<%BANNER%>

Production of Organic Pigment Nanoparticles by Stirred Media Milling

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 E20101123_AAAAAJ INGEST_TIME 2010-11-23T06:22:55Z PACKAGE UFE0011903_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 5625 DFID F20101123_AAAGUF ORIGIN DEPOSITOR PATH hamey_r_Page_57thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
85aa020a67ac581034e190f180a86a3c
SHA-1
e91516956637ace1f9ef072d4425cb3a652f2997
15418 F20101123_AAAGKK hamey_r_Page_44.jp2
9f449ac29a60038ac96990726e4a6c04
d3bfd191ab7f678165029a487626aadb51c90282
100930 F20101123_AAAGFN hamey_r_Page_38.jpg
75ccb9f111ea979057984848c30ea38d
aa9fc89135cf1aa6eacdfb7957918f8dc3d89849
2116 F20101123_AAAGPI hamey_r_Page_15.txt
3c8523f741601c774f77ff8360532b7a
c72bdd7529e761b8496104f93f9c4bea76cbbc99
19821 F20101123_AAAGUG hamey_r_Page_57.QC.jpg
a05cfb5b43cb624a867acf92483f6756
47ebc7d223914c474599901e24ee5ccc201d9811
86537 F20101123_AAAGKL hamey_r_Page_45.jp2
3ba481026b8041816ed2ff3cc6b4b66f
c6717bd07a65ad84e670e63695713f0cf4b81da2
1913 F20101123_AAAGFO hamey_r_Page_18.txt
f2482dae8dad1304993b86f39b00db10
eddfde65ccb8fd775a2dbe007c931cea76406b0e
1825 F20101123_AAAGPJ hamey_r_Page_16.txt
52743ac3fdd29fa8b400957b7f182fe1
30e4edb67d47fcb2f4597af4b99b7ffee85e6877
4455 F20101123_AAAGUH hamey_r_Page_58thm.jpg
667fc29ffffad42e2155986433862943
07bd1337c3687b36a571eb17543df45408a2b5a6
80286 F20101123_AAAGKM hamey_r_Page_46.jp2
cf7409cd6d69a6458fe04e2a31459a33
b0c77022c2f9adea11c8b317830dbfcb9b5ee9f9
2195 F20101123_AAAGFP hamey_r_Page_01thm.jpg
5ee68d4c8c314bc693c8892ce6a34238
6abc3ba22f8dd5bb1ad0adc9aaadafd61aaf4bf6
2102 F20101123_AAAGPK hamey_r_Page_19.txt
d7c6473f4f7f24aa27a61f32874bef60
3cb189722f15fa9c63d8000c28445c6a402640bb
14531 F20101123_AAAGUI hamey_r_Page_58.QC.jpg
eb9f06ce31158008d8fee92628bf8510
ec4e56848d273199e0e15e78536d73a1edbede0d
107156 F20101123_AAAGKN hamey_r_Page_48.jp2
f868f2231112b135307191a5bfca0f5a
d23485b4e52ece70d6bcd35438c490ff7e3f3b62
37777 F20101123_AAAGFQ hamey_r_Page_46.pro
24f539713ed14c4b63e6c4457a8e0a06
da5d6cbc208c9acf9e1beb80dac815687af18996
1870 F20101123_AAAGPL hamey_r_Page_21.txt
fe4d46f3dc61f3b4b65f037205e838aa
b9a939dcd2e285c4bea69135c0d7599db8ed40ac
29020 F20101123_AAAGUJ hamey_r_Page_59.QC.jpg
cf562f5a6b6a0ef57d412a74127f4041
1ddaf74f47005fe044d47430a9336af5398e4f1a
949260 F20101123_AAAGKO hamey_r_Page_49.jp2
8fc91785fb4e0cdde6d3710df0ee26f6
62a1ef1d0ec8fcd91664a5ace7a6ffbd17072539
95643 F20101123_AAAGFR hamey_r_Page_40.jpg
9d44b48ae5cdf8dce7575da865618814
8c231c00dbbbeacdf462f7aeef0a311d053878bd
2014 F20101123_AAAGPM hamey_r_Page_22.txt
a9951e23dd1a0195c79fdf9897c7d6bb
fcb3f7ae311b0b7c7cc5f820e90a530e6c8455f7
8503 F20101123_AAAGUK hamey_r_Page_60thm.jpg
6e72afc36d209584a1a558d70877ff80
694b96f6b4eeb8efa45f10efb13a5dc90be09a35
88281 F20101123_AAAGKP hamey_r_Page_51.jp2
3a7535d24c1cd56c62fc4e3111664e33
3e136788ecfe179ff307a6e55d90761f73340ea6
1787 F20101123_AAAGFS hamey_r_Page_46.txt
e27175d46b62b9487c169aa3897222db
de634ef7c23362a12d8213d472ea96f7517eca5f
1954 F20101123_AAAGPN hamey_r_Page_23.txt
d7d05357f21896dbc1e824a087e6f046
e276411ef43c2432be61fe3718b9ba1f3c7d5212
33622 F20101123_AAAGUL hamey_r_Page_60.QC.jpg
7ac84b89c7ee9fe35737305b80be3ea7
fd815b2d892dbd9bf7714f37e89100caf5d47d27
65905 F20101123_AAAGKQ hamey_r_Page_52.jp2
93d6defa8457a488c313dffb7ba647b7
74ddd3a61d19543e2374c3696dd11037266dcbdb
1469 F20101123_AAAGFT hamey_r_Page_09thm.jpg
9dbf21fddb08b796205ca5a9140e88a1
c950010af7e342a52d1d1921a529374c95758966
8269 F20101123_AAAGUM hamey_r_Page_61thm.jpg
2550fd2b92e96395c25fa2ef614a7c78
7c35205bc203fc96a5032c12f9a32a623d2be637
92614 F20101123_AAAGKR hamey_r_Page_53.jp2
7847f6140a88b285198095952b77e1f1
2389ad27adb459da71f8ec3a7e61e66bf7b49837
1781 F20101123_AAAGFU hamey_r_Page_08.txt
e60d5885d9dad399522d32664e277032
e2ab0fc6c13e5cfa1fa3a54d962cae34b2582333
1677 F20101123_AAAGPO hamey_r_Page_27.txt
15f6efbf0455251242f73c2cf1bcc6ac
9198ab324e7600aacc48f4360d03055f6032e5f1
32159 F20101123_AAAGUN hamey_r_Page_61.QC.jpg
f87f5be91d7ee06cf442c34eee26d6b5
c8844ade3193d511ab28d5c861d070ceb580fa4e
93949 F20101123_AAAGKS hamey_r_Page_56.jp2
1adf65eefe1829f017707c05a4461936
d4ca650f6f190b22ad1a06076507753dd300dcb7
2132 F20101123_AAAGFV hamey_r_Page_24.txt
13611bda6fe857e9acbbcbf201b737a4
298182e3001c4fee030dfb30f0fdfbdaf2f57dd1
1404 F20101123_AAAGPP hamey_r_Page_28.txt
7e4ca6e037bdcf391d29c90e60b4c564
94f106138c3082152f666dce0d728fcfc68dbf90
8060 F20101123_AAAGUO hamey_r_Page_64thm.jpg
03858e0bc64fc98d7d68386f645de019
542fa2fdae145a8d3a4a4b4da36606e371a2ce8b
57786 F20101123_AAAGKT hamey_r_Page_57.jp2
66f1649cf676adbabc223cdf370c102b
591d57b25533dee85e757abe704758249e2670b5
109 F20101123_AAAGFW hamey_r_Page_02.txt
0d9ecee878ed4161ea889718efc4bb47
820af6e4ef902e216a93ed14a553277c9b98be29
542 F20101123_AAAGPQ hamey_r_Page_29.txt
cd041ae36d4b5ecc7a4cab9e49f95475
713c4d451960d21d170a02998e9e4d27de7d5be0
32060 F20101123_AAAGUP hamey_r_Page_64.QC.jpg
19300b300e8534e6a6d9682d353bd884
e635d649cd572dca7f3f2c3b8295952d7bc611ab
43162 F20101123_AAAGKU hamey_r_Page_58.jp2
2f2be1383bcd6425e57b83495435abb5
1ffdbed46505512db6d0f1bb2e3f0aed52eddfe1
95721 F20101123_AAAGFX hamey_r_Page_34.jp2
dd83618f66128c4dbebbbad659d76ed0
b02ddee6e270431f9bc907758fecaf15b9c53e4f
1941 F20101123_AAAGPR hamey_r_Page_30.txt
195361c209e3c739fa49fefb02027850
847c85926fdad5cd157464afc45da0c96dcfa063
1654 F20101123_AAAGUQ hamey_r_Page_65thm.jpg
ea542b9b041e5d793a8bf6493e2e1615
5b47f2615f61d9a4d1fab9d23c53c5ccb8a0a06d
107213 F20101123_AAAGKV hamey_r_Page_59.jp2
471d6a8d3b4b1f918e57da5d34d0c3f9
22491108d10b78afcc49f2b018a433cb684a56dd
113438 F20101123_AAAGDA hamey_r_Page_24.jp2
94792922f7bcba63faf2216e13524879
3a49a0b9624d3cf99f49acc3766cafc124a24daa
1053954 F20101123_AAAGFY hamey_r_Page_54.tif
fed0cd4dcf2b493f89dfccb93a055eb6
6d244b68c054262eafe52fa4ab2eb5dc0c7edb38
2064 F20101123_AAAGPS hamey_r_Page_31.txt
b4df78e161b1dd18a8ce57af4a198890
a266a11c8712c9f4243d6015de2444a7a7ee293b
5988 F20101123_AAAGUR hamey_r_Page_65.QC.jpg
b97376cba517f1177b2e33fc4ff9e399
1d5bbdab903088a55adf541f2e59f8ed59bc051c
118552 F20101123_AAAGKW hamey_r_Page_60.jp2
885e476b8def1f217a4242aa36fd7aca
98e4a67e36dcce7ed059dfb2cb8485e8f1816856
1819 F20101123_AAAGDB hamey_r_Page_26.txt
6fd9f460007428bfa7efd4de55ae3309
15328ba0929ee850ea6f0b128513bacd86b06761
29484 F20101123_AAAGFZ hamey_r_Page_03.pro
f4dcb8577f2ebd6061d3e540094ebc23
7e5b2cf7cda5aaad992f9af0edf5fbc476f55045
2000 F20101123_AAAGPT hamey_r_Page_32.txt
337d16a9d1eebe77496e503e46493131
069481767b5ef747ab511665a84743292d3ed2fe
2970 F20101123_AAAGUS hamey_r_Page_66thm.jpg
4fbe049d5bbbeb029d499be20a1c17f8
2dc25e872e6f5a43044d02e41020659a11729cb1
118864 F20101123_AAAGKX hamey_r_Page_61.jp2
257e20e8d16bec35dc3daca02c889bb3
d3620501b4db8af0660c28ebab943a1933090230
320 F20101123_AAAGPU hamey_r_Page_33.txt
733bba6cd2784aac95db6c6541a57e21
bd94b69bb54b8f98aecd8b638b079d4419f60849
114783 F20101123_AAAGKY hamey_r_Page_64.jp2
51ae940d308219ceba3298ceed60314f
37dc24e3d134f86c556b6627cd4eef456af0fc55
33858 F20101123_AAAGDC hamey_r_Page_32.QC.jpg
87a0f00de6e5cf5dd86bdeb790a9f50a
938bbc333fba012f42e9dcc5ba334910dbb0d3fa
1850 F20101123_AAAGPV hamey_r_Page_34.txt
68a6d7ea41ab4a521be81403bfc544e2
270499bb8ee0346d6a9fd57b58c3578f4eb40986
66577 F20101123_AAAGIA hamey_r_Page_27.jpg
754248d19cca6cb1de88d2db4cbad2a3
0bc729e2c7ccb40f4945c99d91bef5830639fe45
11470 F20101123_AAAGUT hamey_r_Page_66.QC.jpg
99fc1012ca1f6b2fe8daa484c7236723
8aec7c409ef0a5db792b8672f4a373e560a7b174
2339 F20101123_AAAGDD hamey_r_Page_63.txt
876d81834c6a701f0699cc9e206be97a
3dfd00b58c1805298be09d4fd8a94e2bd5304b51
962 F20101123_AAAGPW hamey_r_Page_35.txt
b0f6e22c824173c1ecaf3bfa0efa6ac3
fc9156894a8f7555e006345f40ebda7519340e7b
98430 F20101123_AAAGIB hamey_r_Page_28.jpg
d3cc3fb3ca2f28e9d59e8f7ed663371a
76c6fa597342c7a7b3e31302126ed65f840b4baf
21457 F20101123_AAAGKZ hamey_r_Page_65.jp2
b893b6bab62fc75abeb540530eb3b3ae
4055c8384745117c95182995ea2580bacb447aba
77664 F20101123_AAAGUU UFE0011903_00001.mets FULL
1cc5db5177e543a32c0926706210a691
3b04d23f8ca39e1ab206b0d038b48d47cf5a3d12
89785 F20101123_AAAGDE hamey_r_Page_37.jpg
e45987500ba336c14f44acf39fc98784
ad4bcf4c79d1ac1b6955aa0c2c800d2a810a830b
1327 F20101123_AAAGPX hamey_r_Page_36.txt
4de9754f429cd296a29a0076bfd8aceb
3f7dcb2978ec6f36927093528c9a87390be087a4
98041 F20101123_AAAGIC hamey_r_Page_30.jpg
cd8b97d6bd3f68b56c709a0f4a79be04
627f5fae50ddb733b7fc84c8cb33de8db0ef87f0
F20101123_AAAGDF hamey_r_Page_64.tif
96736a885a8878072930da2f851965bf
6daa987ecc3a082b61d211e7d34a8cfac6aa2461
F20101123_AAAGNA hamey_r_Page_65.tif
5f17ed3cae3f4e9b279018be5b156f9b
c00bbd1210c7e3dd9add8fe97f190033f329015a
3211 F20101123_AAAGPY hamey_r_Page_38.txt
0093a3f69203becea3fca687295024c8
b9a71411462df32e7f1c3153df34fb25c02f6460
104351 F20101123_AAAGID hamey_r_Page_31.jpg
6c94aa5633d12d71edb275d0b42c61ab
5ee8af4518cb477ee649e330fc71a79439284863
F20101123_AAAGDG hamey_r_Page_16.tif
55b28d28b56ccdf2f74596acd2ac72d6
2d11980fe24eac1e2f3f8ef44535c6e05382703c
F20101123_AAAGNB hamey_r_Page_66.tif
91db69a0b01a7d4961a6c925673b2a1c
4f23d814c0f29548bde565bd3a4d3de058462b55
1628 F20101123_AAAGPZ hamey_r_Page_39.txt
98fad75c5420bd680bcb6daa8e0e8067
a2c3400ff81bba711c5dd8b10fdbf394882493bc
103617 F20101123_AAAGIE hamey_r_Page_32.jpg
7c874eaf0f94fa079b4a11821e11092f
257064d7dca898721eb736c26c02ce54cb0f0f01
6962 F20101123_AAAGDH hamey_r_Page_45thm.jpg
a9f6d5cd26e3722442dce9b06c8c8600
1cf926ddf0672314c8d7b62c5b6f60be720d1c15
7802 F20101123_AAAGNC hamey_r_Page_01.pro
702886a51b42efb405a62d9be4121b4f
2204d3e2b3f941a3c7df1ce94438d582aef1b828
17331 F20101123_AAAGIF hamey_r_Page_33.jpg
0cb6452294682d93394ca7cc3a404fa8
139174434e530ecf7c45ef6a33c1b1075b86796d
F20101123_AAAGDI hamey_r_Page_11.tif
f9c994d06b0dc7970a13862cc22c1aba
d701e004fac595e4f056119920458d81960ff99d
83545 F20101123_AAAGND hamey_r_Page_04.pro
7cc7ab40279517789e0846873625a0f7
9c5bc72b852c21a5b19757ca44ad7a4ff641547e
92039 F20101123_AAAGIG hamey_r_Page_34.jpg
5fb3ebe4778a416d95ba57a8d7e9a39d
088ea7b149cb697eab1d1919df9df2c4f45d95e4
8093 F20101123_AAAGSA hamey_r_Page_20thm.jpg
95541a2622aa40eaa18e303c5aa4c75a
a35a088d3643f50751129a32d3db4c55cf78f15a
8211 F20101123_AAAGDJ hamey_r_Page_38thm.jpg
754e92422fc7f4a130b1cace56a722b7
5937d308930ec2e6f524bff6244611a5c6bbc449
60499 F20101123_AAAGNE hamey_r_Page_05.pro
0ee5d4ecd8744f226eef2922a83db1c6
2b913e083d9b6eadc9ccdcdadc2280b3a2b2c8c3
31667 F20101123_AAAGSB hamey_r_Page_20.QC.jpg
515dd8358a0c95f671a8245c55b7493c
758d90599184fb0576f9cae48fc24d22c0170c44
7779 F20101123_AAAGDK hamey_r_Page_01.QC.jpg
61006ee7bef71db9e077eaad189bc4ef
3e4b70bfb55cdc301cc1ab7599145f07e95f109f
14725 F20101123_AAAGNF hamey_r_Page_06.pro
f6e9cd3097e7745c8e284ede3f09a8f4
de40920fdce14f631858000f4e7b26a38adb5465
81905 F20101123_AAAGIH hamey_r_Page_39.jpg
80913f6aea6378fd1eb48f969ba0f952
fd581fcc333e4a942c433a42dd6ebcd6abf8222e
8179 F20101123_AAAGSC hamey_r_Page_22thm.jpg
eb9d77cbe72f8c1eb0b7ce46dd9ad361
4c7a3042eb1b90749afbe0692338daaff55e3f8d
6861 F20101123_AAAGDL hamey_r_Page_39thm.jpg
dd3b337ad1df7270b44d3047af82b040
b29c257889993caa7bd53169628fcd7e2c3ae9d6
5645 F20101123_AAAGNG hamey_r_Page_09.pro
315b55e3f143fe85631ac613cdc4835a
d821f5a046d1ee56722e7512ada81b0412222c34
97921 F20101123_AAAGII hamey_r_Page_41.jpg
a8e2e5d90b72154d0229d569816975b4
323eb57d5203da5dfc809c6e4d8fbfd6f88f26b4
33444 F20101123_AAAGSD hamey_r_Page_22.QC.jpg
a6ba9251b6bf6f88f96eedd9f92d6ea6
09e04070f54ac79704af95062812960efc9a4520
43257 F20101123_AAAGNH hamey_r_Page_10.pro
652e6dd9b2c033f11b8d3475209c61ce
0b4e1f104cbfd913b1a404a44943e84ec6200546
69681 F20101123_AAAGIJ hamey_r_Page_42.jpg
5444887395fdbc31e4f847be0d791ddd
7c855b7a57c9eeecc53491826b54804cf6334ca9
8090 F20101123_AAAGSE hamey_r_Page_23thm.jpg
53500ea786c23622320da5cef9648c47
d4d5749dac981e491efbcfce828bac247ccb25a8
1991 F20101123_AAAGDM hamey_r_Page_14.txt
3c5bc4d1a3fbe97f373d2827ed68bcfd
ae45d0a815eca5a6b0f569f88b5cb096a0f7b332
48782 F20101123_AAAGNI hamey_r_Page_11.pro
b8514dc91eea72e6e7b9e45cbab39800
7898a138834800818a5f6ed147f26befb6229f9f
77862 F20101123_AAAGIK hamey_r_Page_43.jpg
d98c6918dfb6ecd4764875e724a2e91c
903d41507af629c461a8eb48cda7200d85a3c228
31643 F20101123_AAAGSF hamey_r_Page_23.QC.jpg
9da3f947910afa9a09dee4f08189dfd1
edfb26e5a755bf0c6a914c038abe8bbb6e940684
35557 F20101123_AAAGDN hamey_r_Page_24.QC.jpg
257716543b02d34075c928102b5f13bb
0c1b354f5bc0651dea61ff30da9096708248b220
51935 F20101123_AAAGNJ hamey_r_Page_13.pro
8afd5634825142ba816e2e9fd3ac3b3f
bb0920ac315451d35dbd2685102505f5b2009c76
82011 F20101123_AAAGIL hamey_r_Page_45.jpg
77fb8206058f150ddd30051f53443937
f0681b70acd1358badbf49f445655722613e7861
8478 F20101123_AAAGSG hamey_r_Page_24thm.jpg
447daf14bb588ecc8984906700ac48dc
9b5324183cd29d338d73209c8a9886a3c20f6810
461 F20101123_AAAGDO hamey_r_Page_01.txt
9631090e00033c8c6a2145919d8b90be
a5d764385085188c0b8f1a6c00d5e96c882eecb9
43304 F20101123_AAAGNK hamey_r_Page_14.pro
e8c250bd6561b75bffc9bed10d6811e9
dd407126b89ccc1d56c6221b6205c0dc5ca1f99f
74151 F20101123_AAAGIM hamey_r_Page_46.jpg
e79e1e96cd72626213921ffe79410006
28dc9146e1f5cd5e1b8588fe1614c71cfb107055
2883 F20101123_AAAGSH hamey_r_Page_25thm.jpg
4d7cb15950fb853fca7224661d86d835
c05021e294ed12a6b883c58e084294a781312286
14482 F20101123_AAAGDP hamey_r_Page_44.jpg
fcf32f36f389d41a08c9c1239ab21be4
30e0ff5cde970ce4efef178c899fa0b01c377477
11764 F20101123_AAAGSI hamey_r_Page_25.QC.jpg
c54079ee6f49d6194a47a925f3d9275c
7b998cca21ef7f3e4990d3ed1d3075f74328dc40
53985 F20101123_AAAGNL hamey_r_Page_15.pro
823f03620ee9af4168cab76cae18030e
fd42bd286836a56f7f2173d2d8af0abb88ee2683
96313 F20101123_AAAGIN hamey_r_Page_47.jpg
adf49344c476efbd6db1da53c8e3c5c0
9fac43b9fb16f7435be5ccd67a6a6b55cd8cadfb
120271 F20101123_AAAGDQ hamey_r_Page_62.jp2
e65c7b4f95780b9331fd807aa23da61a
c513b4bfedf1feb708aec7f970c588d4d5d527b2
7376 F20101123_AAAGSJ hamey_r_Page_26thm.jpg
8e235b6af718e598c6e101fb8d1e9606
9fdcc15b5e6da6df0a41bbd44b45765318f0b84f
83301 F20101123_AAAGIO hamey_r_Page_49.jpg
190ad822137205842d804eff715f4e82
79af9c0fcb6ba4ae2ffec640e39d1bd8f8fc37fc
25271604 F20101123_AAAGDR hamey_r_Page_05.tif
773ba1bdfc3302be9ebf70c0b6e8ca2a
b70575afc797a77b87379234d3835461bac6198f
28560 F20101123_AAAGSK hamey_r_Page_26.QC.jpg
0aaf5b961313d77bd408068c7f1b3553
abf066c3bc3b175b4cf40df1f5150b0977d210ca
44530 F20101123_AAAGNM hamey_r_Page_16.pro
69d3790b083f94711bf6380c9fabd290
7d4be3f2d6315628d47eeede27193d7e9204b8a0
92450 F20101123_AAAGIP hamey_r_Page_50.jpg
ed97a861172408f300a17f2c9adab309
ae6c9db4b4a93fbb2a362f9fbf52c279232abbda
112940 F20101123_AAAGDS hamey_r_Page_60.jpg
4a7f9af83562e0bcfa1a37e148e152d0
69a513a1371bc595c3e33b8ef96d588b2ce332f5
5854 F20101123_AAAGSL hamey_r_Page_27thm.jpg
9b960c1b1dc940d6aa3fd5d57c71aa7c
f84930e48d729168e147b672c1ccd22968c87eb1
41059 F20101123_AAAGNN hamey_r_Page_17.pro
236c79f6330dd9a9e32e25dee5411aae
f8edd488422682f76170d1540fc00b102a633b4e
85907 F20101123_AAAGIQ hamey_r_Page_51.jpg
6e6ddc2a237a6687aea94f4313aa2557
a04b1c4d20e99289600ac40a426a8ad19c73f647
39630 F20101123_AAAGDT hamey_r_Page_45.pro
588f82a562e63c6cfae8912bf798454d
26e6476b6ac1dda960eb422e62af9b7b21ca7432
21505 F20101123_AAAGSM hamey_r_Page_27.QC.jpg
60d5b4777b3b67528a2d550edfa45b5c
dfebcd7670d8ad7df483040f9a0662d62e4eb09d
46282 F20101123_AAAGNO hamey_r_Page_18.pro
37ae58d694897b2f3ae4e125377b9cdf
e4b4eeb5252989de5ff984eecc85f55c06121509
62104 F20101123_AAAGIR hamey_r_Page_52.jpg
0d8ba59d72281891b3096cec66178482
972e5c2d52ded143a0d7a5f21a2b0ced1576a764
96411 F20101123_AAAGDU hamey_r_Page_37.jp2
b1e2ace39628a3f407a2e58a67da7ee9
646de4b90d98c12730fdcf4479aff8a21e461ead
32464 F20101123_AAAGSN hamey_r_Page_28.QC.jpg
dc3d9425edee1b62ffee7720a32f5520
60c83b08a40fbcf976931cf301e1fd3f741cc3b3
48152 F20101123_AAAGNP hamey_r_Page_20.pro
38406119cb8d72c97fcf3329d69b72a6
ee425ad1302b36731889b2fce6c4697cb8d73a83
106644 F20101123_AAAGIS hamey_r_Page_54.jpg
07cf5995dc1ef9e96b3127335bbcb706
c29eebd39e01b738df993a5e893bb1cf5ef549a4
43714 F20101123_AAAGDV hamey_r_Page_12.pro
4f65f64ab9bd33eb54e689982d8e2390
36ba3265788e1a1621e7fee7fef29a2fde776280
5733 F20101123_AAAGSO hamey_r_Page_29thm.jpg
383827746eb2e1216f0076b6c50082aa
3ba12d24451f3fb8027987dcda87298a13d13629
51243 F20101123_AAAGNQ hamey_r_Page_22.pro
b54aa59fc8aed24a4b2f81cfc07ae063
0e5f313391ae4f33055b5f9a327ba8be485c75d3
100662 F20101123_AAAGIT hamey_r_Page_55.jpg
4d0133d3cdc45f92b95ff8e5afad3fa3
244e8e372df5425be5f25e678a6365798e8b9979
86291 F20101123_AAAGDW hamey_r_Page_53.jpg
72319bf8bb7315b9028ac1359d6a737d
c8a48444d89b192d811c27deaa5d3fee92e213e8
20379 F20101123_AAAGSP hamey_r_Page_29.QC.jpg
455419f65140b7149aae4a42afa4c731
a99037661781c48b00a261d622c7e832f90d8bb6
48634 F20101123_AAAGNR hamey_r_Page_23.pro
049781494e67865b307488a20b4fee54
54f2cf6bf681981637345984a5a51fac96bcd078
87312 F20101123_AAAGIU hamey_r_Page_56.jpg
1a62224ae99b00666657923194fce071
202eaa40a2973e2b0fff58a96973e519d2f14cff
7183 F20101123_AAAGDX hamey_r_Page_53thm.jpg
279d4b20d555b1c3a8fdccd35cfbba5f
457f5ba627ebc4c46e59bb8aa36a03ceb839e08f
F20101123_AAAGSQ hamey_r_Page_30thm.jpg
a31a8469b19eeb78420aa07cc3f15d8f
cd97362db7fdad244d8976b90b960a97b2856f2c
53662 F20101123_AAAGNS hamey_r_Page_24.pro
d876b04db06aeeaade12c1b0698df844
329a975b3279d91741da459e365d738dacc4ffc3
61128 F20101123_AAAGIV hamey_r_Page_57.jpg
9feb292b6bde467f52bc49761300fff0
a4e2c7cb6c74e8728cfddca522d163b448c29d91
30568 F20101123_AAAGDY hamey_r_Page_21.QC.jpg
1cce6b0dacf9b8f22e5f2666c93b7015
ef10f6e2fac257d7b11ff0ce275d0d63ed3ab700
15457 F20101123_AAAGNT hamey_r_Page_25.pro
54e6983c1811eb839a9b4eba66f014e2
76d29235559831d427d4e53b06f03b3b0dcdf695
45080 F20101123_AAAGIW hamey_r_Page_58.jpg
f6c5c49e84f03d0036adbb9a14e9f133
2b5b71d3775cd58deae59d44763daa7f9cbc1917
102102 F20101123_AAAGDZ hamey_r_Page_47.jp2
ef744a162160e0dc82af46b66b73da1b
900055d4c3e41497fabc60dddfb42effe19d6bdc
30981 F20101123_AAAGSR hamey_r_Page_30.QC.jpg
049a29b8f4f1c7ed155cba087808b246
ca87d8a427907aec1c00bbab929aa1de1ce1a81e
43618 F20101123_AAAGNU hamey_r_Page_26.pro
b4a063ea8428bdf8746e39b18700996a
3830adad8b9bc33b67cc9b99c4553ff25431aaa7
96806 F20101123_AAAGIX hamey_r_Page_59.jpg
58e62a132555bca7854f4b18d92bd118
a852125af12a47014921a6a233a09b51b84dfdf2
8358 F20101123_AAAGSS hamey_r_Page_31thm.jpg
d1107bbab7f0caf8394570fb754ca435
0aad9654c186aecda38e8e0c69790f9041783dcb
F20101123_AAAGGA hamey_r_Page_41.tif
3b7329fddb3a20feec6f2b824475399e
2b9ef6e0485aec61ff032efa40bcafcd15777191
113123 F20101123_AAAGIY hamey_r_Page_61.jpg
d2cf7830df2ccbecd7f22d31dc2b697b
625bcb94f93b5440e2faf09fe22eda76c8255ca7
34391 F20101123_AAAGNV hamey_r_Page_27.pro
e9d4cfaf5f71b78f494738699ffd50b8
f16ca2fde26f9ea1191a74fdcc80c9932aba0d47
8359 F20101123_AAAGST hamey_r_Page_32thm.jpg
249733840b6e5795d1d6a75a8634fed7
f444966099902838ecea0d9a0fd04a581cba089b
94903 F20101123_AAAGGB hamey_r_Page_36.jpg
4d327b59912ff4d38a664b538a2ee642
d2029f86c4418eeb6171b959a2f644835f5c7a8d
113126 F20101123_AAAGIZ hamey_r_Page_62.jpg
fd05219911a41009d1f38a2a63696734
54d6c61b991dcae9190d93edfd7dd72168b5793f
34755 F20101123_AAAGNW hamey_r_Page_28.pro
4bdf64ba77826bc021843ddbd4f4564d
dc482b31d086a2455bb1279a6d14889c71afcfdb
6155 F20101123_AAAGSU hamey_r_Page_33.QC.jpg
0427f8dd5bbec3ad6c6e59b6518348ba
97204df33a0acd0d13cc0b3abd25f09690c19ba6
8726 F20101123_AAAGGC hamey_r_Page_28thm.jpg
e774b07b5e3ef61128580305b03b0238
78b9bcf1911a8c96a3190a7f7778bbe05c8db9ef
11579 F20101123_AAAGNX hamey_r_Page_29.pro
40582ff6ab2fa648bbc2eb21b0519731
41c0c91ae787641695d6941a4ecb20cb32cfdafd
7235 F20101123_AAAGSV hamey_r_Page_34thm.jpg
0aa8ab0edb22c09caba748cb275c26c7
9408895b6efa064f4e71d7a3943bd6a15e73c0b7
22356 F20101123_AAAGGD hamey_r_Page_04.QC.jpg
cac02c6cf07a4a6f0bd9b35f1ecbb16d
e17a6590580fb9381147261aac676e7cb36d8d12
36485 F20101123_AAAGLA hamey_r_Page_66.jp2
f8b0f3d93d3e4bf386cf8afdb7c98ebe
43e567e026cc330fb1f34497e3639314a23afd78
52017 F20101123_AAAGNY hamey_r_Page_31.pro
f402a4cff2dfc98861cd6bce8f8976a3
dd4acede4da73332d8c6433c9ab644822ccb654d
30054 F20101123_AAAGSW hamey_r_Page_34.QC.jpg
d7d60c3a0f99ce771822269ada9370ce
da2c94d9d0b3c241921abc651f52c751fe87fea9
40076 F20101123_AAAGGE hamey_r_Page_39.pro
de71c343e5adde022463c651120917e8
87db4c5b614be79cf4ff343e3f94b64365ec1e10
F20101123_AAAGLB hamey_r_Page_01.tif
ee3284bb99f7c973b84d0ff9394620ec
c6e8d219f2124c370c57fb220007bd49001dc5eb
50955 F20101123_AAAGNZ hamey_r_Page_32.pro
86f60a6a2c043cbb8a6fae3d1bd59f90
245bef9780c126eeea94afaf6f3d18a6a86815fa
7202 F20101123_AAAGSX hamey_r_Page_35thm.jpg
4d19a0278345016d5dba3eda1aa9aceb
cc1641aa29acb45e5effc74a26445a222ac698c4
F20101123_AAAGLC hamey_r_Page_02.tif
71afb6dc06e24f49037ec14e707f9c83
152fe982f7bbbfa8afc742c2d8c73441407ba7b3
25583 F20101123_AAAGSY hamey_r_Page_35.QC.jpg
e6d9a8a4dd17e8d7f8f5d55179cf40b0
7e3914c61a6aa4613ec7b4e92709c3a62e865b58
26280 F20101123_AAAGGF hamey_r_Page_17.QC.jpg
75b048e418272d7b641c8187243431c6
7064a8bdb7eb3af99ff89e26afd82abe84c82a49
3328 F20101123_AAAGQA hamey_r_Page_40.txt
3111cfdf17d5887b5b4a1865e632a12b
26b1a809f1467563981c0bd1f9d87abdd57ef8b9
F20101123_AAAGLD hamey_r_Page_03.tif
cb09cd13aa0570df5b51216970deb5ec
e6776e3d230001694703117a37b36b394d0fed71
7765 F20101123_AAAGSZ hamey_r_Page_36thm.jpg
d368944c1bf3a5dd4c0b0fe203fdfc71
58da901f0ffe143395e40c2c9e1b7fe98cdf0504
67347 F20101123_AAAGGG hamey_r_Page_27.jp2
da0aa3dc186ec7a2678fde0d653bd893
3b9d75c4f4ce7d7819bd74519d6177566dbac886
1989 F20101123_AAAGQB hamey_r_Page_41.txt
63a4f23286e5097bf14bd18bea2bd03b
2814ce58b7237fc842f85a82dbe9d8444572b3d3
F20101123_AAAGLE hamey_r_Page_04.tif
8d33568d77327194c610846e4454dfba
f61ee2fad29640a2d5994fe1b673444a4cc34f3c
F20101123_AAAGGH hamey_r_Page_48thm.jpg
01cbc511257900f8fc848c57c1106a40
5aa336d04b53d61c0ed1a80c60668002a726d03c
1426 F20101123_AAAGQC hamey_r_Page_42.txt
2cc67933edf4fd70e3eec91480f3d251
95942ce5f838eb334129c21eb02f263df44b75de
F20101123_AAAGLF hamey_r_Page_06.tif
8e57815ef14ec48889b8f8f24af39d70
e69064565a1183f3f8ceff0a132219755096d301
1563 F20101123_AAAGQD hamey_r_Page_43.txt
03d49ccfd645f5b80625bcb0b3e40bbf
b73c4af99434c1f6e12fb9973a412883df9c29c6
F20101123_AAAGLG hamey_r_Page_07.tif
f228b38c627a9a6799081b8f09394f59
3c8705c42f539e614f4e15b952a350d71ff59397
33799 F20101123_AAAGGI hamey_r_Page_63.QC.jpg
08f82e13d990b93f34486786c90acf7a
0a2d6480621399ebb07c1655a3ab8ff25f619700
273 F20101123_AAAGQE hamey_r_Page_44.txt
1033fc20c4bdc1a2878045b769b25163
b9eb696647d7353b8cd77963540cfe308658dfb9
F20101123_AAAGLH hamey_r_Page_08.tif
eb091b580d904403638cd271a17df066
8bc4cbecb5606dcd2320018d89c99d4e3612d719
8407 F20101123_AAAGGJ hamey_r_Page_63thm.jpg
b6168779407500d658fda8dffb7b7317
acd4ec7f6f0770746775ea3b3cdbdd84799eb521
1794 F20101123_AAAGQF hamey_r_Page_45.txt
d5cff51125213d09271551f54432ba9f
0fd8ec95a49f996cbd089465d1c2245ceffd40e8
F20101123_AAAGLI hamey_r_Page_09.tif
1dfe214ccb823553697c12c1f3a0c300
697a67ed318c14f0ed6c30e4bb0e7da6cfbe533b
1923 F20101123_AAAGGK hamey_r_Page_20.txt
80032a23e815ea3577a413ed5826352d
693a3cff2bb9180adf7fab76a8cf7d159330bd57
2004 F20101123_AAAGQG hamey_r_Page_47.txt
e496a2ace9799b7e611bfa6795a08009
bdf5977dc5a3c71943942f677f68caa060053793
F20101123_AAAGLJ hamey_r_Page_10.tif
1f5f2023230d3d3b06bfdacc51f91537
dbeba541b6f175083e13eb3d297b90cf3f4ea3b7
40067 F20101123_AAAGGL hamey_r_Page_08.pro
593bc4efd6cd1a9209cb1661c71d8d12
450669787ff8788b6919439da13657954fc6a5ab
2013 F20101123_AAAGQH hamey_r_Page_48.txt
4e368df2e112bf36fa780be61f951320
c2b58d6041a067909bf7fb5ba74b51bc2f67430c
655 F20101123_AAAGGM hamey_r_Page_25.txt
10c4a84abe58b8ea96818d473aed616b
625cd9008dd1e104b131228f38e42fc751596b26
2252 F20101123_AAAGQI hamey_r_Page_50.txt
6131fec745d492cff9aad0d9ac6e5dc4
1b8425ab6549dc88ff66e2a573ce9a64ca47d550
F20101123_AAAGLK hamey_r_Page_12.tif
c970dc2bd93d3121157cbe2f08648124
21aec4cee3f0b81df2a26a8dd6468f3694c35bf4
61525 F20101123_AAAGGN hamey_r_Page_40.pro
3d90436f4de2981b428fda3c7395711b
b776cc4a5f8b8a14b18a995e77bbb1f8aa60899d
1868 F20101123_AAAGQJ hamey_r_Page_51.txt
80295f6c0acb045347ef1e4d110d4bb7
adf956bb3014e2255459cb38ea1e7a3ab5d46913
F20101123_AAAGLL hamey_r_Page_13.tif
ad095f0b865ace315b39bbc8ab081785
9d92bdcca6208d57ed7c6623fb0e48117d0a0a49
8409 F20101123_AAAGGO hamey_r_Page_62thm.jpg
f0fe34bf228f3a0298d5aa3399e7c9a5
9c5920168011eef4a8ee4766806cc97f2eaa20f0
1203 F20101123_AAAGQK hamey_r_Page_52.txt
2bf9076f85348d7be4007ce250eafbeb
402541ebcba31c5d7da54c170cd8550f53b1c5c3
F20101123_AAAGLM hamey_r_Page_14.tif
76f21e519a7d7206bcc3a3c806dfbe0c
9cb8ebfa71a62840f8ab5128b540775be5a21711
1851 F20101123_AAAGGP hamey_r_Page_17.txt
fbcbf652bb5b622aba50cf7339f72ea9
be1a46efd2a72c30185d5275a1fc11c5c40f6877
1777 F20101123_AAAGQL hamey_r_Page_53.txt
2e1b01330d645e194fa83117c8aaee45
366fd85e5f3768973b7c4381c2f5580f2f28fcd1
F20101123_AAAGLN hamey_r_Page_15.tif
fe51aadb9f81a8575962d71efb7e3b99
b05858ebb6e82e9c397804c8eab08f4f1c94ad4c
2516 F20101123_AAAGGQ hamey_r_Page_07.txt
960ca7b6cc754d9f07716bafd6278f8f
e03ea3939bd2171cec80dd300bba9b8653c33959
2113 F20101123_AAAGQM hamey_r_Page_54.txt
abccb3ea29ec1422e1c3cda54cb7782f
991317009a1c559d41d49a51ea78c015383a9f62
F20101123_AAAGLO hamey_r_Page_18.tif
1b694ae2baeb11b07bbddabb346c52a2
ac47c7b88e2e11b8894fdba805adb6b189a1f1e5
2076 F20101123_AAAGGR hamey_r_Page_59.txt
8349ae3211d393d2e3f674e65465c0c7
5f37aea8a59370687cd98380b26d9a9469aa1a66
F20101123_AAAGQN hamey_r_Page_55.txt
9301d7c7d8d31a1dcd18e819632a18e9
10be735e49ca3f1ef7a9f3d0673c79fc53190e79
F20101123_AAAGLP hamey_r_Page_19.tif
09e027d2051437054f003d88900e7261
4d38bacb67a324e237ceb42156b2319098f9b80c
F20101123_AAAGGS hamey_r_Page_22.tif
003cd3382fe1be657adedbf226b3fb07
918590dc66d8335376c3c8c481506db8096c51f8
1703 F20101123_AAAGQO hamey_r_Page_56.txt
45cf250e2fe0b6b9b9aa8bda4fc5c534
01911a3a34a7641f18eac12231f0cd8235c8ceb1
F20101123_AAAGLQ hamey_r_Page_20.tif
8bc327967f04027aae454616439f5167
fe9227132694aeb181f4b0ac7680b484111ed70b
93887 F20101123_AAAGGT hamey_r_Page_16.jp2
57bcbb21075b922873e1ffe2be9ffa03
28c64d4f2e10d8bb669808ae0e46128a26028ac6
F20101123_AAAGLR hamey_r_Page_21.tif
e887884b83b8f6ad95a322d245ce01ad
46a5aefd5ff3eaf7f056d3a609e2e43778048bc5
1768 F20101123_AAAGGU hamey_r_Page_12.txt
d6094f6823a190420ddd23c0c9ee4abe
a9ed5a45e9596560f49801cf9d8a19f278f1f312
1051 F20101123_AAAGQP hamey_r_Page_57.txt
36d6dfda2273a29da1546f7db08b2451
c3561a2dfbc396482fd5674a7edef28484e36a39
F20101123_AAAGLS hamey_r_Page_23.tif
00f9586dc2afce3b135d9d8f52c9bb1a
9e62a7d775c42db93ca3a2fafc23cff0ded1a2cc
8350 F20101123_AAAGGV hamey_r_Page_14thm.jpg
5e577db097a141dcb2927bd04ba0942a
0ac98e570a5305169072c5fa3494abbb763fe326
832 F20101123_AAAGQQ hamey_r_Page_58.txt
1602cf016ab32fc571350d4f064b7837
7c090c0637d6498dbf7b827e782b56afc4dfcaac
F20101123_AAAGLT hamey_r_Page_24.tif
ac3edc64a4fbe7a3806c65a79fecb39b
2cdbb3958dcb9fcb6abea3d89425bfc21b96e896
1251 F20101123_AAAGGW hamey_r_Page_02.pro
d117689864cf054f573af0f897f6812f
8f043c47f80544f8ffa66008f7235033f2ee66ab
F20101123_AAAGLU hamey_r_Page_25.tif
1227f29782d50412fa42b8a0ae652323
65fc429a6e012e07829b1383a735f63095fc31e8
31131 F20101123_AAAGGX hamey_r_Page_42.pro
971e9904c99d6d3ce812b42224c8467e
fc1dec25ba91e0e8df525df973f6b8d88b2d8d4e
2262 F20101123_AAAGQR hamey_r_Page_61.txt
bc9843f1aa94aab43e44fa82a886cf2f
9cd0eb6e820830fbce3fa61f2f82ecf98440be59
F20101123_AAAGLV hamey_r_Page_26.tif
d12bdb50acb23bc8b7572730008abb80
76dcfdd4dbba49b170391c2ac6fe845d43932581
1240 F20101123_AAAGEA hamey_r_Page_03.txt
0bcfcb1cb1c90da22b2bcdf558996f06
7df21f0576183fd833b1dada491ad8b3d03f0c16
7599 F20101123_AAAGGY hamey_r_Page_18thm.jpg
0c1ae45e843f0d664e532fcf8e550477
2ad7b63d71cc193420eb3751071dadc090b1bdd4
2382 F20101123_AAAGQS hamey_r_Page_62.txt
d95758fb6b32979731fae1b4fe35d75c
5e0783efa1211ab9051c5b990f5ea5a31a2a1346
F20101123_AAAGLW hamey_r_Page_27.tif
3b1890c838aec15f4ae383417ff310cf
97758ae1ef6d5ee7eb8982d63dac0bce11c17d2d
112088 F20101123_AAAGEB hamey_r_Page_54.jp2
03f4d802d79f4e0e19659314514fbc9d
d3704e885a7ed324673ed052c9169b8af323ff0a
F20101123_AAAGGZ hamey_r_Page_49.txt
76ae68f101ae7dd4476aeae5c74f5e6e
e62d97f576fccd42d30145fed4c4dea96129fcc4
386 F20101123_AAAGQT hamey_r_Page_65.txt
f2d778f5ad3fab3d28967e34a1649031
2508f33bf7b699d330d80f41fefae382decafe5e
F20101123_AAAGLX hamey_r_Page_28.tif
79713d13a9770a1791ac1960ff28c99a
e9a339586ab03a3508ad563ea9ee9280ac6295e0
61654 F20101123_AAAGEC hamey_r_Page_07.pro
a48b42ae2704c11ed7a2fe0559defdc5
f51053ad2864ef8c47b1db4bd4e8158dc1426072
637 F20101123_AAAGQU hamey_r_Page_66.txt
cdfb655f5a48a74bf2e498f33b6c93f3
e54774c532a1262437b012404dba2776ee303473
119529 F20101123_AAAGJA hamey_r_Page_63.jpg
6ebe8c952c72ef2c7835c8df723be4c9
fee6e15b084616bc620daa96af5d237a745a432a
F20101123_AAAGLY hamey_r_Page_29.tif
7a1a4cc1497e75fbcb456afe68da4b91
8d3b476c123505c65d14deb75cd7cfc6f61c0d91
1395264 F20101123_AAAGQV hamey_r.pdf
89094989a457cbac8d5206b0b4196c80
c989584457c7181f2e67938ec41751666e31bc0e
21337 F20101123_AAAGJB hamey_r_Page_65.jpg
706e2f5be7f40b1471f6cf9f180158d8
c8f261ba1331232d944098b9f0ca2409821f0156
F20101123_AAAGLZ hamey_r_Page_30.tif
46dde6ec2b6d40cebe9e0a2c874a5635
5fd735ccf94561d05b07aec8b5707292206cc698
F20101123_AAAGED hamey_r_Page_48.tif
517de3a5f9054f89693d3e486eb1bdcb
44ef220fbaad06fe0d15d028df5fbc4b74a7d92b
618 F20101123_AAAGQW hamey_r_Page_02thm.jpg
8c85f0e3104c76f3da6b141b27754fae
2e809edfdf256b5777e6524e6b15ba4dd49968fd
34498 F20101123_AAAGJC hamey_r_Page_66.jpg
4fb07826e197ad9e7a7ad8ff28190273
de69999e1863ab91fdd95ee47074876862d4e11b
65138 F20101123_AAAGEE hamey_r_Page_29.jpg
27ec25729a4367ef278012a9028ed924
22040d87d7c93183e0e267f5db59dc762d846261
1595 F20101123_AAAGQX hamey_r_Page_02.QC.jpg
13e75fe15a2b7a8b15f68c7dcb04c7a8
d3cb5c6b6199c480f56b258e45572d628cbcd152
23656 F20101123_AAAGJD hamey_r_Page_01.jp2
3bac1abc8b64a020d8a0b1603db84e29
6a72bc6b2bd337ecd90d22f0fbcf873012368ca6
7486 F20101123_AAAGEF hamey_r_Page_50thm.jpg
75d7be5eb767b0cdd40d0bffa5a3025e
aadf003817ae05c44bca0d49fc2112abac367e37
6968 F20101123_AAAGOA hamey_r_Page_33.pro
049798694292947b08ab40e0ea56ead7
380eac1cf3ae16770847d43631f0cadfee214320
5574 F20101123_AAAGQY hamey_r_Page_03thm.jpg
fa08a23fc5ec879331d7f3fe12fd9417
2164b8fe292983228ea06d16efc66159cf5f52f1
6141 F20101123_AAAGJE hamey_r_Page_02.jp2
75e79b45a0ddabf36bcc30ab77948d8f
2cc772517c28823a3e8c37010d238bec8fc4c205
8543 F20101123_AAAGEG hamey_r_Page_54thm.jpg
a655e1dcbba94cc7b8ca94f9b1fbd1bd
dda5a73a1d9994fe0a4ee080c62cdf30a7c6af80
44687 F20101123_AAAGOB hamey_r_Page_34.pro
721c671ce56eb3bca66b3399c21c4384
4facb2119ad055f870eba61be8827787b937ef9a
5141 F20101123_AAAGQZ hamey_r_Page_04thm.jpg
bab6c5a9cea2d7053d154a4b04f02228
b30e40427b23ad1db492b637a0b4c9eef5d7b9bd
32461 F20101123_AAAGEH hamey_r_Page_62.QC.jpg
428cc0b6bf8a480da4f5c18b000d0bbb
6e4de5d34354c2291cf2f800a6cbd1aa943cc44f
22956 F20101123_AAAGOC hamey_r_Page_35.pro
606851ba2838ef5d153fccaa526cd44f
0e5164edae93b390ffbd5bcdbd18b9f851b38a37
65992 F20101123_AAAGJF hamey_r_Page_03.jp2
1696aae98234c7a3afbf3e271d488de6
581a525d31865aa8cc4bf78cf276373797ad81e1
28527 F20101123_AAAGTA hamey_r_Page_36.QC.jpg
ef21f12400ddafae3f9765ff7b12bf41
612d6431e74747ef652e0d6138a50819022fa085
F20101123_AAAGEI hamey_r_Page_55.tif
9887cd2fa052810a11b8e54f97f5b321
98248287eb2e12509fd19554a890877b50750d24
30290 F20101123_AAAGOD hamey_r_Page_36.pro
43ab5f861b5effc7edf0aa004cf3ef01
c2d9268f62342013b631b4b25dc95ba929ef6110
1051975 F20101123_AAAGJG hamey_r_Page_04.jp2
4baa9b00b6cce334a0b53679184361c3
a6dad2670ee93fbe108f74b472230bd23fa28211
7353 F20101123_AAAGTB hamey_r_Page_37thm.jpg
3770bf873d3307c1dd07e230c01a8969
a96613240e4a82a89be276e3d9f472e6918297ba
F20101123_AAAGEJ hamey_r_Page_61.tif
1fe6c5cce8a6f4059fe530d8b514108f
e127b320442446ce0870ce7b4a7325a32575f835
44368 F20101123_AAAGOE hamey_r_Page_37.pro
3921177a58c3bf406a918a6e7c408971
30c502705ed11165e6ec6cac068a0c54a012b2f3
1051973 F20101123_AAAGJH hamey_r_Page_05.jp2
4c1fcc27ca6db44f8f13ca6db3a0caf2
603de5df68e5ff28aeb0fd506e719df7334f5b8d
30227 F20101123_AAAGTC hamey_r_Page_37.QC.jpg
3513f7e2a3072f23b249821ae700daa6
2fc1d8649f20aabf2db1a1fc4d12d4c35c9f355c
7733 F20101123_AAAGEK hamey_r_Page_21thm.jpg
0febf7bf32d8d824a39d9ce23ffeedd7
bec8773abcafeee16364591fdd130ecbcf1e7d3d
61625 F20101123_AAAGOF hamey_r_Page_38.pro
bd7919db0fdc3cb799ecd154129512d0
03a6b79b58308908b94ccd4393fc0910ccfb83e3
31475 F20101123_AAAGTD hamey_r_Page_38.QC.jpg
5f166d1950a5b1eb518763361b1fbaa9
161afb0fde6a1051a3d6893268321a2d30ab39d2
7661 F20101123_AAAGEL hamey_r_Page_59thm.jpg
462beead14d094446236ebb1bfcefe4c
e7934432d5789d026e5a325d748d06887606dfba
49002 F20101123_AAAGOG hamey_r_Page_41.pro
ca581187debf297c039cbef6adf6bbc6
2aae5bbd5548bfb74695daede02657791c45dd8a
1051930 F20101123_AAAGJI hamey_r_Page_07.jp2
6e2cf3f21fb68c83af50cf838614d30c
b112e6fb368c94de21a1c702d1839e8e25cef572
27404 F20101123_AAAGTE hamey_r_Page_39.QC.jpg
4e947b9f8e5df4fc28e897bee5eb5d22
588a0e2c24b9166709bc809401db4c79bc2a35d5
509004 F20101123_AAAGEM hamey_r_Page_06.jp2
eb2bfcde103d7721872b643a21304343
bd5f64027307bb6a00a16973931e6bc5d1490ea7
33284 F20101123_AAAGOH hamey_r_Page_43.pro
dbcda4b7e5860fed98698e583e9b5500
8cb4ed2c641fbdafcbe32fd4482bd6e781743cb9
88778 F20101123_AAAGJJ hamey_r_Page_08.jp2
b0232e45c2ef3c13677622bdc38ce7df
09def05fec14d8e657dd7d41b3061ecad7a3da32
8112 F20101123_AAAGTF hamey_r_Page_40thm.jpg
6ea49aca5d7d72f13fcf9db933ec999b
22472bd65c62fd77cd8820100301f92453d87baf
121486 F20101123_AAAGEN hamey_r_Page_63.jp2
67b3f6a163280ed813c30bf5e7611850
fb86e6626ca5ba7aeeccaaf031cccc14cf1917bb
5757 F20101123_AAAGOI hamey_r_Page_44.pro
ebb2274d839a1f7abe9c20b1058f49ca
9d219f1bbe3e22e750dfc6b3b35cc52e71bcdade
94156 F20101123_AAAGJK hamey_r_Page_10.jp2
7aee0df9a12fec8015224a62006dc11c
c08378533cfe20e489389e832327263d6af8998e
30652 F20101123_AAAGTG hamey_r_Page_40.QC.jpg
f18a241dfec6f49135ca6e95bf93e252
65a1032c516ab9f40bf7c53d6f02dbcf5355a083
91615 F20101123_AAAGEO hamey_r_Page_18.jpg
71a9f7eb716a160fde98244cc34cd204
69df61657d4a769809006829ad0f2bcc2e655b42
48373 F20101123_AAAGOJ hamey_r_Page_47.pro
18fba0b0adb6c930bfcbb4c7a0441c75
982b999959bb09a06e6230bda38257e29b52265a
104502 F20101123_AAAGJL hamey_r_Page_11.jp2
1b33be1000835336e9bb7503212e000b
abecb1ea88683d0e5086015327b4acf4dff1f151
7862 F20101123_AAAGTH hamey_r_Page_41thm.jpg
d997cdcc7ac95f33a4e71248392dc61e
40cba256d35a66a6982acd0f436d6db5ede8bdcc
2258 F20101123_AAAGEP hamey_r_Page_60.txt
c983c2694aa237c6f1ddf93809e1ab4d
abc327207884af748ba03d4047570507b2c308c2
50837 F20101123_AAAGOK hamey_r_Page_48.pro
6c40b3e315930403e0ecbe968b827366
1d1176daefa69c2927bc8b26118ae0450b242a40
95884 F20101123_AAAGJM hamey_r_Page_12.jp2
d5cb0493997a1efd0831f7e012997f78
eb5195115b7810597cfe6b29df31095c941e3c82
33004 F20101123_AAAGTI hamey_r_Page_41.QC.jpg
3e61b03e42fa07ceaf6c394ec5d2479d
fb896321d47c8ce9511e4a0f49615a7900fae004
5343 F20101123_AAAGEQ hamey_r_Page_44.QC.jpg
a1313c6c810e3e310b0a20ef88ee9111
4445921fc63d30ee45fa35a962aa288150af77e2
50428 F20101123_AAAGOL hamey_r_Page_50.pro
90d3b894deb607e213de73d51aef86bd
1c79bb3086aecc729588fa2d2fe1b3cf15e8e7f3
109603 F20101123_AAAGJN hamey_r_Page_13.jp2
4c8efc69429eff28c90eb873d1eb1d77
09d529d0c2303e62c9b5a7191a456435dd8a4a6d
F20101123_AAAGTJ hamey_r_Page_42thm.jpg
167107b55754753e04bdc8f4a89600e2
c13fba32fccab5ec7b4a782c3b902131d09260f3
53735 F20101123_AAAGER hamey_r_Page_19.pro
9c86d317841729bea93a57a06e085212
c8fa1a366666a68f78ca8ec9d4c0075cfbf15be6
F20101123_AAAGOM hamey_r_Page_51.pro
4062bd7b413ca51d1c2bc956ce3ad397
8e5212df5eb85316965f7b43d6e8917072337a36
964553 F20101123_AAAGJO hamey_r_Page_14.jp2
13731b2e85e7f8b45e15261af78d89c4
3677b3e62a727f896ec803264a9adca93fc62692
23976 F20101123_AAAGTK hamey_r_Page_42.QC.jpg
bb080d717df4d33d1694b21264a7606b
a45f59cb7274c87d2a3b94c247e63b7cceda67a5
33595 F20101123_AAAGES hamey_r_Page_49.pro
4a4a1e430aa10960502277b309c9e2e9
30b8d27175610c9bf8a1e806ff954091bee8fbd4
112567 F20101123_AAAGJP hamey_r_Page_15.jp2
c1a33ec3ba74fe584c15bf3638431128
8da2ee533c8436e3e2210b0342badcc3cb2e40c1
7236 F20101123_AAAGTL hamey_r_Page_43thm.jpg
18509d318bf855c621b8a156fe77828b
5a93b9103b35d559917715981938311b5632a8ab
102899 F20101123_AAAGET hamey_r_Page_23.jp2
3f04badffe68be971c6558fc9399c53a
6c435c34853443a55c15bf7adf71c06c81649ec4
29879 F20101123_AAAGON hamey_r_Page_52.pro
9dffb1e3f440a0782d5088de5eb3cece
ea7984ddf136a2fd14c3a2f49fca41f3abf74e30
86993 F20101123_AAAGJQ hamey_r_Page_17.jp2
5212f5d10ef7ee429f1821fd99d616fa
10a1f0b78b4a6d1dae82908c33995e44557b4fab
1449 F20101123_AAAGTM hamey_r_Page_44thm.jpg
6cf4c1607812f1f24efbe1f57c43fb97
c1076de5f28fd35127076847ce8468372d5e9a2e
F20101123_AAAGEU hamey_r_Page_13.txt
e0c25a4cc0546241d500818c8e4a5885
9ef84f9355a21051c0b1da5175e3a6e96d24ef69
42334 F20101123_AAAGOO hamey_r_Page_53.pro
3e1c1bb7eae9158f98996be374e18ba7
6e8e618971a67885133ed03f3bf2f2471e92c7f3
99247 F20101123_AAAGJR hamey_r_Page_18.jp2
e36223a23dc0387d4edf09828fd3eb1b
d4dbb6cc7aefeff9aff31b5ac1800565c12911fe
25349 F20101123_AAAGTN hamey_r_Page_45.QC.jpg
d389cf2bc46f995a5a87ace12fbc3c2f
fed4c26884f53859d3634efb5eb3f1663e4089fe
93092 F20101123_AAAGEV hamey_r_Page_26.jp2
18f8527caa9d1d6c478b32277d6f1b5c
70f5cac36b1a7aed0651b5d8e84c921682a9661f
52797 F20101123_AAAGOP hamey_r_Page_54.pro
834ea3200daf16cf69bbb469ea7eedd8
fb3dfc080930f6a36d7084744baa22f330a172e5
113988 F20101123_AAAGJS hamey_r_Page_19.jp2
62188cfff6e07b22ba39909d126ec053
fe66cbcc11ca352357bb3ef563c5356cbfa596ab
6402 F20101123_AAAGTO hamey_r_Page_46thm.jpg
f5d81a0029fb59363bf14f3b7ccb6113
dfa203c5d4beb5417715a6a6dfb16bc89155ddce
80879 F20101123_AAAGEW hamey_r_Page_17.jpg
0b307eec153eac084addd8ce672aca11
70c093eddab82f2c12eecc92917b9f24e1b00e42
49511 F20101123_AAAGOQ hamey_r_Page_55.pro
501bd4a4e75da067a068c4b4d5321595
ad8632ba9edd3822e59dbe7dddafc5029efb6183
101877 F20101123_AAAGJT hamey_r_Page_20.jp2
0a520d914bb24b39e864c39608fbb189
f13080de17651c61b8e4517f7f44db621b068ad6
25060 F20101123_AAAGTP hamey_r_Page_46.QC.jpg
61e62b2e8a75fc4172db0ab68e027e99
c3a022162db814bae42401b5598897bf2f6d7bd9
16155 F20101123_AAAGEX hamey_r_Page_09.jp2
39f23e7c1c085c03ed5f86f2e46d37c1
3550fa230e54e04a6d0fbf1c8ebc1581b0670150
42806 F20101123_AAAGOR hamey_r_Page_56.pro
c6418c2beb5df79cfc924b5d57051e5c
7e50afa7b6bb1a372d58b142c3af7efe4a2b492f
101231 F20101123_AAAGJU hamey_r_Page_21.jp2
e5a755686c2c53e66c4203fa219cc60c
c9d73f5bfc1d0dc82ff6e55a1b156024489cab81
8106 F20101123_AAAGTQ hamey_r_Page_47thm.jpg
a0f6383f5a378d8d2aa31d7c02072a4a
593748fc4504f90fbbfe4bb3e4c6371d297da40d
F20101123_AAAGEY hamey_r_Page_17.tif
f9a3550197d7efa5caf2339567e9bfce
dcea4a29374e3c87f02bccebadf61affd2658006
20203 F20101123_AAAGOS hamey_r_Page_57.pro
ee303be45a3ee32a6cde80ca5cda4d46
c999236b43a1322b9f00456d3493af91767b696a
109735 F20101123_AAAGJV hamey_r_Page_22.jp2
28f4e83ac3ef42493e023e555eae0bbd
95709b4aca21331df4fc4e950e06162de09bb07c
31585 F20101123_AAAGTR hamey_r_Page_47.QC.jpg
9f37dbfe02cacbe4b1f15830be0db45e
3618649a59b35d88f284b3c279ca62b39e36f248
1695 F20101123_AAAGEZ hamey_r_Page_33thm.jpg
0ac7f74d69c21207209399181a73975e
3b4e6184080cd0ffcb8975deee3b790d9b38897c
16760 F20101123_AAAGOT hamey_r_Page_58.pro
41315c2363156af640e59450a01813c9
4f062045027704d11a6b58930b5c647411fa34ce
36258 F20101123_AAAGJW hamey_r_Page_25.jp2
c78d3d069eeb4a1ef572cf371553591c
ca63ddde07f49c0bf0c2abb2f06d7b40935b2854
49921 F20101123_AAAGOU hamey_r_Page_59.pro
86a6692c7c78bd00075a4d38781bb63d
ae529a4db02f459aa2c1550efa7e03f4eab6fb89
935599 F20101123_AAAGJX hamey_r_Page_29.jp2
ba555fb998f8a1719e9744cc495375ba
12b7eabf0c5d36dcfe6dfffe540a1f471b5524b3
7151 F20101123_AAAGTS hamey_r_Page_49thm.jpg
9bae9af90d3ddcca1bd7dc5aa71858ec
9344396c62318e9016fd355d8dd02141bf8536f7
54852 F20101123_AAAGOV hamey_r_Page_60.pro
a1fe638a16d7bf6c5140198211b5a19c
27448b07852cf862035623939518444212d19e55
26649 F20101123_AAAGHA hamey_r_Page_43.QC.jpg
e849e0d029e0085520354493ec113ab6
6b6e7ec26c7da7767a14e2273a5acfb6a92f083f
104589 F20101123_AAAGJY hamey_r_Page_30.jp2
4a92ceb33ec017f5457e15cdfed38cff
568a41be1fbe6f053567cd1fb9446e34282606d0
27621 F20101123_AAAGTT hamey_r_Page_49.QC.jpg
ce6a0064db7f132fd478c8f396484460
ac41abc767458e7d74dbd3417ebb4fabd2bbc4c4
54407 F20101123_AAAGOW hamey_r_Page_61.pro
40cffe7cfc60aaff543efb5741330aa0
45c46aa5f5766855dcfb5e2e6f55118cca5ade6e
1785 F20101123_AAAGHB hamey_r_Page_37.txt
93df34186ea9cc49568e5dcd9e46819e
bc88aa171a7f2747372fe2916c124df0cf30b0fa
110961 F20101123_AAAGJZ hamey_r_Page_31.jp2
01e5c351ed7d3a5e9c73e99f1bedaa54
38c3218cfd977a2416d57c7fe79022ed86ca006f
30669 F20101123_AAAGTU hamey_r_Page_50.QC.jpg
ded48f3f95d00a4fc63931b61e88dd65
1c3919e60ef53b32b63087cf36335ea1f74bfe0c
57114 F20101123_AAAGOX hamey_r_Page_62.pro
451961a1ffe33910e7701f5ea009ddd9
cb1e8c5f503c1951ce06537a709af1e1de7803e7
1042385 F20101123_AAAGHC hamey_r_Page_28.jp2
f784b0675f546f2560e8fec5aa34ea7a
c93c08698a7bc33e0d82936002ac78a1c2dfb43c
F20101123_AAAGTV hamey_r_Page_51thm.jpg
9bd5a0baa3e8ab41dec6aebcdbb9f3e0
b64a7aeea8ed2e96d59b9921675f2c595653b68e
F20101123_AAAGMA hamey_r_Page_31.tif
d48cd7d6ba61d9b22adade885927aa8e
808576faddb3f3b28fc0353bae7f1376a29e99f6
56761 F20101123_AAAGOY hamey_r_Page_63.pro
3f567f1715679d736953910f4d0dac44
3aff750070650edb7205371c07492189fec2ccde
93711 F20101123_AAAGHD hamey_r_Page_21.jpg
6e51d080e4a64305bceea572e72bf1c0
56aaed59b659bb030ae05a3a7e3cdde70651c429
29029 F20101123_AAAGTW hamey_r_Page_51.QC.jpg
285790fd7976131294563be78216da88
a5554bcd602cc0f2fc1c9135c5cc5eb16bb0cb74
53020 F20101123_AAAGOZ hamey_r_Page_64.pro
0b513cb9dabfc7b8112adeb38460a7c9
ed634ce4131a8e76e4f9a992fc4e71c12f250609
100334 F20101123_AAAGHE UFE0011903_00001.xml
842954c59ed613f15bee8a327762ad37
4be5ee4c62ec1b3702982a31d3a9d5a1269e1849
F20101123_AAAGMB hamey_r_Page_32.tif
5736efa9078909676312353e6f34fcc7
2676270d71d573be4ea1ed45e46fa690531330fd
5243 F20101123_AAAGTX hamey_r_Page_52thm.jpg
304d6c07422aa3fe87c51892a51abd1f
57ab6f5befc822bbf3d0251262deb4eafd5e70f9
F20101123_AAAGMC hamey_r_Page_33.tif
2076974c483043aff66c229b044dde7d
db51b5e048df7ec5cf64fdf4bffb680a7b40069d
19506 F20101123_AAAGTY hamey_r_Page_52.QC.jpg
7386a42f11606a7e4af59b50cd4ccd0d
5da6b4e234489f440350fbb009ecb912cd65a676
4797 F20101123_AAAGRA hamey_r_Page_05thm.jpg
1d6379e6e67fa906f745a2cc8677d22f
533c0fc102afb0b52ea9b07bf95af936ad1dccb7
F20101123_AAAGMD hamey_r_Page_34.tif
23bb8dee4cfb6996a16ef4fcd7bc0b0e
16563146b51aed064d8cac5104688b9f443ab595
27143 F20101123_AAAGTZ hamey_r_Page_53.QC.jpg
7f537c000d160d992ab1a76e92944af4
10f396770d6787d31c6d9c2571050ad345325536
18635 F20101123_AAAGRB hamey_r_Page_05.QC.jpg
4e1c52e6b7205a92d8281ff7793add65
d243cfa803d44f127e11021f11c30533e304b84d
F20101123_AAAGME hamey_r_Page_35.tif
b6024d16b760582ade481acf14a727ec
f0713fb9133a8f9a57b0f2786bf3d4294747c09d
24848 F20101123_AAAGHH hamey_r_Page_01.jpg
906ce10c8e4565ff67b5c60618b8463c
2a3b0c07ec65ae19b3af65e0dde461da2a80a0e0
2586 F20101123_AAAGRC hamey_r_Page_06thm.jpg
8aaa911cec6a164742e570ffd950ec96
d3c648f83cd77d2d5fc2f99754570471fdaf3846
F20101123_AAAGMF hamey_r_Page_36.tif
474de833f7232c7e4e94772c5f34a94a
df56dd3de54838b5115e5cf92e9f29edd485dcbd
4694 F20101123_AAAGHI hamey_r_Page_02.jpg
adf0f019454379c6c48a5f633d4cadf4
f8feefc63fa1660aded3388dbafe6bb3d96c0aa0
9077 F20101123_AAAGRD hamey_r_Page_06.QC.jpg
aa32be57bea2a6f1b64e37627b12c7d7
c00c4131f6e673f406a7cee025fc538017737712
F20101123_AAAGMG hamey_r_Page_37.tif
197108f445df46fb7dd09ed6f55553de
b90ee5fabbf3c9a8a6fd1d2107b38a97dd00176c
64095 F20101123_AAAGHJ hamey_r_Page_03.jpg
67941c3e12ff9bc378bc9c4f060ba341
d037c4623fd26050a7331937da9d66871a2af02a
7748 F20101123_AAAGRE hamey_r_Page_07thm.jpg
bd69286098f874c131cd6d317d8cad2a
53934a93e95eb218560a0aab275acb20d11965a1
F20101123_AAAGMH hamey_r_Page_38.tif
59be99001199cb6666daa0975574d978
f71a418e0d11c9974584d389330b91d7eebe4150
99091 F20101123_AAAGHK hamey_r_Page_04.jpg
81e2550297f52a325828d34aef8d665a
3fbe473a64b69131948b5ab50f3dc0f7f58b2166
31987 F20101123_AAAGRF hamey_r_Page_07.QC.jpg
c89d684cc49f2fd362b0851291f4bcd0
c0fb18c356582bacb32f9b8a8b930d551b204681
F20101123_AAAGMI hamey_r_Page_39.tif
45fc712a44f1f4ce76e5e74eb92f6b7d
697285413de50e8e9c9e517f046fcd4fd5a05b7f
79693 F20101123_AAAGHL hamey_r_Page_05.jpg
9232c770af8cd6c0903c4bedd75140cb
860a74a5e7f18c8fa77f88512d67ee8fac7ec127
6705 F20101123_AAAGRG hamey_r_Page_08thm.jpg
d3b08b7f8c5cfdf03e838abf4e7794ab
5b9fee751e609d3b3f44b2f8044746e927546c0a
F20101123_AAAGMJ hamey_r_Page_40.tif
a5bb9e15e3e3ebc725630ce1241a21fc
5a51a21d32416b08443e6c4a13f96b9ffddc41c8
118558 F20101123_AAAGHM hamey_r_Page_07.jpg
d2f5b1c4de11d560ea57e73102af10d3
66a232e56189ca558d3f9f5441b877e10c40a346
26340 F20101123_AAAGRH hamey_r_Page_08.QC.jpg
66215618a544c7474d197fd54759cb98
aadfa55fad91346b5e4bf859a1ba4431ede78661
F20101123_AAAGMK hamey_r_Page_42.tif
d22f8e5e0b4cc5abcbcfd94b55743f39
09c4e7e1dc969a0562e7001c2a27b86ce3d7a6f2
85306 F20101123_AAAGHN hamey_r_Page_08.jpg
f97894937816dcfcb2bb01db35d19f42
774c55f8598f14b20ecb23f242d11d3b6c5ed8e6
5307 F20101123_AAAGRI hamey_r_Page_09.QC.jpg
c8e30d69aeb3f33d539dd9b2e8fdef47
a4cfa64c7365653edbfd666d828d4800769e8628
14851 F20101123_AAAGHO hamey_r_Page_09.jpg
f5f7bf7fb031069fe0e83cdc08bffea6
989315b4fb52bc9b60743d3f80ab0dd53785cce9
7194 F20101123_AAAGRJ hamey_r_Page_10thm.jpg
2cb6888299e69dd6bd8aad8c4e829f64
1e81923a77dca3b695d5a71accd3b07691972c7a
F20101123_AAAGML hamey_r_Page_44.tif
07cd489a33cf73efe85c33fa666c34be
33aa62e2167fb9c14942e149ccf784421134d4f3
89210 F20101123_AAAGHP hamey_r_Page_10.jpg
9f2d8ebc9bdcd274b91ab8165f7b1b33
46d664bbaee7810fc05e69fd64e081dc8b622601
F20101123_AAAGRK hamey_r_Page_10.QC.jpg
63cd3b70e3e4a8e6b501d5966bd959a8
eefd5a4ad8ccd70644c92fbed04117b98481f8fa
F20101123_AAAGMM hamey_r_Page_45.tif
2564a76d3374726ef46099756289720e
e9229e970f6725edf6cec6e93ab14910bf377559
98228 F20101123_AAAGHQ hamey_r_Page_11.jpg
09fd1f8101b10ec393c0ac8ecaabf7ea
f2c0e26cf4b601cd01b09bba2ea56204ac94cfda
7861 F20101123_AAAGRL hamey_r_Page_11thm.jpg
ef1b64d6f86e0ae7df504e55410cb071
4603c4f5aee0da607db7ebf27146c57dd407c33a
F20101123_AAAGMN hamey_r_Page_46.tif
f9f2d05f90cd8458025ed4de1a606e3e
530ce33c61250fb04d242c1f33751e96d623d1b6
89360 F20101123_AAAGHR hamey_r_Page_12.jpg
6da8e83e42f493a24f758e8b8d5ea659
052319f9e5a29543b027e4c4605b4683c0b74c32
31425 F20101123_AAAGRM hamey_r_Page_11.QC.jpg
285a089eb2c936e7502a30cc71b373b6
45e0f39e7fc9fa7c937dab8093190b3e7d9e4cbe
F20101123_AAAGMO hamey_r_Page_47.tif
d4e4e718d96550dabd568e56261b0ca8
5fd950deefd594232efbebd137e6574b229f00a5
103072 F20101123_AAAGHS hamey_r_Page_13.jpg
7a015160ff6ea8b903ec684aef5afdbd
8f069ce540e9b8258542f3c816ac8aa1241e9f57
7243 F20101123_AAAGRN hamey_r_Page_12thm.jpg
c9f21c869ad67ef10e3d8c2cad82fc16
5462c96fa61909709e07f2b06a99a3e6be483693
8423998 F20101123_AAAGMP hamey_r_Page_49.tif
fb4e0d5fa23d4829d55544ad28372c70
bf6b28f02fba0abf0949ba29bb23cfb0216be6cc
93816 F20101123_AAAGHT hamey_r_Page_14.jpg
546110cac6f38393012a2dbed6960362
60e961f6da6579a955e72bb2a35d0e2abb5d1f53
28978 F20101123_AAAGRO hamey_r_Page_12.QC.jpg
b9c1beab5b3f9dd3dbb92c04ad85f1e2
a52994b26d3760decdeb6f5080763d2ecdf5b461
110860 F20101123_AAAGCW hamey_r_Page_64.jpg
5afc4dafc6d19152caf83bc26e082810
d3536057d95a7bd5c647ecfd14beea061ada2983
F20101123_AAAGMQ hamey_r_Page_50.tif
5fc1e3f83b4d0b9d340bc1fbaabf9222
5296e50a0dbed5370805a7715e387ef99563753b
105031 F20101123_AAAGHU hamey_r_Page_15.jpg
8fb232aafbbc4747ea56d60402d8dd6a
85baae0831224a5c0a90c641c369201dbce00fe3
8230 F20101123_AAAGRP hamey_r_Page_13thm.jpg
7b58e143c754522cead01a576adcd651
a79d7de56d1495685d16163d4963dfab8ecc83a9
20916 F20101123_AAAGCX hamey_r_Page_03.QC.jpg
4144e4c68f1d84016846bb0380eb1c2f
86a27bd985f5e302d0f9aa228e3a3ef974ced0ac
F20101123_AAAGMR hamey_r_Page_51.tif
c8479f92935a406b6eea530930fa6b4b
58fb7cf32b5fe9164d7dbd9bbc02f96097802b6d
88582 F20101123_AAAGHV hamey_r_Page_16.jpg
7d23f97f20ba6500e965b0b1e77a3b90
dcf3ae21665b34681549e41897ad7c9f4949c170
47326 F20101123_AAAGCY hamey_r_Page_21.pro
7f332fe32e8c71c592bc52fe3b153ad4
7057788e85f5b0393ef8a84635f5bcfb297521ca
F20101123_AAAGMS hamey_r_Page_52.tif
0a46c2bc1274f2024be1a8b2a5cec4ee
45d326f4dca998efb089048ecbc4be57de062335
106163 F20101123_AAAGHW hamey_r_Page_19.jpg
16d117e1094da51d0231cc7af94e9b3e
7eeb90d76bb3f65f73462820087f0392ef64d6f4
31687 F20101123_AAAGRQ hamey_r_Page_13.QC.jpg
b007306620cad83c4785bddd5f008d0d
09f3bb3ecd56bc2408223a7c423c7a5085773453
107379 F20101123_AAAGCZ hamey_r_Page_55.jp2
66b9fd4a9105c0f4be6f0451ce7e26ed
8ce30d05b0cc67aa8d3869105a1549ef593282d8
F20101123_AAAGMT hamey_r_Page_53.tif
9c04b1f4ff3f9f20b1d163b6980f3c14
71de1099bab2ef84fba67594abf215c319508d68
96064 F20101123_AAAGHX hamey_r_Page_20.jpg
81e7d6b30b1f92a4e4a084a77a6cf76c
a945628f2ce6fe0dd344d9c8030be61b4b82608c
31798 F20101123_AAAGRR hamey_r_Page_14.QC.jpg
d13d53ed8f374811248cbf50214b6c33
4ac7bc3ec564d35a90fd246b286d969166c9e4db
F20101123_AAAGMU hamey_r_Page_56.tif
1d3847e0c6743e78488468fb43dbf041
22c63156f940656463175a3a7cf35d25134e8a52
95681 F20101123_AAAGHY hamey_r_Page_23.jpg
bbd2c8848c7eaccfd854e255c7371ddc
40de919d6b79bd7c80b308b35f88eac1a90cef88
8467 F20101123_AAAGRS hamey_r_Page_15thm.jpg
242a72ae806e60429ec0f35d935306c2
42c5456ab5d73ecbfc548451946f81c6b467fdea
F20101123_AAAGMV hamey_r_Page_57.tif
d1b1fce0b6e63726556bbb4792bf011a
7d6f04de6bcf49c1c0f85da6972c285831b6a368
48304 F20101123_AAAGFA hamey_r_Page_30.pro
1c5921a165423dc004fb1ce46de83186
ab5ccf9ac4091a333b0b29a0d76dc099b4ade782
89802 F20101123_AAAGHZ hamey_r_Page_26.jpg
1567856e5ff1086f7a661edbcb80509f
ed5583f389411505893e4fdac47e82be18601885
34284 F20101123_AAAGRT hamey_r_Page_15.QC.jpg
56b3514414e80a80bf4f2e2de589c0ab
70cc5d7195e02befb0c4148eca7ae35a45bd3ad6
F20101123_AAAGMW hamey_r_Page_58.tif
8444c508b3108ec38359703e43031553
257ad3c694c1d4bd2923b3251b5f9192706bd003
F20101123_AAAGFB hamey_r_Page_43.tif
e869049b6c8f7a1911485fdff402a2b9
df05da3dfb3d12f0a82c72381d40886811f7d9e1
7595 F20101123_AAAGRU hamey_r_Page_16thm.jpg
b51d0cd98385a83de6b726495f534748
36aa34cf9723b7259b8f622467083bb4d3a63b68
F20101123_AAAGMX hamey_r_Page_60.tif
6f275ffa1ed6744dc085b44b17bb9cc2
55bf5b2e7fbd8c629e0b03d82bf2ed7b1afa3054
31284 F20101123_AAAGFC hamey_r_Page_06.jpg
b44bc3966dae872d6c1d8d153428a34a
87f6e135feb0cf9e8ee75e8c812900bd8d47294d
29281 F20101123_AAAGRV hamey_r_Page_16.QC.jpg
c8a2a3f9b84b3d605398758e72c8f6ed
1183925f55d42be863ac2d745822605cd35957fd
110030 F20101123_AAAGKA hamey_r_Page_32.jp2
afaae45a267a9e5edb7d6b94ac426079
5de79d1729f34c2137698168e8ab604f617b9ace
F20101123_AAAGMY hamey_r_Page_62.tif
a4ddfe8d2cb0ee37f9b1c16de5230aac
341edd0c8eb1da9dd99b3b4d0454f2fcb4bdfb01
33121 F20101123_AAAGFD hamey_r_Page_31.QC.jpg
4bdd92f5019d7e02540d2b00938f5eab
f06a470f25f914e7e2cce3f99f4c029fa4ac5932
7089 F20101123_AAAGRW hamey_r_Page_17thm.jpg
8af59db049d954d98059cad95ee9e961
0ffd9689a7397c9d1721b213279531b39bb16510
19051 F20101123_AAAGKB hamey_r_Page_33.jp2
04ab29d3bf1ce393301962e746ab8e1a
3fdd7f0e86333085b718dc23c459b1b5d7378d0d
F20101123_AAAGMZ hamey_r_Page_63.tif
7b7688c58f4dac8594b26adc9f8049c6
0ad9ecbfa9817c19725a4403044ef4e0c482a674
30661 F20101123_AAAGRX hamey_r_Page_18.QC.jpg
7ee3ea69495c8ae3cd63018e331f44e1
886ace1ed0a1acef40cd8ab94d5171446b2bdf28
1051906 F20101123_AAAGKC hamey_r_Page_35.jp2
35965036716b7ca9acd9de0ae735c155
b1aaf93060a98419411ef640f27e9b5f911d4d8f
97480 F20101123_AAAGFE hamey_r_Page_50.jp2
0f529832a8a09748dab554ac0a21732e
c21c450c51ef902165ae7c14b9993376857a28a2
8566 F20101123_AAAGRY hamey_r_Page_19thm.jpg
13f8ae736ff5e90afe710e2c2a251fd2
0427bf3b7964ab42e3148edc3b69be35e730de0f
98296 F20101123_AAAGKD hamey_r_Page_36.jp2
7fd56c72135e4ab2aa7cb1763d448932
3f2dc222a6bf3d0f736c31b0d39f55d59739939c
82754 F20101123_AAAGFF hamey_r_Page_35.jpg
96ce6794d08da6524a6cd2b8c0fcc676
4657037b22987ce8c78a6f9df013d93289471221
8313 F20101123_AAAGPA hamey_r_Page_65.pro
4439e6d571554e906edce94f44f0395f
3134a078e8390c1b11d8723f98d7ff35ce9cf5f6
35758 F20101123_AAAGRZ hamey_r_Page_19.QC.jpg
e18e686891214712b2442de4ee35c2a5
bbecc79639a32b8d5caf985913af342e84ca131b
105789 F20101123_AAAGKE hamey_r_Page_38.jp2
a27b9f549befe4d301cb1216e2d4854f
266903da0f05795ba4ce02ee89371c7c1ce4ac8b
101572 F20101123_AAAGFG hamey_r_Page_48.jpg
7c19e63b0f4d974cb4699a1b5e6a446b
d073e99dcbc88ae58552951249c30bfb5d19b9f5
14794 F20101123_AAAGPB hamey_r_Page_66.pro
13c69a8836c1e26987a1914de6885405
87289e651116df7a48e3c16cebb6c8e357c03cce
87928 F20101123_AAAGKF hamey_r_Page_39.jp2
bab33f8bc167b996eb95c699d943cca7
34613387716943b9bf28e55ecc2633a97c35061e
F20101123_AAAGFH hamey_r_Page_59.tif
dd1ba00c1b2f01bba12eb9b0d5f4261c
ce3ec0a50012165cc903d975279aede25ca3f720
3521 F20101123_AAAGPC hamey_r_Page_04.txt
303fb3ce6548897b32e4032d0eeed8c2
f56cda7094b67d0500ac80d3b296141e3f997eab
35345 F20101123_AAAGUA hamey_r_Page_54.QC.jpg
0fdb7486cb2722694c90749d51b3d95e
30060528dc4b21303516834ffad0bf509c074064
102030 F20101123_AAAGKG hamey_r_Page_40.jp2
70c6354ab2a1f08efa044a0f27054580
3ae87beec815f9753a56d0bd5f7a64cd3ad93a3f
32733 F20101123_AAAGFI hamey_r_Page_48.QC.jpg
705ed063b52175d4f046e43419adf2a8
f25be1fb50b1f483ae641df50bb37f3523170496
2499 F20101123_AAAGPD hamey_r_Page_05.txt
81562d7ff5e75215e9ec1cccb2326fb0
bbc9ab327131956f493e79a4ddaf3e0d6bb9ce29
8022 F20101123_AAAGUB hamey_r_Page_55thm.jpg
c7e90e2c2989bcc1fe09eae91b03ab1f
7e41c10bed9af7b741b98e53b5e0f3df32ff3f98
104933 F20101123_AAAGKH hamey_r_Page_41.jp2
55e3e805d729587133f34bd718f6db88
8fbc41254db70a2836d0b750c4472292d3e91f7e
2201 F20101123_AAAGFJ hamey_r_Page_64.txt
e7bb3041aa705111d5d8821f8dcde374
86c7bdf136cd40970e6d65bfaa1c601237bf0110
664 F20101123_AAAGPE hamey_r_Page_06.txt
5ccbf46a0f3bd7003e9d7a4fc95e2907
54f3cc239623c57a57962cb860f2e8dbe2c52844
F20101123_AAAGUC hamey_r_Page_55.QC.jpg
51989fd15121a1c24208e993be167180
5ac29bc95e2ccb1b98c330eff5f6d02f412e4e03
71711 F20101123_AAAGKI hamey_r_Page_42.jp2
c0bdb51edcad1b187468941c6486b1f0
ff71682bb07d28b9f7cdae8323e1e5f2de7c43e1
103349 F20101123_AAAGFK hamey_r_Page_22.jpg
06d68969040bc8f09ea45f9cec66df6e
72e64a38e47d5135964ec3def3e1e332b4ed547b
228 F20101123_AAAGPF hamey_r_Page_09.txt
53d24e49a86387555c214b0730b87e6b
672e90c0db225bab5c2169594148d3c9f043fb18
6922 F20101123_AAAGUD hamey_r_Page_56thm.jpg
6f9a2baad8341407491b503353d08946
ccb491aabd480b777801f697f8f9ea9638e94e55
107911 F20101123_AAAGFL hamey_r_Page_24.jpg
a7035170742c00a54e43965e22459527
533d17cfff5b16ef971855ee2147eb6cee6827c3
1814 F20101123_AAAGPG hamey_r_Page_10.txt
d7e85f6b19020dd7322a8eba7100cb51
894f75b4517c0622a50a354171e6c82c88668611
28225 F20101123_AAAGUE hamey_r_Page_56.QC.jpg
ced4aae980a045f4ca58bc43173dbc65
c9417bf953ef7a6b1a8a23293216a0e032149e41
79232 F20101123_AAAGKJ hamey_r_Page_43.jp2
355c415674a85ca43b2532bfdb6f603f
883f6087b5cdbe8645f65f0d55960f922d16b5ba
33818 F20101123_AAAGFM hamey_r_Page_25.jpg
1bdd2a5bec429794138fbcf7717bf48b
62bacf1d463fc39a467daafe93ceb15ce2bf781c
1920 F20101123_AAAGPH hamey_r_Page_11.txt
622cd76d5ad1dde7d75c987b4112f75c
ef8847f7f3b658aca025be1273fd8d9ee47c10a6



PAGE 1

PRODUCTION OF ORGANIC PIGMENT NA NOPARTICLES BY STIRRED MEDIA MILLING By RHYE GARRETT HAMEY A MASTER THESIS PRESENTE D TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Rhye Garrett Hamey

PAGE 3

iii ACKNOWLEDGMENTS The author would like to acknowledge the fr iendship and the wisdom of the late Professor Brian Scarlett, whose instruction and support were th e reasons for the pursuit of this study. Dr. Brian Scarlett taught the author how to be a better en gineer and a better person. The author is grateful for the help of Dr. Brij Moudgil for his countless hours of guidance through out this study. The author woul d also like to thanks Dr. Hassan El-Shall and Dr. Abbas Zaman for their insightful discussions on milling and serving as committee members. The author would like to acknowledge the e fforts of Dr. Ecevit Bilgili, Dr. Dimitri Eskin, and Dr. Olysea Zupanska for their collaborative efforts in understanding the particle and system physics in this study. The author would also like to acknowledge the financial support of the Particle Engineering Research Center (PERC) at the University of Florida, The National Science Foundation (NSF) (Grant EEC-94-02989), and th e Industrial Partners of the PERC for support of this research. Thanks are ex tended to Eastman Kodak for providing the polymeric media.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.....................................................................................................................vi ii CHAPTER 1 INTRODUCTION........................................................................................................1 1.1 Particle Production..................................................................................................1 1.2 Particle Breakage....................................................................................................3 1.2.1 Introduction to Particle Breakage.................................................................3 1.2.2 State of Stress...............................................................................................3 1.2.3 Cracks and Defects.......................................................................................4 1.2.4 Breakage Mechanisms..................................................................................5 1.2.5 Materials.......................................................................................................6 1.2.6 Theoretical Models of Breakage..................................................................7 1.2.6.1 Griffith’s theory of brittle fracture.....................................................7 1.2.6.2 Irwin’s and Orowan’s theories of ductile fracture..............................7 1.2.6.3 Rittinger’s and Kick’s laws on particle breakage...............................8 1.2.6.4 Distribution based models..................................................................9 1.2.7 Measuring Breakage of Particles..................................................................9 1.2.8 Environment...............................................................................................10 1.3 Stirred Media Milling...........................................................................................11 1.3.1 Media Mills................................................................................................11 1.3.2 Grinding Aids.............................................................................................12 1.3.3 Media Loading............................................................................................13 1.3.4 Media Size..................................................................................................13 1.3.5 Other Milling Parameters...........................................................................14 1.4 Milling Kinetics....................................................................................................15 2 EXPERIMENTAL PROCEDURE.............................................................................17 2.1 Equipment.............................................................................................................17 2.2 Materials...............................................................................................................18

PAGE 5

v 2.3 Procedure..............................................................................................................21 2.4 Sampling and Characterization.............................................................................22 3 EXPERIMENTAL RESULTS AND DISCUSION...................................................25 3.1 Base Experiments.................................................................................................25 3.2 Effects of Grinding Aid Concentration.................................................................27 3.3 Effects of Media Concentraion.............................................................................28 3.4 Effects of Media Size............................................................................................30 3.5 Breakage Mechanisms..........................................................................................32 3.6 Grinding Mechanism............................................................................................32 3.7 Grinding Kinetics..................................................................................................33 4 POPULATION BALANCE MODEL........................................................................36 4.1 Method..................................................................................................................36 4.2 Experimental Findings by P opulation Balance Modeling....................................39 5 SUMMARY, CONCLUSIONS AND FUTURE WORK..........................................44 5.1 Summary...............................................................................................................44 5.2 Conclusions...........................................................................................................45 5.3 Future Work..........................................................................................................46 APPENDIX A REPEATABILITY RUNS OF M103.........................................................................48 B CALCULATIONS OF CMC......................................................................................49 LIST OF REFERENCES...................................................................................................50 BIOGRAPHICAL SKETCH.............................................................................................57

PAGE 6

vi LIST OF TABLES Table page 2.1 Operating variables of the stirred media milling process.........................................18 4.1 Parameters of the Kapur’s G–H model for Run M103............................................41 4.2 The breakage distribution matrix for size class 1 through 9 form top to bottom and left to right (b6,3= 0.958)....................................................................................42 4.3 Selection function for experime nt M103 population balance model.......................43

PAGE 7

vii LIST OF FIGURES Figure page 1.1 Attrition, cleavage, and fracture, and particle size distribution for attrition, cleavage, and fracture, respectively...........................................................................5 2.1 Optical microscope image of th e polystyrene beads (mean size, 40.3 m). Marker size is 50 m................................................................................................19 2.2 Images of the pigment pa rticles: (a) SEM image under 50,000 magnifications; and (b) TEM image. Marker size is 100 nm for (a) and 500 nm for (b)..................20 3.1 SEM images of pigment particles in Run M103 samples collected at various milling times: (a) after 10 min; (b) after 16 h; (c) after 24 h....................................26 3.2 A, B, C, and D show the effects of surfactant con centration (0.0, 0.02, 0.037, and 0.0937M) on particle size at grinding tim es of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively.........................................................................................................27 3.3 A, B, C, and D show the effects of media concentration (0, 60, 120, and 200 g) on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively...............................................................................................................29 3.4 A, B, C, and D show the effects of media size (9, 21, 40, and 402 m) on particle size at grinding ti mes of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively....31 3.5 Plot of change in particle size with time for varying media concentrations that describes the grinding rate........................................................................................33 3.6 Plot of change in particle size with time for varying media sizes that describes the grinding rate.......................................................................................................34 4.1 Fit of the Kapur’s G–H model (4.7) to the cumulative volume fraction oversize– time data. Legends indicate the lower edge of each size class i ...............................40 A.1 Plot of the volume fraction versus pa rticles size of M103 including the standard deviations at varying run times................................................................................48 B.1 Measure of surface tension versus surf actant concentration for OMT surfactant....49

PAGE 8

viii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science PRODUCTION OF ORGANIC PIGMENT NA NOPARTICLES BY STIRRED MEDIA MILLING By Rhye Garrett Hamey August 2005 Chair: Brij Moudgil Cochair: Hassan El-Shall Major Department: Materials Science and Engineering Stirred media mills are widely used in i ndustry for fine and ultra-fine grinding of various materials and dispersion of aggregates There is a growing interest in the largescale production of nano-size particles, and wet stirred media milling appears to be a promising technology in this regard. In the pr esent study, a stirred me dia mill operating in the wet-batch mode was used to disperse a nd grind pigment aggregates. To disperse the initial aggregates and to prevent re-agglomer ation of the fine particles, the pigment aggregates and a polymeric media were suspende d in an aqueous solution of a dispersant. As the size reduction progresses, variations of the particle size distribution with time were determined using the photon corre lation spectroscopy. The product pigment particles were also examined under a SEM (scan ning electron microscope). The effects of operating variables such as surfactant concen tration, media loading, and media size were investigated. We demonstrate that the stir red media mill with a polymeric media is capable of producing nano-particles in the 10 to 50 nm size range. Experimental and

PAGE 9

ix population balance model data suggests a fa st initial dispersion of the strong agglomerates followed by cleavage of primary particles. Further evidence for nano-size grinding is supplied by the SEM images.

PAGE 10

1 CHAPTER 1 INTRODUCTION 1.1 Particle Production This study was begun in effort to unders tand and enhance a method for producing nanosized pigment particles de scribed in a series of Koda k patents. The Kodak patents describe milling process in which pigment par ticles of a primary pa rticle size of 150 nm are reduced to less than 50 nm.1, 2 However, the Kodak patents do not describe the mechanisms by which particles breakage and milling occurs. The patents also neglected to describe how milling parameters affect these mechanisms. This study will evaluate how changing milling parameters will affect th e breakage kinetics and particle size. It will also determine the particle breakage m echanism through inspecti on of particle size distribution (PSD) data and a breakage di stribution function ca lculated from the population balance model (PBM). Many industries are realizing the benefits of nanosized materials. Some of the benefits are derived form the high specifi c surface area of nanoparticles. In the pharmaceutical industry poorly water soluble dr ugs have shown an increase in solubility as particle size decreases.3-5 This decrease in size will further increase the bioavailability of the drug due to large amount of availabl e surface area for reacti ons to take place at. This increase in bioavailability could l ead to new types of non water-soluble drug formulations.3 The ceramic industry has found that nanoparticles c ould increase the strength of composite materials. Nanosize or ganic pigments have shown an improvement in color quality, lightfastness and durability.6, 7 So far production of many of these

PAGE 11

2 materials has been limited to small-scale pro cess making the cost of these materials very high. Most nanosize particles are pr oduce in a growth process, from either a liquid phase such as crystallization or from a gas phase like flame synthesis.8-12 Both of these methods have drawbacks in that they can only be a pplied to a limited number of materials. In crystallization processes materials are limit by their solubility in a solvent. In plasma synthesis process materials are limited by thei r ability to with stand high temperatures thus making this process mainly restrict ed to producing metal oxides and carbon nanotubes. The scalability of bot h processes is limited due to control issues that arise when working with large volumes (i.e. te mperature homogeneity inside the vessel or inside the flame 12). To reduce the cost of nanosized materials a process capable of produce large amounts of many different types of materials is need. Milling is a process that is capable of produce larg e amounts of materials. Many researchers have studied the milling of particles of the size of a few inches to the submicron size range. However, little work has been performed in unde rstanding nanoparticle production through a grinding process.13-15 This study will further the unde rstanding of nanopa rticle production in a milling process by identifying methods of particle breakage and relating them to milling mechanisms. Much of the research conduct in a milling process has showed that media size and media concentration can affect the rate of milling and the fineness of the particle size for macroscopic milling processes.16-19 For submicron and nanomilling processes reagglomeration during the milling process can reduce the effectiveness of particle size reduction.14 This study will evaluate the eff ects of surfactant concentration,

PAGE 12

3 media size, and media concentration on the fi neness of the product and rate of particle size reduction in a nanomilling process. 1.2 Particle Breakage 1.2.1 Introduction to Particle Breakage Understanding particle breakage is impor tant in many industries such as the pharmaceutical and mineral processing. Breakage can lead to problems in processing and conveying equipment. In a milling process unde rstanding particle breakage is necessary for obtaining a desired product. In both situations, better knowledge of breakage behavior will lead to significant savings in time and money as well as an overall improvement of product quality. Single partic le impact, and multiple particle impacts on breakage have been studied by many researchers.20-24 Many parameters determine how or if particles will breakage (such as material properties, impact velocity, environmental effects, and history or fatigue).25 This section will discuss some of these conditions and will try to explain how they lead to breakage. 1.2.2 State of Stress There are many ways to apply a force that will result in particle breakage. Rumpf suggested four modes by which a particle can be stressed that could lead to breakage.26 (1) Compression-shear stressing (2) Impact stressing (3) Stressing in shear flow (4) Stressing by other methods of energy transm ission (electrical, chemical, or heat) The first two are caused by contact forces and are the only modes important for milling. Mode 3 is important in dispersion of materi als under shear, and only exerts enough force to destroy weak agglomerates, which will be ne glected in this study will fail. There are

PAGE 13

4 three modes by which a material can fail, which are a function of the material as well as the stress applied to the material. (1) Opening mode that correspondin g to an applied tensile load (2) Sliding mode where the force is applie d in the plane of the defect or crack (3) Tearing mode where a force is applied in a direction perpendicular to the plane of the crack or defect For most particles, breakage is considered to take place in the first mode of failure. Which may be due to artifacts in the techniques developed to measure particle breakage. If a normal load is applied to a particle a load equal in magnitude is acting internally in the direction perpendicular to the load and can then be considered to be a tensile load. Tensile failure is more likely then co mpressive failure for many materials. 1.2.3 Cracks and Defects Understanding crack formation and propaga tion is of fundamental importance to the study of particle breakage. Particle break age is preceded by crack growth, which is preceded by crack formation. Crack forma tion begins around local defects in the material. Defects are separated in to four classes, which correspond to their dimensions. (1) Point defects (vacancies, inters titials, or substitutionals) (2) Line dislocations (screw or edge) (3) Interfacial (4) Volume The strength of a material is related to the number, size and types of defects present. Most materials are only a fraction as strong as th e theoretically strength would predict due to the presence of defects. Cracks are initiated when the applied stress exceeds the strength of the material and form at a defect locati on. Following crack initia tion is crack growth; crack growth will lead to the breakage of the particle. Cracks can grow along the surface of the particle as in the case of lateral crack s. Lateral cracks will lead to wearing away of the particle surface as in the case of attrition. Cracks can al so propagate through a particle

PAGE 14

5 by radial and median type crack s which lead to fragmentation of the particle. The rate and type of crack growth are highly depende nt on the applied stress and the type of material. Shnert extensively studied crack gr owth and found that crack velocity can vary widely depending on the material, environment and state of stress.27 An empirical factor called the fracture toughness can be used to predict when a material will fail. The fracture toughness is dependent upon th e size and geometry of the crack. 1.2.4 Breakage Mechanisms Particle breakage can take place through many different mechanisms of loading. These mechanisms of loading as well as th e material being stresse d can affect how the particle will break. There are three ways by which a pa rticle can break, which are fracture, cleavage or attrition. During fracture a mother part icle will break into daughter particles or fragments that have a wider size distribution. During cleavage a mother particle will break into smaller daughter particles of roughly the same sizes. Figure 1.1 Attrition, cleavage, and fracture, a nd particle size distribution for attrition, cleavage, and fracture, respectively. Attrition is where a mother particle will have surface asperities removed from the particle. This leads to much smaller daughter particles. Figure 1 represent the three Attrition Attrition Fracture Cleavage

PAGE 15

6 different mechanisms of particle breakage.28 Particle breakage is dependent on the type of stress or load which the particle experiences If a brittle material experiences a normal load such as impaction in a milling pro cess it would undergo a fr acture or attrition mechanism of breakage.29 If a ductile material experiences a shear loading or forces it is most likely to fail by a cleavage mechanism.30 1.2.5 Materials The mechanical breakage of materials is a ffected by how the material is subjected to a load, the load applied, and by the materi al properties. Materi al properties are the elasticity, hardness and fracture toughness. Particle size and shape can also affect breakage. Particle size has been shown to ha ve an affect on the hardness of a material.3133 It has been shown that as particle size de creases the number of flaws in the material also decreases, leading to higher fracture toughness. Particle size can also affect the elasticity of the material cha nging a large size partic le of brittle non-deformable material in to a small size particle of plastically deformable ductile material.30, 33 This change is called the brittle–plastic transition, which usua lly takes place in very fine particles. Failure of a material can be characterized as brittle, semi-brittle, or ductile. Bond ruptures usually dominate the brittle failu re mode, whereas semi-brittle fails through dislocation mobility and bond ruptur e. In ductile materials disl ocation mobility is usually the dominant mechanism that leads to fa ilure. Brittle materials fail through the propagation of cracks by internal flaws, su rfaces flaws, and shear deformations. These flaws lead to particle breakage by attrition a nd fracture. For brittle failure the stress field is independent of the strain ra te, which means the rate of de formation does not play a role in whether the material will fracture. It also means that brittle fracture is usually very rapid and independent of fatigue Crack propagation and failure in semi-brittle materials

PAGE 16

7 is preceded by plastic flow. Semi-brittle mate rials can have propagation of cracks in the radial, median, and lateral directions. Thes e systems of crack propagations lead to a breakage by fracture and attrition. Semi-br ittle materials also show a strain rate dependency on failure. Ductile materials fa il through tensile and shear loading. Ductile failure is highly dependent on fatigue, and st rain rate and usually fails through a crack opening or tearing modes. This type of failure usually lead s to a cleavage mechanism of breakage in particles. 1.2.6 Theoretical Models of Breakage 1.2.6.1 Griffith’s theory of brittle fracture If a material is brittle it will fractu re by crack formation and crack propagation. Griffith proposed a theory that thermodynamica lly predicts the energy required for brittle fracture. In Griffith’s equa tion for plane stress, eq. 1, c is the stress required for crack propagation, E is the elastic modulus, s is the surface energy and a is the crack length.34 a Es c 2 (1.1) In the case of plane strain a factor of (1is added to account for the confinement in the direction of the thickness. Roesler’s experiments in conical indentation on silicate glass further verified Griffith’s theory.35, 36 1.2.6.2 Irwin’s and Orowan’s theories of ductile fracture Little work has been conducted to further the understanding of ductile failure. Irwin and Orowan proposed a theory that incorporat es the plastic deformation into material failure.25 Orowan’s equation, eq. 2, relates the rate of strain to the dislocation density ( ), Burger’s vector (b), and the change in the average distance a disloc ation can travel with

PAGE 17

8 time l; k is a correction factor for the non-uni form movements of dislocations. The response of materials to external stresses can change during processing. They may go dt l d b k dt dp (1.2) from behaving as a brittle material to being completely plastic or from being plastic to behaving as a brittle material. These changes can be due to the processing environment, the change in size, or by fatigue of the material. 1.2.6.3 Rittinger’s and Kick’s laws on particle breakage Two of the most common theories used to model particle breakage are Rittinger’s and Kick’s Laws. Both of these theories are used to predic t the amount of energy or stress required to break a particle. Ri ttinger’s law, Eq (3), is an empirical model that relates energy to the difference of product size particles, x2, less the feed size particles, x1: 1 2 11 1 x x K E (1.3) Rittinger’s law states that as the particle size decreases the energy required to fracture the particle increases. This equa tion has been found to be more suited for fine grinding. If Rittinger’s law were completely true it would be impossible to fracture a nanosize particle, due to the energy require d to fracture the particle (as x2 goes to zero E goes to infinity). Another law relating particle si ze to energy is Kick’s law, Eq. (4): 2 1 3log x x K E (1.4) Kick’s law has been found to work well in describing the energy required for the breakage of coarse particle sizes. Both laws have some capability for predicting particle

PAGE 18

9 m cx x Q exp 1size. But as particle size decr eases the energy required to fr acture the particles goes to infinity that is obviously erroneous. 1.2.6.4 Distribution based models Reid proposed an empirical model that defi nes a specific breakage rate to describe a grinding process.37 The model describes the breakage as a first-order rate process. Broadbent expanded this model and now it is referred to as a population balance model.38 The model uses parameters such as the sele ction and breakage function to describe the mechanisms of failure. The selection func tion represents the fractional number of particles of some size that are broken. The br eakage function represents the rate at which particles in a size class are broken. From this model it is possible to interpret a particle size distribution mathematically. A correla tion between these pa rameters and the mechanisms of particle breakage such as fracture, cleavage, and attrition can be determined. The Weibull distribut ion equation is another model th at uses the particle size distribution data to obtain an empirical representation of particle breakage. Here Q is the cumulative fraction of fragment size x, while m and xc are fitting parameter. The Weibull equation has been used to describe the fracture of brittle materials su ch as soda-lime glass (1.5) spheres.39 1.2.7 Measuring Breakage of Particles The most common technique to measure particle breakage is by a single impact test. Where a particle of some size is either s hot at a target with some velocity or a load is applied to a stati onary particle at some velocit y. The breakage distribution of the particle can then be determined by m easuring the size of the broken particles.25 A

PAGE 19

10 drawback of this method is that the only way to apply a stress is by a normal load. Other methods use indentation techniques to determ ine the effects of stra in on the breakage of particles. However, similar to the impact t echnique, the indentation method also subjects particles to a normal load. Both these met hods have proved invaluable for understanding single particle breakage under a normal loa d. Schnert studied crack propagation and crack behavior in many types of materials.25 He showed that materi al properties, stain rate, impact velocity, particle shape, and particle size effect how particles break in single impact and compression tests. It is known that for real sy stems particles of some size distribution will show significantl y different breakage behavior.23 Particles of small sizes will not break as easily as particles of larger sizes. Also particles of the same material and of the same size still can differ in strength. To account for these disc repancies in particle breakage many researchers have devised techniques to measure multiply particle breakage. These techniques are similar to th e impact test. They are usually performed by either dropping some load onto a bed of par ticles or by propelling a mass of particles at some target. From the particle size distribution data obtained after the te st it is possible to calculate a fractional number of breakages by fitt ing this data with some sort of empirical model (i.e. population balance model). 1.2.8 Environment Environmental effects have been shown to be negligible on the fracture toughness of a material. Most research suggests that environmental effects for a milling process are limited to changes in the slurry viscosity that reduces the energy dissipated as heat through viscous flow.40 However, Griffith’s theory includes a surface energy term that relates to energy required for crack pr opagation. It is known that additives and temperature can change the surface tension of a material. Reducing the surface tension in

PAGE 20

11 Griffith's equation reduces the stress required to propagate a crack. This only holds for particles with an existing crack. It has been suggested that all mate rials have defects in the form of small cracks or micro cracks. Rebinder showed that lowering the surface tension of a material can great ly reduce the fracture toughness of the material and in turn leads to easier breakage.41 Rebinder’s experiments were performed in dry systems where the only additive was water. Rehbinder’s th eory has been challenged by a number of researchers who found that the reduction of surface tension does not play a dominant role in breakage. 1.3 Stirred Media Milling 1.3.1 Media Mills Stirred media mills have been in use for over 70 years, Szegvari invented the first mill in 1928. However, media mills did not become popular until the 1950’s when they were used as disperser in the pigment and paint industry.42 Stirred media mills come in two designs, which are, either a vertical or horizontal mounted vessel. Stirred media mills can be run in either the wet or dr y milling conditions. In many respects stirred media mills are similar to ball mills. The ma jor difference between them lies in the way the energy is supplied to mill and the material Ball mills operate by rotating the complete mill volume; the contents are then broken by the impaction of the media colliding in a cascade mechanism. In a stirred media mill media and materials are placed in the vessel and then agitated. This reduces the amount of energy needing supplied to the mill since only the agitator has to be rotated and not the whole volume of the mill. Another difference is the velocity of the contacting balls or grinding media. In a ball mill the media velocity is limited by the diameter of the mill and the rotational rate of mill

PAGE 21

12 resulting in low media velocities. In a stirre d media mill the velocities are only limited by the rate of rotation of the stirrer. Stirred media mills are also known as attr itors or high-energy mills. An attritor is stirred media mill run at high solids loadi ng where the milling takes place through a shear mechanism. A high-energy mill is where the mill is run at very high rpms or tip speeds in excess of 10 meters per second. It has been es timated that there are over 44 parameters that can affect a stirred media mills performance.17 1.3.2 Grinding Aids The previous section mentioned how envi ronmental affects the strength of a material. This effect is also present in a grinding system. Many researches have investigated the effects of grinding aids on the efficiency of a milling process.40, 43-46 The majority of work has been focused on dry ba ll milling processes. It has been proposed that in these processes that a ddition of grinding aids acts as a lubricant, which reduces the viscosity of the system thus increasing breakag e of the particles. In other work grinding aids were proposed to act as dispersants that prevent particles from agglomerating which can lead to more effective contact between media and materials. Some researchers suggest that at high concentra tions the grinding aids can a dd a protective coating to the particle and decrease the milling efficiency.45, 47 Other effects include the Rehbinder effect discussed previously where the reduction in surface energy can decrease the stress required to fracture a material. This effect would be more pronounced in a nano-milling system where the number of def ects per particle is very low. In this study we milled at low solids loading allowing us to ignore the effect of surfactant on slurry viscosity.

PAGE 22

13 1.3.3 Media Loading Arguably the most important parameter in a stirred media milling process is media loading. Media loading can affect a milling pr ocess in a number of ways. The higher the media loading the larger the number of cont acts which lead to a higher milling rate.18, 19 Also the higher the media loading the highe r the viscosity and the larger the energy needed to operate the mill.48 It is apparent that an optim um must be found to determine the extent to which a mill should be loaded. However, media loading is not limited to just these two effects. Media loading also effects the contamination or wear of grinding media and mill. It affects breakage mechanism taking place, and the flow field inside the mill. Hashi and other extensively studied the flow fi eld as well as wear of grinding media with respect to media loading and found optim um media loadings for specific milling conditions.13, 49 The effects of media loading beco me even more complicated when coupled with rotational rate, media size and media density. Media density and rotational rate can easily change any optimum establishe d for efficiency with respect to energy. We varied media loading at a constant media de nsity; media size and rotational rate in order to study just the effect of media loading in the context of number of vary the number of contacts and not the momentum of the media. 1.3.4 Media Size The final parameter studied was media size and its effect on particle size reduction and rate of particle size reduc tion. It has often been stated to mill to fine sizes small media is needed.17, 18, 50, 51 It is easy to imagine that as media size decreases the number of contacts increase. However, as the size of the media decreases the maximum stress that can be imparted on the particle also decrea ses. An optimum must be present for media size similar to the optimum for media load ing. The optimum for media loading should

PAGE 23

14 change with particle size. This change is due to change of the strength of the material with particle size and the necessary stress to fr acture the material. It is important to note that the momentum of the media is related to the size and mass of the media times the velocity of the media. Therefore, media of larger size has a larger momentum and is capable of imparting more energy on a particle. 1.3.5 Other Milling Parameters It is important to note that in addition to the parameters discussed above there are other parameters that can infl uence a grinding process. Stirre r speed directly affects the amount of energy consumed in the process as well as the maximum velocity that the grinding media can achieve, and thus increa sing the momentum of grinding media. Media density also directly affects the momentum of the media and the force that can be imparted on a particle. Kwade formulated a simple equation to mathematically relates media size, media density, and rotati onal rate to the stress intensity.48, 52 Stress intensity (SI) is directly proportional to circumferential velocity v, media size dp, and media density m less the fluid density f: 2 2 f p pd SI (1.6) This expression is limited in that it only describes the influence of media size, media density and rotational rate on the milling process and it is not n ecessary true to all case. In our study media density and fluid density ar e almost the same, which would mean SI would equal zero. Inhomogeneous regions in th e mill may also be present for large dense medias at low rotational rates.13, 49, 53 For the most part over years media hardness have only been looked at as how it affects media wear. Most conventiona l grinding media has a very high hardness (high density zirconia ar ound 9 on Mohr’s scale). It has been found

PAGE 24

15 that the harder the media the less the wear that takes place.54 Many industries have found that by increasing the media hardness the can save significantly on the life of the media. However, in a nanomilling process where contamination from media wear may significantly affect the product material even a small amount of contamination could be bad.55 In the present study polymeric media to mill particles to nanosize, and this is believed to be possible because of the softne ss of the pigment material. The softness of the media may be so high that it prevents it from wearing away the mill and thus leading to no contamination of the product. This is a contradiction to what the researchers above have found for media’s hardness relationship to wear; however, in this case the polymeric media is an order of magnitude less that that of the softest media studied. A secondary effect may arise when the media comes in contact with one another. 1.4 Milling Kinetics Milling kinetics is the measure of the time it takes to break a particle to a specific size. Many researchers have investigated breakage rates in a milling process. Vogel has constructed a master curve for kinetics of different materials.56 Researchers have studied the effect of environment on breakage rate, the effect of milling parameters and the effect of material56-61 This study will discuss kinetic on two levels. The first will be macroscopic representation of the grinding ra te determined by the change in the mean particle size with time.62 This representation will show th e nonlinearity of grinding over long time periods. This nonlinearity has been observed by several researchers and has been explained by the accumulation of fine pa rticles in the milling zone which reduces the amount of energy imparted on part icles as the milling time increases.57, 63-67 The second approach to breakage kinetics will be with a popul ation balance model (PBM). Population balance models have been used to describe both the breakage rate and the

PAGE 25

16 breakage distribution in a milling process.57, 63, 64, 68, 69 The breakage distribution obtained from the PBM will be then used to evalua te possible particle breakage mechanisms taking place in the mill at specific times. Th ese mechanisms will then be used to determine the dominant milling mechanism in the process. From this information it is then possible to optimize the mill through the complete understanding of particle breakage, milling parameters, and grinding ra te, which can give significant predictability of product quality and the energy need to produce that product.

PAGE 26

17 CHAPTER 2 EXPERIMENTAL PROCEDURE 2.1 Equipment This study employed the use of the Micr os Superfine Grinder manufactured by NARA Machinery Corp. of Japan. The Micros mill is an attrition mill designed for ultrafine grinding of particles to the micron and submicron size range. The mill was originally designed with zircon ia rings mounted on six equally distant vertical columns to grind the material against the inner zirconia wall of the mill vessel. The volume of the vessel is 500 mL and a cooling jacket surrounds it to maintain a desired temperature. In this study, the zirconia rings were removed, and the vertical columns we re used to agitate the milling media. Thus, the mill was modified to operate as a vertical, batch stirred media mill. The rotation rate was set to the maximum possible, which is 2000 rpm in order to facilitate the most rapid and en ergetic grinding conditions possible. This corresponds to an equivalent tangential speed of 11 m s-1. The effects of the stirrer speed were not investigated in this study. Eastman Kodak provided the patented polym eric grinding media used in this study. The Kodak media patent discusses a method fo r manufacturing large quantities of vary narrow size distribut ion polymeric media.70, 71 Kodak provided several media sizes to be used in this study. The grinding media sizes were measured by a laser scattering instrument (Coulter LS230). The volume-mean diameter and standard deviation of each fraction used in the experiments are pres ented in Table 2.1 along with other operating conditions of the milling process. The optical microscope image of the beads (Fig. 2.1)

PAGE 27

18 illustrates the sphericity of the beads. The gri nding media had a density of approximately 1 g/cm3. 2.2 Materials The material used in this study is Ma genta 122 which is a quinacridone organic pigment used in inkjet ink formulations Magenta 122 is a water-insoluble, semicrystalline pigment; with a molecular formula of C22H16N2O2 it has a structure that resembles two naphthalene molecules joined by a cyclohexane molecule. The pigment used was manufactured by Sun Chemical Company and has a density of 1.24 g cm-3. Magenta 122 is hydrophobic and has a contact angle of approximately 1540 as measured by a goniometer. Table 2.1 Operating variables of the stirred media milling process.72 Run ID Surfactant concentration /Molar Mass of polymeric grinding media/ grams Mean size and standard deviation of the beads / m M100 0.037 M101 120 40.3, +/8.7 M102 0.02 120 40.3, +/8.7 M103 0.037 120 40.3, +/8.7 M104 0.0937 120 40.3, +/8.7 M201 0.037 60 40.3, +/8.7 M202 0.037 200 40.3, +/8.7 M301 0.037 120 9.0, +/6.0 M302 0.037 120 21.5, +/14.2 M303 0.037 120 402.0, +/173.0

PAGE 28

19 Figure. 2.1 Optical microscope image of the polystyrene beads (mean size, 40.3 m). Marker size is 50 m.72 The dry pigment particles consist of primary semi-crystalline particles, agglomerates, and aggregates. The SEM images (Figs. 2.2(a) and the TEM image (Fig. 2.2(b)) illustrate various structures of th e pigment particles including rod-like primary particles as well as agglomerates. Several primar y particles seem to cluster or fuse at their faces, which is possibly due to growth a nd sintering of individual crystals during manufacturing via precipitation.10, 11 Although primary particles of about 50 nm width and up to few hundred nanometers in length ca n be identified from Fig. 2.2, the initial particle size distribution is entirely dependent on the method of dispersion and deagglomeration. The use of different susp ension liquids and different dispersion techniques results in vastly different initial particle size distributions.73 The Kodak patent for the production of the nanosized pigment us es the anionic surfactant sodium N-methylN-oleyl taurate (OMT) for dispersion a nd stabilization of the pigment during the process.1, 2 In this study we chose to focus on surfactant concentrations and not on surfactant types so we employed the same surfactant Kodak in their patents.

PAGE 29

20 Figure 2.2 Images of the pigment particles: (a) SEM image under 50,000 magnifications; and (b) TEM image. Ma rker size is 100 nm for (a) and 500 nm for (b).72 The surfactant is a long carbon chain surf actant with a molecular weight of 454 g/mol and a critical micelle concentra tion (CMC) of 0.037 M as measured by the Wilhelmy Plate method at 250 C. (a) (b) 500 nm

PAGE 30

21 2.3 Procedure Surfactant was mixed with nanopure water to make 250 mL solutions at the concentrations listed in table 2.1. The pigment particles were suspended in the solution to prepare a slurry with a fixed volumetric pigm ent loading of 1.27%. This relatively dilute loading was chosen in order to suppress th e tendency to aggregation, and in order to study the breakage kinetics. A dilute suspen sion also limits the effect of increasing viscosity as particle size d ecreases, allowing us to study the effects of surfactant on dispersion and not as grinding aid that reduces the viscosity of the system. The milling vessel was filled with the beads, to the masses corresponding in Table 2.1. The remaining volume was filled with the 250 mL, 1.27% pigm ent slurry, which was poured into the vessel. The concentration of the milling media in the mill is an important in determining the breakage mechanism and kinetics of a milling process. The mass of grinding media was varied in the range 0~200 grams, where th e bead-to-bead collisions are expected to be the dominant stressing mechanism.74 Keeping all other conditi ons the same, the higher the bead loading; the higher the specific energy expenditure will be.48 It is important to note that Kwade’s calculations, chapter 1 e q. 1.6, for specific energy with respect to solids loading are for dense grinding media, which may not be correct for media with a density close to that of the surrounding flui d. Eskin’s approach for calculating energy and stress are more likely to be correct bei ng that they take in to consideration the hydrodynamics of the system.74 More importantly, the fric tion and associated heating may lead to excessive wear and softening of the beads. The torque and specific energy consumption were not measured in this series of experiments.

PAGE 31

22 2.4 Sampling and Characterization Particle size measurements have many connot ations that can be described by size, shape, volume, surface area, length, number, and any other definition which is usually process specific.75 We chose to represent the particle size distribution by the volume of a particle so we could use the mass balance e quation incorporated in the PBM model for measuring particle breakage. The particle si ze distribution of the pigment was monitored with time; measurements were performed usi ng the Microtrac ultrafine particle analyzer (UPA) 150 instrument, which is capable of meas uring particle sizes in the range of 3.2 nm to 6.5 m. The UPA measures particle throu gh the technique of photon correlation spectroscopy (PCS) or dynamic light scatteri ng. Particles in the na nosize range undergo a large amount of movement due to Browni an motion. The Brownian motion causes interference in the scattered light, which result s in a fluctuation in the recorded intensity with time. The fluctuations are then fit to a correlation function that equates the fluctuations to particle size. Filella revi ewed the PCS techniques used by different instruments and discussed some of the problems associated with them.76 One of the largest problems associated with a PCS measurement technique is colloidal suspension concentration, which can result in agglomera tion of particles and multiple scattering of the signal. By representing a particle size on a volume basis there is a weighting of the distribution towards the large h eavy particle sizes. Even one large particle can lead to a significant amount of error in th e particle size measurement. Our data shows that after 10 min of milling all the pigment particles are reduced to a size of less then 1 m. To assure that the samples are free from particles of above 2 m, which would lead to significant error in the measurements, the samples are removed and filtered using a 2 m syringe

PAGE 32

23 filter. Freud measured particle size distributions of nanosize particles at varying concentrations on the UPA and found that th e error in particle size is linear for concentrations above 0.1%.77 The pigment concentration in the mill is approximately 1.27%, therefore, after filtration the samples ar e diluted to approximately 0.1% using the appropriate concentration of surfactant soluti on to eliminate any instability by diluting in water that could lead to aggl omeration during the measurement. To be assured that no agglomeration was taking place during the measurement, an experiment was run where samples were measured at sampling times 30 sec., 1 min, 3 min, 5 min, 9 min, 15 min, 30 min, and 1 hr. This experiment showed there was no variation in the samples with time. Approximately 1 ml of sample was removed from the mill after 10 minutes and then every 4 hours after that to reduce the reduction in mill volume with time due to sampling. For the PBM study, samples were taken every 2hours to increase th e number of data points and reduce the er ror in the fit. In addition to monitoring the time variati on of the particle size distribution, the size and shape change of the pigments before and after milling was also monitored using scanning electron microscopy (SEM), which was also used to check whether or not the beads had been damaged during the milling process. The diluted and dried samples were ion beam coated prior to examination w ith the SEM (JEOL JSM6330F). The elemental composition of the samples imaged via transmission electron microscopy (TEM) was determined using energy dissipative X-ray spec troscopy (EDS). Elemental analysis of the product slurry was performed using induc tively coupled plasma spectroscopy (ICP). SEM, TEM, EDS, and ICP measurements were used to check if the pigment slurry had become contaminated with zirconia or stainless steel from the mill. Gas chromatography

PAGE 33

24 (GC) was used to measure the contamination from the polymeric media; this method lacked the mass spectroscopy capability so we were only able to determine that the samples were qualitatively free from polym eric media, which agreed with the SEM results.

PAGE 34

25 CHAPTER 3 EXPERIMENTAL RESULTS AND DISCUSION 3.1 Base Experiments An experiment was run to determine the e ffect of dispersion on the final pigment particle size without the presence of milling media. As mentioned previously different dispersion methods revealed largely varying particle size distribu tions. Experiment M100 was carried out by dispersing the pigment unde r the milling conditions listed in table 2.1. The results of the particle size obtained from this run can be seen in fig. 3.3. The particle size distribution is bi-modal and consists of a minor coarse fraction in the size range 1.6~4.6 m and a major fraction in the size rang e 80~800 nm. The particle size evolved to a stable state that indicates dispersion to primary particles and small fraction of hard agglomerates and coarse particles was achieve d. It has been show n in literature that breakage of the soft agglomerates is a prer equisite for wetting and subsequent adsorption of dispersants and stabilizati on of the particle suspension.78 The intent of this experiment was to provide and estimate of the initial pa rticle size distribution, which is important in formulating the population balance model (p rovides a basis for the mass balance equation). Another base case experiment, M103, was carried to determine the effectiveness of the polymeric media on the rate of particle size reduction and on th e final particle size obtained. A fit of the population balance model to the base case experiment is presented in chapter 4. SEM images were taken of samples collected from M103 after 10 minutes, 16 hours and 24 hours of mill can be seen in Fi g. 3.1, which shows that a fast dispersion

PAGE 35

26 of the soft agglomerates takes place, forming rod-like and irregular pigment particles with sizes ranging from 40 to 400 nm. These pa rticles are a mixture of primary particles and hard agglomerates. After 16 hours, more rounded particles in the size range 20~200 nm were observed. Spheroid nanoparticles w ith sizes 10~60 nm can be observed after 24 hours. Figure 3.1 SEM images of pigment particles in Run M103 samples collected at various milling times: (a) after 10 min; (b) after 16 h; (c) after 24 h.72 SEM images suggest a transition from irregular or rod-lik e aggregates and agglomerates to rounded nanoparticles. The EDS was performed on these particles to prove that they were not composed of any i norganic contaminants th at may be present in the mill. SEM images of the indicated that surface of the beads were neither chipped nor abraded. (a) (b) (c)

PAGE 36

27 3.2 Effects of Grinding Aid Concentration Experiments M101, M102, M103 and M104 we re conducted to study the effects of surfactants on the rate of pa rticle size reduction as well as the effectiveness of particle size reduction. Fig 3.2 illustrate s the particle size distribut ion (PSD) of samples from these experiments at the indicated sampling tim es. One way a grinding aid can affect a Figure 3.2 A, B, C, and D show the eff ects of surfactant co ncentration (0.0, 0.02, 0.037, and 0.0937M) on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively. milling process is by acting as a dispersant where particle are either held apart by electrostatic repulsion or thr ough steric repulsion. The agglomer ates must first be broken apart to allow the surfactant to adsorb on the surface and stabilize the particles. The PSD shifts towards coarser size with time in the absence of a disp ersant, indicating the Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.02 M OMT 0.037 M OMT 0.0937 M OMT A)B) D) C) Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.0 M OMT 0.02 M OMT 0.037 M OMT 0.0937 M OMT Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 0.02 M OMT 0.037 M OMT 0.0937 M OMT A)B) D) C)

PAGE 37

28 aggregation of particles. Even for the lo w pigment concentration under consideration (1.27 vol.%), collisions of the pigment partic les due to turbulent agitation and Brownian motion led to aggregation when no surfactant is present. Another effect of the grinding aid is to reduce the pulp viscosity, thus allowing higher solids loading in wet milling.43 However, depending on the size and material properties of the solid to be ground, dispersants may have little effect on the milli ng efficiency and dynamics when the initial feed is very dilute.44, 45 In the current work, at low pigm ent concentration, the stabilizing action of the surfactant is th e most important as nanopartic les have a high tendency to aggregate. When sufficient surfactant is added (0.02, 0.037 and 0.0937 M), the PSD is observed to move monotonically toward finer sizes. The PSDs for surfactant concentrations of 0.02 and 0.037 M cases are si milar within the expe rimental accuracy (Appendix A); almost all of th e particles are in the nano-si ze range 9~100 nm after 24 hours of milling. However, excessive use of the surfactant (0.0937 M) reduced the breakage rate. These experiments indicate an optimum surfactant concentration for this pigment loading in order to achieve finer particles within a s horter milling period. 3.3 Effects of Media Concentraion With an increase in the concentration of the polymeric media (Runs M201, M103 and M202), the formation of nanoparticles is ach ieved faster (Fig. 3.3). It is important to note that at zero media concentration (M100) the particle size does not deviate from the primary particle size of approximately 150 nm over the entire milling time. It should also

PAGE 38

29 Figure 3.3 A, B, C, and D show the effects of media concentration (0, 60, 120, and 200 g) on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively. be observed that after 10 minutes of milling a ll the concentrations have almost the same particle size distribution, whic h supports our initial finding th at we rapidly disperse the agglomerated pigment particles to there primary particle size. With an increase in the concentration of the media th e production of nanoparticles is achieved faster. This is intuitively expected because the number of bead–bead collisions and consequently the rate of stressing of the pigmen t increase as the bead concentration increase. It is difficult to predict from these results whether it is the number of increasing bead-bead collisions that result in the increased ra te of grinding or if a change from an impaction mechanism of grinding to a shear mechanism of grindi ng is what led to th e higher grinding rate. Experiments at higher media loadings wh ere shear would obviously have been the Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 no media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g A) C)D) B) Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 no media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 no media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 No media 60 g 120 g 200 g A) C)D) B)

PAGE 39

30 grinding mechanism taking place failed due to the large pressure build up in the mill at high rotational rates. From a fracture mechanics perspective a particle undergoing a shear mechanism of stress is more likely to result in a cl eavage or fracture mechanism of breakage as opposed to an attrition mechanism. An attrit ion mechanism of breakage is most likely to result from an impaction mechanism of stress.27 The distributions above suggest that a cleavage or fracture mechanism of breakage is taking place, which would mean shear is the dominant mechanism of breakage. As stated in the first chapter cleavage mechanism is mostly to take place for ductile mate rials and a fracture mechanism for brittle materials. Considering that pigment is a soft organic material it would be safe to assume that a cleavage is the most likely mechanism of breakage. It is important to note that these conditions of failure a nd loading are valid for macrosc opic brittle material and may not always be true for nanosized materials, which have proven to show significantly different material properties, wh ich may change during processing.30, 32, 33 3.4 Effects of Media Size Experiments M301, M302, M103 and M303 were pe rformed in order to investigate the effects of the size of the grinding media. In Fig. 3.4, it is seen that this is a vital factor. It is apparent from these figures that media size can have a significant effect on the rate of milling as well as the effectiveness of milling. The 9 micron media is unable to reduce

PAGE 40

31 Figure 3.4 A, B, C, and D show the effects of media size (9, 21, 40, and 402 m) on particle size at grinding ti mes of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively the particles to less then the primary particle size due to th e insufficient amount of force exerted by the media. The 21 and 40 and to some extent 402 microns are very effective in reducing the particles. These observations coin cide with those of other workers regarding the existence of an optimum media size.17, 48, 51 For a given beads loading and stirrer speed, large beads have more momentum due to their mass than small ones. Therefore, it can be suggested that if grinding we re occurring purely by impaction grinding mechanism then after some extended grinding ti me reach a smaller particle size would be achieved. On the other hand, small beads experi ence more collisions than the large ones due to their relatively high number. This shoul d result in a much more rapid rate of particle size reduction for smaller grinding me dia due to the much higher numbers. The Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um A) D) B) C) Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000 Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um Particle Size (nm) 101001000Cumulative Fraction Undersized 0.0 0.2 0.4 0.6 0.8 1.0 9 um 21 um 40 um 402 um A) D) B) C)

PAGE 41

32 effects of momentum and number of contacts can both be seen in the figure above. For example, the 21 micron media decreases the pa rticle size much more rapidly then the 40 or 402 um grinding media. However as grinding time increase the 40 micron media yields a slightly smaller part icles then 21 micron media due to the large amount of force exerted by the 40 micron media. The existence of the larger particle size distribution for the 402 micron media at a grinding time of 24 hrs is due to the fact that the 402 micron media did not have sufficient time to mill the particles to or below the size obtained by the 21 or 40 micron media. 3.5 Breakage Mechanisms In the first chapter possibl e breakage mechanism were discussed these mechanism are fracture, cleavage, and at trition. Many authors have disc ussed the resultant PSD from different breakage mechanism, as illustrated in chapter 1, fig 1.1. 26, 79 From the distributions above it is obvious that an attrition mechanism is not taking place. If this were the case a small fraction of 10’s nanomet er size particles would be present at the short grinding times. It is not possible to distinguish between a cleavage and a fracture mechanism from the present data. However in the next chapter the data from the experiments is fitted to a population ba lance model (PBM) and it is possible to quantitatively prove that a cleavage mechanis m is clearly the mech anism of breakage for the base case experiment. 3.6 Grinding Mechanism Grinding mechanisms are difficult to predict in a stirred media milling process. It has been shown in this chapter that media lo ading has a significant effect on the grinding mechanism. It was also shown that certain mechanisms are more likely depending on the response of the particle size distribution to th e size of the media or change in momentum

PAGE 42

33 (stress intensity 48). Considering that a stirred media milling process is an extremely complex process it is most likely that more then one grinding mechanism is responsible for the size reduction in media milling. 3.7 Grinding Kinetics Grinding kinetics is more simply referred to as the rate of part icle size reduction. It is possible to discuss grinding kinetics on two di fferent levels. The first level is the time it takes to reduce a particle to a specific size. Figure 3.5 Plot of change in particle size with time for varying medi a concentrations that describes the grinding rate. The second is the rate at which particles of a specific size class are ground to a subsequently smaller size class. The first level is a macroscopic approach to grinding kinetics and can simply be de scribed by a plot of d50 vs. time where the slope is the rate of grinding. Fig 3.5 is a plot of d50 vs. time for varying media concentration. At grinding Time (minutes) 02004006008001000120014001600 D50 (microns) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 60 g 120 g 200 g

PAGE 43

34 times up to 250 minutes the data obeys first or der grinding kinetics, which is similar to result reported in litera ture for grinding kinetics.68 Figure 3.6 Plot of change in particle size with time for varying media sizes that describes the grinding rate. However as grinding time increases the ra te becomes nonlinear. Austin suggested that nonlinear kinetics is due to the cushioning effect of fine particles on the breakage of large particles.64, 65, 68 For the purpose of the present study the nonlinear kinetics are not discussed. It is important to see from fig 3.4 that the rate of grinding at high media concentrations is more rapid then that at lower media concentrations. Fig 3.6 shows the rate of particle size reduc tion vs. time for varying medi a sizes. For experiments M302 and M103 the rate of grinding to the grinding limit is more ra pid then when larger media was used (M303). This result is expected due to the larger number of beads present in M302 and M301.A more quantitative description of grinding kinetics will be presented in Time (minutes) 02004006008001000120014001600 D50 (microns) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 9 um 21 um 40 um 402 um

PAGE 44

35 the following chapter. Population balance mode ling will be used to describe the breakage of particles from one size class to a smaller size class and will also discuss the nonlinear observation in the data.

PAGE 45

36 CHAPTER 4 POPULATION BALANCE MODEL 4.1 Method For over 50 years researchers have modele d grinding and other particulate process through the use a populati on balance model (PBM).63, 64, 68, 80 Bass was the first to derive an integro-differential mass balance equation for a batch grinding system.81 For a grinding process it is generally assumed that only breakage occurs, thus allowing simplifications to be made to the integrodifferential mass balance equation. Due to this assumption the accumulation term (agglomeration) can be left out. The system that we have chosen to model uses a surfactant to prevent particles from reagglomerating.78 Considering that Bass’s equation does not have a simple analytical solution, Reid suggested a finite difference a pproximation to Bass’s equation.37 The fractional weight function is discrete in size x and continuous in time t in the form where ()iqtis a set of n size classes numbered from the coarsest downwards, ijb is a volume fraction of the broken material finer than size j x (breakage distribution function) that is retained on class i, where ij x x and consequently ij () ()i i idqt k qtdt is a fractional rate of breakage (also called a selection function) assumed to be independent of time. 1 1() ()(),1i i ijjjii jdqt bkqtkqtin dt (4.1) This is the basic set of ordinary diffe rential equations that form the population balance model (PBM) of grinding. Generall y, the solution of the system (4.1) is

PAGE 46

37 reproduced after determinat ion in advance (from experimental data on monoclass samples) breakage and selection functions which is becoming problematic for the submicron particles. That is why we use the so-called "inverse formulation" of the problem. Parameters ijb and ik of breakage and selection function are determined from experimental data for particle size distribu tions obtained during the whole experiment. They have to be selected in such a way th at solutions of (4.1) match the experimental data as closely as possible. The best possi ble approximation of the experimental data by curves ()iqt we understand in the least squares se nse is represented in eq. (4.2): 2 expth 1min()()kn eor ikik tiqtqt, (4.2) where exp()ikqt and th()eor ikqt are the experimental and th eoretical points respectively. Besides providing the best possible fit of our solution to the experimental data, the parameters of the system (4.1) have to obey the law of mass conservation. From the structure of the equations in system (4.1) it follows that parameters ijb must satisfy the following constraints: 11n ij ijb (4.3) 0,ijbij (4.4) In addition to (4.3) and (4.4) fractional ra te of breakage is nonnegative as long as we consider comminution ( no agglomeration) process a nd therefore we have extra constraints 0,1,1ikin (4.5) 0,ikin (4.6)

PAGE 47

38 Constraint (4.6) reflects th e fact that there is no size reduction in the finest class number n. Thus, the problem of determining the para meters in system (4.1) is reduced to finding the minimum of function (4.2) subject to the set of constraints (4.3)-(4.6). These sets of equations make up the backbone of a ny PBM. This set of equations is known as a nonlinear model subject to linear constraint s. Being nonlinear it means that no exact solution exist, however, many authors have solved this set of equations for there specific systems and applied there own fitting parameters.57, 61, 65, 66, 69, 82 In this study we will use the populati on balance to describe the breakage mechanism taking place dur ing the milling process.83 By inspection of the bij, the breakage distribution function it is possible to distinguish between an attrition, fracture, or cleavage mechanism. Most PBM used t oday are in the form listed above. The novelty in a PBM solutions exist in the method used to solve and optimize the equations listed above (solving equations of the form 4.2) Instead of using th e least squares fit approximation listed above, eq 4.2, for solv ing eq.4.1, in our st udy we will obtain an approximate solution to equation 4.1 through the use of Kapur’s G-H method.84 Kapur twice applied a Cauchy-Picar d iteration to arrive at the eqs 4.7, 4.8: 2 exp2 0t H t G q t qi i i i (4.7) ij j ij j ij i j i j i j ij i i j i j ij i iB k B k q q G G H q q k G 1 1 1 1 0 0 1 1 0 0, (4.8) The G-H method has been found to give co mparable data to that of the exact solution listed above with only a fraction of the computational as well as mathematical complexity necessary to so lve for the exact solution.85 The G-H method has also been shown to give accurate da ta at long grinding times.86 For most milling experiments it has

PAGE 48

39 been found that only one term is ne eded to describe the process (Hi=0). Due to the extended grinding times used in this study we wi ll use Kapur’s original two term model. The particle size range in this study is divided into nine classes that were arranged in geometric progression of 2, beginning with 344 nm as the lower edge of the largest size class ( i =1).87 Size class 9 (+9~30 nm) represents the smallest class or sink where no particles can be broke n out of that size class.87 For each size class, qi( t ) was generated using the particle size distribution data at pre-selected sampling times. Then, the model (Eq.(4.7)) was fitted to the experimental data to find the values of Gi and Hi and to check the applicability of the first-order breakage kinetics. The model fit was carried out with Matlab 6. 1 subroutine LSQCURVEFIT that utilizes the Marquardt– Levenberg optimization algorithm. 4.2 Experimental Findings by Population Balance Modeling Kapur’s G–H model in Eq. (4.7) was used to fit the experimental data sequentially from size class 1 to class 9. Table 4.1 lists Kapur’s G and H parameters for their respective size classes. A graphi cal representation of the fit to the experimental data can be seen in fig. 4.1. The sum-of-squared resi duals and the standard error of the estimate were found to be 0.1304 and 0.0441. The G–H model explains th e temporal variation of the cumulative volume fraction oversize we ll for the size range +86–578 nm. However, the model deviates from the experimental data for the size range +30–86 nm, as can be inferred from the standard error of the estima tes in Table 4.1. The inverse of the first Kapur coefficient Gi–1 as seen in table 4.1 may be re garded as a characteristic time constant for a given size class. From the char acteristic time constant it can be seen that to mill to finer sizes (G9) the time constant is larger than the time constant at G1.

PAGE 49

40 Figure 4.1 Fit of the Kapur’s G–H model (4.7) to the cumulative volume fraction oversize–time data. Legends indicate the lower edge of each size class i The large difference in the time constants for larger size particles and small size particles agrees well with other theories (Rittinger and Kick) that state milling to finer sizes takes an increasing amount of time/ener gy. For sizes larger than +86 nm the ch aracteristic time

PAGE 50

41 constants are nearly equal this can be due to the fact that agglomerates (+86–578 nm) generally break faster than primary particles (-86 nm), a pproximately the particle size obtained after 10 minutes of milling as s een in the fig. 2.2 in chapter 2. Table 4.1 Parameters of the Kapur’s G–H model for Run M103 Size class in the model Size range of the class (nm) First Kapur coefficient – Gi (h–1) Second Kapur coefficient – Hi 103 (h–2) Standard error of the estimate 102 (-) 1 +344 1.477 10–1 0* 0.535 2 +243–344 1.455 10–1 4.568 0.638 3 +172–243 1.747 10–1 0.855 1.118 4 +122–172 1.590 10–1 0.037 1.557 5 +86–122 1.109 10–1 1.047 2.021 6 +61–86 5.229 10–2 3.179 2.941 7 +43–61 1.735 10–8 5.448 4.205 8 +30–43 1.212 10–13 3.065 9.951 *By definition The delay time in the size range +30–86 nm may be due to fatigue failure of primary nanoparticles. Since the G–H method provides an approximate solution, albeit sufficiently accurate, one may argue that the observed deviations from the experimental data may be due to the approximate nature of the G–H method in addition to its exclusion of the non-first-order effects. The breakage distribution function will be used to discuss particle breakage and particle breakage mechanisms. Table 4.3 is the particle breakage distribution matrix. The breakage distribution matrix, b, details quant itatively all the information regarding the volume transfer among size classes. For example, b32 = 0.159 and b42 = 0.841 mean that 15.9% of the volume broken from particles in size class 2 goes to size class 3 and 84.1% goes to size class 4. It is not ed that b is highly sparse, wh ich precludes massive fracture as a breakage mechanism. If a fracture m echanism were taking place fractional quantities

PAGE 51

42 would be present throughout the table (all values above the 00 diagonal must be 0 due to the initial assumption that no aggl omeration is taking place). Table 4.2 The breakage distribut ion matrix for size class 1 through 9 form top to bottom and left to right (b6,3= 0.958) If an attrition mechanism were the mech anism of breakage, pa rticles would break from size class 1 to one of the finest size class such as 6-8 (par ticle sizes of 344 nm would break to 86-30 nm). Parent particles ma inly break into daughter particles in a few size classes. Thus, cleavage of agglomerates and primary particles appear to be the dominant breakage mechanism. Primary partic les in size classes 6–8 break into daughter particles in adjacent size cl asses. The data above proves quantitatively that a cleavage mechanism is the dominant mechanism of breakage throughout the milling process for the base case experiments. From fracture mechanics it is known that ductile materials will tend to fracture through a cl eavage mechanism and that a cleavage mechanism most commonly occur from a tensil e loading of material whic h would evolve from a shear mechanism of milling. 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 1 0 0 0 0 019 0 0 0 0 0 0 0 0 00 0 1 0 0 958 0 0 .0 0 0 0 0 0 0 00 0 1 023 0 0 0 0 0 0 0 0 0 0 00 0 0 841 0 0 0 0 0 0 0 0 0 00 159 0 0 1 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 00 b

PAGE 52

43 The population balance model also explains how fast particles of a specific size class break, this is known as the breakage ra te or selection functi on. Table 4.3gives the selection function for this experiment. It is difficult to consider the value of the selection function having a physical or real meaning, due to it not being defined. Comparing selection functions of different experiments may not be possible. Table 4.3. Selection function for ex periment M103 population balance model However, comparison of selection functions with in a set of data is possible. The selection functions in Table 4. 3 are all of the same order of magnitude and the values do not vary much. This would indicate that the ra te of breakage is linear with time, however, in chapter 3 it was shown that the rate br eakage in non-linear and Rittinger’s law also states the rate of breakage should decrease exponentially with particle size. It may be possible to correct the population balance model by introducing a tim e-variant selection function, leading to more re liable breakage rates. 1h 0 193 0 184 0 229 0 198 0 238 0 293 0 173 0 152 0ik

PAGE 53

44 CHAPTER 5 SUMMARY, CONCLUSIONS AND FUTURE WORK 5.1 Summary This study’s objectives were to determine the effects of surfact ant concentration, media loading, and media size on the rate a nd fineness of the ground product. Additional objectives of this study were to determine the particle breakage mechanism and the most probable milling mechanism. The experiments performed at varying surf actant concentration showed that at no surfactant concentration the particle si ze increased as milling time increased. The experiments also showed that as surfactant c oncentration increased the rate of particle size reduction decreased. However, increasing surfactant concentration did not appear to have an affect on the fineness of the final produc t. In order to optimize the rate of particle size reduction with respect to surfactant c oncentration it may be necessary to add the surfactant as the milling process progresses in order to match the increase in surface area as the particle size decreases. As media concentration increased the rate of particle size reduction increased; this is due to the increase in bead-bead col lisions. When no media was present the mill dispersed the pigment to the primary particle size of 150 nm. Experiments at high media concentration where a shear mechanism of mill should have been dominant were unable to determine difference between a shea r and impaction mechanism of milling. Varying media size was shown to affect the fineness of product material and the rate of particle size reducti on. It was observed that the sm allest media was only able to

PAGE 54

45 disperse the pigment particles to their primar y particle size and was not able to reduce the particle size. This was due to the media not having sufficient force to cause the particles to break. Experiment with the largest media size, showed that the rate of particle size reduction was reduced. This was attributed to the reduction in the number of grinding media and the number of bead-bead collisions due to the increase in grinding media size. The final particle size obtained for the largest media was less after 24 hours due to the slow rate of particle size reduction, however, if milling time was increased this experiment should give a finer particle size due to the larger amount of force exerted by the larger grinding media. The rate of partic le size reduction was the most rapid for the for the 21 micron media due to the fact that it had the highest number of bead-bead collisions with sufficient force to cause the particles to break. Th e final particle size obtained for the 21 and 40 micron media were approximately the same showing that the force that the media exerted on the pigment particles was equivalent. Particle breakage mechanisms were dete rmined through inspection of the graphical particle size distribution data and interpre tation of the breakage distribution function obtained from the population balance model. The PSD data from the base case experiment proved that an at trition mechanism of breakag e was not taking place. From the PBM breakage distribution function we were able to quantitatively determine into what size class a mother particle broke into. From this we were able to show that cleavage of particles was the dominant mech anism of breakage. The PBM also showed that the rate of pa rticle size reduction became nonlinea r at extended grinding time. 5.2 Conclusions This study proved that we were capable of reducing agglomerate pigment particles of greater that 344 nm in size to particles of less than 10 nm in a stirred media milling

PAGE 55

46 process using polymeric media. We have also shown that it is possible to determine particle breakage mechanism and particle br eakage rates through the use of PSD data and a PBM. We were unable to differentiate between a shear and impaction milling mechanism due to a lack of experimental data. 5.3 Future Work The above work proved that it possible to show which particle breakage mechanisms are dominant for the milling of pi gment particles. In our current study we were not able to show any changes in the particle breakage mechanisms with varying mill conditions. Future work on this topic woul d include milling of other materials and determine if the breakage mechanism is materi al dependent or an artifact of the milling experiments. It is also im portant to apply the PBM to allow the experimental data collected to be able to quan titatively determine the differe nces in breakage mechanisms and breakage rates of different materi als and different milling conditions. This study assumed that the material prope rties of the pigment stated the same throughout the entire experiment and the lim it in the particle size achieved was due to insufficient energy being applie d to the particles to cause breakage. To better understand this limit it would be advantageous to m easure the hardness and elasticity of the ground material before and after the milling experiments. Gerberich, who measured the strength of a 50 nm silica particle, showed the feasibil ity of measuring the ha rdness and elasticity of nanosize materials.33 One of the principle objectives of this study was to determine the milling mechanism. However, we were unable to prove which mechanism was dominant due to a lack of experimental data. Future work woul d include a statistical design of experiments based of media loading, media density, ro tational rate, time, and media sizes, with

PAGE 56

47 response variables of rate of pa rticle size reduction and final particle size. This type of study would allow us to focus on the variables that most effect th e energy exerts on a particle and the milling regime present. By using the statistical design we would reduce the number of experiments needing to be pe rformed, and obtain a correlation to the response we are measuring. By measuring the material propertie s discussed above and performing this study we would also be able to determine what parameters (whether material or process) affects the grinding limit. The current study determined the effect of surfactant on the disp ersion and rate of particle size reduction. We ignor ed in the present study any e ffect that the surfactant may have on the viscosity of the system and the a ny effect it may have in facilitating particle breakage. Future work should emphasis understa nding the effect of su rfactants on particle breakage as well as monitoring the viscosity of the system as a function of surfactant concentration and solids loading. At high solid s loading the viscosity of the system may changes as the agglomerated particles are disp ersed this could be s een as a kind of shear thinning effect in this type of suspension. An understandi ng of surfactant concentration on particle breakage may be obtained by measur ing the change in surface tension of the particles with respect to surf actant type (anionic, nonionic, cationic, l ong or short chain). Performing experiments and measur ing the breakage rate and pa rticle size with respect to surfactant type would determ ine the surfactants effect on particle breakage.

PAGE 57

48 APPENDIX A REPEATABILITY RUNS OF M103 Figure A.1 Plot of the volume fraction ve rsus particles size of M103 including the standard deviations at varying run times. Figure A.1 gives the standard deviation in volume fraction of at specific particle sizes for the sum of three runs at the base cas e conditions listed in chapter 2 Table 2.1. It is important to remember that the particle size axis is in log scale so the volume change at lower particle size is much more dramatic wh ich is why such a larg e amount of deviation exist at the lower particle sizes. 0 5 10 15 20 25 30 0.010.11Dp (um)Volume Fraction Ave 10min Ave 490min Ave 730min Ave 1210min

PAGE 58

49 APPENDIX B CALCULATIONS OF CMC Fig. B.1 gives the surface te nsion of surfactant OMT with respect to concentration of OMT. The intersection of a tangent through the rapidl y decreasing slope at low concentration and a tangent through the zero slope at high concentrations is the CMC concentration of the surfactant. All measurements were performed using a Wilhelmy plat technique at 250 C. Figure B.1 Measure of surface tension ve rsus surfactant concentration for OMT surfactant. 37.00 38.00 39.00 40.00 41.00 42.00 43.00 00.020.040.060.080.10.120.14 Concentration (M)Tension (dyne/cm^2) Surface Tension

PAGE 59

50 LIST OF REFERENCES 1. Czekai A.D., Seaman L.P., inventor Eastman Kodak Co., assignee. Comminution with Small Particle Milling Media. U. S. Pat. 5500331 patent U.S. Pat. 5500331. 1996. 2. Czekai A.D., Seaman L.P., Smith, D.E., inventor Eastman Kodak Co., assignee. Process for Milling and Media. U.S. Pat. 5662279 patent U.S. Pat. 5662279. 1997. 3. Merisko-Liversidge E, Liversidge G.G., Cooper E.R. Nanosizing: A Formulation Approach for Poorly-Water-Solubl e Compounds. European Journal of Pharmaceutical Sciences 2003;18(2):113-120. 4. Lee J. Drug Nanoand Microparticles Processed into Solid Dosage Forms: Physical Properties. Journal of Phar maceutical Sciences 2003;92(10):2057-2068. 5. Liversidge G.G., Cundy K.C., Particle-Size Reduction for Improvement of Oral Bioavailability of Hydrophobic Drugs .1. Absolute Oral Bioavailability of Nanocrystalline Danazol in Beagle Dogs. International Journal of Pharmaceutics 1995;125(1):91-97. 6. Mendel J.B., Particle Generation and Ink Particle Size Effects in Pigmented Inkjet Inks part 1. J. Nanopart. Res., 1999;1:419. 7. Mendel J.B., Bermel, A.D., Particle Ge neration and Ink Particle Size Effects in Pigmented Inkjet Inks part 2. J. Nanopart. Res. 1999;1:421-424. 8. Hyeon T., Chemical Synthesis of Magne tic Nanoparticles. Th e Royal Society of Chemistry 2002:927-934. 9. Zhu W.P., Pratsinis S.E., Flame S ynthesis of Nano-size Powders. In ACS Symposium series 622, Nanotechnology. Washington: American Chemical Society 1996. 10. Mather R.R. The Degree of Crystal A ggregation in Organic Pigments. Dyes and Pigments 1999;42(1):103-106. 11. Mather R.R., Evaluation by Nitrogen Adsorption of Crystal Aggregation in Organic Pigments. Dyes and Pigments 2000;47(1-2):17-21. 12. Kammerler H.M., Pratsinis, S.E., Flam e Synthesis of Nanoparticles. Chemical Engineering Technology 2001;24(6):583-596.

PAGE 60

51 13. Hashi Y.S., Grinding Efficiency of an Agitation-Beads Mill (1). Micromeretic 1990:732-737. Translated by N. Nagouchi 2005 14. Mende S., Stenger F., Peukert W., Schwedes J., Mechanical Production and Stabilization of Submicron Particles in Stirred Media Mills. Powder Technology 2003;132(1):64-73. 15. Mende S., Stenger F., Peukert W., Schwedes J., Production of Sub-micron Particles by Wet Comminution in Stirre d Media Mills. Journal of Materials Science 2004;39(16-17):5223-5226. 16. Tanaka T., Optimum Design for Fine and Ultrafine Grinding Mechanisms Using Grinding Media. KONA 1995;13:19-24. 17. Jankovic A., Variables Affecting the Fi ne Grinding of Minerals Using Stirred Mills. Minerals Engine ering 2003;16(4):337-345. 18. Kwade A.,Schwedes J., Wet Commi nution in Strirred Media Mills. KONA 1997;52:91-101. 19. Kwade A., Wet Comminution in Stirred Media Mills Research and its Practical Application. Powder T echnology 1999;105(1-3):14-20. 20. Salman AD, Gorham DA. The Fracture of Glass Spheres. Powder Technology 2000;107(1-2):179-185. 21. Shipway P.H., Hutchings I.M., Attriti on of Brittle Spheres by Fracture under Compression and Impact Loading. Powder Technology 1993;76(1):23-30. 22. Ghadiri M, Zhang Z., Impact Attrition of Particulate Solids. Part 1: A Theoretical Model of Chipping. Chemical Engi neering Science 2002;57(17):3659-3669. 23. Bemrose C.R., Bridgwater J., A Review of Attrition and Attr ition Test Methods. Powder Technology 1987;49(2):97-126. 24. Zhang Z., Ghadiri M., Impact Attrition of Particulate Solids. Part 2: Experimental Work. Chemical Engineeri ng Science 2002;57(17):3671-3686. 25. Prasher C.L., Crushing and Gri nding Process Handbook: wiley; 1987. 26. Rumpf H., Particle Technology: Chapma n and Hall; 1990. Translated by A. Bull 27. Schnert K.U., H. Kleman, In: Proceedi ngs of the 2nd Interna tional Conference of Fracture; 1969; Brighton, England; 1969. p. 474-482. 28. Cho H.C., Austin L.G., An Equation fo r the Breakage of Particles under Impact. Powder Technology 2003;132(2-3):161-166.

PAGE 61

52 29. Schonert K., About the Particle Breaka ge and the Possibities and Limits for Improving Size-Reduction Processes. In : The International Symposium on Powder Technology; 1981; Kyoto, Japan: Powder Technology; 1981. 30. Kendall K., Impossibility of Comminuting Small Particles by Compression. Nature 1978;272(5655):710-711. 31. Weertman J.R., MRS Bulletin Physic Society B 1955:68. 32. Gerberich W.W., Dauskardt R.H ., Fracture Nanomechanics Preface. International Journal of Fracture 2003;119(4-2) 33. Gerberich W.W., Mook W.M., Perrey C.R ., Carter C.B., Baskes M.I., Mukherjee R., Superhard Silicon Nanospheres. Journal of the Mechanics and Physics of Solids 2003;51(6):979-992. 34. Griffith A.A., The Phenomena of Ru pture and Flow in Solids. Philosophical Transactions of the Roya l Society A 1921;221:163-198. 35. Roesler F.C., Brittle Fractures near Equilibrium. Proceedings of the Physical Society of London Section B 1956;69(10):981 36. Roesler F.C., Indentation Hardness of Glass as an Energy Scaling Law. Proceedings of the Physical Societ y of London Section B 1956;69(1):55-60. 37. Reid K.J., A Solution to Batch Grindi ng Equation. Chemical Engineering Science 1965;20(11):953 38. Broadbent R.A., Callcott T.G., Coal Breakage Processes: 2. A Matrix Representation of Breakage. Journal of Institute of Fuel 1956;29:528-539. 39. Cheong Y.S., Salman A.D., Hounslow M.J ., Effect of Impact Angle and Velocity on the Fragment Size Distribution of Glass Spheres. Powder Technology 2003;138(2-3):189-200. 40. Klimpel R.R., Austin L.G., Chemical Additives for Wet Grinding of Minerals. Powder Technology 1982;31(2):239-253. 41. Rebinder P.A., Shchunkin E.D., Surface Ph enomena in Solids During the Course of their Deformation and Failure. Soviet Physics Us pekhi 1973:533-669. 42. Hochberg S., inventor Process For Dispersing Pigments In Film-Forming Materials. USA patent 2581414. 1948. 43. Klimpel R.R., The Selection of Wet Gr inding Cemical Additives Based on Slurry Rheology Control. Powder Technology 1999;105(1-3):430-435.

PAGE 62

53 44. Frances C., Laguerie C., Mazzarotta B., Veccia T., On the Analysis of Fine Wet Grinding in a Batch Ball Mill. Chem ical Engineering Journal 1996;63(3):141147. 45. Fuerstenau D.W., Gr inding Aids. KONA 1995;13:5-17. 46. Bernhardt C., Reinsch E., Husemann K., The Influence of Suspension Properties on Ultra-fine Grinding in Stirred Ball Mills. Powder Technology 1999;105(13):357-361. 47. El-Shall H.E., Chemical Additives in a Grinding Process. New York: Columbia; 1980. 48. Kwade A., Blecher L., Schwedes J., Motion and Stress Inte nsity of Grinding Beads in a Stirred Media Mill .2. Stress Intensity and its Effect on Comminution. Powder Technology 1996;86(1):69-76. 49. Hashi Y., Senna M., Grinding Effici ency of an Agitation-Beads Mill (2). Micromeretic 1990:726-731. Tran slated by N. Naguchi 2005 50. Jankovic A., Media Stress Intensity Analys is for Vertical Stir red Mills. Minerals Engineering 2001;14(10):1177-1186. 51. Zheng J., Harris C.C., Somasundaran P., A Study on Grindi ng and Energy Input in Stirred Media Mills. Powd er Technology 1996;86(2):171-178. 52. Blecher L., Kwade A., Schwedes J., Motion and Stress Intensity of Grinding Beads in a Stirred Media Mill .1. Energy Density Distribution and Motion of Single Grinding Beads. Powd er Technology 1996;86(1):59-68. 53. Hashi Y., Senna M., Motion of Grinding Media in Axial Direction and its Effect on Comminution in an Agitati on Bead Mill. KONA 1996;14:176-181. 54. Herbst J.A., Lo Y.C., Financial Implications of Comminution System Optimization. Intern ational Journal of Mineral Processing 1996;44-5:209-221. 55. De Castro C.L., Mitchell B.S., The Use of Polymeric Milling Media in the Reduction of Contamination during Mechani cal Attrition. Journal of Materials Research 2002;17(12):2997-2999. 56. Vogel L., Peukert W., Breadage Behaviour of Different Materials Construction of Mastercurve for the Breakage Pr obability. Powder Technology 2003;129(13):101-110. 57. Fuerstenau D.W., Abouzeid A.Z., Effect of Fine Part icles on the Kinetics and Energetics of Grinding Co arse Particles. International Journal of Mineral Processing 1991;31(3-4):151-162.

PAGE 63

54 58. Choi W.S., Grinding Rate Improvement Using Composite Gr inding Balls in an Ultra-fine Grinding Mill. Kinetic Analysis of Grin ding. Powder Technology 1998;100(1):78-78. 59. Choi W.S., Chung H.Y., Yoon B.R., Kim S.S., Applications of Grinding Kinetics Analysis to fine Grinding Characteristi cs of Some Inorganic Materials Using a Composite Grinding Media by Planet ary Ball Mill. Powder Technology 2001;115(3):209-214. 60. Tavares L.M., King R.P., Modeling of Particle Fracture by Repeated Impacts Using Continuum Damage Mechanic s. Powder Technology 2002;123(2-3):138146. 61. Verma R., Rajamani R.K., Environm ent-Dependent Breakage Rates in BallMilling. Powder Tech nology 1995;84(2):127-137. 62. Strazisar J., Runovc F., Kinetics of Co mminution in Microand Sub-Micrometer Ranges. International Journal of Mineral Processing 1996;44-5:673-682. 63. Austin L.G., Luckie P.T., Methods fo r Determination of Breakage Distribution Parameters. Powder Technology 1972;5(4):215. 64. Austin L.G., Bagga P., An Analysis of Fine Dry Grinding in Ball Mills. Powder Technology 1981;28(1):83-90. 65. Austin L.G., Shah J., Wang J., Gallagher E., Luckie P.T., An Analysis of Balland-Race Milling .1. The Hardgrove Mill. Powder Technology 1981;29(2):263275. 66. Fuerstenau D.W., Kapur P.C., De A., Modeling Breakag e Kinetics in Various Dry Comminution Systems. KONA 2003;21:121-131. 67. Rajamani R.K., Guo D., Acceleration and Deceleration of Breakage Rates in Wet Ball Mills. International J ournal of Mineral Proce ssing 1992;34(1-2):103-118. 68. Austin L.G., Introduction to Mathematical Description of Grinding as a Rate Process. Powder Technology 1971;5(1):1. 69. Gao M.W., Forssberg E., Prediction of Product Size Distributions for a Stirred Ball Mill. Powder Technology 1995;84(2):101-106. 70. Nair M., Peirce Z., DiPrima D., inve ntor Eastman Kodak Company, assignee. Colloidally Stabilized Suspension Process. USA patent 4965131. 1990. 71. Nair M., Peirce Z., Sreekumar, C., inve ntor Eastman Kodak Company, assignee. Polymeric Powder Having a Predetermine d Controlled Size Distrubution. USA patent 4833060. 1989.

PAGE 64

55 72. Bilgili E., Hamey R., Scarlett B., Pro duction of Pigment Nanoparticles using a Wet Stirred Mill with Polymeric Medi a. China Particuology 2004;2(3):93-100. 73. Schliebitz F., Grinding of Pigment Particles to Nanoscale by Media Milling [Diploma thesis]: Technical University of Munich.; 2002. 74. Eskin D., Zhupanska O., Hamey R ., Moudgil B., Scarlett B., Microhydrodynamic Analysis of Nanogrinding in Stir red Media Mills. AIChE Journal 2005;51(5):1346-1358. 75. Plantz P., Correlation Among Particle Size Methods. Application Note Microtrac Inc. 76. Filella M.Z., Newman J.M., Jacques B., Analytical App lications of Photon Correlation Spectroscopy for Size Dist rubution Measurements of Natural Colloidal Suspensions: Capabilities a nd Limitations. Colliods and Surfaces A: Physicochemical and Engin eering Aspects 1996;120:27-46. 77. Trainer M., Freud P., High-Concentra tion Submicron Partic le Size Distrubution by Dynamic Light Scattering. Appl ication Note Microtrac Inc 2003. 78. Greenwood R., Rowson N., Kingma n S., Brown G., A New Method for Determining the Optimum Dispersant Concentration in Aqueous Grinding. Powder Technology 2002;123(2-3):199-207. 79. Hogg R., Dynys A.J., Cho H., Fine Gr inding of Aggregated Powders. Powder Technology 2002;122(2-3):122-128. 80. Randolph A.D., Larson M.A., Theory of Particulate Processes. San Diego: Academic Press 1988:72-73. 81. Bass L., Zur Theorie Der Mahlvorgange Z. Angew. Math. Phys. 1954;5:283-292. 82. Austin LG., A Treatment of Impact Br eakage of Particles. Powder Technology 2002;126(1):85-90. 83. Varinot C., Hiltgun S., P ons M.N., Dodds J., Identific ation of the Fragmentation Mechanisms in Wet-phase Fine Grindi ng in a Stirred Bead Mill. Chemical Engineering Science 1997;52(20):3605-3612. 84. Kapur P.C., Agrawal P.K., Approximate Solutions to Discretized Batch Grinding Equation. Chemical Engineering Science 1970;25(6):1111. 85. Kapur P.C., Kinetices of Batch Grin ding part B. Tran s. AIME 1970(247):309313.

PAGE 65

56 86. Berthiaux H., Varinot C., Dodds J ., Approximate Calculation of Breakage Parameters from Batch Grinding Test s. Chemical Engineering Science 1996;51(19):4509-4516. 87. Austin L.G, Shah I., A Method for In ter-Conversion of Micr otrac and Sieve Size Distributions. Powder T echnology 1983;35(2):271-278.

PAGE 66

57 BIOGRAPHICAL SKETCH Rhye Garrett Hamey was born in Medi na, Ohio. He atte nded Cloverleaf High School. Following high school he obtained a de gree in chemical engineering at The Ohio State University. While at Ohio State, R hye attended a class on particle technology instructed by visiting Professors Brian S carlett and Brij Moudgil. Through continued contact with Professor Scarle tt, Rhye became interested in particle technology and chose to pursue a graduate degree at the Engineeri ng Research Center fo r Particle Technology at the University of Florida.


Permanent Link: http://ufdc.ufl.edu/UFE0011903/00001

Material Information

Title: Production of Organic Pigment Nanoparticles by Stirred Media Milling
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011903:00001

Permanent Link: http://ufdc.ufl.edu/UFE0011903/00001

Material Information

Title: Production of Organic Pigment Nanoparticles by Stirred Media Milling
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011903:00001


This item has the following downloads:


Full Text












PRODUCTION OF ORGANIC PIGMENT NANOPARTICLES BY STIRRED MEDIA
MILLING















By

RHYE GARRETT HAMEY


A MASTER THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005





























Copyright 2005

by

Rhye Garrett Hamey















ACKNOWLEDGMENTS

The author would like to acknowledge the friendship and the wisdom of the late

Professor Brian Scarlett, whose instruction and support were the reasons for the pursuit of

this study. Dr. Brian Scarlett taught the author how to be a better engineer and a better

person.

The author is grateful for the help of Dr. Brij Moudgil for his countless hours of

guidance through out this study. The author would also like to thanks Dr. Hassan El-Shall

and Dr. Abbas Zaman for their insightful discussions on milling and serving as

committee members.

The author would like to acknowledge the efforts of Dr. Ecevit Bilgili, Dr. Dimitri

Eskin, and Dr. Olysea Zupanska for their collaborative efforts in understanding the

particle and system physics in this study.

The author would also like to acknowledge the financial support of the Particle

Engineering Research Center (PERC) at the University of Florida, The National Science

Foundation (NSF) (Grant EEC-94-02989), and the Industrial Partners of the PERC for

support of this research. Thanks are extended to Eastman Kodak for providing the

polymeric media.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ......... .................................................................................... iii

LIST OF TABLES ......... ........ ................................... .......... .... ............ vi

L IST O F F IG U R E S .... ...... ................................................ .. .. ..... .............. vii

A B STR A C T ..................... ................................... ........... ... .............. viii

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

1.1 P article P rodu action .......... ............................................................... ........ .. .. ...
1.2 P article B reakage .................... .... .. ... .................................. .... ........... 3
1.2.1 Introduction to Particle Breakage............... .............................. ...............3
1.2 .2 State of Stress .............................................. ..................... 3
1.2.3 Cracks and Defects .............................................. ...... .. .............. .4
1.2.4 B reakage M echanism s............................................ ........... ............... 5
1.2.5 M materials .................. .................................. 6
1.2.6 Theoretical M odels of B reakage ........................................ .....................7
1.2.6.1 Griffith's theory of brittle fracture ........................ ............... 7
1.2.6.2 Irwin's and Orowan's theories of ductile fracture..........................7
1.2.6.3 Rittinger's and Kick's laws on particle breakage ..............................8
1.2.6.4 D distribution based m odels ................. .... ........................................ 9
1.2.7 M easuring Breakage of Particles.............................................................. 9
1.2.8 Environm ent ............................................. .. ....... .............. .. 10
1.3 Stirred M edia M killing ............................................... .... .. .......... .. ........ .... 11
1.3.1 M edia M ills .............................................................. .. ......... 11
1.3.2 G rinding A ids ................................................... .. ................. 12
1.3.3 M edia Loading................................................ .. .. .. .............. ... 13
1.3.4 M edia Size .................................................. ......................... ... 13
1.3.5 O their M killing Param eters ........................................ ........ ............... 14
1.4 M killing K inetics .................................................................. .... .. ......... .15

2 EXPERIMENTAL PROCEDURE................................... .................................... 17

2 .1 E qu ip m en t .............................................................................................. ..... 17
2 .2 M a te ria ls ...............................................................................................................1 8










2 .3 P ro cedu re ........................................................................................................ 2 1
2.4 Sam pling and Characterization....................................... .......................... 22

3 EXPERIMENTAL RESULTS AND DISCUSION .................................................25

3.1 B ase E xperim ents ....................................................................... ....................25
3.2 Effects of Grinding Aid Concentration............. ............ ....... ............. 27
3.3 Effects of M edia Concentraion .................................................. ..... .......... 28
3.4 E effects of M edia Size................................ .......... ............... ............... 30
3.5 Breakage M echanism s .............................. .............. ........................ 32
3.6 G rinding M echanism .................................................. .............................. 32
3 .7 G rinding K inetics........... ...... ........................................................ .. .. .... .... .. 33

4 POPULATION BALANCE MODEL............................................. ............... 36

4.1 M ethod ............. ..... .... ...... ....... .............. ..... .................36
4.2 Experimental Findings by Population Balance Modeling............................... 39

5 SUMMARY, CONCLUSIONS AND FUTURE WORK .......................................44

5 .1 S u m m a ry ......................................................................................................... 4 4
5 .2 C o n c lu sio n s ..................................................................................................... 4 5
5 .3 F u tu re W o rk .................................................................................................... 4 6

APPENDIX

A REPEATABILITY RUNS OF M103 ........................................ ............... 48

B CALCU LA TION S OF CM C ...............................49...........................

LIST O F R EFEREN CE S ................................................................................................... 50

B IO G R A PH IC A L SK E T C H ....................................................................................... 57




















v
















LIST OF TABLES

Table pge

2.1 Operating variables of the stirred media milling process............... ....................18

4.1 Parameters of the Kapur's G-H model for Run M103 ..........................................41

4.2 The breakage distribution matrix for size class 1 through 9 form top to bottom
and left to right (b6,3= 0.958) ........................................................ ............... 42

4.3 Selection function for experiment M103 population balance model .....................43















LIST OF FIGURES


Figure page

1.1 Attrition, cleavage, and fracture, and particle size distribution for attrition,
cleavage, and fracture, respectively. ........................................ ....... ............... 5

2.1 Optical microscope image of the polystyrene beads (mean size, 40.3 [tm).
M arker size is 50 tm ............................................... ........ .. ........ .... 19

2.2 Images of the pigment particles: (a) SEM image under x50,000 magnifications;
and (b) TEM image. Marker size is 100 nm for (a) and 500 nm for (b). .................20

3.1 SEM images of pigment particles in Run M103 samples collected at various
milling times: (a) after 10 min; (b) after 16 h; (c) after 24 h..................................26

3.2 A, B, C, and D show the effects of surfactant concentration (0.0, 0.02, 0.037,
and 0.0937M) on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24
h rs resp ectiv ely ................................................. ................ 2 7

3.3 A, B, C, and D show the effects of media concentration (0, 60, 120, and 200 g)
on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs
respectively..................................... ................................ ........... 29

3.4 A, B, C, and D show the effects of media size (9, 21, 40, and 402 |tm) on
particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively ....31

3.5 Plot of change in particle size with time for varying media concentrations that
describes the grinding rate............................................... ............................. 33

3.6 Plot of change in particle size with time for varying media sizes that describes
the grinding rate. ......................................................................34

4.1 Fit of the Kapur's G-H model (4.7) to the cumulative volume fraction oversize-
time data. Legends indicate the lower edge of each size class i..............................40

A. 1 Plot of the volume fraction versus particles size of M103 including the standard
deviations at varying run tim es. ........................................ ......................... 48

B. 1 Measure of surface tension versus surfactant concentration for OMT surfactant.... 49















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

PRODUCTION OF ORGANIC PIGMENT NANOPARTICLES BY STIRRED MEDIA
MILLING

By

Rhye Garrett Hamey

August 2005

Chair: Brij Moudgil
Cochair: Hassan El-Shall
Major Department: Materials Science and Engineering

Stirred media mills are widely used in industry for fine and ultra-fine grinding of

various materials and dispersion of aggregates. There is a growing interest in the large-

scale production of nano-size particles, and wet stirred media milling appears to be a

promising technology in this regard. In the present study, a stirred media mill operating in

the wet-batch mode was used to disperse and grind pigment aggregates. To disperse the

initial aggregates and to prevent re-agglomeration of the fine particles, the pigment

aggregates and a polymeric media were suspended in an aqueous solution of a dispersant.

As the size reduction progresses, variations of the particle size distribution with time

were determined using the photon correlation spectroscopy. The product pigment

particles were also examined under a SEM (scanning electron microscope). The effects of

operating variables such as surfactant concentration, media loading, and media size were

investigated. We demonstrate that the stirred media mill with a polymeric media is

capable of producing nano-particles in the 10 to 50 nm size range. Experimental and









population balance model data suggests a fast initial dispersion of the strong

agglomerates followed by cleavage of primary particles. Further evidence for nano-size

grinding is supplied by the SEM images.














CHAPTER 1
INTRODUCTION

1.1 Particle Production

This study was begun in effort to understand and enhance a method for producing

nanosized pigment particles described in a series of Kodak patents. The Kodak patents

describe milling process in which pigment particles of a primary particle size of 150 nm

are reduced to less than 50 nm. 1,2 However, the Kodak patents do not describe the

mechanisms by which particles breakage and milling occurs. The patents also neglected

to describe how milling parameters affect these mechanisms. This study will evaluate

how changing milling parameters will affect the breakage kinetics and particle size. It

will also determine the particle breakage mechanism through inspection of particle size

distribution (PSD) data and a breakage distribution function calculated from the

population balance model (PBM).

Many industries are realizing the benefits of nanosized materials. Some of the

benefits are derived form the high specific surface area of nanoparticles. In the

pharmaceutical industry poorly water soluble drugs have shown an increase in solubility

as particle size decreases.3 This decrease in size will further increase the bioavailability

of the drug due to large amount of available surface area for reactions to take place at.

This increase in bioavailability could lead to new types of non water-soluble drug

formulations.3 The ceramic industry has found that nanoparticles could increase the

strength of composite materials. Nanosize organic pigments have shown an improvement

in color quality, lightfastness and durability.6 7 So far production of many of these









materials has been limited to small-scale process making the cost of these materials very

high.

Most nanosize particles are produce in a growth process, from either a liquid phase

such as crystallization or from a gas phase like flame synthesis.8-12 Both of these methods

have drawbacks in that they can only be applied to a limited number of materials. In

crystallization processes materials are limit by their solubility in a solvent. In plasma

synthesis process materials are limited by their ability to with stand high temperatures

thus making this process mainly restricted to producing metal oxides and carbon

nanotubes. The scalability of both processes is limited due to control issues that arise

when working with large volumes (i.e. temperature homogeneity inside the vessel or

inside the flame 12). To reduce the cost of nanosized materials a process capable of

produce large amounts of many different types of materials is need. Milling is a process

that is capable of produce large amounts of materials. Many researchers have studied the

milling of particles of the size of a few inches to the submicron size range. However,

little work has been performed in understanding nanoparticle production through a

grinding process.13-15 This study will further the understanding of nanoparticle production

in a milling process by identifying methods of particle breakage and relating them to

milling mechanisms. Much of the research conduct in a milling process has showed that

media size and media concentration can affect the rate of milling and the fineness of the

particle size for macroscopic milling processes.1619 For submicron and nanomilling

processes reagglomeration during the milling process can reduce the effectiveness of

particle size reduction.14 This study will evaluate the effects of surfactant concentration,









media size, and media concentration on the fineness of the product and rate of particle

size reduction in a nanomilling process.

1.2 Particle Breakage

1.2.1 Introduction to Particle Breakage

Understanding particle breakage is important in many industries such as the

pharmaceutical and mineral processing. Breakage can lead to problems in processing and

conveying equipment. In a milling process understanding particle breakage is necessary

for obtaining a desired product. In both situations, better knowledge of breakage

behavior will lead to significant savings in time and money as well as an overall

improvement of product quality. Single particle impact, and multiple particle impacts on

breakage have been studied by many researchers.20-24 Many parameters determine how or

if particles will breakage (such as material properties, impact velocity, environmental

effects, and history or fatigue).25 This section will discuss some of these conditions and

will try to explain how they lead to breakage.

1.2.2 State of Stress

There are many ways to apply a force that will result in particle breakage. Rumpf

suggested four modes by which a particle can be stressed that could lead to breakage.26

(1) Compression-shear stressing
(2) Impact stressing
(3) Stressing in shear flow
(4) Stressing by other methods of energy transmission (electrical, chemical, or heat)
The first two are caused by contact forces and are the only modes important for milling.

Mode 3 is important in dispersion of materials under shear, and only exerts enough force

to destroy weak agglomerates, which will be neglected in this study will fail. There are









three modes by which a material can fail, which are a function of the material as well as

the stress applied to the material.

(1) Opening mode that corresponding to an applied tensile load
(2) Sliding mode where the force is applied in the plane of the defect or crack
(3) Tearing mode where a force is applied in a direction perpendicular to the plane
of the crack or defect
For most particles, breakage is considered to take place in the first mode of failure.

Which may be due to artifacts in the techniques developed to measure particle breakage.

If a normal load is applied to a particle a load equal in magnitude is acting internally in

the direction perpendicular to the load and can then be considered to be a tensile load.

Tensile failure is more likely then compressive failure for many materials.

1.2.3 Cracks and Defects

Understanding crack formation and propagation is of fundamental importance to

the study of particle breakage. Particle breakage is preceded by crack growth, which is

preceded by crack formation. Crack formation begins around local defects in the

material. Defects are separated in to four classes, which correspond to their dimensions.

(1) Point defects (vacancies, interstitials, or substitutionals)
(2) Line dislocations (screw or edge)
(3) Interfacial
(4) Volume
The strength of a material is related to the number, size and types of defects present. Most

materials are only a fraction as strong as the theoretically strength would predict due to

the presence of defects. Cracks are initiated when the applied stress exceeds the strength

of the material and form at a defect location. Following crack initiation is crack growth;

crack growth will lead to the breakage of the particle. Cracks can grow along the surface

of the particle as in the case of lateral cracks. Lateral cracks will lead to wearing away of

the particle surface as in the case of attrition. Cracks can also propagate through a particle








by radial and median type cracks which lead to fragmentation of the particle. The rate

and type of crack growth are highly dependent on the applied stress and the type of

material. Sh6nert extensively studied crack growth and found that crack velocity can vary

widely depending on the material, environment and state of stress.27 An empirical factor

called the fracture toughness can be used to predict when a material will fail. The

fracture toughness is dependent upon the size and geometry of the crack.

1.2.4 Breakage Mechanisms

Particle breakage can take place through many different mechanisms of loading.

These mechanisms of loading as well as the material being stressed can affect how the

particle will break. There are three ways by which a particle can break, which are

fracture, cleavage or attrition. During fracture a mother particle will break into daughter

particles or fragments that have a wider size distribution. During cleavage a mother

particle will break into smaller daughter particles of roughly the same sizes.


0 At Attriti
a o .0 Cleavage
SFracture


bC Z ___
(D I

D Z Attrition Particle size
(C>=\ b 13 Attrition
< Initial si:


Figure 1.1 Attrition, cleavage, and fracture, and particle size distribution for attrition,
cleavage, and fracture, respectively.

Attrition is where a mother particle will have surface asperities removed from the

particle. This leads to much smaller daughter particles. Figure 1 represent the three


ion


ze









different mechanisms of particle breakage.28 Particle breakage is dependent on the type of

stress or load which the particle experiences. If a brittle material experiences a normal

load such as impaction in a milling process it would undergo a fracture or attrition

mechanism of breakage.29 If a ductile material experiences a shear loading or forces it is

most likely to fail by a cleavage mechanism.30

1.2.5 Materials

The mechanical breakage of materials is affected by how the material is subjected

to a load, the load applied, and by the material properties. Material properties are the

elasticity, hardness and fracture toughness. Particle size and shape can also affect

breakage. Particle size has been shown to have an affect on the hardness of a material.31

33 It has been shown that as particle size decreases the number of flaws in the material

also decreases, leading to higher fracture toughness. Particle size can also affect the

elasticity of the material changing a large size particle of brittle non-deformable material

in to a small size particle of plastically deformable ductile material.30'33 This change is

called the brittle-plastic transition, which usually takes place in very fine particles.

Failure of a material can be characterized as brittle, semi-brittle, or ductile. Bond

ruptures usually dominate the brittle failure mode, whereas semi-brittle fails through

dislocation mobility and bond rupture. In ductile materials dislocation mobility is usually

the dominant mechanism that leads to failure. Brittle materials fail through the

propagation of cracks by internal flaws, surfaces flaws, and shear deformations. These

flaws lead to particle breakage by attrition and fracture. For brittle failure the stress field

is independent of the strain rate, which means the rate of deformation does not play a role

in whether the material will fracture. It also means that brittle fracture is usually very

rapid and independent of fatigue. Crack propagation and failure in semi-brittle materials









is preceded by plastic flow. Semi-brittle materials can have propagation of cracks in the

radial, median, and lateral directions. These systems of crack propagations lead to a

breakage by fracture and attrition. Semi-brittle materials also show a strain rate

dependency on failure. Ductile materials fail through tensile and shear loading. Ductile

failure is highly dependent on fatigue, and strain rate and usually fails through a crack

opening or tearing modes. This type of failure usually leads to a cleavage mechanism of

breakage in particles.

1.2.6 Theoretical Models of Breakage

1.2.6.1 Griffith's theory of brittle fracture

If a material is brittle it will fracture by crack formation and crack propagation.

Griffith proposed a theory that thermodynamically predicts the energy required for brittle

fracture. In Griffith's equation for plane stress, eq. 1, oc is the stress required for crack

propagation, E is the elastic modulus, Ys is the surface energy and a is the crack length.34


-c -- (1.1)


In the case of plane strain a factor of (1-v2) is added to account for the confinement

in the direction of the thickness. Roesler's experiments in conical indentation on silicate

glass further verified Griffith's theory.35' 36

1.2.6.2 Irwin's and Orowan's theories of ductile fracture

Little work has been conducted to further the understanding of ductile failure. Irwin

and Orowan proposed a theory that incorporates the plastic deformation into material

failure.25 Orowan's equation, eq. 2, relates the rate of strain to the dislocation density (p),

Burger's vector (b), and the change in the average distance a dislocation can travel with









time ; k is a correction factor for the non-uniform movements of dislocations. The

response of materials to external stresses can change during processing. They may go

d7P dl
d- = kpb (1.2)
dt dt

from behaving as a brittle material to being completely plastic or from being plastic to

behaving as a brittle material. These changes can be due to the processing environment,

the change in size, or by fatigue of the material.

1.2.6.3 Rittinger's and Kick's laws on particle breakage

Two of the most common theories used to model particle breakage are Rittinger's

and Kick's Laws. Both of these theories are used to predict the amount of energy or stress

required to break a particle. Rittinger's law, Eq (3), is an empirical model that relates

energy to the difference of product size particles, x2, less the feed size particles, xi:


E=K, 1 (1.3)


Rittinger's law states that as the particle size decreases the energy required to fracture the

particle increases. This equation has been found to be more suited for fine grinding. If

Rittinger's law were completely true it would be impossible to fracture a nanosize

particle, due to the energy required to fracture the particle (as x2 goes to zero E goes to

infinity). Another law relating particle size to energy is Kick's law, Eq. (4):


E =K3og1 xi (1.4)


Kick's law has been found to work well in describing the energy required for the

breakage of coarse particle sizes. Both laws have some capability for predicting particle









size. But as particle size decreases the energy required to fracture the particles goes to

infinity that is obviously erroneous.

1.2.6.4 Distribution based models

Reid proposed an empirical model that defines a specific breakage rate to describe

a grinding process.37 The model describes the breakage as a first-order rate process.

Broadbent expanded this model and now it is referred to as a population balance model.38

The model uses parameters such as the selection and breakage function to describe the

mechanisms of failure. The selection function represents the fractional number of

particles of some size that are broken. The breakage function represents the rate at which

particles in a size class are broken. From this model it is possible to interpret a particle

size distribution mathematically. A correlation between these parameters and the

mechanisms of particle breakage such as fracture, cleavage, and attrition can be

determined. The Weibull distribution equation is another model that uses the particle size

distribution data to obtain an empirical representation of particle breakage. Here Q is the

cumulative fraction of fragment size x, while m and xc are fitting parameter. The Weibull

equation has been used to describe the fracture of brittle materials such as soda-lime glass


Q = 1- exp -
L (1.5)

spheres.39

1.2.7 Measuring Breakage of Particles

The most common technique to measure particle breakage is by a single impact

test. Where a particle of some size is either shot at a target with some velocity or a load

is applied to a stationary particle at some velocity. The breakage distribution of the

particle can then be determined by measuring the size of the broken particles.25 A









drawback of this method is that the only way to apply a stress is by a normal load. Other

methods use indentation techniques to determine the effects of strain on the breakage of

particles. However, similar to the impact technique, the indentation method also subjects

particles to a normal load. Both these methods have proved invaluable for understanding

single particle breakage under a normal load. Schonert studied crack propagation and

crack behavior in many types of materials.25 He showed that material properties, stain

rate, impact velocity, particle shape, and particle size effect how particles break in single

impact and compression tests. It is known that for real systems particles of some size

distribution will show significantly different breakage behavior.23 Particles of small sizes

will not break as easily as particles of larger sizes. Also particles of the same material and

of the same size still can differ in strength. To account for these discrepancies in particle

breakage many researchers have devised techniques to measure multiply particle

breakage. These techniques are similar to the impact test. They are usually performed by

either dropping some load onto a bed of particles or by propelling a mass of particles at

some target. From the particle size distribution data obtained after the test it is possible to

calculate a fractional number of breakages by fitting this data with some sort of empirical

model (i.e. population balance model).

1.2.8 Environment

Environmental effects have been shown to be negligible on the fracture toughness

of a material. Most research suggests that environmental effects for a milling process are

limited to changes in the slurry viscosity that reduces the energy dissipated as heat

through viscous flow.40 However, Griffith's theory includes a surface energy term that

relates to energy required for crack propagation. It is known that additives and

temperature can change the surface tension of a material. Reducing the surface tension in









Griffith's equation reduces the stress required to propagate a crack. This only holds for

particles with an existing crack. It has been suggested that all materials have defects in

the form of small cracks or micro cracks. Rebinder showed that lowering the surface

tension of a material can greatly reduce the fracture toughness of the material and in turn

leads to easier breakage.41 Rebinder's experiments were performed in dry systems where

the only additive was water. Rehbinder's theory has been challenged by a number of

researchers who found that the reduction of surface tension does not play a dominant role

in breakage.

1.3 Stirred Media Milling

1.3.1 Media Mills

Stirred media mills have been in use for over 70 years, Szegvari invented the first

mill in 1928. However, media mills did not become popular until the 1950's when they

were used as disperser in the pigment and paint industry.42 Stirred media mills come in

two designs, which are, either a vertical or horizontal mounted vessel. Stirred media

mills can be run in either the wet or dry milling conditions. In many respects stirred

media mills are similar to ball mills. The major difference between them lies in the way

the energy is supplied to mill and the material. Ball mills operate by rotating the complete

mill volume; the contents are then broken by the impaction of the media colliding in a

cascade mechanism. In a stirred media mill media and materials are placed in the vessel

and then agitated. This reduces the amount of energy needing supplied to the mill since

only the agitator has to be rotated and not the whole volume of the mill. Another

difference is the velocity of the contacting balls or grinding media. In a ball mill the

media velocity is limited by the diameter of the mill and the rotational rate of mill









resulting in low media velocities. In a stirred media mill the velocities are only limited by

the rate of rotation of the stirrer.

Stirred media mills are also known as attritors or high-energy mills. An attritor is

stirred media mill run at high solids loading where the milling takes place through a shear

mechanism. A high-energy mill is where the mill is run at very high rpms or tip speeds in

excess of 10 meters per second. It has been estimated that there are over 44 parameters

that can affect a stirred media mills performance.17

1.3.2 Grinding Aids

The previous section mentioned how environmental affects the strength of a

material. This effect is also present in a grinding system. Many researches have

investigated the effects of grinding aids on the efficiency of a milling process.40 43-46 The

majority of work has been focused on dry ball milling processes. It has been proposed

that in these processes that addition of grinding aids acts as a lubricant, which reduces the

viscosity of the system thus increasing breakage of the particles. In other work grinding

aids were proposed to act as dispersants that prevent particles from agglomerating which

can lead to more effective contact between media and materials. Some researchers

suggest that at high concentrations the grinding aids can add a protective coating to the

particle and decrease the milling efficiency.45'47 Other effects include the Rehbinder

effect discussed previously where the reduction in surface energy can decrease the stress

required to fracture a material. This effect would be more pronounced in a nano-milling

system where the number of defects per particle is very low. In this study we milled at

low solids loading allowing us to ignore the effect of surfactant on slurry viscosity.









1.3.3 Media Loading

Arguably the most important parameter in a stirred media milling process is media

loading. Media loading can affect a milling process in a number of ways. The higher the

media loading the larger the number of contacts which lead to a higher milling rate.8, 19

Also the higher the media loading the higher the viscosity and the larger the energy

needed to operate the mill.48 It is apparent that an optimum must be found to determine

the extent to which a mill should be loaded. However, media loading is not limited to just

these two effects. Media loading also effects the contamination or wear of grinding media

and mill. It affects breakage mechanism taking place, and the flow field inside the mill.

Hashi and other extensively studied the flow field as well as wear of grinding media with

respect to media loading and found optimum media loadings for specific milling

conditions.13,49 The effects of media loading become even more complicated when

coupled with rotational rate, media size and media density. Media density and rotational

rate can easily change any optimum established for efficiency with respect to energy. We

varied media loading at a constant media density; media size and rotational rate in order

to study just the effect of media loading in the context of number of vary the number of

contacts and not the momentum of the media.

1.3.4 Media Size

The final parameter studied was media size and its effect on particle size reduction

and rate of particle size reduction. It has often been stated to mill to fine sizes small

media is needed.17' 1, 50, 51 It is easy to imagine that as media size decreases the number of

contacts increase. However, as the size of the media decreases the maximum stress that

can be imparted on the particle also decreases. An optimum must be present for media

size similar to the optimum for media loading. The optimum for media loading should









change with particle size. This change is due to change of the strength of the material

with particle size and the necessary stress to fracture the material. It is important to note

that the momentum of the media is related to the size and mass of the media times the

velocity of the media. Therefore, media of larger size has a larger momentum and is

capable of imparting more energy on a particle.

1.3.5 Other Milling Parameters

It is important to note that in addition to the parameters discussed above there are

other parameters that can influence a grinding process. Stirrer speed directly affects the

amount of energy consumed in the process as well as the maximum velocity that the

grinding media can achieve, and thus increasing the momentum of grinding media. Media

density also directly affects the momentum of the media and the force that can be

imparted on a particle. Kwade formulated a simple equation to mathematically relates

media size, media density, and rotational rate to the stress intensity.48'52 Stress intensity

(SI) is directly proportional to circumferential velocity v, media size dp, and media

density pm less the fluid density pf:

SI = dp( -p pf)2 (1.6)

This expression is limited in that it only describes the influence of media size, media

density and rotational rate on the milling process and it is not necessary true to all case. In

our study media density and fluid density are almost the same, which would mean SI

would equal zero. Inhomogeneous regions in the mill may also be present for large dense

medias at low rotational rates.13' 49'53 For the most part over years media hardness have

only been looked at as how it affects media wear. Most conventional grinding media has

a very high hardness (high density zirconia around 9 on Mohr's scale). It has been found









that the harder the media the less the wear that takes place.54 Many industries have found

that by increasing the media hardness the can save significantly on the life of the media.

However, in a nanomilling process where contamination from media wear may

significantly affect the product material even a small amount of contamination could be

bad.55 In the present study polymeric media to mill particles to nanosize, and this is

believed to be possible because of the softness of the pigment material. The softness of

the media may be so high that it prevents it from wearing away the mill and thus leading

to no contamination of the product. This is a contradiction to what the researchers above

have found for media's hardness relationship to wear; however, in this case the polymeric

media is an order of magnitude less that that of the softest media studied. A secondary

effect may arise when the media comes in contact with one another.

1.4 Milling Kinetics

Milling kinetics is the measure of the time it takes to break a particle to a specific

size. Many researchers have investigated breakage rates in a milling process. Vogel has

constructed a master curve for kinetics of different materials.56 Researchers have studied

the effect of environment on breakage rate, the effect of milling parameters and the effect

of material56-61 This study will discuss kinetic on two levels. The first will be

macroscopic representation of the grinding rate determined by the change in the mean

particle size with time.62 This representation will show the nonlinearity of grinding over

long time periods. This nonlinearity has been observed by several researchers and has

been explained by the accumulation of fine particles in the milling zone which reduces

the amount of energy imparted on particles as the milling time increases.57' 63-67 The

second approach to breakage kinetics will be with a population balance model (PBM).

Population balance models have been used to describe both the breakage rate and the






16


breakage distribution in a milling process.57 63 64,68,69 The breakage distribution obtained

from the PBM will be then used to evaluate possible particle breakage mechanisms

taking place in the mill at specific times. These mechanisms will then be used to

determine the dominant milling mechanism in the process. From this information it is

then possible to optimize the mill through the complete understanding of particle

breakage, milling parameters, and grinding rate, which can give significant predictability

of product quality and the energy need to produce that product.














CHAPTER 2
EXPERIMENTAL PROCEDURE

2.1 Equipment

This study employed the use of the Micros Superfine Grinder manufactured by

NARA Machinery Corp. of Japan. The Micros mill is an attrition mill designed for

ultrafine grinding of particles to the micron and submicron size range. The mill was

originally designed with zirconia rings mounted on six equally distant vertical columns to

grind the material against the inner zirconia wall of the mill vessel. The volume of the

vessel is 500 mL and a cooling jacket surrounds it to maintain a desired temperature. In

this study, the zirconia rings were removed, and the vertical columns were used to agitate

the milling media. Thus, the mill was modified to operate as a vertical, batch stirred

media mill. The rotation rate was set to the maximum possible, which is 2000 rpm in

order to facilitate the most rapid and energetic grinding conditions possible. This

corresponds to an equivalent tangential speed of 11 m-s1. The effects of the stirrer speed

were not investigated in this study.

Eastman Kodak provided the patented polymeric grinding media used in this study.

The Kodak media patent discusses a method for manufacturing large quantities of vary

narrow size distribution polymeric media.70'71 Kodak provided several media sizes to be

used in this study. The grinding media sizes were measured by a laser scattering

instrument (Coulter LS230). The volume-mean diameter and standard deviation of each

fraction used in the experiments are presented in Table 2.1 along with other operating

conditions of the milling process. The optical microscope image of the beads (Fig. 2.1)









illustrates the sphericity of the beads. The grinding media had a density of approximately

1 g/cm3.

2.2 Materials

The material used in this study is Magenta 122 which is a quinacridone organic

pigment used in inkjet ink formulations. Magenta 122 is a water-insoluble, semi-

crystalline pigment; with a molecular formula of C22H16N202 it has a structure that

resembles two naphthalene molecules joined by a cyclohexane molecule. The pigment

used was manufactured by Sun Chemical Company and has a density of 1.24 g.cm3

Magenta 122 is hydrophobic and has a contact angle of approximately 1540 as measured

by a goniometer.

Table 2.1 Operating variables of the stirred media milling process.72
Surfactant Mass of polymeric Mean size and standard
Run ID
concentration /Molar grinding media/ grams deviation of the beads /pm

M100 0.037

M101 120 40.3, +/- 8.7

M102 0.02 120 40.3, +/- 8.7
M103 0.037 120 40.3, +/- 8.7
M104 0.0937 120 40.3, +/- 8.7
M201 0.037 60 40.3, +/- 8.7
M202 0.037 200 40.3, +/- 8.7
M301 0.037 120 9.0, +/- 6.0
M302 0.037 120 21.5, +/- 14.2
M303 0.037 120 402.0, +/- 173.0

























Figure. 2.1 Optical microscope image of the polystyrene beads (mean size, 40.3 [tm).
Marker size is 50 tm.72

The dry pigment particles consist of primary semi-crystalline particles,

agglomerates, and aggregates. The SEM images (Figs. 2.2(a) and the TEM image (Fig.

2.2(b)) illustrate various structures of the pigment particles including rod-like primary

particles as well as agglomerates. Several primary particles seem to cluster or fuse at their

faces, which is possibly due to growth and sintering of individual crystals during

manufacturing via precipitation.10, 11 Although primary particles of about 50 nm width

and up to few hundred nanometers in length can be identified from Fig. 2.2, the initial

particle size distribution is entirely dependent on the method of dispersion and

deagglomeration. The use of different suspension liquids and different dispersion

techniques results in vastly different initial particle size distributions.73 The Kodak patent

for the production of the nanosized pigment uses the anionic surfactant sodium N-methyl-

N-oleyl taurate (OMT) for dispersion and stabilization of the pigment during the

process. 1,2 In this study we chose to focus on surfactant concentrations and not on

surfactant types so we employed the same surfactant Kodak in their patents.










(a)

















(b)


i L



Figure 2.2 Images of the pigment particles: (a) SEM image under x50,000
magnifications; and (b) TEM image. Marker size is 100 nm for (a) and 500
nm for (b).72

The surfactant is a long carbon chain surfactant with a molecular weight of 454

g/mol and a critical micelle concentration (CMC) of 0.037 M as measured by the

Wilhelmy Plate method at 250 C.


ERC SO 10.OkV X,50.000 100r 7m WD 15.3mm









2.3 Procedure

Surfactant was mixed with nanopure water to make 250 mL solutions at the

concentrations listed in table 2.1. The pigment particles were suspended in the solution to

prepare a slurry with a fixed volumetric pigment loading of 1.27%. This relatively dilute

loading was chosen in order to suppress the tendency to aggregation, and in order to

study the breakage kinetics. A dilute suspension also limits the effect of increasing

viscosity as particle size decreases, allowing us to study the effects of surfactant on

dispersion and not as grinding aid that reduces the viscosity of the system. The milling

vessel was filled with the beads, to the masses corresponding in Table 2.1. The remaining

volume was filled with the 250 mL, 1.27% pigment slurry, which was poured into the

vessel.

The concentration of the milling media in the mill is an important in determining

the breakage mechanism and kinetics of a milling process. The mass of grinding media

was varied in the range 0-200 grams, where the bead-to-bead collisions are expected to

be the dominant stressing mechanism.74 Keeping all other conditions the same, the higher

the bead loading; the higher the specific energy expenditure will be.48 It is important to

note that Kwade's calculations, chapter 1 eq. 1.6, for specific energy with respect to

solids loading are for dense grinding media, which may not be correct for media with a

density close to that of the surrounding fluid. Eskin's approach for calculating energy and

stress are more likely to be correct being that they take in to consideration the

hydrodynamics of the system.74 More importantly, the friction and associated heating

may lead to excessive wear and softening of the beads. The torque and specific energy

consumption were not measured in this series of experiments.









2.4 Sampling and Characterization

Particle size measurements have many connotations that can be described by size,

shape, volume, surface area, length, number, and any other definition which is usually

process specific.7 We chose to represent the particle size distribution by the volume of a

particle so we could use the mass balance equation incorporated in the PBM model for

measuring particle breakage. The particle size distribution of the pigment was monitored

with time; measurements were performed using the Microtrac ultrafine particle analyzer

(UPA) 150 instrument, which is capable of measuring particle sizes in the range of 3.2

nm to 6.5 itm. The UPA measures particle through the technique of photon correlation

spectroscopy (PCS) or dynamic light scattering. Particles in the nanosize range undergo a

large amount of movement due to Brownian motion. The Brownian motion causes

interference in the scattered light, which results in a fluctuation in the recorded intensity

with time. The fluctuations are then fit to a correlation function that equates the

fluctuations to particle size. Filella reviewed the PCS techniques used by different

instruments and discussed some of the problems associated with them.76 One of the

largest problems associated with a PCS measurement technique is colloidal suspension

concentration, which can result in agglomeration of particles and multiple scattering of

the signal. By representing a particle size on a volume basis there is a weighting of the

distribution towards the large heavy particle sizes. Even one large particle can lead to a

significant amount of error in the particle size measurement. Our data shows that after 10

min of milling all the pigment particles are reduced to a size of less then 1 |tm. To assure

that the samples are free from particles of above 2 |tm, which would lead to significant

error in the measurements, the samples are removed and filtered using a 2 |tm syringe









filter. Freud measured particle size distributions of nanosize particles at varying

concentrations on the UPA and found that the error in particle size is linear for

concentrations above 0.1%.77 The pigment concentration in the mill is approximately

1.27%, therefore, after filtration the samples are diluted to approximately 0.1% using the

appropriate concentration of surfactant solution to eliminate any instability by diluting in

water that could lead to agglomeration during the measurement. To be assured that no

agglomeration was taking place during the measurement, an experiment was run where

samples were measured at sampling times 30 sec., 1 min, 3 min, 5 min, 9 min, 15 min, 30

min, and 1 hr. This experiment showed there was no variation in the samples with time.

Approximately 1 ml of sample was removed from the mill after 10 minutes and then

every 4 hours after that to reduce the reduction in mill volume with time due to sampling.

For the PBM study, samples were taken every hours to increase the number of data

points and reduce the error in the fit.

In addition to monitoring the time variation of the particle size distribution, the size

and shape change of the pigments before and after milling was also monitored using

scanning electron microscopy (SEM), which was also used to check whether or not the

beads had been damaged during the milling process. The diluted and dried samples were

ion beam coated prior to examination with the SEM (JEOL JSM6330F). The elemental

composition of the samples imaged via transmission electron microscopy (TEM) was

determined using energy dissipative X-ray spectroscopy (EDS). Elemental analysis of the

product slurry was performed using inductively coupled plasma spectroscopy (ICP).

SEM, TEM, EDS, and ICP measurements were used to check if the pigment slurry had

become contaminated with zirconia or stainless steel from the mill. Gas chromatography






24


(GC) was used to measure the contamination from the polymeric media; this method

lacked the mass spectroscopy capability so we were only able to determine that the

samples were qualitatively free from polymeric media, which agreed with the SEM

results.














CHAPTER 3
EXPERIMENTAL RESULTS AND DISCUSSION

3.1 Base Experiments

An experiment was run to determine the effect of dispersion on the final pigment

particle size without the presence of milling media. As mentioned previously different

dispersion methods revealed largely varying particle size distributions. Experiment M100

was carried out by dispersing the pigment under the milling conditions listed in table 2.1.

The results of the particle size obtained from this run can be seen in fig. 3.3. The particle

size distribution is bi-modal and consists of a minor coarse fraction in the size range

1.6-4.6 |tm and a major fraction in the size range 80-800 nm. The particle size evolved

to a stable state that indicates dispersion to primary particles and small fraction of hard

agglomerates and coarse particles was achieved. It has been shown in literature that

breakage of the soft agglomerates is a prerequisite for wetting and subsequent adsorption

of dispersants and stabilization of the particle suspension.78 The intent of this experiment

was to provide and estimate of the initial particle size distribution, which is important in

formulating the population balance model (provides a basis for the mass balance

equation).

Another base case experiment, M103, was carried to determine the effectiveness of

the polymeric media on the rate of particle size reduction and on the final particle size

obtained. A fit of the population balance model to the base case experiment is presented

in chapter 4. SEM images were taken of samples collected from M103 after 10 minutes,

16 hours and 24 hours of mill can be seen in Fig. 3.1, which shows that a fast dispersion









of the soft agglomerates takes place, forming rod-like and irregular pigment particles

with sizes ranging from 40 to 400 nm. These particles are a mixture of primary particles

and hard agglomerates. After 16 hours, more rounded particles in the size range 20-200

nm were observed. Spheroid nanoparticles with sizes 10-60 nm can be observed after 24

hours.


























Figure 3.1 SEM images of pigment particles in Run M103 samples collected at various
milling times: (a) after 10 min; (b) after 16 h; (c) after 24 h.72

SEM images suggest a transition from irregular or rod-like aggregates and

agglomerates to rounded nanoparticles. The EDS was performed on these particles to

prove that they were not composed of any inorganic contaminants that may be present in

the mill. SEM images of the indicated that surface of the beads were neither chipped nor

abraded.












3.2 Effects of Grinding Aid Concentration


Experiments M101, M102, M103 and M104 were conducted to study the effects


of surfactants on the rate of particle size reduction as well as the effectiveness of particle


size reduction. Fig 3.2 illustrates the particle size distribution (PSD) of samples from


these experiments at the indicated sampling times. One way a grinding aid can affect a


Particle Size (nm)


Particle Size (nm)


10


08 aD
N

06 D
o
0
04

E
02 D
0


10 100 1000 10 100 1000
Particle Size (nm) Particle Size (nm)
Figure 3.2 A, B, C, and D show the effects of surfactant concentration (0.0, 0.02, 0.037,
and 0.0937M) on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and
24 hrs respectively.


milling process is by acting as a dispersant where particle are either held apart by


electrostatic repulsion or through steric repulsion. The agglomerates must first be broken


apart to allow the surfactant to adsorb on the surface and stabilize the particles. The PSD


shifts towards coarser size with time in the absence of a dispersant, indicating the









aggregation of particles. Even for the low pigment concentration under consideration

(1.27 vol.%), collisions of the pigment particles due to turbulent agitation and Brownian

motion led to aggregation when no surfactant is present. Another effect of the grinding

aid is to reduce the pulp viscosity, thus allowing higher solids loading in wet milling.43

However, depending on the size and material properties of the solid to be ground,

dispersants may have little effect on the milling efficiency and dynamics when the initial

feed is very dilute.44'45 In the current work, at low pigment concentration, the stabilizing

action of the surfactant is the most important as nanoparticles have a high tendency to

aggregate. When sufficient surfactant is added (0.02, 0.037 and 0.0937 M), the PSD is

observed to move monotonically toward finer sizes. The PSDs for surfactant

concentrations of 0.02 and 0.037 M cases are similar within the experimental accuracy

(Appendix A); almost all of the particles are in the nano-size range 9-100 nm after 24

hours of milling. However, excessive use of the surfactant (0.0937 M) reduced the

breakage rate. These experiments indicate an optimum surfactant concentration for this

pigment loading in order to achieve finer particles within a shorter milling period.

3.3 Effects of Media Concentraion

With an increase in the concentration of the polymeric media (Runs M201, M103

and M202), the formation of nanoparticles is achieved faster (Fig. 3.3). It is important to

note that at zero media concentration (M100) the particle size does not deviate from the

primary particle size of approximately 150 nm over the entire milling time. It should also











Particle Size (nm) Particle Size (nm)
10 100 1000 10 100 1000
10 10

A) B)
1 08 0 8 a)






-A- no media
S060 No media
E E
D 02 ---- 120 g -- 00 2
S-0-200g -0-120 g O
--0--- 200 g

10 10 1 0 10 10 10








C) D)Particle Size (nm) Particle Size (nm)
S08 8 08


06 06








be observed that after 10 minutes of milling all the concentrations have almost the same
04 04
SNo media
-0- 60 g No media E
E of stressing 120of the pigment increase as the bead concentration increase. It is difficult
S02 g O6 g 02 Z














to predict from these results whether it is the number of increasing bead-bead collisions120
-0- 200
0 0 00
10 100 1000 10 100 1000
Particle Size (nm) Particle Size (nm)

Figure 3.3 A, B, C, and D show the effects of media concentration (0, 60, 120, and 200 g)
on particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs
respectively.


be observed that after 10 minutes of milling all the concentrations have almost the same


particle size distribution, which supports our initial finding that we rapidly disperse the


agglomerated pigment particles to there primary particle size. With an increase in the


concentration of the media the production of nanoparticles is achieved faster. This is


intuitively expected because the number of bead-bead collisions and consequently the


rate of stressing of the pigment increase as the bead concentration increase. It is difficult


to predict from these results whether it is the number of increasing bead-bead collisions


that result in the increased rate of grinding or if a change from an impaction mechanism


of grinding to a shear mechanism of grinding is what led to the higher grinding rate.


Experiments at higher media loadings where shear would obviously have been the









grinding mechanism taking place failed due to the large pressure build up in the mill at

high rotational rates.

From a fracture mechanics perspective a particle undergoing a shear mechanism

of stress is more likely to result in a cleavage or fracture mechanism of breakage as

opposed to an attrition mechanism. An attrition mechanism of breakage is most likely to

result from an impaction mechanism of stress.27 The distributions above suggest that a

cleavage or fracture mechanism of breakage is taking place, which would mean shear is

the dominant mechanism of breakage. As stated in the first chapter cleavage mechanism

is mostly to take place for ductile materials and a fracture mechanism for brittle

materials. Considering that pigment is a soft organic material it would be safe to assume

that a cleavage is the most likely mechanism of breakage. It is important to note that

these conditions of failure and loading are valid for macroscopic brittle material and may

not always be true for nanosized materials, which have proven to show significantly

different material properties, which may change during processing.30' 32, 33

3.4 Effects of Media Size

Experiments M301, M302, M103 and M303 were performed in order to investigate the

effects of the size of the grinding media. In Fig. 3.4, it is seen that this is a vital factor. It

is apparent from these figures that media size can have a significant effect on the rate of

milling as well as the effectiveness of milling. The 9 micron media is unable to reduce












Particle Size (nm) Particle Size (nm)
10 100 1000 10 100 1000
10 10

A) B)
)08 -08



0 0
S06 06


04 04
tNj 9um N
--A- 9um -- 21 um D
F 02 -- 21 um -- 40um 02
-0- 40 um --402um
-0- 402 um

OobC K XXN . . 00
10 10

C) D)
S08 08 a)
N N

06 06
0 0

W 04 04
S--A-- 9u 9 um
-A-- 9 um -*-2um
E 2 um -- 40 um E
S02 -0- 402um 02


00Y 100
10 100 1000 10 100 1000
Particle Size (nm) Particle Size (nm)

Figure 3.4 A, B, C, and D show the effects of media size (9, 21, 40, and 402 plm) on
particle size at grinding times of 10 min., 8 hrs, 16 hrs, and 24 hrs respectively


the particles to less then the primary particle size due to the insufficient amount of force


exerted by the media. The 21 and 40 and to some extent 402 microns are very effective in


reducing the particles. These observations coincide with those of other workers regarding


the existence of an optimum media size.17, 48, 51 For a given beads loading and stirrer


speed, large beads have more momentum due to their mass than small ones. Therefore, it


can be suggested that if grinding were occurring purely by impaction grinding


mechanism then after some extended grinding time reach a smaller particle size would be


achieved. On the other hand, small beads experience more collisions than the large ones


due to their relatively high number. This should result in a much more rapid rate of


particle size reduction for smaller grinding media due to the much higher numbers. The









effects of momentum and number of contacts can both be seen in the figure above. For

example, the 21 micron media decreases the particle size much more rapidly then the 40

or 402 um grinding media. However as grinding time increase the 40 micron media

yields a slightly smaller particles then 21 micron media due to the large amount of force

exerted by the 40 micron media. The existence of the larger particle size distribution for

the 402 micron media at a grinding time of 24 hrs is due to the fact that the 402 micron

media did not have sufficient time to mill the particles to or below the size obtained by

the 21 or 40 micron media.

3.5 Breakage Mechanisms

In the first chapter possible breakage mechanism were discussed these mechanism

are fracture, cleavage, and attrition. Many authors have discussed the resultant PSD from

different breakage mechanism, as illustrated in chapter 1, fig 1.1. 26, 79 From the

distributions above it is obvious that an attrition mechanism is not taking place. If this

were the case a small fraction of 10's nanometer size particles would be present at the

short grinding times. It is not possible to distinguish between a cleavage and a fracture

mechanism from the present data. However in the next chapter the data from the

experiments is fitted to a population balance model (PBM) and it is possible to

quantitatively prove that a cleavage mechanism is clearly the mechanism of breakage for

the base case experiment.

3.6 Grinding Mechanism

Grinding mechanisms are difficult to predict in a stirred media milling process. It

has been shown in this chapter that media loading has a significant effect on the grinding

mechanism. It was also shown that certain mechanisms are more likely depending on the

response of the particle size distribution to the size of the media or change in momentum










(stress intensity 48). Considering that a stirred media milling process is an extremely

complex process it is most likely that more then one grinding mechanism is responsible

for the size reduction in media milling.

3.7 Grinding Kinetics

Grinding kinetics is more simply referred to as the rate of particle size reduction. It

is possible to discuss grinding kinetics on two different levels. The first level is the time it

takes to reduce a particle to a specific size.

0.30


0.25


0.20- -0- 60 g
-0- 120 g

E 0.15-


S0.10 -


0.05 -


0.00 -...
0 200 400 600 800 1000 1200 1400 1600

Time (minutes)
Figure 3.5 Plot of change in particle size with time for varying media concentrations that
describes the grinding rate.

The second is the rate at which particles of a specific size class are ground to a

subsequently smaller size class. The first level is a macroscopic approach to grinding

kinetics and can simply be described by a plot of d50 vs. time where the slope is the rate

of grinding. Fig 3.5 is a plot of d50 vs. time for varying media concentration. At grinding










times up to 250 minutes the data obeys first order grinding kinetics, which is similar to

result reported in literature for grinding kinetics.68

0.30


0.25 --*- 9 um
-0- 21 um
-A- 40 um
00 -0- 402 um
S0.20 -
Cl)
0

E 0.15
0

0.10


0.05


0.00 -
0 200 400 600 800 1000 1200 1400 1600

Time (minutes)
Figure 3.6 Plot of change in particle size with time for varying media sizes that describes
the grinding rate.

However as grinding time increases the rate becomes nonlinear. Austin suggested

that nonlinear kinetics is due to the cushioning effect of fine particles on the breakage of

large particles.64'65'68 For the purpose of the present study the nonlinear kinetics are not

discussed. It is important to see from fig 3.4 that the rate of grinding at high media

concentrations is more rapid then that at lower media concentrations. Fig 3.6 shows the

rate of particle size reduction vs. time for varying media sizes. For experiments M302

and M103 the rate of grinding to the grinding limit is more rapid then when larger media

was used (M303). This result is expected due to the larger number of beads present in

M302 and M301.A more quantitative description of grinding kinetics will be presented in






35


the following chapter. Population balance modeling will be used to describe the breakage

of particles from one size class to a smaller size class and will also discuss the nonlinear

observation in the data.














CHAPTER 4
POPULATION BALANCE MODEL

4.1 Method

For over 50 years researchers have modeled grinding and other particulate process

through the use a population balance model (PBM).63 64,68, 80 Bass was the first to derive

an integro-differential mass balance equation for a batch grinding system.81 For a

grinding process it is generally assumed that only breakage occurs, thus allowing

simplifications to be made to the integro-differential mass balance equation. Due to this

assumption the accumulation term (agglomeration) can be left out. The system that we

have chosen to model uses a surfactant to prevent particles from reagglomerating.7

Considering that Bass's equation does not have a simple analytical solution, Reid

suggested a finite difference approximation to Bass's equation.37 The fractional weight

function is discrete in size x and continuous in time t in the form where q, (t) is a set of

n size classes numbered from the coarsest downwards, b, is a volume fraction of the

broken material finer than size x, (breakage distribution function) that is retained on


class i, where x, < x, and consequently i > j, k, = (t) is a fractional rate of
q, (t)dt

breakage (also called a selection function) assumed to be independent of time.

dq, (t) -1
d-t = bjkjqj(t)-kq,(t), 1 dt =1

This is the basic set of ordinary differential equations that form the population

balance model (PBM) of grinding. Generally, the solution of the system (4.1) is









reproduced after determination in advance (from experimental data on monoclass

samples) breakage and selection functions, which is becoming problematic for the

submicron particles. That is why we use the so-called "inverse formulation" of the

problem. Parameters b, and k, of breakage and selection function are determined from

experimental data for particle size distributions obtained during the whole experiment.

They have to be selected in such a way that solutions of (4.1) match the experimental

data as closely as possible. The best possible approximation of the experimental data by

curves q, (t) we understand in the least squares sense is represented in eq. (4.2):


min (q, qk theo)- ))2 (4.2)
tk 1-
h e a eor

where qexp () and qtheo (k) are the experimental and theoretical points respectively.

Besides providing the best possible fit of our solution to the experimental data, the

parameters of the system (4.1) have to obey the law of mass conservation. From the

structure of the equations in system (4.1) it follows that parameters b, must satisfy the

following constraints:


b k =1, (4.3)
i=j+1

b, >0, Vi> j. (4.4)

In addition to (4.3) and (4.4) fractional rate of breakage is nonnegative as long as

we consider comminution (no agglomeration) process and therefore we have extra

constraints

k, >0, Vi=l,n-1, (4.5)


k = 0, i =n.


(4.6)









Constraint (4.6) reflects the fact that there is no size reduction in the finest class

number n. Thus, the problem of determining the parameters in system (4.1) is reduced to

finding the minimum of function (4.2) subject to the set of constraints (4.3)-(4.6). These

sets of equations make up the backbone of any PBM. This set of equations is known as a

nonlinear model subject to linear constraints. Being nonlinear it means that no exact

solution exist, however, many authors have solved this set of equations for there specific

systems and applied there own fitting parameters.57'61 65 66 69,82

In this study we will use the population balance to describe the breakage

mechanism taking place during the milling process.83 By inspection of the b,, the

breakage distribution function it is possible to distinguish between an attrition, fracture,

or cleavage mechanism. Most PBM used today are in the form listed above. The novelty

in a PBM solutions exist in the method used to solve and optimize the equations listed

above (solving equations of the form 4.2). Instead of using the least squares fit

approximation listed above, eq 4.2, for solving eq.4.1, in our study we will obtain an

approximate solution to equation 4.1 through the use of Kapur's G-H method.84 Kapur

twice applied a Cauchy-Picard iteration to arrive at the eqs 4.7, 4.8:

q, (t)= q, 0 exp(G,t + H, t2/2) (4.7)


G =-k+ A ,0 H = Aj(G -G)J, =kB -kB (4.8)
J=1 qo0 J=1 qlO

The G-H method has been found to give comparable data to that of the exact

solution listed above with only a fraction of the computational as well as mathematical

complexity necessary to solve for the exact solution.85 The G-H method has also been

shown to give accurate data at long grinding times.86 For most milling experiments it has









been found that only one term is needed to describe the process (Hi=0). Due to the

extended grinding times used in this study we will use Kapur's original two term model.

The particle size range in this study is divided into nine classes that were

arranged in geometric progression of 2 beginning with 344 nm as the lower edge of

the largest size class (i=1).87 Size class 9 (+9-30 nm) represents the smallest class or sink

where no particles can be broken out of that size class.87 For each size class, q,(t) was

generated using the particle size distribution data at pre-selected sampling times. Then,

the model (Eq.(4.7)) was fitted to the experimental data to find the values of G, and H,

and to check the applicability of the first-order breakage kinetics. The model fit was

carried out with Matlab 6.1 subroutine LSQCURVEFIT that utilizes the Marquardt-

Levenberg optimization algorithm.

4.2 Experimental Findings by Population Balance Modeling

Kapur's G-H model in Eq. (4.7) was used to fit the experimental data sequentially

from size class 1 to class 9. Table 4.1 lists Kapur's G and H parameters for their

respective size classes. A graphical representation of the fit to the experimental data can

be seen in fig. 4.1. The sum-of-squared residuals and the standard error of the estimate

were found to be 0.1304 and 0.0441. The G-H model explains the temporal variation of

the cumulative volume fraction oversize well for the size range +86-578 nm. However,

the model deviates from the experimental data for the size range +30-86 nm, as can be

inferred from the standard error of the estimates in Table 4.1. The inverse of the first

Kapur coefficient G,-I as seen in table 4.1 may be regarded as a characteristic time

constant for a given size class. From the characteristic time constant it can be seen that to

mill to finer sizes (G9) the time constant is larger than the time constant at G1.






40




1.0 -
1.o +344 nm
N A +243 nm
S 0.8 +172 nm
()
>- +122 nm

0 +344 nm
U \ ----- +243 nm
S0.4 +172 nm
0 0.4
S -A +122 nm
> +86 nm
4--


S 0.2 + -1 "




S1.06




S:-. -- ----.+ v n
) 0.06 -
0
06
4--









6 5 16 15 26 25

Time, 1 (h)
LL 0.4 +61 nm
( 0.4 +4361 nm









an increasing amount of time/energy. For sizes larger than +86+3 nm the characteristic time
So +30 nm
0.2 +61 nm
E nm------___



0 5 10 15 20 25

Time, f (h)
Figure 4.1 Fit of the Kapur's G-H model (4.7) to the cumulative volume fraction
oversize-time data. Legends indicate the lower edge of each size class i.

The large difference in the time constants for larger size particles and small size particles

agrees well with other theories (Rittinger and Kick) that state milling to finer sizes takes

an increasing amount of time/energy. For sizes larger than +86 nm the characteristic time









constants are nearly equal this can be due to the fact that agglomerates (+86-578 nm)

generally break faster than primary particles (-86 nm), approximately the particle size

obtained after 10 minutes of milling as seen in the fig. 2.2 in chapter 2.

Table 4.1 Parameters of the Kapur's G-H model for Run M103


Size class Size range of First Kapur Second Kapur Standard error of
in the the class coefficient coefficient the estimate
model (nm) -G, (h 1) -H,x103 (h2) x102 (-)
1 +344 1.477x101 0* 0.535
2 +243-344 1.455x101 4.568 0.638
3 +172-243 1.747x101 0.855 1.118
4 +122-172 1.590x101 0.037 1.557
5 +86-122 1.109x10 1 1.047 2.021
6 +61-86 5.229x10 2 3.179 2.941
7 +43-61 1.735x10 8 5.448 4.205
8 +30-43 1.212x1013 3.065 9.951
*By definition
The delay time in the size range +30-86 nm may be due to fatigue failure of

primary nanoparticles. Since the G-H method provides an approximate solution, albeit

sufficiently accurate, one may argue that the observed deviations from the experimental

data may be due to the approximate nature of the G-H method in addition to its exclusion

of the non-first-order effects.

The breakage distribution function will be used to discuss particle breakage and

particle breakage mechanisms. Table 4.3 is the particle breakage distribution matrix. The

breakage distribution matrix, b, details quantitatively all the information regarding the

volume transfer among size classes. For example, b32 = 0.159 and b42 = 0.841 mean that

15.9% of the volume broken from particles in size class 2 goes to size class 3 and 84.1%

goes to size class 4. It is noted that b is highly sparse, which precludes massive fracture

as a breakage mechanism. If a fracture mechanism were taking place fractional quantities








would be present throughout the table (all values above the 00 diagonal must be 0 due to

the initial assumption that no agglomeration is taking place).

Table 4.2 The breakage distribution matrix for size class 1 through 9 form top to bottom
and left to right (b6,3= 0.958)

00 0 0 0 0 0 0 0 0
0.0 00 0 0 0 0 0 0 0
1.0 0.159 00 0 0 0 0 0 0
0.0 0.841 0.0 00 0 0 0 0 0
b= 0.0 0.0 0.023 1.0 00 0 0 0 0
0.0 0.0 0.958 0.0 1.0 00 0 0 0
0.0 0.0 0.019 0.0 0.0 1.0 00 0 0
0.0 0.0 0.0 0.0 0.0 0.0 1.0 00 0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 1

If an attrition mechanism were the mechanism of breakage, particles would break

from size class 1 to one of the finest size class such as 6-8 (particle sizes of 344 nm

would break to 86-30 nm). Parent particles mainly break into daughter particles in a few

size classes. Thus, cleavage of agglomerates and primary particles appear to be the

dominant breakage mechanism. Primary particles in size classes 6-8 break into daughter

particles in adjacent size classes. The data above proves quantitatively that a cleavage

mechanism is the dominant mechanism of breakage throughout the milling process for

the base case experiments. From fracture mechanics it is known that ductile materials will

tend to fracture through a cleavage mechanism and that a cleavage mechanism most

commonly occur from a tensile loading of material which would evolve from a shear

mechanism of milling.









The population balance model also explains how fast particles of a specific size

class break, this is known as the breakage rate or selection function. Table 4.3gives the

selection function for this experiment. It is difficult to consider the value of the selection

function having a physical or real meaning, due to it not being defined. Comparing

selection functions of different experiments may not be possible.

Table 4.3. Selection function for experiment M103 population balance model

k =[0.152 0.173 0.293 0.238 0.198 0.229 0.184 0.193 0] (h')
However, comparison of selection functions with in a set of data is possible. The

selection functions in Table 4.3 are all of the same order of magnitude and the values do

not vary much. This would indicate that the rate of breakage is linear with time, however,

in chapter 3 it was shown that the rate breakage in non-linear and Rittinger's law also

states the rate of breakage should decrease exponentially with particle size. It may be

possible to correct the population balance model by introducing a time-variant selection

function, leading to more reliable breakage rates.














CHAPTER 5
SUMMARY, CONCLUSIONS AND FUTURE WORK

5.1 Summary

This study's objectives were to determine the effects of surfactant concentration,

media loading, and media size on the rate and fineness of the ground product. Additional

objectives of this study were to determine the particle breakage mechanism and the most

probable milling mechanism.

The experiments performed at varying surfactant concentration showed that at no

surfactant concentration the particle size increased as milling time increased. The

experiments also showed that as surfactant concentration increased the rate of particle

size reduction decreased. However, increasing surfactant concentration did not appear to

have an affect on the fineness of the final product. In order to optimize the rate of particle

size reduction with respect to surfactant concentration it may be necessary to add the

surfactant as the milling process progresses in order to match the increase in surface area

as the particle size decreases.

As media concentration increased the rate of particle size reduction increased; this

is due to the increase in bead-bead collisions. When no media was present the mill

dispersed the pigment to the primary particle size of 150 nm. Experiments at high media

concentration where a shear mechanism of mill should have been dominant were unable

to determine difference between a shear and impaction mechanism of milling.

Varying media size was shown to affect the fineness of product material and the

rate of particle size reduction. It was observed that the smallest media was only able to









disperse the pigment particles to their primary particle size and was not able to reduce the

particle size. This was due to the media not having sufficient force to cause the particles

to break. Experiment with the largest media size, showed that the rate of particle size

reduction was reduced. This was attributed to the reduction in the number of grinding

media and the number of bead-bead collisions due to the increase in grinding media size.

The final particle size obtained for the largest media was less after 24 hours due to the

slow rate of particle size reduction, however, if milling time was increased this

experiment should give a finer particle size due to the larger amount of force exerted by

the larger grinding media. The rate of particle size reduction was the most rapid for the

for the 21 micron media due to the fact that it had the highest number of bead-bead

collisions with sufficient force to cause the particles to break. The final particle size

obtained for the 21 and 40 micron media were approximately the same showing that the

force that the media exerted on the pigment particles was equivalent.

Particle breakage mechanisms were determined through inspection of the graphical

particle size distribution data and interpretation of the breakage distribution function

obtained from the population balance model. The PSD data from the base case

experiment proved that an attrition mechanism of breakage was not taking place. From

the PBM breakage distribution function we were able to quantitatively determine into

what size class a mother particle broke into. From this we were able to show that

cleavage of particles was the dominant mechanism of breakage. The PBM also showed

that the rate of particle size reduction became nonlinear at extended grinding time.

5.2 Conclusions

This study proved that we were capable of reducing agglomerate pigment particles

of greater that 344 nm in size to particles of less than 10 nm in a stirred media milling









process using polymeric media. We have also shown that it is possible to determine

particle breakage mechanism and particle breakage rates through the use of PSD data and

a PBM. We were unable to differentiate between a shear and impaction milling

mechanism due to a lack of experimental data.

5.3 Future Work

The above work proved that it possible to show which particle breakage

mechanisms are dominant for the milling of pigment particles. In our current study we

were not able to show any changes in the particle breakage mechanisms with varying mill

conditions. Future work on this topic would include milling of other materials and

determine if the breakage mechanism is material dependent or an artifact of the milling

experiments. It is also important to apply the PBM to allow the experimental data

collected to be able to quantitatively determine the differences in breakage mechanisms

and breakage rates of different materials and different milling conditions.

This study assumed that the material properties of the pigment stated the same

throughout the entire experiment and the limit in the particle size achieved was due to

insufficient energy being applied to the particles to cause breakage. To better understand

this limit it would be advantageous to measure the hardness and elasticity of the ground

material before and after the milling experiments. Gerberich, who measured the strength

of a 50 nm silica particle, showed the feasibility of measuring the hardness and elasticity

of nanosize materials.33

One of the principle objectives of this study was to determine the milling

mechanism. However, we were unable to prove which mechanism was dominant due to a

lack of experimental data. Future work would include a statistical design of experiments

based of media loading, media density, rotational rate, time, and media sizes, with









response variables of rate of particle size reduction and final particle size. This type of

study would allow us to focus on the variables that most effect the energy exerts on a

particle and the milling regime present. By using the statistical design we would reduce

the number of experiments needing to be performed, and obtain a correlation to the

response we are measuring. By measuring the material properties discussed above and

performing this study we would also be able to determine what parameters (whether

material or process) affects the grinding limit.

The current study determined the effect of surfactant on the dispersion and rate of

particle size reduction. We ignored in the present study any effect that the surfactant may

have on the viscosity of the system and the any effect it may have in facilitating particle

breakage. Future work should emphasis understanding the effect of surfactants on particle

breakage as well as monitoring the viscosity of the system as a function of surfactant

concentration and solids loading. At high solids loading the viscosity of the system may

changes as the agglomerated particles are dispersed this could be seen as a kind of shear

thinning effect in this type of suspension. An understanding of surfactant concentration

on particle breakage may be obtained by measuring the change in surface tension of the

particles with respect to surfactant type (anionic, nonionic, cationic, long or short chain).

Performing experiments and measuring the breakage rate and particle size with respect to

surfactant type would determine the surfactants effect on particle breakage.













APPENDIX A
REPEATABILITY RUNS OF M103

30

25 -- Ave 10min
= -E-- Ave 490min
S20
S\ -A- Ave 730min
S15 --e- Ave 1210min
E
10
0 1/-_




0.01 0.1 1
Dp (urn)

Figure A. 1 Plot of the volume fraction versus particles size of M103 including the
standard deviations at varying run times.

Figure A. 1 gives the standard deviation in volume fraction of at specific particle

sizes for the sum of three runs at the base case conditions listed in chapter 2 Table 2.1. It

is important to remember that the particle size axis is in log scale so the volume change at

lower particle size is much more dramatic which is why such a large amount of deviation

exist at the lower particle sizes.
















APPENDIX B
CALCULATIONS OF CMC

Fig. B. 1 gives the surface tension of surfactant OMT with respect to concentration

of OMT. The intersection of a tangent through the rapidly decreasing slope at low

concentration and a tangent through the zero slope at high concentrations is the CMC

concentration of the surfactant. All measurements were performed using a Wilhelmy plat

technique at 250 C.


43.00


42.00


41.00


40.00


39.00


38.00


37.00


0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Concentration (M)
Figure B.1 Measure of surface tension versus surfactant concentration for OMT
surfactant.


SSurface Tension
















LIST OF REFERENCES


1. Czekai A.D., Seaman L.P., inventor Eastman Kodak Co., assignee. Comminution
with Small Particle Milling Media. U.S. Pat. 5500331 patent U.S. Pat. 5500331.
1996.

2. Czekai A.D., Seaman L.P., Smith, D.E., inventor Eastman Kodak Co., assignee.
Process for Milling and Media. U.S.Pat. 5662279 patent U.S. Pat. 5662279. 1997.

3. Merisko-Liversidge E, Liversidge G.G., Cooper E.R. Nanosizing: A Formulation
Approach for Poorly-Water-Soluble Compounds. European Journal of
Pharmaceutical Sciences 2003;18(2):113-120.

4. Lee J. Drug Nano- and Microparticles Processed into Solid Dosage Forms:
Physical Properties. Journal of Pharmaceutical Sciences 2003;92(10):2057-2068.

5. Liversidge G.G., Cundy K.C., Particle-Size Reduction for Improvement of Oral
Bioavailability of Hydrophobic Drugs .1. Absolute Oral Bioavailability of
Nanocrystalline Danazol in Beagle Dogs. International Journal of Pharmaceutics
1995;125(1):91-97.

6. Mendel J.B., Particle Generation and Ink Particle Size Effects in Pigmented Inkjet
Inks part 1. J. Nanopart. Res., 1999;1:419.

7. Mendel J.B., Bermel, A.D., Particle Generation and Ink Particle Size Effects in
Pigmented Inkjet Inks part 2. J. Nanopart. Res. 1999; 1:421-424.

8. Hyeon T., Chemical Synthesis of Magnetic Nanoparticles. The Royal Society of
Chemistry 2002:927-934.

9. Zhu W.P., Pratsinis S.E., Flame Synthesis of Nano-size Powders. In ACS
Symposium series 622, Nanotechnology. Washington: American Chemical
Society 1996.

10. Mather R.R. The Degree of Crystal Aggregation in Organic Pigments. Dyes and
Pigments 1999;42(1):103-106.

11. Mather R.R., Evaluation by Nitrogen Adsorption of Crystal Aggregation in
Organic Pigments. Dyes and Pigments 2000;47(1-2):17-21.

12. Kammerler H.M., Pratsinis, S.E., Flame Synthesis of Nanoparticles. Chemical
Engineering Technology 2001;24(6):583-596.









13. Hashi Y.S., Grinding Efficiency of an Agitation-Beads Mill (1). Micromeretic
1990:732-737. Translated by N. Nagouchi 2005

14. Mende S., Stenger F., Peukert W., Schwedes J., Mechanical Production and
Stabilization of Submicron Particles in Stirred Media Mills. Powder Technology
2003;132(1):64-73.

15. Mende S., Stenger F., Peukert W., Schwedes J., Production of Sub-micron
Particles by Wet Comminution in Stirred Media Mills. Journal of Materials
Science 2004;39(16-17):5223-5226.

16. Tanaka T., Optimum Design for Fine and Ultrafine Grinding Mechanisms Using
Grinding Media. KONA 1995;13:19-24.

17. Jankovic A., Variables Affecting the Fine Grinding of Minerals Using Stirred
Mills. Minerals Engineering 2003;16(4):337-345.

18. Kwade A.,Schwedes J., Wet Comminution in Strirred Media Mills. KONA
1997;52:91-101.

19. Kwade A., Wet Comminution in Stirred Media Mills Research and its Practical
Application. Powder Technology 1999; 105(1-3):14-20.

20. Salman AD, Gorham DA. The Fracture of Glass Spheres. Powder Technology
2000;107(1-2):179-185.

21. Shipway P.H., Hutchings I.M., Attrition of Brittle Spheres by Fracture under
Compression and Impact Loading. Powder Technology 1993;76(1):23-30.

22. Ghadiri M, Zhang Z., Impact Attrition of Particulate Solids. Part 1: A Theoretical
Model of Chipping. Chemical Engineering Science 2002;57(17):3659-3669.

23. Bemrose C.R., Bridgwater J., A Review of Attrition and Attrition Test Methods.
Powder Technology 1987;49(2):97-126.

24. Zhang Z., Ghadiri M., Impact Attrition of Particulate Solids. Part 2: Experimental
Work. Chemical Engineering Science 2002;57(17):3671-3686.

25. Prasher C.L., Crushing and Grinding Process Handbook: wiley; 1987.

26. Rumpf H., Particle Technology: Chapman and Hall; 1990. Translated by A. Bull

27. Schonert K.U., H. Kleman, In: Proceedings of the 2nd International Conference of
Fracture; 1969; Brighton, England; 1969. p. 474-482.

28. Cho H.C., Austin L.G., An Equation for the Breakage of Particles under Impact.
Powder Technology 2003; 132(2-3): 161-166.









29. Schonert K., About the Particle Breakage and the Possibities and Limits for
Improving Size-Reduction Processes. In: The International Symposium on
Powder Technology; 1981; Kyoto, Japan: Powder Technology; 1981.

30. Kendall K., Impossibility of Comminuting Small Particles by Compression.
Nature 1978;272(5655):710-711.

31. Weertman J.R., MRS Bulletin Physic Society B 1955:68.

32. Gerberich W.W., Dauskardt R.H., Fracture Nanomechanics Preface.
International Journal of Fracture 2003; 119(4-2)

33. Gerberich W.W., Mook W.M., Perrey C.R., Carter C.B., Baskes M.I., Mukherjee
R., Superhard Silicon Nanospheres. Journal of the Mechanics and Physics of
Solids 2003;51(6):979-992.

34. Griffith A.A., The Phenomena of Rupture and Flow in Solids. Philosophical
Transactions of the Royal Society A 1921;221:163-198.

35. Roesler F.C., Brittle Fractures near Equilibrium. Proceedings of the Physical
Society of London Section B 1956;69(10):981

36. Roesler F.C., Indentation Hardness of Glass as an Energy Scaling Law.
Proceedings of the Physical Society of London Section B 1956;69(1):55-60.

37. Reid K.J., A Solution to Batch Grinding Equation. Chemical Engineering Science
1965;20(11):953

38. Broadbent R.A., Callcott T.G., Coal Breakage Processes: 2. A Matrix
Representation of Breakage. Journal of Institute of Fuel 1956;29:528-539.

39. Cheong Y.S., Salman A.D., Hounslow M.J., Effect of Impact Angle and Velocity
on the Fragment Size Distribution of Glass Spheres. Powder Technology
2003;138(2-3):189-200.

40. Klimpel R.R., Austin L.G., Chemical Additives for Wet Grinding of Minerals.
Powder Technology 1982;31(2):239-253.

41. Rebinder P.A., Shchunkin E.D., Surface Phenomena in Solids During the Course
of their Deformation and Failure. Soviet Physics Uspekhi 1973:533-669.

42. Hochberg S., inventor Process For Dispersing Pigments In Film-Forming
Materials. USA patent 2581414. 1948.

43. Klimpel R.R., The Selection of Wet Grinding Cemical Additives Based on Slurry
Rheology Control. Powder Technology 1999;105(1-3):430-435.









44. Frances C., Laguerie C., Mazzarotta B., Veccia T., On the Analysis of Fine Wet
Grinding in a Batch Ball Mill. Chemical Engineering Journal 1996;63(3):141-
147.

45. Fuerstenau D.W., Grinding Aids. KONA 1995;13:5-17.

46. Bernhardt C., Reinsch E., Husemann K., The Influence of Suspension Properties
on Ultra-fine Grinding in Stirred Ball Mills. Powder Technology 1999;105(1-
3):357-361.

47. El-Shall H.E., Chemical Additives in a Grinding Process. New York: Columbia;
1980.

48. Kwade A., Blecher L., Schwedes J., Motion and Stress Intensity of Grinding
Beads in a Stirred Media Mill .2. Stress Intensity and its Effect on Comminution.
Powder Technology 1996;86(1):69-76.

49. Hashi Y., Senna M., Grinding Efficiency of an Agitation-Beads Mill (2).
Micromeretic 1990:726-731. Translated by N. Naguchi 2005

50. Jankovic A., Media Stress Intensity Analysis for Vertical Stirred Mills. Minerals
Engineering 2001;14(10):1177-1186.

51. Zheng J., Harris C.C., Somasundaran P., A Study on Grinding and Energy Input
in Stirred Media Mills. Powder Technology 1996;86(2): 171-178.

52. Blecher L., Kwade A., Schwedes J., Motion and Stress Intensity of Grinding
Beads in a Stirred Media Mill .1. Energy Density Distribution and Motion of
Single Grinding Beads. Powder Technology 1996;86(1):59-68.

53. Hashi Y., Senna M., Motion of Grinding Media in Axial Direction and its Effect
on Comminution in an Agitation Bead Mill. KONA 1996;14:176-181.

54. Herbst J.A., Lo Y.C., Financial Implications of Comminution System
Optimization. International Journal of Mineral Processing 1996;44-5:209-221.

55. De Castro C.L., Mitchell B.S., The Use of Polymeric Milling Media in the
Reduction of Contamination during Mechanical Attrition. Journal of Materials
Research 2002; 17(12):2997-2999.

56. Vogel L., Peukert W., Breadage Behaviour of Different Materials Construction
of Mastercurve for the Breakage Probability. Powder Technology 2003;129(1-
3):101-110.

57. Fuerstenau D.W., Abouzeid A.Z., Effect of Fine Particles on the Kinetics and
Energetics of Grinding Coarse Particles. International Journal of Mineral
Processing 1991;31(3-4):151-162.









58. Choi W.S., Grinding Rate Improvement Using Composite Grinding Balls in an
Ultra-fine Grinding Mill. Kinetic Analysis of Grinding. Powder Technology
1998;100(1):78-78.

59. Choi W.S., Chung H.Y., Yoon B.R., Kim S.S., Applications of Grinding Kinetics
Analysis to fine Grinding Characteristics of Some Inorganic Materials Using a
Composite Grinding Media by Planetary Ball Mill. Powder Technology
2001;115(3):209-214.

60. Tavares L.M., King R.P., Modeling of Particle Fracture by Repeated Impacts
Using Continuum Damage Mechanics. Powder Technology 2002; 123(2-3):138-
146.

61. Verma R., Rajamani R.K., Environment-Dependent Breakage Rates in Ball-
Milling. Powder Technology 1995;84(2):127-137.

62. Strazisar J., Runovc F., Kinetics of Comminution in Micro- and Sub-Micrometer
Ranges. International Journal of Mineral Processing 1996;44-5:673-682.

63. Austin L.G., Luckie P.T., Methods for Determination of Breakage Distribution
Parameters. Powder Technology 1972;5(4):215.

64. Austin L.G., Bagga P., An Analysis of Fine Dry Grinding in Ball Mills. Powder
Technology 1981;28(1):83-90.

65. Austin L.G., Shah J., Wang J., Gallagher E., Luckie P.T., An Analysis of Ball-
and-Race Milling .1. The Hardgrove Mill. Powder Technology 1981;29(2):263-
275.

66. Fuerstenau D.W., Kapur P.C., De A., Modeling Breakage Kinetics in Various Dry
Comminution Systems. KONA 2003;21:121-131.

67. Rajamani R.K., Guo D., Acceleration and Deceleration of Breakage Rates in Wet
Ball Mills. International Journal of Mineral Processing 1992;34(1-2): 103-118.

68. Austin L.G., Introduction to Mathematical Description of Grinding as a Rate
Process. Powder Technology 1971;5(1):1.

69. Gao M.W., Forssberg E., Prediction of Product Size Distributions for a Stirred
Ball Mill. Powder Technology 1995;84(2):101-106.

70. Nair M., Peirce Z., DiPrima D., inventor Eastman Kodak Company, assignee.
Colloidally Stabilized Suspension Process. USA patent 4965131. 1990.

71. Nair M., Peirce Z., Sreekumar, C., inventor Eastman Kodak Company, assignee.
Polymeric Powder Having a Predetermined Controlled Size Distrubution. USA
patent 4833060. 1989.









72. Bilgili E., Hamey R., Scarlett B., Production of Pigment Nanoparticles using a
Wet Stirred Mill with Polymeric Media. China Particuology 2004;2(3):93-100.

73. Schliebitz F., Grinding of Pigment Particles to Nanoscale by Media Milling
[Diploma thesis]: Technical University of Munich.; 2002.

74. Eskin D., Zhupanska O., Hamey R., Moudgil B., Scarlett B., Microhydrodynamic
Analysis of Nanogrinding in Stirred Media Mills. AIChE Journal
2005;51(5):1346-1358.

75. Plantz P., Correlation Among Particle Size Methods. Application Note Microtrac
Inc.

76. Filella M.Z., Newman J.M., Jacques B., Analytical Applications of Photon
Correlation Spectroscopy for Size Distrubution Measurements of Natural
Colloidal Suspensions: Capabilities and Limitations. Colliods and Surfaces A:
Physicochemical and Engineering Aspects 1996; 120:27-46.

77. Trainer M., Freud P., High-Concentration Submicron Particle Size Distrubution
by Dynamic Light Scattering. Application Note Microtrac Inc 2003.

78. Greenwood R., Rowson N., Kingman S., Brown G., A New Method for
Determining the Optimum Dispersant Concentration in Aqueous Grinding.
Powder Technology 2002; 123(2-3): 199-207.

79. Hogg R., Dynys A.J., Cho H., Fine Grinding of Aggregated Powders. Powder
Technology 2002; 122(2-3): 122-128.

80. Randolph A.D., Larson M.A., Theory of Particulate Processes. San Diego:
Academic Press 1988:72-73.

81. Bass L., Zur Theorie Der Mahlvorgange. Z. Angew. Math. Phys. 1954;5:283-292.

82. Austin LG., A Treatment of Impact Breakage of Particles. Powder Technology
2002;126(1):85-90.

83. Varinot C., Hiltgun S., Pons M.N., Dodds J., Identification of the Fragmentation
Mechanisms in Wet-phase Fine Grinding in a Stirred Bead Mill. Chemical
Engineering Science 1997;52(20):3605-3612.

84. Kapur P.C., Agrawal P.K., Approximate Solutions to Discretized Batch Grinding
Equation. Chemical Engineering Science 1970;25(6):1111.

85. Kapur P.C., Kinetices of Batch Grinding part B. Trans. AIME 1970(247):309-
313.






56


86. Berthiaux H., Varinot C., Dodds J., Approximate Calculation of Breakage
Parameters from Batch Grinding Tests. Chemical Engineering Science
1996;51(19):4509-4516.

87. Austin L.G, Shah I., A Method for Inter-Conversion of Microtrac and Sieve Size
Distributions. Powder Technology 1983;35(2):271-278.















BIOGRAPHICAL SKETCH

Rhye Garrett Hamey was born in Medina, Ohio. He attended Cloverleaf High

School. Following high school he obtained a degree in chemical engineering at The Ohio

State University. While at Ohio State, Rhye attended a class on particle technology

instructed by visiting Professors Brian Scarlett and Brij Moudgil. Through continued

contact with Professor Scarlett, Rhye became interested in particle technology and chose

to pursue a graduate degree at the Engineering Research Center for Particle Technology

at the University of Florida.