Hg 0 confirming that the Hg source in this study was only Hg 0 In P eriod B when the gas passed through the reactor, the outlet Hg concentration first dropped by approximately 10%, probably due to the physical adsorption by the porous SiO 2 TiO 2 pellets. However, it quickly recovered to the same level of the inlet concentra tion indicating the adsorption was saturated. A significant decrease in Hg concentration was detected only when the UV light was turned on in Period C. The concentration of Hg T dropped to 59% of the inlet level during the first 10 min of UV irradiation and slowly decreased to 53% in the next 80 min. At 100 min, the rate of the decrease of Hg T was approaching zero, and thus it was assumed that the performance of the catalyst reached a relatively stable level at this point. At 100 min, the outlet concentratio n of Hg 0 decreased to 34% of the inlet Hg 0 level while Hg(II) slowly increased to 19%. The amount of Hg captured on the pellets can be expressed as Hg cap = Hg T in Hg T out (4 6) where Hg cap represents captured Hg, and Hg T in and Hg T out represent Hg T at the inlet and outlet of the reactor, respectively. Since the inlet Hg source is 100% Hg 0 and negligible Hg 0 capture was observed without UV, it is reasonable to assume that the captured Hg species under UV irradiation was only Hg(II) due to the photocatal ytic oxidation. Hence, the total amount of Hg 0 oxidized can be expressed as Hg oxi = Hg 0 in Hg 0 out (4 7) where Hg oxi represents oxidized Hg, and Hg 0 in and Hg 0 out represent Hg 0 at the inlet and outlet of the reactor, respectively. It should be noted t hat Hg T in was equal to Hg 0 in in this study. In the

PAGE 63

63 baseline test, the percentage of oxidized Hg was calculated to be 66%, while 47% was captured by the pellets and 19% penetrated through the reactor. In other word, the inlet Hg 0 partial pressure was around 9.5 ppb and approximately 6.3 ppb (66%) of Hg 0 was oxidized to form HgO over the SiO 2 TiO 2 nanocomposite. The saturated vapor pressure of HgO at 135 C, calculated according to Borderieux et al. (2 004) is 6.0 ppb. The measured partial pressure of HgO (6.3 ppb) is very close to the saturated vapor pressure, indicating that most of the HgO formed would stay in the gas phase at 135 C. Then a fraction of the gaseous HgO (47%) was captured by the nano composite, and the rest (19%) penetrated through the reactor. E ffects of Individual Flue Gas Component s The effects of individual flue gas components were examined and the results were compared with the baseline. At least two runs were performed at each o f the experimental condition listed in Table 1. The average values of Hg capture and oxidation efficiencies are shown in Figure 4 3, where the error bars represent the envelope of minimum and maximum values. An inhibitory effect of water vapor on Hg remov al was observed as shown in Figure 4 3a. Experiments were first conducted in a relatively dry condition (<0.1%) by bypassing the water bubbler and then in humid conditions with an increasing water vapor concentration. In the dry condition, the efficiencies of both Hg capture and oxidation reached over 99%. As the water vapor concentration increased from 4% to 16% (baseline was 8%), the Hg capture efficiency decreased from 73% to 18%, and the Hg oxidation efficiency decreased from 88% to 32 %. The inhibitory effect of water vapor is very likely due to its competitive adsorption with Hg 0 on the active sites, and as shown in Figure 4 3a, the extent of inhibition on Hg removal is proportional to the concentration of water vapor. This trend agrees with the results obtained at room temperature and low water vapor concentrations (< 2.3 %) (Li and Wu, 2006) which indicates

PAGE 64

64 that even at a higher temperature (135 C in this work) the competitive adsorption of water vapor is still significant. It has been reported that Hg 0 is not adsorb ed (or is only weakly adsorbed) on the surface of sorbents including unburned carbon and selective cata lytic reduction (SCR) catalysts (Niksa and Fujiwara, 2005a, 2005b) T he weakly bonded Hg 0 can even be desorbed fr om the surface of SiO 2 TiO 2 composite by water vapor at high concentrations (Li and Wu, 2006) The desorption process would also lead to a lower efficiency of Hg 0 removal. Since the concentration of water vapor in this study was seven to eight orders of magnitude higher th an that of Hg 0 the inhibitory effect of water vapor c ould be very significant even at higher temperatures. It should also be noted that penetration of oxidized Hg from the reactor (i.e. the difference between Hg oxidized and captured) occurred in humid co nditions (4 16% H 2 O) but not in the dry condition. This can be explained by the competitive adsorption of water vapor with the gas phase oxidized Hg. Since Hg 0 is not adsorbed (or is only weakly adsorbed) on the sorbent surface (Niksa and Fujiwara, 2005a, 2005b) it is very possible that a portion of the oxidized Hg, which is the product of the reaction between Hg 0 and OH radicals, existed in the gas phase in the vicinities of the reaction sites. In the dry conditio n, the oxidized Hg in the gas phase was adsorbed and thus there was no penetration. However, in humid conditions, water vapor competes with the oxidized Hg and consequently not all oxidized Hg in the gas phase can be adsorbed. The superhydrophilic surface of TiO 2 after exposure to UV irradiation can further enhance the adsorption of water vapor (Li and Wu, 2006) but reduce the capture of oxidized Hg. As a result, penetration of oxidized Hg was usually observed in humid conditions in this work The effect of HCl on Hg remov al was found to be promotional (Figure 4 3b). In the range of 10 to 50 ppm HCl, Hg capture efficiency increased to approximately 75% and Hg oxidation

PAGE 65

65 efficiency increased to over 90%. However, the extent of promotion was not apparently related to the HCl c oncentration in the range studied. The promotional effect of HCl is consistent with the literature that HCl promotes heterogeneous Hg oxidation (Presto and Granite, 2006) It has been reported that in the presence of an appropriate catalyst (e.g. metal oxides), a Deacon process (Pan et al., 1994) could convert HCl in flue gas to Cl 2 at high temperatures (300 400 C), thereby enhancing Hg 0 oxidation (or chlorination). Niksa et al. proposed that Hg oxidation occurs via an Eley Rideal mechanism, where adsorbed HCl reacts with gas phase (or weakly adsorbed) Hg 0 (Niksa and Fujiwara, 2005a, 2005b) The mechanism was also consistent with the observation of enhanced Hg 0 sorption to halogen promoted sorbents and fly ashes in literature (Granite et al., 2000; Maroto Valer et al., 2005) In this work, the Deacon process was less likely to occur because of the relatively low flue gas temperature (135 C). Instead, it is more likely to follow the Eley Rideal mechanism. In the Eley Rideal mechanism, HCl may first be adsorbed on the surface of SiO 2 TiO 2 nanocomposite, and then react with gas phase Hg 0 This, together with the Hg 0 oxidatio n by OH radicals, can result in a higher Hg removal efficiency. Further investigation is needed to confirm the reaction mechanism. As shown in Figure 4 3c, SO 2 was found to have a promotional effect on Hg capture and oxidation and the promotion was proport ional to the concentration of SO 2 in the range of 0~1200 ppm. The Hg capture and oxidation efficiencies reached 73% and 91% respectively at 1200 ppm SO 2 The effect of SO 2 on heterogeneous Hg oxidation was not conclusive in literature. It has been reported that SO 2 competes with HCl for sites on activated carb on and fly ash sorbents and thus inhibit s mercury oxidation and adsorption in flue gas (Laudal et al., 2000; Laumb et al., 2004) Carey et al. reported that the adsorption capability of a Darco FGD carbon for both Hg 0 and HgCl 2 decreased as the SO 2 concentration increased from 0 to 500 ppm but neither capacity

PAGE 66

66 changed significantly above 500 ppm SO 2 (Carey et al., 1998) However, in some cases, SO 2 appears to enhance Hg 0 oxidation (Eswaran and Stenger, 2005; Norton et al., 2003) Eswaran and Stenger reported a promotional effect of SO 2 on Hg 0 oxidation over a selective catalytic reduction (SCR) catalyst (Eswaran and Stenger, 2005) The mechanism was proposed as: (4 8) (4 9) Sim ilar mechanism can be used to explain the promotional effect of SO 2 in this work, where SO 3 was formed through the oxidation of SO 2 by OH radicals (Bai et al., 2006) which were generated on the SiO 2 TiO 2 pellets under UV irradiation. The effect of NO on Hg removal was found to be inhibitory at a significant but relatively constant level in the concentration range of 50~300 ppm NO (Figure 4 3d). The Hg capture and oxidation efficiencies both decreased to around 10% in the presence of NO. NO has been reported as an inhibitor for heterogeneous Hg 0 oxidation on fly ash (Norton et al., 2003) but the mechanism was not clear. In this study, it is very likely that the scavenging of OH radicals by NO hindered the photocatalytic oxidation of Hg 0 The inhibition occurred via (Niksa et al., 2001) (4 10) In the presence of 10~30 ppm NO 2 the efficiencies of Hg capture and o xidation were slightly lower than those in the baseline (Figure 4 3e). However, t he effect of NO 2 in this range can be considered as insignificant compared to other flue gas components It has been reported in literature that NO 2 can enhance hete rogeneous oxidation of Hg 0 in the presence of fly ash (Norton et al., 2003) or iron oxides (Borderieux et al., 2004; Galbreath et al., 2005) though this effect is often cons idered of minor importance compared to chlorination.

PAGE 67

67 Hg Removal in Simulated Flue Gases The performance of the SiO 2 TiO 2 nanocomposite was finally tested in two simulated flue gases, the compositions of which were in line with those reported in literature (Pavlish et al., 2003) Flue gas 1 (FG1, Set 7) represents burning of high rank (bituminous) coals that contain higher chlorine and sulfur contents. Flue gas 2 (FG2, Set 8) represents burning of low rank (subbituminous and lignite) coals, which contain less chlorine and sulfur but more moisture. As shown in Figure 4 3f, Hg removal in Flue Gas 1 was close to that in baseline, indicating that the prohibitory effect of 300 ppm NO counteracted the promotional effects of 30 p pm HCl and 1200 ppm SO 2 Hg removal in Flue Gas 2 was less than in Flue Gas 1, very likely due to the higher concentration of H 2 O and lower concentrations of HCl and SO 2 Hence, high rank coals are preferable to low rank coals for the application of the Si O 2 TiO 2 nanocomposite. Minimizing the adverse effect of NO so as to improve the overall performance of the catalyst would be an important task for future research. Summary A novel SiO 2 TiO 2 nanocomposite has been synthesized to removal Hg 0 from simulated c oal fired power plant flue gas. The flue gas components were found to have significant effects on Hg removal efficiency in a fixed bed study. HCl and SO 2 promoted Hg oxidation and capture, while H 2 O and NO inhibited Hg removal and the effect of NO 2 was not significant. Experiments of Hg removal in simulated flue gases showed that high rank coals are preferable to low rank coals because of the lower moisture and higher HCl and SO 2 concentrations in the flue gas. It is essential, however, to minimize the adve rse effect of NO to improve the catalytic performance of the SiO 2 TiO 2 nanocomposite.

PAGE 68

68 Table 4 1. Experimental conditions for investigation of the flue gas effects H 2 O (%) HCl (ppm) SO 2 (ppm) NO (ppm) NO 2 (ppm) Set 1 (baseline) 8 Set 2 0, 4, 12, 16 Set 3 8 10, 30, 50 Set 4 8 400, 800, 1200 Set 5 8 50, 100, 300 Set 6 8 10, 20, 30 Set 7 (FG 1) 8 30 1200 300 10 Set 8 (FG 2) 12 10 400 300 10 Note: All the conditions contained 4% O 2 12% CO 2 75~80 g m 3 Hg 0 (inlet), and balanced with N 2 ; the temperature was controlled at approximately 135 C

PAGE 69

69 Figure 4 1. Photocatalytic reaction system under flue gas conditions

PAGE 70

70 Figure 4 2. Hg speciation at the outlet of t he reactor in the baseline test

PAGE 71

71 a) b) c) d) e) f) Figure 4 3 Effects of flue gas components on Hg capture and oxidation under various conditions of a) H 2 O, b) HCl, c) SO 2 d) NO, e) NO 2 and f) simulated flue gases

PAGE 72

72 CHAPTER 5 DEVELOPMENT OF SILIC A/VANADIA/TITANIA CO MPOSITES FOR REMOVAL OF ELEMENTAL MERCURY FR OM FLUE GAS Background As indicated in Chapter 4, the performance of the SiO 2 TiO 2 nanocomposite is significantly affected by the flue gas composition. NO, particularly, has a dramatically inh ibitory effect on Hg oxidation. To minimize the adverse effect of NO, there are two feasible ways. One way is to remove NO from the flue gas before passing the gas through the photocatalytic bed. Actually, in industrial practices, NO removal from the flue gas is usually achieved by selective catalytic reduction (SCR) of NO with ammonia (NH 3 ) (Parvulescu et al., 1998) However, the results in Chapter 4 show that even at a relatively low concentration (i.e. 50 ppm), NO still greatly inhibits Hg oxidation and capture by the SiO 2 TiO 2 nanoco mposite. This undesired impact implies that unless NO can be completely or nearly completely removed from the flue gas, the adverse effect of NO is inevitable and a larger amount of the catalyst must be used to compensate the effect. Apparently, this is no t a cost effective method. The other way to enhance the catalytic performance in Hg oxidation is to modify the composition of the catalyst or even develop a new catalyst that is more effective under the flue gas conditions. Recently, it has been reported t hat the SCR catalyst is capable of oxidizing Hg 0 in addition to its ability of removing NO (Benson et al., 2005; Lee, Srivastava et al., 2004; Lee et al., 2006; Niksa and Fujiwara, 2005b; Senior, 2006) The extent of Hg 0 oxidation through SCR processes varies under different operating conditions burning different types of coal. It was reported that Cl species in the flue gas promote the Hg 0 oxidation across SCR (Lee et al. 2006; Senior, 2006) although the exact mechanism is still poorly understood. In contrast, the mechanisms of the reduction of NO with NH 3 by various SCR catalysts have been extensively investigated (Parvulescu et al., 1998; Weckhuysen and Keller, 2003) Bosch and Janssen (198 8)

PAGE 73

73 reported a broad survey of metal oxide catalysts active for the reduction of NOx with NH 3 and they indicated vanadium oxide (V 2 O 5 ) to be the most active and selective catalyst. The active sites on industrial SCR catalysts are V 2 O 5 species supported on TiO 2 (Busca et al., 1998) The nature of the support, on the other hand, is also an important factor for the catalytic activity (Weckhuysen and Keller, 2003) TiO 2 as a support normally has drawbacks such as low surface area and low resistance to sintering, and thus, a common practice is to use a silica support coated with TiO 2 (Kobayashi et al., 2005; Martra et al., 2000; Tesser et al., 2004) In addition, Shikada et al. (1981) reported that the order of activity of NO reduction for supported V 2 O 5 is SiO 2 TiO 2 Al 2 O 3 > SiO 2 The SiO 2 TiO 2 composite used in this research for Hg removal has a high surface area and is in line with the composition of the high activity support reported by Shikada et al. (1981) Hence, it is specula ted that an addition of V 2 O 5 to the SiO 2 TiO 2 composite could further enhance the capability of oxidizing Hg. In this chapter, V 2 O 5 was doped onto the SiO 2 TiO 2 matrix to form a SiO 2 TiO 2 V 2 O 5 composite. The catalytic ability of the new material in Hg re moval was compared to that of the SiO 2 TiO 2 composite. In addition, in order to investigate whether V 2 O 5 is the active species for Hg oxidation, SiO 2 V 2 O 5 composite was also synthesized and tested for its catalytic performance. Since SiO 2 is generally cons idered as an inert support, the hypothesis is that the activity of the SiO 2 V 2 O 5 composite would represent the nature of the doped V 2 O 5 species. Therefore, the objectives of this chapter are: (1) to synthesize and characterize the SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 composites; (2) to test the catalytic abilities of those composites in Hg removal in a fix bed reactor; and (3) to investigate the reaction mechanisms of the catalytic removal of Hg.

PAGE 74

74 Materials and Methods Catalyst Preparation The details of the proced ure of synthesizing the SiO 2 TiO 2 composite are described in Chapter 2. When synthesizing the SiO 2 TiO 2 V 2 O 5 composite, the same procedure was used but with one more step of adding V 2 O 5 Vanadium triisopropoxide oxide (VTPO) (Alfa Aesar) was used as the p recursor of V 2 O 5 A known amount of VTPO was first dissolved in well stirred ethanol to form an orange brown solution. It was then added dropwise to the prepared silica sol under vigorous stirring. TiO 2 nanoparticles (P25 Degussa) were finally added to the mixture before it started to gel. When synthesizing the SiO 2 V 2 O 5 composite, the step of adding TiO 2 nanoparticles was skipped. In this study, the synthesized SiO 2 V 2 O 5 composite had a weight fraction of V 2 O 5 ranging from 2% to 10%. The synthesized SiO 2 T iO 2 V 2 O 5 composite had 12 wt.% of TiO 2 and varied contents of V 2 O 5 The composites were originally made in the pellet form (3 mm in diameter and 5 mm in length). Powder form of the composites was also obtained by grinding the pellets and sieving through me shes. Both the pellets and powders of the composites were tested in a fix bed study. The powders used have a mesh size of 40100 (425 150 m). The names of the catalysts were abbreviated by way of STxVy, where S represents SiO 2 T represents TiO 2 V repre sents V 2 O 5 and x and y represent the weight percentages of the TiO 2 and V 2 O 5 respectively. Catalyst Characterization Techniques The BET (Brunauer, Emmett, and Teller equation) surface areas of the catalysts were measured using a Quantachrome NOVA 1200 Ga s Sorption Analyzer (Boynton Beach, FL). The powder samples were outgassed at 180 C for 3 hours before the analysis. X ray diffraction (XRD) patterns of the powders were recorded with a Philips APD 3720 diffractometer using Cu ) with a step size of 0.02.

PAGE 75

75 Catalyst Activity Measurement As reported in Chapter 4, the SiO 2 TiO 2 composite needs activation by UV light for catalytic oxidation of Hg. Thus, the necessity of UV light activation for V 2 O 5 based composites was first investigated. Pellet form of the catalysts was then used because the space between the pellets allows penetration of UV light so that a maximum exposure of the catalyst to the light can be achieve d. In contrast, for a catalyst bed densely packed with powders, it is difficult for the UV light to reach the central part of the bed. However, if tests indicate that the UV light is unnecessary, powders are preferable to pellets because of better contact of the gas with the inner pore surfaces of the catalyst. Table 5 1 lists the experimental parameters for measuring the photocatalytic activities using pellet catalysts. The three catalysts (SiO 2 TiO 2 SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 ) were tested following th e same procedure as described in Chapter 4 using the reactor system previously illustrated in Figure 4 1. Experiments were conducted under two flue gas conditions, FG1 and FG2, as listed in Table 4 1 in Chapter 4. FG1 has relatively higher concentrations o f HCl and SO 2 and lower concentration of water vapor, which represents flue gas burning high rank coals. In contrast, FG2 represents flue gas burning low rank coals. Next, the powder catalysts were tested with a modified reactor as shown in Figure 5 1. Th e experimental parameters are listed in Table 5 2. No UV irradiation was supplied in this modified system. The U tube quartz reactor (13 mm ID) was immersed in an oil bath heated by a hotplate to a constant temperature at 135 C ( 0.5 C). The catalyst po wders were packed in between glass wools in the reactor. The flue gas (FG3) in this series of experiments contained 4% O 2 12% CO 2 8% H 2 O, 10 ppm HCl, 400 ppm SO 2 300 ppm NO, 10 ppm NO 2 and balanced with N 2 The inlet Hg concentration was maintained in a relatively constant range of 15.0~16.5 ppb. The total flow rate was controlled at 1.5 L/min. Finally, to explore the reaction mechanisms and

PAGE 76

76 the roles of the flue gas components in the catalytic reactions, the catalytic activity of a fixed amount of cata lyst (SV5) was examined with introduction of individual flue gas components. Results and Discussion Characterization of the C atalysts The BET specific surface areas of the catalysts are listed in Table 5 3. All the catalysts exhibit high surface areas (> 2 50 m 2 /g). Without any doping, the pure silica gel had the highest surface area, 341.8 m 2 /g. The inclusion of 12% TiO 2 to the silica gel (ST12) slightly reduced the surface area to 319.4 m 2 /g. The doping of V 2 O 5 (2 10%) to the silica gel moderately reduce d the surface area, but all the SiO 2 V 2 O 5 catalysts had the similar level of surface areas and did not exhibit any clear dependence on the V 2 O 5 loading. The surface areas of the SiO 2 TiO 2 V 2 O 5 catalysts were close to those of the SiO 2 V 2 O 5 catalysts. The X RD patterns of the catalysts are shown in Figure 5 2. No visible crystal phase of V 2 O 5 (peak at 2 = 26.1) was detected for SV2 and SV5, which indicated that the vanadium contents were highly dispersed on these catalysts (Kobayashi et al., 2006) A very small peak of crystalline V 2 O 5 was detected for SV8, and SV10 showed a relatively br oader and more prominent peak of crystalline V 2 O 5 The molecular structures of the vanadium oxides at different surface loadings have been widely reported in literature (Parvulescu et al., 1998; Rodella et al., 2001; Weckhuysen and Keller, 2003) At low surface vanadia concentration s mainly monomeric vanadyl (V 4+ ) species are formed containing one terminal V=O bond and three bridging V O support bonds. As t he vanadia loading increases, the monomeric species react to form polymeric vanadates (V 5 + ) which consist of a terminal V=O bond with one V O support and two bridging V O V bonds The presence of these monomeric and polymeric vanadium oxide species has been identified by Raman and/or Infrared (IR) spectroscopy. As the vanadia loading further increases, a fraction of the vanadia aggregates to form amorphous and crystalline V 2 O 5 clusters

PAGE 77

77 The XRD results in this study indicated that crystalline V 2 O 5 begins to grow as the vanadia loading increases to somewhere between 5 and 8%. The XRD pattern of ST12 showed a strong anatase phase (peaks at 2 = 25.3 and 38.0) and a weak rutile phase (peak at 2 = 27.6) of TiO 2 This agrees with the composition of the TiO 2 nanoparticles (P25 Degussa A G) which is an 80/20 mixture of anatase/rutile phases. For the catalyst of ST12V5, no crystal phase of V 2 O 5 was detected because of the relatively low vanadia loading. It has been reported that TiO 2 anatase is metastable and tends to convert to the thermod ynamically stable form rutile and that V 2 O 5 favors this anantase to rutile phase transformation (Busca et al., 1998) This may explain the finding that the peak of anatase TiO 2 in ST12V5 was lower than in ST12. However, no increase in the rutile phase in ST12V5 was observed. Mercury Removal Using Pellet C atalysts Mercury removal tests were first conducted using pellet catalysts to investigate their photocatalytic activities. T he results a re summarized in Figure 5 3. It was found that the catalytic abilities of SV2 and ST12V2 on Hg removal were almost the same with or without UV irradiat ion. Thus, only the results without UV are demonstrated in Figure 5 3 for SV2 and ST12V2. In contrast, ST12 had to be activated under UV irradiation and without UV there was negligible removal of Hg. Apparently, the addition of V 2 O 5 is advantageous. It sim plifies the system by eliminating the UV devices and reduces the cost by saving the energy of UV irradiation. Figure 5 3 also compares the behavior of the catalysts under different flue gas conditions and mass loadings. As described in Chapter 4, the oxidi zed Hg is calculated as the difference between inlet and outlet Hg 0 concentrations, and the captured Hg is calculated as the difference

PAGE 78

78 between the inlet and outlet Hg T concentrations. In flue gas 1 (FG1), 8 g of ST12V2 demonstrated a very h igh efficiency of Hg capture/oxidation (~ 92%), compared to only about 50% capture/oxidation of Hg by 8g of ST12. This increase in Hg removal is obviously caused by the addition of the 2% V 2 O 5 When reducing the mass loading of the ST12V2 from 8 g to 4 g, the efficiency of Hg oxidation decreased a little to 73% but that of Hg capture dramatically decreased to 27%. The reduction in Hg capturing ability might result from the decrease in total available surface area due to lower mass loading It is interesting to find that c ompared to 4 g of SV2, 4 g of S T12 V2 had a higher Hg oxidation efficiency but a lower Hg capturing efficiency. The enhanced oxidizing ability of ST12V2 is in line with the literature where the SiO 2 TiO 2 supported V 2 O 5 has a higher activity than SiO 2 suppor ted V 2 O 5 for reduction of NO (Shikada et al., 1981) It has also been reported that the V O support bonds are the most critical structures for the catalytic oxidation of methanol to formaldehyde (Weckhuysen and Keller, 2003) Thus, it is possible that the V O Ti bonds have superior activities than th e V O Si bonds, which resulted in a higher oxidation of Hg on the SiO 2 TiO 2 V 2 O 5 composite than on the SiO 2 V 2 O 5 composite. The reason for the penetration of oxidized Hg from ST12V2 may be related to the selectivity of the reactions, i.e. ST12V2 may have a higher capacity of converting Hg 0 to volatile oxidized Hg species such as Hg(NO 3 ) 2 This agrees with the suggestions by other researchers that the V O support bonds are determinant for activity and selectivity of different reactions (Tesser et al., 2004; Weckhuysen and Keller, 2003) More discussion relevant to the formation of Hg(NO 3 ) 2 is provided later in this chapter. In flue gas 2 (FG2), the behavior of the catalysts followed a similar pattern as in FG1. For 4 g of ST12V2 and 4g of SV2, no obvious change in the Hg removal efficiency was observed i n FG2 compared to in FG1. For the case of 8 g ST12V2, the oxidized Hg was at the same level as

PAGE 79

79 in FG1 but the captured Hg was much lower than that in FG1. The difference may be due to the higher water vapor concentration in FG2 which competes with oxidized Hg for the adsorption sites. It is also possible that a larger fraction of certain volatile Hg compounds (such as Hg(NO 3 ) 2 ) were produced in FG2. However, identification of the oxidized Hg species in the current system is technically difficult due to thei r trace amounts. Mercury Removal Using Powder C atalysts The previous experiments using pellet catalysts showed no necessity of UV light activation for the SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 catalysts. Hence, no UV light was used for the study of powder catalyst s. Tests were first performed to verify that the glass wools (used as the support of powders) and pure silica powders were inert to Hg removal. Then, with a feed of approximately 16 ppb Hg, each flue gas component (4% O 2 8% H 2 O, 50 ppm HCl, 1200 ppm SO 2 300 ppm NO, or 30 ppm NO 2 ) balanced with N 2 was individually introduced to the system. In all cases, the change of the Hg concentration was observed to be within 5%. Since the Hg permeation tube has an error of 2%, the result indicates that homogeneous o xidation of Hg in the gas phase was negligible. The first set of experiments was carried out using 500 mg of SiO 2 V 2 O 5 catalysts (corresponding to a bed height of 17 mm in average) but with different V 2 O 5 loadings (2 10 wt.%). The results are shown in F igure 5 4. In the beginning, the flue gas bypassed the reactor to obtain the inlet Hg concentration. It then passed through the reactor and the extent of the catalytic oxidation was recorded. After a 6 hr period, the flue gas bypassed the reactor again and the inlet Hg concentration was checked. It was observed that the Hg removal efficiency increased as the V 2 O 5 loading increased from 2 to 8%. For SV2 (Figure 5 4a) and SV5 (Figure 5 4b), the outlet concentration of Hg T and Hg 0 were at the same level, indic ating that all the oxidized Hg has been captured. For SV2, the Hg removal efficiency was initially around 22%

PAGE 80

80 and slowly decreased to 12% in 6 hours. For SV5, the Hg removal efficiency was initially around 65% and slowly decreased to 45% in 6 hours. For SV 8, as shown in Figure 5 4c, the outlet Hg T concentration was initially dropped to a very low level (corresponding to 93% capture), and then it increased and was relatively stabilized at around 32% of the inlet Hg concentration (i.e. 68% capture). The outle t Hg 0 concentration was found to be always lower than the Hg T concentration, indicating that a portion of the oxidized Hg penetrated the reactor. The Hg oxidation efficiency at the end of the 6 hr test was around 77%, higher than that of SV2 and SV5. For S V10, as shown in Figure 5 4d, a different pattern of Hg removal was observed. The outlet Hg T concentration slowly dropped to 40% of the inlet level in the 6 hr period, unlike the previous cases where the outlet Hg T concentrations dropped to their minimum l evels immediately after the flue gas passed through the reactor. The outlet Hg 0 concentration was relatively stable at 30% of the inlet level during the last 3 hours. Both the Hg capturing and oxidation efficiencies of SV10 were lower than those of SV8. N ext, 500 mg of ST12V5 was tested for its catalytic activity and the result is shown in Figure 5 4e. The outlet Hg 0 concentration remained almost constantly low at 15% of the inlet level (i.e. 85% oxidation). The outlet Hg T concentration initially dropped t o 11% of the inlet level but quickly increased and stabilized around 65% (i.e. 35% capture). Similar to the findings observed in the study of pellet catalysts (Figure 5 3), ST12V5 powder has a higher oxidation but lower capturing efficiency compared to SV5 The amounts of Hg captured/oxidized on the various catalysts in the 6 hr test are summarized in Table 5 4. To better compare the effectiveness of the catalysts, the capabilities of the catalysts were normalized to unit mass of catalyst and unit mass of V 2 O 5 (the active phase), respectively. Table 5 4 showed that the amount of Hg captured per gram of the SiO 2 V 2 O 5

PAGE 81

81 catalyst increased as the V 2 O 5 loading increased from 2 to 8% but decreased as the V 2 O 5 loading further increased to 10%. When normalized to pe r gram of V 2 O 5 the capacity of SV5 was the highest; either lower (2%) or higher (8% and 10%) V 2 O 5 loadings reduced the capacity. These results suggested that the optimal V 2 O 5 loading for an optimal catalytic activity is somewhere between 5 and 8%. This is coincident with the XRD result that the maximum loading of V 2 O 5 without the formation of crystalline V 2 O 5 is somewhere between 5 and 8%. It should be noted that ST12V5 exhibited an even greater ability of oxidizing Hg than SV5, although its Hg capturing a bility was lower than SV5. Applications of ST12V5 can be beneficial to coal fired power plants equipped with wet scrubbers where oxidized Hg can be easily captured. Therefore, ST12V5 surpasses SV5 in terms of the total potential for Hg removal. As shown i n Figure 5 4, the Hg removal efficiencies slowly decreased in the 6 hr test for SV2, SV5, and SV8. To investigate whether a 100% breakthrough would happen, a smaller amount (250 mg) of SV5 was subject to a test with a longer period of time, as shown in Fig ure 5 5. The outlet Hg T (or Hg 0 ) concentration initially dropped to 65% of the inlet level when the flue gas passed the bed. Then the Hg concentration slowly increased to 80% in a 9 hr period. Thereafter, the Hg removal efficiency remained relatively stabl e at approximately 20% for another 3 hr (in Day 1). With the same batch of catalyst under the same flue gas conditions, the experiment was continued in Day 2 after a pause of 40 hours. Only the Hg T concentration was recorded in Day 2 since Hg 0 was found to be at the same level of Hg T in Day 1. This time when the flue gas passed through the bed, the Hg T concentration first dropped to 68%, very close to the lowest level of Hg T concentration (65%) in Day 1. Then the Hg T concentration increased to 81% in 2 hr a nd remained at this level for another 4 hr. This 20 hr test suggested that a steady state catalytic activity was reached (approximately 20% Hg removal) using either fresh (Day 1) or

PAGE 82

82 used (Day 2) catalyst. This transition to a steady state activity is in ag reement with the reported kinetics of oxidative dehydrogenation of propane over V 2 O 5 /TiO 2 in literature (Grabowski et al., 2002; Sloczynski, 1996) Those researchers observed that the rate of reduction is limited by th e adsorption of reductant at the surface of the vanadium phase and that reoxidation of the reduced catalysts proceeds via two stages: (i) a quick surface reaction, and (ii) a slow process limited by bulk diffusion through the growing coat of the oxidation product. They also indicated that the surface reoxidation rate is considerably higher than that of the reduction. According to these theories and findings, the rate of Hg oxidation in this work should quickly reach a maximum point at the beginning and then gradually decrease due to the reduced number of available vanadium sites and/or the increasing difficulty of O 2 diffusing through the coat of oxidized Hg. However, this decrease in the catalytic activity proceeded so slowly that the catalyst behaved like reaching a steady state activity for a relatively long time (Grabowski et al., 2002) This agrees with the result that the captured Hg on the 250 mg SV5 during the 20 hr test was calculated to account for only 0.2 mol % of the total van adium sites. Future study is needed to investigate the long term performance of the catalyst. Mercury Removal M echanisms To explore the Hg removal mechanisms on the SiO 2 V 2 O 5 catalyst, it is important to understand the role of the flue gas components in t he catalytic reactions. Hence, experiments were conducted by mixing Hg with individual flue gas components (e.g. HCl, SO 2 and NOx) and/or in combination with O 2 The role of water vapor was also examined. 250 mg of fresh SV5 was used in each test since 5% V 2 O 5 was previously found to be close to the optimal loading. R ole of O 2 The role of O 2 on Hg removal was first investigated because O 2 is an important oxidant in flue gas. A background test was conducted using high purity N 2 (>99.995%, Airgas) as the

PAGE 83

83 car rier gas and around 5% of Hg removal was detected, as shown in Figure 5 6. Considering that the carrier gas contained a maximum of 50 ppm impurity very likely consisting of O 2 the 5% Hg removal might be contributed by the trace amount of O 2 When 4% O 2 wa s introduced, the Hg removal increased to about 15%. Hg removal further increased to 26% as O 2 increased to 20%. Granite et al. (2000) studied various metal oxides for catalytic Hg removal from flue gas and they proposed that lattice oxygen of the metal oxides can serve as the oxidant of Hg, forming mercuric oxide (HgO). It has also been reported in literature that lattice ox ygen is the most abundant reactive intermediates that are responsible for oxidative dehydrogenation of alkanes over V 2 O 5 based catalysts (Argyle et al., 2004; Grabowski et al., 2002) Gas phase O 2 on the other hand, reoxidizes the reduced metal oxides, replenishing the latt ice oxygen (Grabowski et al., 2002; Granite et al., 2000) In this work, a similar redox cycle is proposed for the catalytic oxidation of Hg on SiO 2 V 2 O 5 i n the presence of O 2 : (5 1) (5 2) The overall reaction can be written as: (5 3) O 2 plays an important role in the redox mechanism, which is supported by finding that H g oxidation on the catalyst is proportional to the concentration of O 2 R ole of HCl HCl was found to enhance Hg oxidation over the SiO 2 V 2 O 5 catalyst (Figure 5 6). 10 ppm HCl (balanced with N 2 ) resulted in 15% Hg removal, and as HCl increased to 50 ppm, t he Hg removal increased to 25%. The combination of 50 ppm HCl with 20% O 2 further improved the Hg removal to 39%. It has been reported in literature that Hg 0 is not adsorbed (or is only weakly

PAGE 84

84 adsorbed) on the surface of unburned carbon and SCR catalysts (Niksa and Fujiwara, 2005a, 2005b) The researchers also proposed that Hg oxidation occurs via an Eley Rideal mechanism, where adsorbed HCl reacts with gas phase (or weakly adsorbed) Hg ; however, the specific reaction pathway was not given. In this work, it was also found that Hg is not (or only weakly) adsorbed on the SiO 2 V 2 O 5 catalyst, which is verified by the result that there was only 5% removal of Hg in the environment of pure N 2 The Eley Rideal mechanism can be used as well to explain the reaction on the SiO 2 V 2 O 5 catalyst in the presence of HCl. The pathway of HCl adsorption can be inferred from literature about HCl adsorption on other types of metal oxide surfaces. Parfitt et al. (1971) conducted an infra red study of HCl adsorption on rutile surface and found an increase in surface hydroxyl (OH) groups due to the introductio n of HCl. The OH groups further react with excess HCl to form Cl and water. Tseng et al. (2003) reported the adsorption of HCl on the CuO surface forming a hydroxychloride (Cu(OH)Cl) intermediate. Similar to the mechanisms reported in those studies, it is proposed that the formation of adsorbed Cl on t he V 2 O 5 surface occurs via: (5 4) (5 5) Actually, the V OH structures are one type of the active sites readily present on the surface of the vanadia based catalysts (Busca et al., 1998; Kantcheva et al., 1994; Parvulescu et al., 1998) Thus, the re action with HCl can directly start from Reaction 5 5. The adsorbed Cl species then react with gas phase Hg through the Eley Rideal mechanism to generate an intermediate HgCl species, which then further reacts with HCl or Cl to form a more stable mercuric c hloride, HgCl 2 The overall reaction can be written as: (5 6)

PAGE 85

85 It should be noted that chlorination of Hg may take place without the presence of O 2 as shown in Figure 5 6. In this case, V 2 O 5 is consumed to form V 2 O 4 The additi on of 20% O 2 to 50 ppm HCl enhanced the total oxidation of Hg, very likely due to the oxidation of V 2 O 4 to V 2 O 5 i.e. the regeneration of the catalyst. R ole of NO 2 The effect of NOx was also found to be promotional in Hg oxidation over the SiO 2 V 2 O 5 catal yst (Figure 5 6). In the presence of 10 ppm NO 2 57% of Hg was oxidized while 50% was captured. It should be noted that the cylinder gas of NO 2 was balanced with both N 2 and O 2 with O 2 concentration three times as much as NO 2 Thus, 30 ppm O 2 was present i n the gas together with 10 ppm NO 2 When 20% O 2 was added, the Hg oxidation remained at a similar level. Increasing the concentration of NO 2 to 30 ppm (with 20% O 2 ) increased the Hg oxidation to 68% and Hg capture to 57%. The results indicated that NO 2 gre atly promoted Hg oxidation with or without O 2 It has been reported in literature that NO 2 significantly improves heterogeneous oxidation of Hg on fly ash (Norton et al., 2003) and on activated carbon based sorbents (Miller et al., 2000) Mercuric nitrate (Hg(NO 3 ) 2 ) was suggested to be the reaction product initiated by NO 2 (Laumb et al., 2004; Miller et al., 2000) Other researchers have reported that adsorption of NO 2 on TiO 2 supported V 2 O 5 catalysts was t he first step in the process of selective catalytic reduction of NOx (Kantcheva et al., 1994; Parvulescu et al., 1998) Kantcheva et al. (1994) indicated two pathways for the NO 2 adsorption on V 2 O 5 involving V=O and V OH sites: (5 7) (5 8) In this work, similar mechanism can be applied to explain the oxidation of Hg in the presence of NO 2 NO 2 is first adsorbed on V 2 O 5 via Reactions 5 7 and 5 8 and then transformed

PAGE 86

86 to adsorbed nitrate species, which react with gaseous Hg to form Hg(N O 3 ) 2 via the Eley Rideal mechanism. The overall reaction can be written as: (5 9) Considering the low melting point of Hg(NO 3 ) 2 79 C (Weast et al., 1983) it is likely that the formed Hg(NO 3 ) 2 is vola tile at the reactor temperature (135 C) and thus part of it may be released from the reactor in the gas phase, as shown in Figure 5 6. This formation of volatile Hg(NO 3 ) 2 initiated by NO 2 is in agreement with the findings by other researchers. Using a car bon based sorbent to remove Hg 0 Miller et al. (2000) observed nearly 100% breakthrough of a volatile oxidized Hg species in a gas mixture of SO 2 and NO 2 In a follow up study conducted by Olson et al. (2002) using both MnO 2 and carbon sorbents, they identified this volati le Hg species to be Hg(NO 3 ) 2 by trapping the effluent in cold acetonitrile followed by analysis using gas chromatography mass spectroscopy (GC MS). Considering the much higher melting/decomposing point of HgCl 2 (277 C) and HgO (500 C) (Weast et al., 1983) the penetration of these two less volatile Hg species is less likely or occurs to a very small extent, as shown in Figure 5 6. R ole of NO As also shown in Figure 5 6, in the presence of 300 ppm NO, there was 29% removal of H g and an addition of 20% O 2 further increased the removal to 48%. The oxidation and capture of Hg increased with the increase of NO concentration. Decreasing the NO concentration to 100 ppm moderately decreased the Hg removal efficiency. The Hg removal may be ascribed to the adsorbed NO species on the V 2 O 5 surface. It is generally agreed that NO adsorbs as nitrosyl and dinitrosyl surface species on reduced vanadia surfaces, whereas it does not adsorb over fully oxidized surfaces (Busca et al., 1998) The adsorbed NO can be oxidized on the surface, giving

PAGE 87

87 rise to species like NO + NO 2 an d NO 3 or it can be reduced by reduced catalyst centers (Busca et al., 1998) It is likely that these adsorbed species derived from NO are responsible for the observed Hg oxidation, but the exact reaction pathways and products are unknown. It has been reported that gas phase NO can first be oxidized by O 2 and then is adsorbed as NO 2 on the SCR catalyst (Parvulescu et al., 1998) This route is less likely to occur on the SiO 2 V 2 O 5 catalyst in this work, because there was no penetration of the oxidized Hg through the reactor in the presence of NO, unlike that observed in the presence of NO 2 R ole of SO 2 Figure 5 6 show s that the effect of SO 2 on Hg removal was insignificant in the concentration range of 400~1200 ppm. When combining 20% O 2 with 400 ppm SO 2 the Hg oxidation/capture efficiency was very close to or slightly lower than that without SO 2 It has been reported that SO 2 competes with HCl for sites on activated carb on and fly ash sorbents and thus inhibit s Hg oxidation and adsorption in flue gas (Laudal et al., 2000; Laumb et al., 2004) Other studies, however, reported that SO 2 can promote heterogeneous oxidation of Hg over fly ash sorbents (Norton et al., 2003) or SCR catalysts (Eswaran and Stenger, 2005) Hence, further investigations involving the combination of SO 2 with other flue gas components (e.g. HCl and NOx) would warrant a more comprehensive understanding of the role of SO 2 for Hg removal on the SiO 2 V 2 O 5 catalyst. R ole of H 2 O It has been well discussed in Chapters 2 to 4 that a high concentration of water vapor can inhibit Hg oxidation or capture on SiO 2 Ti O 2 in room or flue gas conditions. In this work, H 2 O was also found to have a dramatic inhibitory effect on Hg removal over SiO 2 V 2 O 5 (Figure 5 7). Under flue gas conditions (FG3) using 250 mg SV5, when the gas was switched from humid (8% H 2 O) to dry, the Hg removal efficiency increased from 20% to 66%. A H 2 O concentration as low

PAGE 88

88 as 0.6% (6000 ppm) also exhibited inhibitory effects. In a test with 10 ppm HCl and 20% O 2 switching the gas from dry to a 0.6% H 2 O caused a decrease in Hg capture from 30% to 19% Similarly, in another test with 10 ppm NO 2 and 20% O 2 the introduction of 0.6% H 2 O caused a decrease in Hg capture from 43% to 24%. The competitive adsorption of water vapor on the active sites may have prohibited the adsorption of reactive species such as HCl and NOx. The findings agree with the literature where Laumb et al. (2004) reported that the absence of water vapor increased the adsorption of Cl compounds on activated carbon in flue gas condition. Summary In this chapter, the active pha se of SCR catalysts, V 2 O 5 was doped on the SiO 2 and SiO 2 TiO 2 supports. Compared to SiO 2 TiO 2 composites improvements in Hg removal from flue gas were observed in fixed bed studies using both pellet and powder forms of the SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 ca talysts. No UV light activation is needed for the vanadia based catalysts, which simplifies the system and reduces the operating cost. For SiO 2 V 2 O 5 catalysts, the Hg removal efficiency increased as the V 2 O 5 loading increased from 2 to 8% but decreased as the V 2 O 5 loading further increased to 10%. The results suggested that the optimal V 2 O 5 loading for an optimal catalytic activity is somewhere between 5 and 8%. The SV5 catalyst reached a steady state activity during a 20 hr test and no deactivation of the catalyst was observed. The SiO 2 TiO 2 V 2 O 5 catalysts have an even greater ability of oxidizing Hg compared to SiO 2 V 2 O 5 which can be advantageous to power plants equipped with wet scrubbers where oxidized Hg can be easily captured. The different supports o f V 2 O 5 may account for the difference in their catalytic activity. The mechanisms of Hg oxidation on the SiO 2 V 2 O 5 catalyst have also been investigated. It was found that the Hg oxidation may follow an Eley Rideal mechanism where HCl, NO, and NO 2 are first adsorbed on the catalyst and then react with gas phase Hg. While HCl, NO, and NO 2 promote Hg oxidation, SO 2 has an insignificant effect and water vapor

PAGE 89

89 dramatically inhibits Hg oxidation. Penetration of oxidized Hg was observed in the presence of NO 2 whi ch is probably due to the formation of volatile Hg(NO 3 ) 2 It was verified in this work that V 2 O 5 is the active center for the Hg oxidation, but further studies are needed to better understand the enhancement of Hg oxidation by the SiO 2 TiO 2 V 2 O 5 catalysts.

PAGE 90

90 Table 5 1. Experimental parameters for activity measurement of the catalysts in pellet form Catalyst Composition Mass (g) Carrier gas UV light ST12 12% TiO 2 + 88% SiO 2 8.0 FG1, FG2 On/Off ST12V2 12% TiO 2 + 2% V 2 O 5 + 86% SiO 2 4.0, 8.0 FG1, FG2 On/Off SV2 2% V 2 O 5 + 98% SiO 2 4.0 FG1, FG2 On/Off Table 5 2. Experimental parameters for activity measurement of the catalysts in powder form Catalyst Composition Mass (mg) Carrier gas SV2 2% V 2 O 5 + 98% SiO 2 500 FG3 SV5 5% V 2 O 5 + 95% SiO 2 500 FG3; individual gases SV8 8% V 2 O 5 + 92% SiO 2 500 FG3 SV10 10% V 2 O 5 + 90% SiO 2 500 FG3 ST12V5 12% TiO 2 + 5% V 2 O 5 + 83% SiO 2 500 FG3 Table 5 3. BET surface areas of the catalysts Sample BET specific surface area (m 2 /g) Silica gel 341.8 SV2 263.4 SV5 283 .2 SV8 273.8 SV10 262.9 ST12 319.4 ST12V2 258.0 ST12V5 262.5 Table 5 4. Amounts of Hg captured and oxidized on the catalysts in a 6 hr test Catalyst Hg Captured (g/g catalyst) Hg Captured (g/g V 2 O 5 ) Hg Oxidized (g/g catalyst) Hg Oxidized (g/g V 2 O 5 ) SV2 (500 mg) 22.5 1130 22.5 1130 SV5 (500 mg) 75.1 1500 75.1 1500 SV8 (500 mg) 83.7 1050 106.3 1320 SV10 (500 mg) 63.1 630 86.6 870 ST12V5 (500mg) 52.2 1040 117.4 2350

PAGE 91

91 Figure 5 1. Experimental system for the f ixed bed study using powder catalysts

PAGE 92

92 Figure 5 2. XRD patterns of (a) SV2, (b) SV5, (c) SV8, (d) SV10, (e) ST12, and ( f ) ST12V5.

PAGE 93

93 Figure 5 3. Catalytic removal of Hg using the pellet catalysts under various conditions

PAGE 94

94 Figure 5 4. O utlet Hg concentration as a function of time using 500 mg powder of (a) SV2, (b) SV5, (c) SV8, (d) SV10, and (e) ST12V5. ( b ) ( c ) ( a )

PAGE 95

95 Figure 5 4. Continued. ( e ) ( d )

PAGE 96

96 Figure 5 5. Outlet Hg concentration as a function of time using 250 mg powder of SV5 ( 20 hr test).

PAGE 97

97 Figure 5 6. The role of flue gas components on Hg removal using 250 mg SV5 powder under dry conditions.

PAGE 98

98 Figure 5 7. The role of water vapor on Hg removal using 250 mg SV5 powder

PAGE 99

99 CHAPTER 6 UV ABSORPTION BASED MEASUREMENTS OF OZONE AND MERCURY : AN INVESTIGATION ON THE IR MUTUAL INTERFEREN CES Background The regulations on Hg emissions and the development of Hg control technologies require that reliable methods be used for accurate Hg measurement. Currently, the EPA accepted m ethods for Hg measurement in the United States are manual procedures based on wet chemistry such as EPA Methods 29 and 101A (for total mercury) and the Ontario Hydro M ethod (for speciated mercury) (Laudal et al., 2004) However, continuous mercury monitors (CMMs) have distinct advantages over these manual methods in that CMMs are able to provide a real time or near real time response for Hg measurements and to perform long t erm emission disadvantage of CMMs lies in their measurement interferences, which may vary depending on the principle of the Hg detection technique. A t omic ab sorption spectrometry ( AAS) is one of the major techniques applied to current CMMs. In the case of AAS, the co ncentration of elemental Hg in a gas sample is determined by measuring the light that is absorbed by Hg atoms at their characteristic wave lengths (usually at the resonance line of 254 nm). Thus, interferences can occ ur when other components of the sample gas possess strong absorption bands near this wavelength (254 nm) Environmental Technology Verification (ETV) program (USEPA, 2001a) identified sulfur dioxide (SO 2 ), nitrogen oxides (NO X ), hydrogen chloride (HCl) and chlorine (Cl 2 ) as interference Although pr etreatment or conditioning systems can be used to remove or negate the effects of Reprinted with permission from Li, Y., Lee, S. R., Wu, C. Y., 2006. UV Absorption Based Measurements of Ozone and Mercury: An Investigation of Their Mutual Interferences. Aerosol Air Qual. Res. 6, 418 429.

PAGE 100

100 these interfering gases prior to the sample delivery to the detectors (Laudal et al., 2004) they may increase the complexity and cost of the instrumentation and impair the real time feature of the CMMs. Consequently, many types of AAS based CMMs do not have pretreatment systems. Since the 254 nm Hg emission line also falls into the absorption spectra of ozone, which is capable of absorbing UV light below 290 nm, the presence of ozone in the sampling environment may impact the Hg measurement by AAS based CMMs. Granite and Pennline (2002) studied photochemical oxidation of Hg and speculated that photosensitized formation of ozone may interfere with Hg measurement by absorbing UV radiation. However, no quantitative data were reported on the magnitude of ozone interference. Therefore, the first objective of this study was to quantitatively investigate the interference of ozone on Hg measurement. This study may be of particular importance to ambient and indoor Hg measurement because ozone and Hg coexist in these conditions. Monitoring ground level of ozone, another significant air pollutant, is also required by US EPA. Control of ozone is expensive, w ith costs estimated in the billions of dollars (USEPA, 2005b) Hence, deployment of accurate ozone measurement is of great importance to demonstrate compliance with the N ational Ambient Air Quality Standard (NAAQS) for ozone. M any methods have been developed for ozone measurement where UV absorption and gas phase chemiluminescence are the major techniques used nowadays. The method of UV absorption is based on the principle that upon exposure to UV light ozone will absorb some of the light and the intensity difference is directly proportional to the concentration of ozone. Frequently the UV light source is a 254 nm emission line from a Hg discharge lamp. Known interferences on this type of ozone detection method include gaseous hydrocarbons with strong absorption at 254 nm, such as aromatic hydrocarbons (i.e., benzene and substituted benzene rings) (NARSTO, 1999)

PAGE 101

101 Since 254 nm is exactly one of the H g absorption lines, it is speculated that even a small amount of Hg in the sample gas may absorb a considerable amount of UV light. The U.S EPA (1999b) reported that at a baseline ozone concentration of approximately 75 ppb, the addition of 0.04 p pb (300 ng m 3 at room temperature) Hg caused an increase in measured ozone concentration by 12.8% at low humidity (RH = 20 ~ 30%) and 6.4% at high humidity (RH = 70 ~80%) using a UV photometric ozone monitor. The interferences of Hg using another two type s of ozone monitors were above 30% at either low or high humidity. However, the interference data were reported only at one Hg level (0.04 ppb). More data at other levels of Hg are needed to determine the relationship between Hg concentration and its corre sponding interference. Therefore, the second objective of this study was to quantitatively investigate Hg interference on ozone measurement. This is of importance in accurate ozone measurement in ambient and indoor conditions. Methods Descriptions of Hg a nd Ozone I nstruments The CMM used in this study is a RA 915+ Hg analyzer (OhioLumex Co.) which is capable of recording Hg concentrations every second It employs Zeeman AAS using High Frequency Modulated light polarization (ZAA S HFM) (Sholupov et al., 2004) which combines the approach of AAS with a simultaneous background correction provided by the Zeeman splitting of the Hg resonance line (254 nm ). In the RA 915+ Hg analyzer the emissions from a Hg discharge lamp are subjected to a strong magnetic field, which causes the three fold splitting ) have identical intensity when Hg is absent in the analytical cell. When Hg is presen t in the cell, the difference between the intensities of the two components is proportional to the Hg concentration. The calibration was conducted by the manufacturer using Dynacal permeation device (VICI

PAGE 102

102 Metronics Inc.) which is certified traceable to NI ST (National Institute of Standards and Technology) standards. Two optical cells are available for different ranges of Hg concentration. A single path cell is available for measuring higher Hg concentrations from 0.5 to 200 g m 3 A multi path cel l with a n effective length of 9.6 m is used to enhance the sensitivity of analysis, and thus the detection limit can reach as low as 2 ng m 3 The RA 915+ Hg analyzer may not be suitable for ambient Hg measurement since the ambient Hg concentration has been report ed to be approximately 1.5 to 1.9 ng m 3 in the northern hemisphere (Ebinghaus et al., 2002; Weiss Penzias et al., 2003) which is lower than t he detection limit of this instrument. However, the RA 915+ Hg analyzer has been used for measurement in st ationary sources or Hg contaminated sites (including indoor areas) (Kinsey et al., 2004; Pogarev et al., 2002; Sholupov et al., 2004) A M146 dynamic gas calibration system (Thermo Electron Instrument) served as the ozon e generating source using its internal ozonator. The precision of the ozone concentration that can be generated is 1 ppb. A M49 UV photometric ozone analyzer (Thermo Electron Instrument) was used to measure the ozone concentration in the sample gas. The UV light source in the ozone analyzer is a 254 nm emission line from a Hg discharge lamp. The full scale of the ozone analyzer was set to be from 0 to 500 ppb. Its precision is 2 ppb while the noise is within 1 ppb. This type of ozone monitor is equipped wi th a standard (manganese dioxide) scrubber, which was reported to suffer the lowest interference from Hg compared to other two types of ozone monitors (USEPA, 1999b) A M49 PS UV photometric ozone calibrator (Thermo Electron Instrument) was applied to calibrate b oth the ozone generator and analyzer. Experimental S etup and P rocedures A schematic diagram of the experimental system is shown in Figure 6 1. All the experiments were carried out at room temperature (25 1 C ). To test the ozone interference on Hg measur ement (Figure 6 1A), zero air was produced using a zero air supplier (M111, Thermo

PAGE 103

103 Electron Instrument) and then passed through the ozone generator to provide designated ozone concentrations. The total air flow rate at the outlet of the ozone generator was controlled to be 5 L/min. The ozone laden air was then divided into two streams and connected to the RA 915+ Hg analyzer and the ozone analyzer respectively. The reading from the Hg analyzer would indicate potential interference caused by ozone. The inter ferences from a designated range of ozone concentrations were measured. Each time before changing the ozone concentration, the entire system was purged by ozone free air until the readings from both Hg and ozone analyzers were zero. This was to minimize th e experimental error from residual ozone in the system. Finally, the sample gas was cleaned through an activated carbon trap before exhausted to the vent hood. Figure 6 1B shows the experimental setup for testing Hg interference on ozone measurement. The incoming zero air with a flow rate of 5 L/min was split into two streams. One stream was passed through the surface of a liquid Hg reservoir, which was placed in an ice water bath to maintain a constant Hg vapor pressure. By doing this, saturated Hg vapor was introduced into the system. The other stream served as dilution air and was used to adjust the Hg concentration. Both air streams were controlled by mass flow controllers ( MFC, Model. FMA 5400/5500, Omega Engineering, Inc.). The RA 915+ Hg analyzer was used to measure the Hg concentration in the sample gas while the ozone analyzer was used to detect the potential interference caused by Hg. Similarly, the interferences from a designated range of Hg concentrations were measured, and each time before chang ing the Hg concentration, the entire system was cleaned until the readings from both Hg and ozone analyzers were zero. Results and Discussion Interference of O zone on Hg M easurement The NAAQS for ozone is 80 ppb for an 8 hour average and 120 ppb for a 1 ho ur average (USEPA, 1997b) Thus, the ozone conc entration generated in this work ranged from 0 to 120

PAGE 104

104 ppb. The corresponding interference on the RA 915+ Hg analyzer is shown in Figure 6 2. The blank test showed that no interference was detected when no ozone was fed into the gas stream. As the ozone co ncentration increased, the reading on the Hg analyzer was almost linearly elevated. At each ozone concentration level, tests were repeated for three times. For ozone concentrations at 80 and 120 ppb, the interferences on Hg measurement reached approximatel y 46 and 63 ng m 3 respectively. The relationship between ozone concentration and its corresponding interference can be approximated as: ( 6 1) where is the equivalent Hg concentration, i.e., the inter ference on Hg analyzer (in unit of ng m 3 ) and is the ozone concentration (in unit of ppb). 1 ng m 3 of Hg is equivalent to 1.22 10 4 ppb at room temperature. For a convenient understanding of the significance of the interferen ce magnitudes, units used in Equation 6 1 and Equation 6 2 are in line with the EPA standards (USEPA, 1997b, 1999a) This observed ozone interference may have an important effect on Hg measurement. For example, Ferrara (1999) used the RA 915+ Hg analyzer to measure the Hg distribution over the area of Idria, where one of the largest European Hg mines was located. The Hg concentration was reported to range from 50 to 170 ng m 3 in the central part of Idria near the Hg mines and dumps. Suppose that 40 ppb ozone existed in that local atmosphere, an overestimate of Hg concentration at about 22 ng m 3 would have been involved according to the results from this work. Indoor Hg measurement can also be affected by ozone interference. Although the indoor ozone levels are typically less than those outdoors (Weschler, 2000) they can be much greater when strong ozone generating sources are present such as photocopiers, electr ostatic filters, and ozone generators (Godish, 2001) In addition, the us e of ozone for the removal of indoor air

PAGE 105

105 contaminants has been widely promoted in the United States (Li et al., 2002) In those areas with elevated ozone concentration s, the interference of ozone on Hg measurement may result in a significant overestimate of Hg concentrations when using RA 915+ or other similar Hg analyzers. It has been indicated by Singhvi et al. (2001) that water vapor can have positive interference on UV Hg analyzers. The magnitude of this interference was not reported, and it could vary for different UV Hg analyzers. If this positive interference of water vapor were applied to the RA 915+ Hg analyzer used in this study, the total interference on Hg measurement (caused by both ozone an d water vapor) could be even larger. The interference of ozone on Hg measurement can also impact the risk assessment of human exposure to Hg. The Reference Concentration ( RfC ) for elemental Hg specified by the EPA is 300 ng m 3 based on central nervous system (CNS) effects in humans (USEPA, 1999a) Hg concentrations above this level may result in a further investigation of hazardous exposure. According to the findings in this work, an ozone concentration in the range of 0 120 ppb can exert a positive bias in Hg measurement up to 21% (63 ng m 3 ) of the EPA RfC (300 ng m 3 ). The interference may be especially critical for conditions where the measured Hg concentration is slightly above the RfC level, because after subtracting the bias caused by ozone interference, the actual Hg concentration may not exceed the RfC any more. Therefore, eliminating ozone from sample gas is essential to obtain accurate Hg con centration and thus is important for risk assessment of human exposure to Hg. Helmig (1997) reviewed ozone removal techniques in sampling of atm ospheric volatile organic trace gases. These techniques may be applicable to ozone removal in Hg sampling as long as they do not tamper with Hg concentrations.

PAGE 106

106 Interference of Hg on Ozone M easurement A designated range of Hg concentrations were fed into t he gas stream to investigate the possible Hg interference on ozone measurement. Due to the limitation of the Hg vapor generating unit used in this work (the dilution ratio was relatively small compared with the saturated Hg vapor concentration), the minimu m Hg concentration introduced was about 2300 ng m 3 In addition, Hg levels were controlled so that the interferences on ozone readings were within the measurement range of the ozone analyzer (0 500 ppb). A blank test showed no interference when no Hg was present. As the Hg concentration in the sample gas increases, the corresponding reading on the ozone analyzer also increases (as shown in Figure 6 3). An approximately linear relationship can be obtained as expressed in the following equation: (6 2) where is the equivalent ozone concentration, i.e., the interference on ozone analyzer (in unit of ppb) and is the Hg concentration (in unit of ng m 3 ). Equation 6 2 implies that Hg c an exert a significant interference on ozone measurement, which is very likely due to the UV absorption by Hg when passing through the ozone analyzer (USEPA, 1999b) Although the Hg level in the ambient and indoor environment is typically lower than the minimum H g concentration (2300 ng m 3 ) tested in this work, the results obtained can be used as a reference to predict the practical conditions. It should be noted that extrapolation beyond the tested concentration range might involve errors that could impact the a ccuracy of the prediction. However, given the high value of R 2 (0.9957) of the regression analysis which has incorporated the origin point, it is suggested that the error associated with the extrapolation to the range of 0 to 2300 ng m 3 may not be signifi cant. Carpi and Chen (2001) reported that the highest Hg concentration measured at 12 indoor sites in New York City was

PAGE 107

107 523 ng m 3 This Hg level would result in 6 1 ppb interference on ozone measurement using a UV absorption based ozone analyzer, provided that Equation 6 2 is valid at lower Hg concentration levels. The interference, 61 ppb, added to the normal ambient ozone concentration (0 50 ppb) (Lim and Turpin, 2002) may have many chances to exceed the NAQQS for ozone (80 ppb). In the case when Hg concentration is equal to the EPA RfC (300 ng m 3 ), an interference of 35 ppb would be involved, which is comparable to the average ambi ent ozone concentration. Since indoor ozone concentration is typically lower than that outdoors (Weschler, 2000) and indoor Hg concentrations generally higher than that outdoors due to various indoor Hg contamination sources (such as accidental spills of Hg from natural gas meters, Hg thermometers, fluorescent light bulbs, etc.) (Carpi and Chen, 2001) the impact of Hg interference on indoor ozone measurement may be much greater. These results indicate that it is essential to eliminate the Hg interference to obtain correct ozone concentration at Hg contaminated places. Now that the linear relationship between Hg concentration and corresponding interference on ozone measurement was successfully established, it is necessary to compare the results with those by the EPA (1999b) where the same type of ozone monitor was used. At a Hg concentration of 0.04 ppb (328 ng m 3 at 25C), Equation 6 2 predicts an ozone interference of 38 ppb, whereas the EPA (1999b) reported approximately 10 ppb at low humidity (RH = 20 to 30%) and 5 ppb at high humidity (RH = 70 to 80%). Since in this work the experiments were conducted using dry zero air, comparison of the above results suggests that higher humidity may di minish the interference caused by Hg. A possible reason is the deposition of water vapor in the optical cell attenuates the incident UV light, causing a negative interference by water vapor itself. The finding also suggests that the reported interference m ay not be so significant in

PAGE 108

108 ambient and indoor air as in the dry air used in this work. Further research is needed to verify this hypothesis. Summary Mutual interferences of UV absorption based measurements of ozone and Hg were investigated in this study It was found that ozone in the range of 0 to 120 ppb can exert an interference of up to 63 ng m 3 on an AAS based Hg analyzer. A linear relationship was established between the ozone concentration and corresponding interference on Hg measurement. On the other hand, it was found that Hg can also result in significant interferences on an ozone analyzer based on UV absorption. Results showed that Hg at a concentration of 300 ng m 3 can potentially cause a bias in ozone measurement of approximately 35 ppb. Th ese mutual interferences may consequently affect the risk assessment of human exposure to both Hg and ozone. It should be noted that the results were obtained from a relatively small set of equipment (only one Hg and one ozone analyzer). There are many dif ferent types of Hg or ozone analyzers based on principle of UV absorption. It is possible that certain types of analyzers are not compromised by the interferences found in this work. T hus, the findings in this work should be considered as analyzer specific Further investigation is needed to determine whe ther the trends observed in this work can be extended to other types of analyzers

PAGE 109

109 Figure 6 1. Experimental setup. A) Ozone interference on Hg measurement. B) Hg interference on ozone measurement. Zero Air Ozone Analyzer Hg Analyzer Carbon Trap Vent V ent A B Ice Water Bath Hg 2 Way Valve MFC Zero Air Ozone Generator Hg Analyzer Carbon Trap Ozone Analyzer

PAGE 110

110 Figure 6 2. Measurement interference of ozone on the RA 915+ Hg analyzer as a function of ozone concentration (The error bars represent one standard deviation).

PAGE 111

111 Figure 6 3. Measurement interference of Hg on the ozone analyzer as a function of Hg concentration.

PAGE 112

112 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS This doctoral research has focused on the fixed bed studies of Hg 0 removal using nanostructured SiO 2 /TiO 2 / V 2 O 5 composites. Three types of composites were tested for their catalyt ic activities in both pellet and powder forms: (1) TiO 2 nanoparticles doped on high surface area SiO 2 gel (SiO 2 TiO 2 ), (2) V 2 O 5 doped on SiO 2 gel (SiO 2 V 2 O 5 ), and (3) V 2 O 5 doped on the SiO 2 TiO 2 support (SiO 2 TiO 2 V 2 O 5 ). Experiments were conducted under bo th room conditions and flue gas conditions. The kinetics of the Hg oxidation was studied, and the interactions between Hg and the flue gas components on the catalyst surfaces were investigated. The catalytic activities of the three catalysts were also comp ared. The following conclusions have been obtained from this research. Conclusion 1: A Langmuir Hinshelwood (L H) model can be used to express the kinetics of photocatalytic oxidation of Hg 0 on the SiO 2 TiO 2 nanocomposite under room conditions. Good agreem ent between the experimental data and the L H model was demonstrated. The model predicted a great potential of the SiO 2 TiO 2 nanocomposite for Hg 0 removal even at very high Hg 0 concentrations. The rate of photocatalytic Hg 0 oxidation increased when the inl et Hg 0 concentration increased and it reached a maximum value in the absence of water vapor. The addition of water vapor was found to inhibit Hg 0 photocatalytic oxidation, which may be explained by the competitive adsorption of water vapor with Hg 0 on the TiO 2 surface. Conclusion 2: The mechanisms of Hg 0 removal and reemission from used catalysts were investigated in a fixed bed reactor at 65 C using air as the carrier gas. Without UV irradiation, Hg 0 adsorption was found to be insignificant, but it could be enhanced by the photocatalytic oxidation product, HgO, possibly due to the high affinity between HgO and Hg 0 Under dry conditions 95% of Hg 0 can be removed; however, increased humidity levels

PAGE 113

113 remarkably suppress both Hg 0 adsorption and photocatalytic o xidation. Introducing water vapor can also result in significant reemission of captured Hg 0 from the nanocomposite, which may be ascribed to the repellant effect of water vapor adsorbed on the superhydrophilic TiO 2 surface. Exposure to UV light was found e ither to prohibit Hg 0 reemission when photocatalytic oxidation of reemitted Hg 0 prevailed or to promote Hg 0 reemission when photocatalytic reduction of Hg O to Hg 0 dominated later on. It is concluded that Hg 0 capture on the SiO 2 TiO 2 nanocomposite in a humi d environment under UV irradiation is controlled by four mechanisms: adsorption, photocatalytic oxidation, desorption, and photocatalytic reduction. The observed inhibitory effect of water vapor is contributed by its competitive occupancy of the available adsorption sites, displacement of adsorbed Hg 0 and participation in the photocatalytic reduction of HgO to Hg 0 Conclusion 3: Hg 0 removal using the SiO 2 TiO 2 nanocomposite has been investigated in simulated flue gas of coal fired power plants. The flue ga s components were found to have significant effects on Hg removal efficiency. H 2 O inhibited Hg oxidation and capture and the inhibitory effect was proportional to the H 2 O concentration. HCl enhanced Hg 0 oxidation probably following an Eley Redeal mechanism where adsorbed Cl species reacts with gas phase Hg 0 SO 2 promoted Hg oxidation, possibly forming mercury sulfate species. NO significantly inhibited Hg removal by scavenging OH radicals that are necessary for Hg 0 oxidation. The effect of NO 2 was found to be insignificant. Experiments in simulated flue gases also showed that high rank coals are preferable to low rank coals because of the lower moisture and higher HCl and SO 2 concentrations in the flue gas. Conclusion 4: A method has been developed to dope V 2 O 5 on the SiO 2 and SiO 2 TiO 2 supports. Improvements in Hg removal from flue gas were observed in fixed bed studies

PAGE 114

114 using both pellet and powder forms of the SiO 2 V 2 O 5 and SiO 2 TiO 2 V 2 O 5 catalysts. No UV light activation is needed for the V 2 O 5 doped catal ysts. For SiO 2 V 2 O 5 catalysts, the Hg removal efficiency increased as the V 2 O 5 loading increased from 2 to 8% but decreased as the V 2 O 5 loading further increased to 10%. The results suggested that the optimal V 2 O 5 loading for a maximum catalytic activity i s somewhere between 5 and 8%. The SiO 2 TiO 2 V 2 O 5 catalysts have an even greater ability of oxidizing Hg compared to SiO 2 V 2 O 5 which can be advantageous to power plants equipped with wet scrubbers where oxidized Hg can be easily captured. It was found that the Hg oxidation on the V 2 O 5 doped catalysts may follow an Eley Rideal mechanism where HCl, NO, and NO 2 are first adsorbed on the catalyst and then react with gas phase Hg. On the contrary, water vapor dramatically inhibits Hg oxidation while SO 2 has an i nsignificant effect. Conclusion 5: Mutual interferences of UV absorption based measurements of ozone and Hg were also investigated in this research. It was found that ozone in the range of 0 to 120 ppb can exert an interference of up to 63 ng m 3 on an AA S based Hg analyzer and the interference is linearly related to the ozone concentration. On the other hand, it was found that Hg can have significant interferences on ozone analyzers that are based on UV absorption. Hg at a concentration of 300 ng m 3 can potentially cause a bias in ozone measurement of approximately 35 ppb, an average ozone concentration in the air under normal conditions. These mutual interferences may consequently affect the risk assessment of human exposure to both Hg and ozone. Based o n the findings of this doctoral research, future research is recommended on the following topics:

PAGE 115

115 1. It has been verified in this research that V 2 O 5 is the active center for the Hg oxidation, but the SiO 2 TiO 2 V 2 O 5 composites exhibit much higher oxidation abi lities than the SiO 2 V 2 O 5 composites. This warrants further studies to understand the enhancement of Hg oxidation on the superior TiO 2 support and to explore the optimal V 2 O 5 / TiO 2 ratio. Finding the optimal V 2 O 5 / TiO 2 loading is of great importance to the a dvancement of a most cost effective catalyst. Surface analysis techniques such as X ray photoelectron spectroscopy (XPS) are recommended to be used to examine the surface structure of the catalysts before and after the Hg oxidation experiments. XPS may pro vide information on the oxidation states of the vanadium and the dispersion of the vanadia on the surface by providing the V/Ti ratio on the surface. This information would be helpful for a better understanding of the catalyst characteristics as well as th e reaction mechanisms. 2. A long term test on the performance of the SiO 2 TiO 2 V 2 O 5 composites on Hg removal is recommended. Although this research demonstrated a stabilized activity of the catalyst in a 20 hr test, a longer testing time, probably a few hund red hours, would be considered sufficient to verify its long term performance. However, some modifications of the current system may be needed for the long term test. 3. Exploring the potential of the SiO 2 TiO 2 V 2 O 5 composites as a multipollutant control stra tegy, i.e. a low temperature SCR catalyst for control of both Hg and NOx. It is promising because the catalyst contains the same active phases as the commercial SCR catalysts. However, the effect of the injected ammonia (NH 3 ) on Hg removal is yet to be inv estigated. 4. A pilot scale study using the SiO 2 TiO 2 V 2 O 5 composites is recommended as a scale up of the current bench scale system. Kinetic studies may be necessary to determine the amount (or total surface area) of the catalyst that is needed for treating a higher flow rate of flue gas. A

PAGE 116

116 good configuration of the packed bed reactor may also be critical to achieve a high efficiency of Hg removal in the pilot scale study.

PAGE 117

117 LIST OF REFERENCES Argyle, M. D., Chen, K. D., Resini, C., Krebs, C., Bell, A. T., I glesia, E., 2004. Extent of reduction of vanadium oxides during catalytic oxidation of alkanes measured by in situ UV visible spectroscopy. J. Phy. Chem. B 108, 2345 2353. Augugliaro, V., Coluccia, S., Loddo, V., Marchese, L., Martra, G., Palmisano, L., S chiavello, M., 1999. Photocatalytic oxidation of gaseous toluene on anatase TiO 2 catalyst: mechanistic aspects and FT IR investigation. Appl. Catal. B Environ. 20, 15 27. Bai, M., Zhang, Z., Bai, M., Yi, C., Bai, X., 2006. Removal of SO 2 from Gas Streams b y Oxidation using Plasma Generated Hydroxyl Radicals. Plasma Chem. Plasma Process. 26, 177 186. Benson, S. A., Laumb, J. D., Crocker, C. R., Pavlish, J. H., 2005. SCR catalyst performance in flue gases derived from subbituminous and lignite coals. Fuel Pro cess. Technol. 86, 577 613. Berglund, F., Bertin, M., 1969. Chemical Fallout. Tomas Publisher, Springfield. Bidstrup, P. C., 1964. Toxicity of Mercury and its Compounds. Elsevier, Amsterdam. Borderieux, S., Wu, C. Y., Bonzongo, J. C., Powers, K., 2004. Con trol of Elemental Mercury Vapor in Combustion Systems Using Fe 2 O 3 Nanoparticles. Aerosol Air Qual. Res. 4, 74 90. Bosch, H., Janssen, F., 1988. Formation and control of nitrogen oxides. Catal. Today 2, 369 379. Brown, T. D., Smith, D. N., Hargis, R. A., O 'Dowd, W. J., 1999. Mercury measurement and its control: What we know, have learned, and need to further investigate. J. Air. Waste Manage. 49, 628 640. Busca, G., Lietti, L., Ramis, G., Berti, F., 1998. Chemical and mechanistic aspects of the selective ca talytic reduction of NOx by ammonia over oxide catalysts: A review. Appl. Catal. B Environ. 18, 1 36. Calvert, J. G., Lindberg, S. E., 2005. Mechanisms of mercury removal by O 3 and OH in the atmosphere. Atmospheric Environment 39, 3355 3367. Canela, M. C. Alberici, R. M., Jardim, W. F., 1998. Gas phase destruction of H 2 S using TiO 2 /UV VIS. J. Photochem. Photobio. A Chem. 112, 73 80. Carey, T. R., Hargrove, C. W., Richardson, C. F., Chang, R., 1998. Factors affecting mercury control in utility flue gas usi ng activated carbon. J. Air Waste Manage. Assoc. 48, 1166 1174.

PAGE 118

118 Carpi, A., Chen, Y. F., 2001. Gaseous elemental mercury as an indoor air pollutant. Environ. Sci. Technol. 35, 4170 4173. Chen, S., Rostam Abadi, M., Chang, R., 1996. Mercury removal from comb ustion flue gas by activated carbon injection: mass transfer effects. Prepr. Pap. Am. Chem. Soc., Div. Fuel Chem. 41, 442 446. Ebinghaus, R., Kock, H. H., Coggins, A. M., Spain, T. G., Jennings, S. G., Temme, C., 2002. Long term measurements of atmospheric mercury at Mace Head, Irish west coast, between 1995 and 2001. Atmos. Environ. 36, 5267 5276. Edwards, J. R., Srivastava, R. K., Kilgroe, J. D., 2001. A study of gas phase mercury speciation using detailed chemical kinetics. J. Air Waste Manage. Assoc. 51 869 877. Eswaran, S., Stenger, H. G., 2005. Understanding Mercury Conversion in Selective Catalytic Reduction (SCR) Catalysts. Energy Fuels 19, 2328 2334. Ferrara, R., 1999. Mercury Contaminated Sites. Springer, Berlin. Flora, J. R. V., Vidic, R. D., Li u, W., Thurnau, R. C., 1998. Modeling powdered activated carbon injection for the uptake of elemental mercury vapors. J. Air Waste Manage. Assoc. 48, 1051 1059. Fujishima, A., Rao, T. N., Tryk, D. A., 2000. Titanium dioxide photocatalysis. J. Photochem. Ph otobio. C: Photochem. Reviews 1, 1 21. Galbreath, K. C., Zygarlicke, C. J., Tibbetts, J. E., Schulz, R. L., Dunham, G. E., 2005. Effects of Fe 2 O 3 Fe 2 O 3 and HCl on mercury transformations in a 7 kW coal combustion system. Fuel Process. Technol. 86, 429 448. Godish, T., 2001. Indoor En vironmental Quality. Lewis Publishers, Boca Raton, Florida. Goodsite, M. E., Plane, J. M. C., Skov, H., 2004. A theoretical study of the oxidation of Hg 0 to HgBr 2 in the troposphere. Environ. Sci. Technol. 38, 1772 1776. Grabowski, R., Pietrzyk, S., Sloczy nski, J., Genser, F., Wcislo, K., Grzybowska Swierkosz, B., 2002. Kinetics of the propane oxidative dehydrogenation on vanadia/titania catalysts from steady state and transient experiments. Appl. Catalysis A Gen. 232, 277 288. Granite, E. J., Pennline, H. W., 2002. Photochemical removal of mercury from flue gas. Ind. Eng. Chem. Res. 41, 5470 5476. Granite, E. J., Pennline, H. W., Hargis, R. A., 2000. Novel sorbents for mercury removal from flue gas. Ind. Eng. Chem. Res. 39, 1020 1029. Guan, K. H., 2005. Rel ationship between photocatalytic activity, hydrophilicity and self cleaning effect of TiO 2 /SiO 2 films. Surface & Coatings Technology 191, 155 160.

PAGE 119

119 Helmig, D., 1997. Ozone removal techniques in the sampling of atmospheric volatile organic trace gases. Atmo s. Environ. 31, 3635 3651. Kantcheva, M., Bushev, V., Klissurski, D., 1994. Study Of The NO 2 NH 3 Interaction On A Titania (Anatase) Supported Vanadia Catalyst. J. Catalysis 145, 96 106. Kim, S. B., Hong, S. C., 2002. Kinetic study for photocatalytic degra dation of volatile organic compounds in air using thin film TiO2 photocatalyst. Appl. Catal. B Environ. 35, 305 315. Kinsey, J. S., Anscombe, F. R., Lindberg, S. E., Southworth, G. R., 2004. Characterization of the fugitive mercury emissions at a chlor alk ali plant: overall study design. Atmos. Environ. 38, 633 641. Kobayashi, M., Kuma, R., Masaki, S., Sugishima, N., 2005. TiO 2 SiO 2 and V 2 O 5 /TiO 2 SiO 2 catalyst: Physico chemical characteristics and catalytic behavior in selective catalytic reduction of NO by NH 3 Appl. Catal. B Environ. 60, 173 179. Kobayashi, M., Kuma, R., Morita, A., 2006. Low temperature selective catalytic reduction of NO by NH 3 over V 2 O 5 supported on TiO 2 SiO 2 MoO 3 Catal. Letters 112, 37 44. Laudal, D. L., Brown, T. D., Nott, B. R., 20 00. Effects of flue gas constituents on mercury speciation. Fuel Process. Technol. 65, 157 165. Laudal, D. L., Thompson, J. S., Pavlish, J. H., Brickett, L. A., Chu, P., 2004. Use of continuous mercury monitors at coal fired utilities. Fuel Process. Techno l. 85, 501 511. Laumb, J. D., Benson, S. A., Olson, E. A., 2004. X ray photoelectron spectroscopy analysis of mercury sorbent surface chemistry. Fuel Process. Technol. 85, 577 585. Lee, C. W., Srivastava, R. K., Ghorishi, S. B., Hastings, T. W., Stevens, F M., 2004. Investigation of selective catalytic reduction impact on mercury speciation under simulated NOx emission control conditions. J. Air Waste Manage. Assoc. 54, 1560 1566. Lee, C. W., Srivastava, R. K., Ghorishi, S. B., Karwowski, J., Hastings, T. W., Hirschi, J. C., 2006. Pilot scale study of the effect of selective catalytic reduction catalyst on mercury speciation in Illinois and powder river basin coal combustion flue gases. J. Air Waste Manage. Assoc. 56, 643 649. Lee, T. G., Biswas, P., Hedric k, E., 2001. Comparison of Hg 0 capture efficiencies of three in situ generated sorbents. AICHE J. 47, 954 961. Lee, T. G., Biswas, P., Hedrick, E., 2004. Overall kinetics of heterogeneous elemental mercury reactions on TiO 2 sorbent particles with UV irradi ation. Ind. Eng. Chem. Res. 43, 1411 1417. Li, T. H., Turpin, B. J., Shields, H. C., Weschler, C. J., 2002. Indoor hydrogen peroxide derived from ozone/d limonene reactions. Environ. Sci. Technol. 36, 3295 3302.

PAGE 120

120 Li, Y., Wu, C. Y., 2006. Role of Moisture in Adsorption, Photocatalytic Oxidation, and Reemission of Elemental Mercury on a SiO 2 TiO 2 Nanocomposite. Environ. Sci. Technol. 40, 6444 6448. Li, Y., Wu, C. Y., 2007. Kinetic Study for Photocatalytic Oxidation of Elemental Mercury on a SiO 2 TiO 2 Nanocomp osite. Environ. Eng. Sci. 24, 3 12. Lim, H. J., Turpin, B. J., 2002. Origins of primary and secondary organic aerosol in Atlanta: Results' of time resolved measurements during the Atlanta supersite experiment. Environ. Sci. Technol. 36, 4489 4496. Maroto V aler, M. M., Zhang, Y. Z., Granite, E. J., Tang, Z., Pennline, H. W., 2005. Effect of porous structure and surface functionality on the mercury capacity of a fly ash carbon and its activated sample. Fuel 84, 105 108. Martra, G., Arena, F., Coluccia, S., F rusteri, F., Parmaliana, A., 2000. Factors controlling the selectivity of V 2 O 5 supported catalysts in the oxidative dehydrogenation of propane. Catal. Today 63, 197 207. McLarnon, C. R., Granite, E. J., Pennline, H. W., 2005. The PCO process for photochem ical removal of mercury from flue gas. Fuel Process. Technol. 87, 85 89. Meites, L., Ed. 1963. Handbook of Analytical Chemistry. New York, McGraw Hill. Meserole, F. B., Richardson, C. F., Miller, S. D., Searcy, K., Chang, R. 2000, Estimated the costs of el ectric utility mercury control using sorbent injection. Proceedings of the Air and Waste Management Association 93rd Annual Meeting, Salt Lake City, UT. Miller, S. J., Dunham, G. E., Olson, E. S., Brown, T. D., 2000. Flue gas effects on a carbon based merc ury sorbent. Fuel Process. Technol. 65, 343 363. Mishima, A., 1992. Bitter Sea: The Human Cost of Minamata Disease. Kosei Pub. Co., Tokyo, Japan. North American Research Strategy for Tropospheric Ozone (NARSTO), 1999. Measurement Methods Compendium Ozone Method. Oak Ridge National Laboratory, Oak Ridge, TN. http://narsto.ornl.gov/Compendium/methods/o3.shtml Date last accessed: June 2007. Niksa, S., Fujiwara, N., 2005a. Predicting extents of mercury oxidation in coal derived flue gases. J. Air Waste Manage. Assoc. 55, 930 939. Niksa, S., Fujiwara, N., 2005b. A predictive mechanism for mercury oxidation on selective catalytic reduction catalysts under coal derived flue gas. J. Air Waste Man age. Assoc. 55, 1866 1875.

PAGE 121

121 Niksa, S., Helble, J. J., Fujiwara, N., 2001. Kinetic modeling of homogeneous mercury oxidation: The importance of NO and H 2 O in predicting oxidation in coal derived systems. Environ. Sci. Technol. 35, 3701 3706. Norton, G. A., Y ang, H. Q., Brown, R. C., Laudal, D. L., Dunham, G. E., Erjavec, J., 2003. Heterogeneous oxidation of mercury in simulated post combustion conditions. Fuel 82, 107 116. Obee, T. N., 1996. Photooxidation of sub parts per million toluene and formaldehyde le vels an titania using a glass plate reactor. Environ. Sci. Technol. 30, 3578 3584. Obee, T. N., Hay, S. O., 1997. Effects of moisture and temperature on the photooxidation of ethylene on Titania. Environ. Sci. Technol. 31, 2034 2038. Olson, E. S., Sharma, R. K., Pavlish, J. H., 2002. On the analysis of mercuric nitrate in flue gas by GC MS. Anal. Bioanal. Chem. 374, 1045 1049. Pal, B., Ariya, P. A., 2004. Gas phase HO Initiated reactions of elemental mercury: Kinetics, product studies, and atmospheric impli cations. Environ. Sci. Technol. 38, 5555 5566. Pan, H. Y., Minet, R. G., Benson, S. W., Tsotsis, T. T., 1994. Process for Converting Hydrogen Chloride to Chlorine. Ind. Eng. Chem. Res. 33, 2996 3003. Parfitt, G. D., Ramsboth.J, Rocheste.Ch, 1971. Infra Red Study of Hydrogen Chloride Adsorption on Rutile Surfaces. Trans. Faraday Society 67, 3100 &. Parvulescu, V. I., Grange, P., Delmon, B., 1998. Catalytic removal of NO. Catal. Today 46, 233 316. Pavlish, J. H., Sondreal, E. A., Mann, M. D., Olson, E. S., Galbreath, K. C., Laudal, D. L., Benson, S. A., 2003. Status review of mercury control options for coal fired power plants. Fuel Process. Technol. 82, 89 165. Pitoniak, E., Wu, C. Y., Londeree, D., Mazyck, D., Bonzongo, J. C., Powers, K., Sigmund, W., 2003 Nanostructured silica gel doped with TiO2 for mercury vapor control. J. Nanopart. Res. 5, 281 292. Pitoniak, E., Wu, C. Y., Mazyck, D. W., Powers, K. W., Sigmund, W., 2005. Adsorption enhancement mechanisms of silica titania nanocomposites for elemental mercury vapor removal. Environ. Sci. Technol. 39, 1269 1274. Pogarev, S. E., Ryzhov, V., Mashyanov, N., Sholupov, S., Zharskaya, V., 2002. Direct measurement of the mercury content of exhaled air: a new approach for determination of the mercury dose receiv ed. Anal. Bioanal. Chem. 374, 1039 1044. Presto, A. A., Granite, E. J., 2006. Survey of Catalysts for Oxidation of Mercury in Flue Gas. Environ. Sci. Technol. 40, 5601 5609.

PAGE 122

122 Raillard, C., Hequet, V., Le Cloirec, P., Legrand, J., 2004. Kinetic study of keto nes photocatalytic oxidation in gas phase using TiO 2 containing paper: effect of water vapor. J. Photochem. Photobio. A Chem. 163, 425 431. Ravichandran, M., 2004. Interactions between mercury and dissolved organic matter a review. Chemosphere 55, 319 3 31. Rodella, C. B., Nascente, P. A. P., Franco, R. W. A., Magon, C. J., Mastelaro, V. R., Florentino, A. O., 2001. Surface characterisation of V 2 O 5 /TiO 2 catalytic system. Phy. Stat. Solidi A Appl. Res. 187, 161 169. Rodriguez, S., Almquist, C., Lee, T. G., 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. Romero, C. E., Li, Y., Bilirgen, H., Sarunac, N., Levy, E. K., 200 6. Modification of boiler operating conditions for mercury emissions reductions in coal fired utility boilers. Fuel 85, 204 212. Rostam Abadi, M., Chen, S. G., Hsi, H. C., Rood, M., Chang, R., Carey, T. R., Hargrove, B., Richardson, C., Rosenhoover, W., M eserole, F. 1997, Novel vapor phase mercury sorbents. Proceedings of the EPRI DOE EPA Combined Utility Air Pollutant Control, Washington, DC. Schroeder, W. H., Munthe, J., 1998. Atmospheric mercury An overview. Atmospheric Environment 32, 809 822. Senio r, C. L., 2006. Oxidation of mercury across selective catalytic reduction catalysts in coal fired power plants. J. Air Waste Manage. Assoc. 56, 23 31. Senior, C. L., Sarofim, A. F., Zeng, T. F., Helble, J. J., Mamani Paco, R., 2000. Gas phase transformatio ns of mercury in coal fired power plants. Fuel Process. Technol. 63, 197 213. Shang, J., Du, Y. G., Xu, Z. L., 2002. Photocatalytic oxidation of heptane in the gas phase over TiO2. Chemosphere 46, 93 99. Shikada, T., Fujimoto, K., Kunugi, T., Tominaga, H. Kaneko, S., Kubo, Y., 1981. Reduction of Nitric Oxide with Ammonia on Vanadium Oxide Catalysts Supported on Homogeneously Precipitated Silica Titania. Ind. Eng. Chem. Prod. Res. Dev 20, 91 95. Sholupov, S., Pogarev, S., Ryzhov, V., Mashyanov, N., Stroga nov, A., 2004. Zeeman atomic absorption spectrometer RA 915+for direct determination of mercury in air and complex matrix samples. Fuel Process. Technol. 85, 473 485. Singhvi, R., Turpin, R., Kalnicky, D. J., Patel, J., 2001. Comparison of field and labora tory methods for monitoring metallic mercury vapor in indoor air. Journal of Hazardous Materials 83, 1 10.

PAGE 123

123 Sloczynski, J., 1996. Kinetics and mechanism of reduction and reoxidation of the alkali metal promoted vanadia titania catalysts. Appl. Catalysis A Gen. 146, 401 423. Sommar, J., Gardfeldt, K., Stromberg, D., Feng, X. B., 2001. A kinetic study of the gas phase reaction between the hydroxyl radical and atomic mercury. Atmospheric Environment 35, 3049 3054. Son, H. S., Lee, S. J., Cho, I. H., Zoh, K. D ., 2004. Kinetics and mechanism of TNT degradation in TiO2 photocatalysis. Chemosphere 57, 309 317. Tesser, R., Maradei, V., Di Serio, M., Santacesaria, E., 2004. Kinetics of the oxidative dehydrogenation of ethanol to acetaldehyde on V2O5/TiO2 SiO2 catal ysts prepared by grafting. Ind. Eng. Chem. Res. 43, 1623 1633. Tseng, H. H., Wey, M. Y., Liang, Y. S., Chen, K. H., 2003. Catalytic removal of SO 2 NO and HCl from incineration flue gas over activated carbon supported metal oxides. Carbon 41, 1079. US EPA 1997a. Mercury Study Report to Congress. EPA 452/R 97 003. US EPA, 1997b. National ambient air quality standards for ozone, final rule. Federal Register 62 (138). US EPA, 1997c. Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources: Hospital/Medical/Infectious Waste Incinerators; Final Rule. Fed. Regist., 62 48347 448391. US EPA, 1999a. Integrated Risk Information System (IRIS) on Elemental Mercury. National Center for Environmental Assessment, Office of Researc h and Development, Washington, DC. US EPA, 1999b. Laboratory Study to Explore Potential Interferences to Air Quality Monitors. EPA 454/C 00 002. US EPA, 2000. Regulatory Finding on the Emissions of Hazardous Air Pollutants from Electric Utility Steam Gener ating Units. Federal Register 65 (245), 79825 79831. US EPA, 2001a. Environmental Technology Verification Program, Verification Reports and Statements for each tested Mercury CEM. Research Triangle Park, NC. http://www.epa.gov/etv/verifications/vcenter1 11.html Date last accessed: June 2007. US EPA, 2001b. Standards of Performance for Large Municipal Waste Combustors; Final Rule and Proposed Rule. Fed. Regist., 66 57823 57828. US EPA, 2003. National Emission Standards for Hazardous Air Pollutants: Mercury Emissions from Mercury Cell Chlor Alkali Plants; Final Rule. Fed. Regist., 68 70903 70946.

PAGE 124

124 US EPA, 2005a. Clean Air Mercury Rule; Standards of Performance for New and Existing Statio nary Sources: Electric Utility Steam Generating Units. 40 CFR Parts 60, 63, 72, and 75. US EPA, 2005b. Regulatory Impact Analyses for the Particulate Matter and Ozone NAAQS and Proposed Regional Haze Rule. Research Triangle Park, NC. http://www.epa.gov/ttn/oarpg/naaqsfin/ria.html Date last accessed: June 2007. Wang, R., Hashimoto, K., Fujishima, A., Chikuni, M., Kojima, E., Kitamura, A., Shimohigoshi, M., Watanabe, T., 1997. Light induced amph iphilic surfaces. Nature 388, 431 432. Weast, R. C., Astle, M. J., Beyer, W. H., Eds. 1983. CRC Handbook of Chemistry and Physics. Boca Raton, FL, CRC Press. Weckhuysen, B. M., Keller, D. E., 2003. Chemistry, spectroscopy and the role of supported vanadiu m oxides in heterogeneous catalysis. Catal. Today 78, 25 46. Weiss Penzias, P., Jaffe, D. A., McClintick, A., Prestbo, E. M., Landis, M. S., 2003. Gaseous elemental mercury in the marine boundary layer: Evidence for rapid removal in anthropogenic pollutio n. Environ. Sci. Technol. 37, 3755 3763. Weschler, C. J., 2000. Ozone in indoor environments: Concentration and chemistry. Indoor Air 10, 269 288. 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. Yang, C. S., Chen, C. J., 2005. Synthesis and characterization of silica capped titania nanorods: An enhanced photocatalyst. Applied Catalysis a General 294, 40 48. Zha ng, F. S., Nriagu, J. O., Itoh, H., 2004. Photocatalytic removal and recovery of mercury from water using TiO 2 modified sewage sludge carbon. J. Photochem. Photobio. A: Chemistry 167, 223 228.

PAGE 125

125 BIOGRAPHICAL SKETCH Ying Li was born in 1977 in Luzhou, Sic huan Province, China. He earned his B.S. degree in thermal engineering in June 1999 at Zhejiang University, China. He continued his master studies in thermal engineering at Zhejiang University and received his M.S. degree in March 2002. Ying Li came to the United States in August 2002. He studied in mechanical engineering at Lehigh University, Pennsylvania, and worked as a research assistant at the Energy Research Center. He earned his second M.S. d egree at Lehigh University in 2004. He joined the research group of Dr. Chang Yu Wu at the University of Florida in August 2004 and started pursuing his Ph.D. degree in the Department of Environmental Engineering Sciences. His doctoral research focused on mercury removal from power plant flue gas using novel catalysts. Ying Li was the president of the student chapter of Air & Waste Management Association (A&WMA) at the University of Florida in 2006 2007. He was awarded the Milton Feldstein Memorial Scholar ship from A&WMA in 2006. He was awarded the Axel Hendrickson Scholarship and the Clair Fancy Scholarship from Florida Section A&WMA in 2005 and 2006, respectively. He won 1 st place in the student paper poster competition (doctoral level) in the 2006 A&WMA Annual Conference. He won 1 st place in the student poster competition (graduate level) in the 2006 Florida A&WMA Annual Conference.