4.,“ ~4. .m. ul 4" kt 4 .r 1.: 4| 4 u ‘. “nun x.-4\.. ._ ...~: .. . .... .\\\-.~~.. 3:3“ “hm“...w...‘ - -.—‘~..v. ...v «.L. 5 .n. . , - w V ~ ‘- ...L.p‘_ ‘ ‘ “3.-.“.‘44 .. .0 4.. “.7 \‘s. Au. .. hykndkk L \.-‘._\,.\ u» 1. 2'- .L V ‘ .m.» cm: «- ~ 4... mm» M». wu. ‘5...“ -. .N...‘ L‘. «a... _ u - -.- a \7‘\ h; ‘2 nkm‘J.‘ .1 k'h ,; z , .1 ‘ “~ L\.;:::~.‘- —~,::'—,;--,~v3 “ .. ‘ .M |.~. M. ‘4‘. ,' "‘ - aid um”. N .. <15. “aim.“ w... . u u .‘ v... «uhuva. . -\.\.v\.u‘ . n-umLu .m . 5| mm. v u ‘ w “5:. mum... | g,m.4 5. H .' ~~ . “.1115“ N ‘ .. 7 4.3... ;\ _’ q :xl-u. .'.t-. x.“ .1‘\«\ Am. . m.‘ ARE - MIME. A. “mm .m. A L.“ ‘ “buy...“ “fl .- a. .MM . a: mu» . .. ..;; 1H :3; ,, cawgaa '1'» m mg g h ‘ "“- m [MIN “inn. ,,.... ‘: ”l“. m. ‘-»A “04.. “r‘ .. Jay.“ ‘4 .hL\\.\...‘.\ .m ”.0.- A;-.‘.‘ , ., . \ I: l r ‘msg—ef I RV ‘ «a.» r... -.ru‘<'\‘-.~ J _.- m~ yum. '4...“ u. - - u .4»... . ~ «Ngugm ~V..“4k \. . u—x ' l,‘ 4 _4 3le1. .\ nu ‘ \ 4 .~.. u 41. _,._... yum" - 4.4 ill-h“. ‘ ..‘_ . .wm‘ A . m. 4 I T . ... h: an“ , : 4‘ ' #5419 , .5 - I. \-.,,'.. ~«u.,«.-‘» ' .. . u ‘1‘“. \M .4 .2: .. ,3," 1.. .1... ..-. km s. . u A u -. “p.07. . k \. . N‘.~<\ m““‘x""‘ w -A. -u:.|.LL- m. w“... n l .L.‘ . P' ‘r . 1.. 4‘ \. x. “In. my... .4 , . v. . -u m.uh...'."fl LN .. \nll. . . ‘ \\.\|.x.h.£.‘ ,.u~~» -. s. 3"“? “tau-h » . “N“. x . -‘M1.\s.4.um “‘ ": :2. :: L». . .. ...~a. .. ‘lflq‘.\4‘1 nikl .M .s , 7 3245'“ x. m :5. an -.-. ‘ _ u. ‘ ~ w L$~ nuns. . .m u;L-.—...~~... . “x. _u. —\.-.- .x. - qua mama. .\ ~ .-.\ u. E: .172.“ . .-.-.-. . ‘. .x-. mm .x'r‘ a. . nu 'mu. 1 n .-.-.- ‘tfi ‘: r m .~ ‘ .3 I‘.’\ ‘ “Nu-3.3.,» it; 3 i V 3.9.49.6, a < x f'\ 5’: “a: 1'4 . Ht 1 -.“. l «,1. L . :I'"-.~:::=n‘;"- 2* ‘1 ‘ u. ~( I \' vu'au y w T”... nu THESIS, 3 1293 01051 7799 This is to certify that the dissertation entitled Hydrolysis Reactions of Inverted 1:1 and Layered 2:1 Silicates presented by Hemamali D.~ Kaviratna has been accepted towards fulfillment of the requirements for Ph.D. degree in Chemistry Major professor Thomas J. Pinnavaia Date November 30, 1993 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative Action/Equal Opportunity Institution «EMMA HYDROLYSIS REACTIONS OF INVERTED 1:1 AND LAYERED 2:1 SILICATES By Hemamali D. Kaviratna A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1993 ABSTRACT HYDROLYSIS REACTIONS OF INVERTED 1:1 AND LAYERED 2:1 SILICATES By Hemamali D. Kaviratna Crystalline porous materials find wide-spread use as catalysts, ion- exchangers and adsorbents. They also are potentially useful as composites and materials for the design of electronic, optical or magnetic devices. The nanoporous regime (1-10 nm) spans the mid-micropore region characteristic of traditional crystalline porous materials (< 2.0 nm) and the lower mesopore size range (2.0-50 nm) typical of amorphous oxides. Regularly ordered nanoporous materials would represent new arenas for chemistry in constrained environments. The present work reports a new approach for synthesizing nanoporous materials. The approach makes use of a layered nonporous material as a template for the formation of a new nanoporous derivative that can not be obtained by using direct crystallization. The viability of this concept is demonstrated for the topochemical acid hydrolysis of antigorite, a silicate with an inverted wave structure. Approximately 70% of the octahedral Mg can be depleted by acid hydrolysis without dramatically changing the crystallographic order of the antigorite. A BET surface area of more than 300 m2 g-1 was obtained. This is a very large increase compared to 6 m2 g-1 for the starting clay. The mechanism for acid hydrolysis of antigorite is considered to have three main steps: (i) Initial acid attack of the octahedral Mg sheet through the eight membered rings of the basal plane of antigorite; (ii) Secondary lateral hydrolysis of the octahedral Mg of already hydrolyzed 001 planes; (iii) A relatively slow, compared to the steps i and ii, edge hydrolysis process. A regular Hemamali D. Kaviratna nanoporous magnesium silicate was synthesized by topochemical hydrolysis of antigorite. The nanopore size varied from a diameter ~ 8 A to ~ 39 A depending on the Mg2+ depletion and the rearrangement of the Si02 sheet. The acid hydrolysis reactions of kaolinite, phlogophite and fluorohectorite also were studied. Kaolinite with aluminum in octahedral sites is not a good candidate for acid hydrolysis reactions as judged by the insignificant change in surface area upon hydrolysis. Acid hydrolysis of phlogophite increases the surface area from 2 m2 g'1 to a maximum value of 77m2 g-1 at 87% Mg depletion. The acid hydrolysis of phlogophite was found to involve an edge hydrolysis diffusional mechanism. Fluorohectorite is more sensitive towards acid hydrolysis than all other silicates studied. The final product exhibits a substantially higher surface area (208 m2 g'l) compared to the starting surface area (3 m2 g'l). Acid attack most likely occurs both through the basal surface (hexagonal cavities) and the edge sites of the layer. Depletion of octahedral Mg occurs starting from edges of the clay particles as it would be expected to occur in talc and mica Iclays. But fluorohectorite, unlike mica or talc, affords an exceptionally high surface area comparable to those of acid hydrolyzed palygorskite, sepiolite and other smectites. A According to the present work and the early work done in area of acid hydrolysis of clay minerals, three different categories of minerals can be identified depending on their behavior towards the acids: (1) Swellable clay minerals including vermiculite give high BET surface areas upon acid treatment; (2) Nonswellable clays afford little or no surface area increase upon acid treatment (e.g. talc, phlogophite); (3) Nonswellable clays with special structural features give high surface areas (e.g. sepiolite, palygorskite, antigorite). TO MY TEACHERS ACKNOWLEDGMENTS I am deeply grateful to Dr. Thomas J. Pinnavaia for his guidance and financial support throughout this work. His knowledge, experience and understanding of chemistry with his affectionate personality made him a great mentor for me. I am grateful to Dr. Harry A. Eick for presiding as my second reader and for the editorial assistance in improving my writing. Department of Chemistry at Michigan State University is greatly appreciated for giving me this opportunity to extend my education. To you all "Clay guys", I do not have words to thank you for the fine time Ihad in Pinnavaia group. I learned chemistry, scientific techniques and instrumentation, multicultural aspects, how to cook international food and group politics from this marvelous group of people. I am deeply thankful to all the group members for their friendship, sincerity and concern. Many endeared friends have helped me survive throughout my graduate student life. You made this long, arduous endeavor a pleasant event. The friendship and numerous helps of you are appreciated. I would like to thank my family, most importantly my parents for their support given to achieve many things. I would also give special thanks for my father and sister Anne for continuously keeping me informed about my family, friends and politics in Sri Lanka. My gratitude to my best friend, Kavi is beyond words. He was with me to share my failures and achievements with continuous encouragement and mountains of love. I also thank God for his wonderful gift , our son Lakmal, who brought me a tremendous amount of happiness. vi TABLE OF CONTENTS Page N 0. LIST OF TABLES .............................................................................. xi LIST OF FIGURES ............................................................................ xv ABBREVIATIONS .............................................................................. xxvi Chapter I Introduction 1.3 Objective and Rationale .................................................................... 1 Nanoporous Materials ............................................................. 1 Silica ....................................................................................... 4 Established Approaches to the Synthesis of Crystallographically Regular N anoporous Materials ............ 6 New Approach to Synthesis of N anoporous Materials ........... 9 Topotactic Reactions .............................................................. 11 Lb Structure and Properties of Clays ..................................................... 13 2:1 Clays ................................................................................. 15 1:1 Clays ................................................................................. Antigorite ................................................................................ I.c Early Studies ..................................................................................... I.d References ........................................................................................ Chapter II Acid Hydrolysis of 1:1 Layered Silicate Structures 11.3 Introduction .................................................................................... II.b Acid Hydrolysis of Serpentine ....................................................... II.b.l. Experimental ....................................................................... Starting Material .................................................................... Acid Hydrolysis ...................................................................... Physical Measurements .......................................................... Alumination Experiments ...................................................... Determination of Crystallinity ................................................ II.b.2. Results and Discussion ....................................................... Characterization of Starting Material .................................... viii 24 28 30 35 44 45 45 45 45 46 48 49 50 50 A". Highly Active Material from Acid Leaching of Antigorite ................................................................................ 54 Micro and Mesoporous Material Obtained from Acid Hydrolysis of Antigorite ......................................................... 82 Electron Micrographic Evidence for Acid Leaching of Anti gorite ................................................................................ 101 Acidity of Acid Hydrolyzed Antigorite .................................. 110 Thermal Stability of Acid Hydrolyzed Antigorite .................. 115 Alumination of Acid Hydrolyzed Antigorite .......................... 115 H.c Acid Hydrolysis of Kaolinite ......................................................... 124 H.c.l Experimental ........................................................................ 124 H.c.2 Results and Discussion ......................................................... 124 II.d Conclusions .................................................................................... 125 [Le References ...................................................................................... l3 1 Chapter III Acid Hydrolysis of 2:1 Layered Silicate Structures III.a Introduction .................................................................................. 136 II.b Acid Hydrolysis of Phlogophite .................................................... 136 ix II.b 1 Experimental ....................................................................... 136 Starting Material and Chemical Reactions ............................. 136 Physical Measurements .......................................................... 139 II.b.2 Results and Discussion ....................................................... 139 III.c Acid Hydrolysis of Fluorohectorite and Pillared Fluorohectorite 155 III.c.l Experimental ...................................................................... 15 5 Starting Material ..................................................................... 156 Physical Measurements .......................................................... 156 III.c. 2 Results and Discussion ...................................................... 157 Acid Hydrolysis of Fluorohectorite ........................................ 157 Acid Hydrolysis of Pillared Fluorohectorite ........................... 164 III.d Conclusions .................................................................................. 168 III.c References .................................................................................... 169 Table 1.1 Table 1.2 Table 1.3 Table 11.1 Table 11.2 TableII.3 LIST OF TABLES Idealized Structural Formulae for representative 2:1 Phyllosilicates. In each Formula the Parentheses and Bracket Define Metal Ions in Tetrahedral and Octahedral Sites Respectively.... Idealized Structural Formulae for representative 1:1 Phyllosilicates. In each Formula the Parentheses and Bracket Define Metal Ions in Tetrahedral and Octahedral Sites Respectively.... Activation energies of octahedral cations ............. Effect of various HCl concentrations on the physical and chemical characteristics of serpentine (acid treatment for 7 h at 75 0C) ......... Crystallinity and magnesium depletion with respect to reaction time for antigorite samples hydrolyzed at 60 0C .............................................. Crystallinity and magnesium depletion with respect to reaction time for antigorite samples hydrolyzed at 80 0C ............................................ xi Page No. 17 25 33 59 60 61 Table 11.4 Table II.5 Table 11.6 Table 11.7 Table 11.8 Table H. 9 Crystallinity and magnesium depletion with respect to reaction time for antigorite samples hydrolyzed at 105 0C ............................................ Effect of various duration of treatment on the physical characteristics of serpentine (acid treatment at 75 0C) ............................................... BET (m2/g) surface area and MAS NMR curve fitting results with respect to Mg depletion for antigorite samples hydrolyzed at 60 0C ................ BET (ml/g) surface area and MAS NMR curve fitting results with respect to Mg depletion for antigorite samples hydrolyzed at 80 0C ................ BET (mZ/g) surface area and MAS NMR curve fitting results with respect to Mg depletion for antigorite samples hydrolyzed at 105 0C .............. 29Si chemical shifts, 5(ppm), relative intensities, nSia and calculated mean SiOSi bond angles, or, of highly dealuminated zeolite ZSM-5 ................. xii Page No. 62 69 70 71 72 77 Table 11.10 Table 11.11 Table 11.12 Table 11.13 Table 11.14 Table 11.15 Table 11.16 Crystallinity, BET surface area, MAS NMR and percentage of magnesium depletion with respect toileaching time for 325 mesh antigorite samples leached at 80 0C. (NOTE: Different batches of starting clay were used in samples with reaction time 3, 4 and 6-12 hours) ...................................... Dependence of the surface area, crystallinity and the percentage of magnesium removed on the Particle size for acid hydrolyzed antigorite .......... Adsorption of different molecules by acid hydrolyzed antigorite (mmoles g-l) ................... Molecular parameters of the adsorbates used ....... Effect of outgassing temperature on the BET surface area (mZg-D of acid hydrolyzed Antigorite .............................................................. Variation of BET surface area of antigorite hydrolyzed with 25 and 40% HCl, according to Girgis and Mourad.3 ............................................. Cation Exchange Capacities and BET surface Area of alurninated acid hydrolyzed samples and antigorite ............................................................... xiii Page No. 83 84 97 98 116 116 123 Table 11.17 Table 111.1 Table 111.2 Table 111.3 Table III. 4 Table 111. 5 Comparison of percentage Al removed and BET surface area to reaction time of kaolinite .............. Percentage of Mg2+ depleted by 20 wt% HCl and the Percentage retained in the phlogophite residues ................................................................. BET surface area of acid hydrolyzed phlogophite samples under different conditions ....................... Free silica content and BET surface area for different acid hydrolyzed reaction products ......... BET surface area of acid hydrolyzed fluorohectorite ( 0.5M HCl at 60 0C ) .................. Basal spacing for different F-hectorite samples... xiv Page No. 125 142 143 154 159 166 Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 Figure 1.5 Figure 1.6 Figure 1.7 Figure 1.8 Figure 1.9 LIST OF FIGURES Page No. Frequently observed pore types ................................. Schematic representation of pillaring or intercalation of layered materials .................................................... Cross-linking of layered materials via functional groups ......................................................................... Stepwise in-situ pillar formation of layered materials"!4b ............................................................... The formation of composite silicon-aluminum- oxygen or silicon-magnesium-oxygen layers. 16 ........ Idealized oxygen framework of clay minerals ........... Schematic structures of palygorskite and sepiolite.71 ............................................................................. Pairing of tetrahedral oxygen atoms (stippled) at the base of an upper 1:1 layer with the octahedral OH groups (large double circles) on the upper surface of the layer below.65 ...................................................... Schematic [010] projections of the three species of serpentine minerals.55b .............................................. XV 3 10 14 16 23 26 27 Figure 1.10 Figure 1.11 Figure H.1 Figure 11.2 Figure 11.3 Figure 11.4a Figure H.4b Figure II.4c Page No. [010] projection of the structure of antigorite; the structure reverses polarity at PP', QQ; and RR'.65 ..... [001] projection of the tetrahedral sheet of antigorite structure ...................................................................... X-ray powder diffraction of antigorite. Inset: X-ray diffraction of antigorite reported by Brindley and Santos, 1971 ............................................................... TEM pictures of antigorite. Top: From Brindley and Santos, 1971. Bottom: From present work ............... 29Si MAS NMR of starting antigorite ....................... Top: IR spectra of antigorite from 3000 cm'1 to 4000 cm-1 (Vander and Beutelspacher, 1976). Bottom: FTIR spectra of antigorite (present work)... Top: IR spectra of antigorite from 700 cm-1 to 1800 cm“1 (Vander and Beutelspacher, 1976). Bottom: FI‘IR spectra of antigorite (present work) ................. Top: IR spectra of antigorite from 400 cm'1 to 700 cm-1 (Vander and Beutelspacher, 1976). Bottom: FI‘IR spectra of antigorite (present work) ................. xvi 29 31 51 52 53 55 56 57 Figure 11.5 Figure 11.6 Figure 11.7 Figure 11.8 Figure 11.9 Figure 11.10 Figure 11.11 Page No. Percent of depletion of magnesium from anti gorite (>100 mesh) versus the reaction time at different temperatures. The initial HCl concentration was 20 wt%. The Hi’IMg2+ was 5/1 ..................................... Relationship between the amount of magnesium depleted and the crystallinity, as determined by method I ................................................................... Typical X-ray powder diffraction patterns of acid hydrolyzed anti gorite at different levels of octahedral Mg depletion ............................................ 24 3 0; 0 6 0 and 0 6 1 diffraction peaks of antigorite and Mg-depleted antigorite ..................................................................... Percentage of Mg depleted from antigorite and the BET surface areas of hydrolyzed products plotted versus reaction time at 60 0C ..................................... Percentage of Mg depleted from antigorite and the BET surface areas of hydrolyzed products plotted versus reaction time at 80 0C ..................................... Percentage of Mg depleted from antigorite and the BET surface areas of hydrolyzed products plotted versus reaction time at 105 0C ................................... xvii 58 63 64 65 66 67 68 Figure 11.12 Figure 11.13 Figure H.14 Figure 11.15 Figure 11.16 Figure 11.17 Figure 11.18 Figure 11.19 Page No. 298i MAS NMR spectra of acid hydrolyzed antigorite samples with different reaction time at 80 0C ......................................................................... 74 29Si MAS NMR and 29Si CP/MAS NMR spectra for Mg depleted antigorite at 80 0C. A. 50% Mg depletion, B. 96% Mg depletion .............................. 75 Scheme for the Mg hydrolysis of octahedral Mg in 1:1 layered silicate lattice ........................................... 76 Schematic representation of antigorite showing disordered structure 4Q on acid hydrolysis ............... 79 FI‘IR bands in the region 2000 cm'1-400 cm'1 for antigorite and acid hydrolyzed antigorite at different levels of Mg depletion .............................................. 80 EUR bands of antigorite and acid hydrolyzed antigorite at different levels of Mg depletion " between 4000 cm“1 and 2500 cm'1 ........................... 81 29Si MAS NMR spectra of acid hydrolyzed antigorite (325 mesh) ................................................ 85 (parts a and b) t-plots of antigorite hydrolyzed at 80' 0C (325 mesh) ................................. - ...................... 86 xviii Figure 11.19 Figure 11.20 Figure 11.21 Figure 11. 22 Figure 11.23 Figure 11. 24 Figure H. 24 Figure 11.25 Figure 11.26 Page No. (part c) t-plots of antigorite hydrolyzed at 80 0C (3 25 mesh) ............................................................... Lecloux-Pirad plot of the nitrogen adsorption data for acid hydrolyzed antigorite with 60% Mg depletion, illustrating the absence of microporosity in the sample .............................................................. Pore size distribution of acid hydrolyzed antigorite samples obtained by BJH method (Reference 13)... Dubinin-Radusikerich plot of hydrolyzed antigorite with 39% Mg depletion .............................................. Nitrogen adsorption-desorption isotherms for acid hydrolyzed antigorite with 39% and 60% Mg depletion ..................................................................... (parts a and b)Derivative of nitrogen adsorption versus log partial pressure .......................................... (parts c and d)Derivative of nitrogen adsorption versus log partial pressure .......................................... Edge view of acid hydrolysis of the octahedral sheets of an antigorite clay particle ......... -.; ................. Stepwise acid hydrolysis of antigorite ....................... xix 87 89 91 92 93 94 95 100 102 Figure 11.27 Figure 11.28 Figure 11.29 Figure 11.30 Figure 11.31 Figure 11.32 Figure 11.33 Page No. HRTEM of antigorite (Mellini, reference 28). Twin planes (001) run from top to bottom, lattice fringes (100) run from left to right ......................................... 103 Schematic drawing of (001) faults. The curved solid line represents the tetrahedral and octahedral layers, and the 6 and 8 refer to the types of reversals. In (a), a mirror perpendicular to 0*, and in (b), two-fold axes parallel to b, the types of reversal that do not change across the fault. In (c) an a-glide perpendicular to 0* produces a change in the type of reversal across the fault .......................... a: .................. 104 Lattice fringes of antigorite ........................................ 105 A TEM image of the starting antigorite used in this work ........................................................................... 106 A TEM picture of acid hydrolyzed sample a (acid hydrolyzed for 3 hours at 80 0C) ............................... 107 Calculated HRTEM images of antigorite as viewed down (010) ................................................................. 108 Calculated HRTEM images of antigorite as a function of crystal tilt ................................................. 109 Figure 11.34 Figure 11.35 Figure 11.36 Figure 11.37 Figure 11.38 Figure 11.39 Figure 11.40 Figure H.41 Page No. A TEM image of acid hydrolyzed sample with 70% Mg depletion (acid hydrolyzed 12 hours at 80 °C).... A 001 TEM view of antigorite ................................... An HRTEM picture of antigorite from Spinnler, 1985 ........................................................................... Temperature Programmed Desorption of ammonia for acid hydrolyzed 325 mesh antigorite with different levels of Mg depletion. Outgassing temperature was 400 °C ............................................. 29Si MAS NMR of aluminated antigorite and acid hydrolyzed samples .................................................... 2951 MAS NMR of acid hydrolyzed samples ............ 27A1 MAS NMR of, 1. Blank, 2. Aluminated antigorite, 3. Aluminated sample with 39% Mg depletion, 4. Aluminated sample 70% Mg depletion ..................................................................... X-Ray Diffraction pattern of aluminated samples. xxi 111 112 113 114 118 119 120 122 Figure 11.42 Figure 11.43 Figure 11.44 Figure 111.1 Figure 111.2 Figure 111.3 Figure 111. 4 Page No. Suggested structures I for future leaching experiments. a. Manganpyrosmalite, b. Bementite, c. apophyllite, d. Hypothetical structure ................. Suggested structures H for future leaching experiments (001 view is shown). a. Greenalite, b. Carlosturanite ............................................................. Suggested structures III for future leaching experiments 001 and 100 views) a. Zussmanite, b. Stilpnomelane ............................................................ Percentage of magnesium depleted from phlogophite by reaction with different acids at 105 °C. The H‘l'lMgZ+ ratio in each case was 5/1 .......................... XRD spectra showing the change in crystallinity upon acid hydrolysis of phlogophite with 20% HC1.. FI‘IR spectra of, Top: Untreated phlogophite, Bottom: Acid hydrolyzed phlogophite formed by reaction with 20% HCl for 6 hours at boiling (105 0C) ............................................................ 298i MAS NMR spectra of acid hydrolyzed phlogophite for 1 hour at 80 0C. a. 5% HCl, b. 10% HCl, c. 30% HCl ........................................... xii 128 129 130 140 145 146 148 Figure 111.5 Figure 111.6 Figure 111.7 Figure 111.8 Figure 111.9 Figure 111.10 Page No. 295i CP/MAS NMR spectrum of acid hydrolyzed phlogophite formed reaction with 20% HCl for 1 hour at boiling temperature (top), 29Si MAS N MR spectrum of acid hydrolyzed phlogophite formed by reaction with 20% HCl for 1 hour at boiling temperature (bottom) ................................................. 29Si MAS NMR spectrum of acid hydrolyzed phlogophite formed by reaction with 20% HCl for 1 hour at boiling temperature. Top inset shows the XRD spectra and the bottom inset shows the FTIR spectrum of the same product .................................... 27A1 MAS N MR spectrum of phlogophite acid hydrolyzed with 20% HCl for 1 hour at boiling temperature ................................................................ Proposed mechanism for acid hydrolysis of phlogophite ................................................................ XRD spectra of fluorohectorite acid hydrolyzed with 0.5M HCl at 60 0C ..................................................... Dependence of BET surface area and CEC of montmorillonite on time of acid hydrolysis.28 .......... xxiii 149 150 151 153 158 160 Figure 111.11 Figure 111.12 Figure 111.13 Figure 111.14 Page N 0. XRD patterns of montmorillonite, following progressive acid treatment. (a) 0, (b) 1, (c) 5 and (d) 15 min.28 ............................................................ 29Si MAS NMR spectra of fluorohectorite acid hydrolyzed with 0.5M HCl at 60 0C .......................... 29Si MAS NMR spectra of (3) untreated and progressively acid-treated [(b) 15 min., (c) 20 hours] montmorillonite.28 ..................................................... A cartoon showing the possible acid attacking sides of fluorohectorite ....................................................... Figure 111. 15 19F MAS NMR spectra of acid hydrolyzed fluorohectorites at different levels of Mg2+ depletion ............................................................. Figure 111. 16 Schematic illustration of solvation effects on Figure 111.17 the textures of silicate products formed in the acid hydrolysis of swelling and non - swelling 2:1 structures. Open circles represent water molecules ............................................................ The idealized structure of acid activated montmorillonite as postulated by Thomas, Hickey and Stecker.31 ............................................................ xxiv 161 162 163 165 166 170 172 Figure 111.18 Schematic illustration of the depletion of octahedral Mg2+ from 2:1 layered silicates by proton edge attack and gallery access me. .L ' 175 XXV BET: BJH: CEC: DCP: GC: ICP: MAS: TCD TEM: TPD: ABBREVIATIONS Brunauer, Emmett and Teller Barrett, Joyer and Halenda Cation Exchange Capacity Direct Current Plasma Fourier Transformed Infra Red Gas Chromatography Inductively Coupled Plasma Magic Angle Spinning Nuclear Magnetic Resonance Thermo Conductivity Detector Transmission Electron Microscopy Temperature Programmed Desorption X-ray Diffraction xxvi , » -3.;‘-.«_-ow‘*j Chapter I Introduction 1.3 Objective and rationale Nanoporous Materials The economy of the United States, indeed the economy of the entire world, depends critically on nanoporous materials. Consider, for instance, the fact that almost all of the petroleum-derived fuels are processed over catalysts that contained a class of aluminosilicates known as zeolites. These open framework structures can adsorb a wide variety of organic molecules on their intracrystal surfaces. Once constrained in the nanoporous space of zeolites, organic reagents can be transformed with unique efficiency into specific reaction products. Thus, many of the fine chemicals that support large volume manufacturing technologies are produced using nanoporous zeolites and related shape-selective materials as catalysts.1'4 In addition to their use as shape selective catalysts, nanoporous solids exhibit adsorption and ion exchange properties useful for a wide variety of advanced technological processes including environmental pollution control, the design of new structural composites and novel electronic, optical and magnetic devices.58 In general terms, nanoporous materials are solids with an accessible open space of 1.0-10 nm range.9 In describing porous materials, the 1UPAC ’ 1 2 has defined three size domains: micropore, <2 nm; mesopore, 2-50 nm; and macropore, >50 nm. Thus, the nanoporous regime spans the traditional mid- micropore to lower -mesopore range. Meso- and macropores are associated with the materials that are either finely divided or structurally highly disordered (amorphous). That is, meso- and macro porosity often are consequences of the texture of a material. Figure 1.1 illustrates the "textural pores" arising , for example, from the random aggregation of platy particles and from the voids formed within the grains in an amorphous solid, such as silica gel. Micropores can also result from the textural properties of materials (e.g., carbon molecular sieves), but, more commonly, micropores are associated with crystalline materials with open framework structures. In zeolites and other molecular sieves, for example, the oxide framework defines open channels and cavities that can accommodate guest molecules. As illustrated in Figure la, these "crystallographic" pores are rigorously regular on an atomic scale. In contrast, textural pores normally exhibit broad size distributions, some over hundreds of nanometers. Until recently, relatively few microporous materials approached the nanoscopic regime. For instance, the faujasitic zeolites used for cracking petroleum are accessed through 12—membered oxygen rings of approximately 0.74 nm diameter.10 Molecules with kinetic diameter substantially larger than 0.74 nm are unable to access the intracrystal surface of these zeolites. Zeolites, VPI-S and cloverite have 18- and 20- ring apertures, respectively“:12 In the case of VPI-S, the pore opening is ~l.2 nm in diameter, but in the case of cloverite the ring is not symmetrical. Crystalline materials with regular pores in the 1-10 nm regime are of considerable current interest because they offer exiting new arenas for a. "Textural Pores" b. "Crystallographic Pores" Figure 1.1 Frequently observed pore types. 4 molecular assembly and chemical reactions. Early in 1992, Kresge and his co-workers at Mobil disclosed in the patent literature13 the synthesis of nanoporous zeolites with channel size of 6.0 nm or more by using liquid crystal templates to direct the crystallization of the aluminosilicate framework.13b, 13c Silica Silica, "Si02" is an almost universally present and is a frequent constituent of clays. It exists in three different basic structural types as amorphous silicon dioxide, quasicrystalline layered silicic acids and crystalline three-dimensional silica Silicon dioxide, especially the amorphous form which can be fabricated in the purest state, is a material of very considerable technological importance. It is used in microelectronics devices containing metal-oxide- semiconductor transistors, optical fibers, etc. It also forms the basis of the glass matrix used for stocking radioactive waste materials.14 Furthermore, silica is used in various other applications such as ceramics and cement industries. Silica supported transition metal complexes serve as heterogeneous catalysts with high stability, good dispersion. Moreover, these catalysts are free of contamination and have high reaction rates. 15 Silica exists in many crystalline forms, the better known being quartz, cristobalite and tridymite. Quartz, cristobalite, tridymite and amorphous silicon dioxide are all built of $104 tetrahedra linked together so that each oxygen atom is common to two tetrahedra giving the composition SiOz. However, the spatial arrangement of the tetrahedral links are quite different in these different forms. In these structures some, often about one -half, of 5 the tetrahedral positions are occupied by aluminum, and rarely, by beryllium. Other positive ions such as Na+, K+ and Ca2+ are present to neutralize the negative charge of the (Si, Al)02 framework. These framework structures and some of their physical properties are more easily understood if they are subdivided based on whether or not they contain polyhedral cavities or tunnels in their structures. Examples of these are feldspars, zeolites and ultramarines. The feldspar structure is relatively compact, but polyhedral cavities or tunnels are found in the other two groups. These cavities or tunnels are filled with water molecules ( in zeolites ) or finite anions (Cl', 8042', 82' etc. in the marines), in addition to the necessary number of cations.16 Zeolites are potentially useful in industrial applications because of their capability to adsorb gases, vapors and liquids and to act as cation exchangers. Nowadays zeolites are widely used as catalysts in photochemistry and in organic synthesis.17 Due to their layered structure, lamellar silicic acids show properties similar to those found in some layered silicates. Some of the interesting properties of the layered silicic acids are interlamellar sorption of water and other polar molecules and their ion exchange capacity. Owing to the presence of interlamellar SiOH groups, which are capable of ligating metal ions, the layered silicic acids are excellent hosts for the immobilization of metal complexes.18 A naturally occurring crystalline silicic acid, silhydrite, was discovered in Trinity county (CA) magadiite deposit.19, 20 Many of the silicic acids are synthesized via the acid treatment of alkaline layered silicates. A large number of naturally occurring alkali metal silicate minerals with layered structures have been discovered in alkaline lakes 6 around the world over the last five decades.21 Also synthetic forms of silicic acids and other alkali metal silicates are readily obtainable.18 Established Approaches to the Synthesis of Crystallographically Regular Nanoporous Materials. Materials with silicate frameworks, such as zeolites, are known to be good adsorbents and catalysts due to their porous structure. The direct crystallization method promises to be a fruitful approach to the synthesis of nanoporous solids, at least for frameworks based on comer-shared $104 and A104 tetrahedra. In addition, there recently has been developed a conceptionally complementary synthetic strategy for the design of nanoporous solids based on the structural modification of layered solids by pillaring reactions. Pillared layered materials also play such an important role in this aspect but not as good as Zeolites}; 23 The porous materials derived from the pillaring of lamellar solids are cationic smectite clay324'28, anionic layered double hydroxidesZ9'32, and many other cationic structures, including layered titanates,33 phosphates and phosphonates,34'40 silicates,41 niobates.42 Pillaring of these materials can be achieved by different experimental methods. In the case of smectite clays and layered double hydroxides, ion exchange intercalation procedures are preferable (Figure 1.2). The metal phosphates are particularly interesting because, unlike smectite clays, the gallery surfaces are chemically functional. Figure 1.3 shows the structure of CHYDRATION * SWELLING PILLAR ION EXCHANGE v v v f' v v \\\\\\ \\\\\\\\ I I I \\\ I I I I I I \ \ \\\ I I II III III III III III III I I I I I I I I I \ \ I \ \\\\\\ I II \ \ \\\\\\\\\\\\\\\ IIIIIIIIIIIIIIIIIIIIIIIII \\\\\\\\ \\\ \ \\\\\\\ 114441111 III I VVV‘VV‘V‘ IIIIIIIIIIII \\\\\\\\\\\\\\\\\ IIIIIIIIIII \\\\\\\\\\\\\ II I I \\\\\\\\\\\ \xxxxpxxpxxx\\\\\x\xxxxx \ IIII I \\\\\\\ III I I \\ \\\ \ \\ \ I I I I I I I I I I IIIIIIIIIIII \ \ Figure 1.2 Schematic representation of pillaring or intercalation of layered materials. OH OH -"'-oA‘\-A\A‘°¢ 1.‘§VEAL ‘VEAL‘V:’4:AL" of OH Figure 1.3 Cross-linking of layered materials via flmctional groups. 9 pillared material obtained by replacing some of the P-OH groups with rigid difunctional phosphate groups to form crosslinking bridges between adjacent layers.43 Another recent approach for designing nanoporous layered materials is to form the pillaring species directly within the galleries of the host (Figure 1.4). This concept was first demonstrated for the hydrolysis and polymerization of tetraethylorthosilicate (TEOS) in the galleries of layered titanates and silicates interlayered by a1kylarnrnoniumions.44 New Approach to Synthesis of Nanoporous Materials The main objective of this work is the synthesis of new layered silicic acids from clay precursors. The significance of this work would be the synthesis of new nanoporous structural types not achievable by direct crystallization. Novel reactivity for catalysis and adsorption would also be advantages. The general strategy is to utilize clay as templates and convert these materials to new layered silicic acids by topotactic acid hydrolysis. The use of concentrated acids to selectively dissolve clay minerals was developed in the early part of the twentieth century as a means for obtaining aluminum solution for the production of aluminum metal. Octahedral cations, such as Al3+, Fe3+r 2+, Mg2+ can be depleted by treating the clay minerals with concentrated acid solutions at elevated temperatures.45'47 In addition to strong acids (e.g. hydrochloric, nitric and sulfuric) weak acids(e.g. acetic), strong bases(e. g. sodium hydroxide), neutral solutions(e. g. sodium chloride) as well as natural sea water have also been used to examine the selective extraction of cations from clay minerals.48 Our approach is the examination of the acid hydrolysis of 1:1 and 2:1 layered silicates and the 10 \w \agi.‘ -‘\- snow .‘. “V‘s. .QSSSQ :\'\'§v . ‘ $§valf I. ‘ .z. .‘l‘f ©: counterion ”~91.“ “\Am\ A\\\A\ \u' s‘g\\ ~\\‘\\9\\ \): \v \‘W 21"{\‘ ‘x..\ MW \v ’\I\ \r.'x’ 31‘ \- ° ' “warms.“ Weswfio ~ \ “‘ ‘ V “\ u - ‘ ““ifism‘ Q ~' .-'. ...... 1.; . ,_:<-_- ----- ‘(fiufis . Vet-x or .- - 8&3, View ' . . \ :3" ' .""'2'.".""?'3' . ~ “‘ \O’T“ - -' ' “... . -’ $39. _ (t\ .. (. ~ ' ' -A\',V\\V~k ’.\'\( .1 .. .‘h i. o o ‘ z$ .g.‘ v\\\Q- -. .;: ...-.... . .3. . .... Figure 1.4 Stepwise in-situ pillar formation of layered materials.44b. 11 characterization of products by using various analytical techniques (e. g. MAS NMR, 1R, XRD, physical sorption of molecules, elemental analysis and TEM). Topotactic Reactions According to Gunter and Oswald“9 the term "topochemistry" was first used by V. Kohlschutter to describe reactions occurring in or at the surface of a solid, which often show specific influences of the substrate on the kinetics and mechanism of the reaction as well as on the properties of the product. The porosity, particle size, shape, surface structure, crystal structure or even chemical composition of the solid substrate can influence topochemically.50 Such reactions have been studied extensively by Kohlschutter and later by Feitknecht.51,52 The growing field of applications of X-ray and electron diffraction revealed a great number of topochemical reactions, in which the property influenced most strikingly by the substrate is the product crystal structure. The necessity of introducing a special term for this type of reaction was first recognized by Lotgering in 1959.53 He proposed the term "topotaxy" for all the solid state chemical reactions that lead to a material with crystal orientations correlated with the crystal orientations of the starting material. At about the same time, Bemal introduced the term "metataxy" for the same reaction, but Lotgering's expression gained the general acceptance. Bemal reduced the wide range of Lotgering's definition in 1960 by requiring a three dimensional accord between original and product crystals for a reaction to be called topotactic.54 He stated that in general the main directions of symmetry of the reacting crystal remain intact in topotactic reactions. This definition was expanded by Mackay to include structural transforrnations.55 Furthermore, according 12 to Mackay the term should only be applied if the majority of the atomic positions remain fixed. This was again altered by Dent-Glasser and co- workers56 who adopted an intermediate position with respect to the requirements of structural accord between the strict requests of Bemal and Mackay and the rather diffuse statement of Lotgering. In 1964, Bemal and Mackay again stressed the importance of dimensional and structural correspondence in the three axial directions.57 Shannon and Rossi attempted a new definition of topotaxy by reducing the requirement of Bemal and Mackay by introducing different degrees of topotaxy, corresponding to the extent of perfection of the preferred orientation found in the product crystallites.58 A clear separation of topotaxy from epitaxy was suggested, the latter term being used for oriented overgrowth and not for transformations. Kleber pointed out that most topotactic reactions do not yield single crystal products, but "topotactic reaction fabrics", assuming only "a structural relation" between parent and product crystals, which may be one-, two- or three dimensional.59 In a review of topotactic reactions, Deschanvres and Raveau cover only phenomena included in Bemal and Mackay's concept, treating phase transformations as well.60 Their classification is according to supposed reaction mechanisms (homogeneous, heterogeneous and others), which, however, are often not established definitely. The most recent paper on questions of terminology for crystallographic orientation relations by Bonev defines topotactic reactions as chemical reactions of a solid leading to a product with defined crystallographic orientation with respect to the original crystal.61 This orientation need not be three-dimensional, but may be only two- or even one-dimensional. An additional requirement is the exchange of components with the surrounding, separating topotaxy from endotaxy (in Bonev's l3 terminology), in which only energy, but no components may be exchanged. This definition is supported by a number of observations by Kleber and mainly based on work of Oswald and Gunter who expressed similar views earlier.62’63,49 In 1975 Gunter and Oswald49 concluded that a chemical reaction of a solid may be called topotactic if the product is formed in one or several crystallographically equivalent orientations relative to the parent crystal, if there has been an exchange of components with the surroundings, and if the reaction can proceed through out the entire volume of the parent crystal.64 It can be concluded that a topochemical reaction is a solid state chemical reaction in which a product is formed with a structure that reflects the structure of the original crystal. If the product is formed in a small number of defined, but not equivalent orientations, the reaction is treated as a case of several different topotactic reactions occurring simultaneously. I.b Structure and Properties of Clays The hydrous layer silicates commonly known as clay minerals are part of a larger family called phyllosilicates. The layer silicates considered in this work contain a continuos two-dimensional tetrahedral sheets of composition M205 (M=tetrahedral cation, generally Si, Al, Fe3+) in which individual tetrahedra are linked with neighboring tetrahedra by sharing three comers each (the basal oxygens) to form a hexagonal mesh pattern (Figure 1.5a). The fourth tetrahedral comer (the apical oxygen) points in a direction normal to the sheet and at the same time forms a part of an immediately adjacent octahedral sheet in which individual octahedra are linked laterally 14 Figure 1.5 The formation of composite silicon-aluminum-oxygen or silicon— rnagnesium-oxygen layers. 16 15 by sharing octahedral edges (Figure 1.5b). The common plane of junction between the tetrahedral and the octahedral sheets consists of the shared apical oxygen and unshared OH groups that lie at the center of each tetrahedral six-fold ring at the same z-level as the apical oxygen. F may substitute for OH in some species. The Octahedral cations normally are Mg2+, A13+, Fe2+ and Fe3 +. The assemblage formed by linking one tetrahedral sheet with one octahedral sheet is known as a 1:1 layer. In such layers the uppermost, unshared plane of anions in the octahedral sheet consists entirely of OH groups. A 2:1 layer links two tetrahedral sheets with one octahedral sheet. In order to accomplish this linkage, the upper tetrahedral sheet must be inverted so that its apical oxygen points down and can be shared with the octahedral sheet below (Figure I.5c and d).16, 65, 66 2:1 Clays The idealized oxygen framework of a 2:1 clay is shown in Figure 16. In a unit cell formed from twenty oxygen and four hydroxyl groups, there are eight tetrahedral sites and six octahedral sites along with four cavities surrounded by a six-membered oxygen ring on the surface. When two thirds of the octahedral sites are occupied by cations, the mineral is classified as a dioctahedral 2:1 phyllosilicate. A trioctahedral 2:1 phyllosilicate has all the octahedral sites filled with cations. Based on the magnitude of the layer charge per unit cell, 2:1 phyllosilicates are divided into five groups; talc- pyrophillite, smectite, vermiculite, mica and brittle mica (Table 1.1)65’ 67. The members of each group are distinguished by the type and location of cations in the oxygen framework. Figure 1.6 Idealized oxygen framework of clay minerals. l7 efiovomoaav 1533253» $2.”... xmovomoaé 1m: 325$: +52“ A? - to u 5 29:8: BEoztoEEo—z ofiooEm eAzovomko_mv_cw2_ 459803933 23. flamingoim 25.653233; 3605302; 3.605385 95on 385—2 53:89.3” mozm 3605300 28 38558. E .83 1302 2:on 8x85 98 momoficemm 2: 2:88”— :omo 5 2028:3235 fim ozgcomamfi .8 22:88”— :aaosbm ©3232 fi— «35,—. 56839 1.325?» 32.”. oiouoocegi xmovomoaé usercwéofa $2.“ afovaoaé 2&5st $2”. 8:593 2:85:02 xmovowoazxammv .32 53» $2.“. 2298012 ice 330mm» .52...“ 289% 2:325 6.280 _.= 03a? 19 «$9825ng REEENS c.2256 afiovoaofizcé SEE BEoonm exmovomofiflaé .3269; 32m 8:22.59» 459824239 3390 Scams—2 «Eovomofifiav 33mm 23832 8:2 22cm 8:2 a: - NA n 5 2:32.23» 6.250 ~.= 035,—. 20 £0va.Qawiafmmxmzova£90mof;rm:mvm $321-wa cacaom 20$:.SamiarmmaazovmEovowormam£61 +m£N$mw§ 238»..me 3:0 39: 3605302; 333.5, n xv c._c_acfi_aa-o=_cacm .3230 fiflo—aarr 21 111 talc, all the tetrahedral and octahedral sites are filled by Si4+ and Mg2+ respectively, and the layers are electrically neutral. In pyrophyllite only two thirds of the octahedral sites are occupied by Al3+ ions leaving a neutral aluminosilicate layer. Therefore, in crystals of these minerals the layers are coupled through relatively weak dipolar and van der Waals forces.68 In contrast to talc and pyrophyllite, the layers in muscovite and phlogopite bear a net charge of 2 e‘ per Si8020 unit due to a positive charge deficiency which results from the substitution of Si4+ by A13+ in tetrahedral positions. The charge deficiency is balanced by interlayer potassium ions which are coordinated to the hexagonal arrays of oxygen atoms at the layer surface.65 Among the 2:1 layered phyllosilicates, the brittle mica group has the highest layer charge, 4 e' per Si8020 unit cell. The layer charge in vermiculite arises from the substitution of Al3+ for Si4+ in the tetrahedral layer. Vermiculite has a varying layer charge depending on the amount of substitution, 1.2 - 1.8 e' per Si8020 unit. The charge on the layers of smectite is intermediate and varies from 0.4—1.2 e‘ per Si8020 unit.69 To balance the layer charge, layers of hydrated cations are intercalated between the clay layers. The moderate layer charge in smectite clays gives physical and chemical properties that are not found in the end members. In montmorillonite, the most familiar and common member of the smectite group, the layer charge originates from the substitution of octahedral Al3+ by Mg2+. Hectorite is also octahedrally charged with Li+ substituting for Mg2+ in the octahedral sheet. Nontronite is a tetrahedrally charged smectite with Al3+ replacing Si4+ in the tetrahedral sheet and Fe3+ replacing Al3+ in the octahedral sheet. Laponite and fluorohectorite are synthetic hectorites and they represent two extremes in 22 particle size and layer charge within the smectite group. Fluorohectorite has a particle size of up to about 2000 A with a layer charge of 1.2 e' per Si8020 unit and originates from the substitution of Li+ for Mgz+ in the octahedral layer while laponite has a particle size of 200 A with a layer charge of 0.4 e' per 318020 unit as a results of substitution of Li+ for Mg2+ in the octahedral layer. In fluorohectorite, all the OH groups have been replaced by F. The layer charge in octahedrally charged clays is distributed over all oxygen in the framework. These clays tend to be turbostratic, that is the layers are randomly stacked with respect to the a-b planes of adjoining layers. The negative charge on the layers of the tetrahedrally charged smectites is more localized, and these derivatives tend to exhibit greater three dimensional order.66r7O The sixth group of 2:1 type silicates is sepiolite-palygorskite (Table 1.1). The structures of these minerals can be regarded as containing ribbons of the 2:1 phyllosilicates structure, one ribbon being linked to the next by inversion of SiO4 tetrahedra along a set of Si-O-Si bonds. Ribbons extend parallel to the x-axis, and they have an average width along Y. The tetrahedral sheet is continuous across ribbons but with apices pointing in different directions in adjacent ribbons, where the octahedral sheet is discontinued. Consequently, with this framework rectangular channels also run parallel to the x-axis between opposing 2:1 ribbons. As the octahedral sheet is discontinuous at each inversion of the tetrahedra, oxygen atoms of the octahedra at the edge of the ribbon are coordinated to cations of the ribbon side only, and coordination and charge balance are completed along the channel by H+, H20 (bound), and a small number of exchangeable 23 ** bzrrex - o ,0:L‘o::‘o’ : . o o, o . .7 §' '. .‘ air 0" or?“ ‘0‘\.»2, /¢'\,91‘3:. a: o.» 0’ N . I l N a = a a H ””3 ”03‘903‘Pc'.‘ PALYGORSKITE _ b : 26.95 X . /,“ it ° ‘ __ ..o/ 0" Z... I ..'..e, 0‘- W ‘ 03:13 "i’d‘k‘. . S E P I O LIT E ' Q+Q=O Q=OH .szo o:Si o=NaMg Figure 1.7 Schematic structures of palygorskite and sepiolite.71 24 cations. In addition, a variable amount of zeolitic water is contained in the channels (Figure 1.7a and b).71 1:1 Clays All the regular structures of the 1:1 layer types are included in the serpentine-kaolin group (Table 1.2). Almost all the 1:1 layer structures have been studied in detail, the position of adjacent layers is determined by the pairing of each oxygen of the basal tetrahedral surface of one layer with an OH of the upper octahedral surface of the layer below (Figure 1.8). This O-- —OH pairing results in the formation of long and relatively weak hydrogen bonds, approximately 2.9-3.0 A between anion centers, that bond the layers together. The O--OH pairing can be obtained by several different ways of superimposing one layer on top of another, this variable represents the second degree of freedom in 1:1 layer structures. Additional electrostatic attraction between layer surfaces may also exist as a result of tetrahedral and octahedral substitutions. Although a 1:1 layer ideally must be electrochemically neutral, coupled substitutions can create balanced negative and positive charges on the tetrahedral and the octahedral sheets, respectively.65 In a unit cell formed from ten oxygen and eight hydroxyl groups, there are four tetrahedral sites and six octahedral sites along with two cavities surrounded by a six-membered oxygen ring on the surface. The dioctahedral, trioctahedral classification is the same as that in 2:1 structure types. The kaolin minerals include all the dioctahedral 1:1 layer silicates. The common unit cell formula for the kaolin group is A148i4010(OH)8.72 25 The serpentine minerals, lizardite, chrysotile, and antigorite represent the trioctahedral 1:1 layer silicates. The crystal morphologies and structures of Table 1.2 Idealized Structural Formulae for representative 1:1 Phyllosilicates. In each Formula the Parentheses and Bracket Define Metal Ions in Tetrahedral and Octahedral Sites Respectively. Mineral group Dioctahedral Trioctahedral Kaolin-Serpentine Kaolinite Anti gorite [A14](Si4)010(0H)8 Lizardite Crysotile [Mg61(Si4)010(0H)8 the serpentines range from planar (lizardite) to alternating waves (antigorite) to cylindrical rolls (chrysotile).65 These three serpentine minerals have similar chemical compositions, with the occupancies at the octahedral sites dominated by magnesium. The lateral dimensions of an ideal magnesium occupied octahedral sheet (b= 9.43 A), are larger than those of an ideal silicon occupied tetrahedral sheet (b= 9.1 A). The misfit between sheets is significant, and leads to the three serpentine structures (Figure 1.9). Each has different solution to the misfit problem. In lizardite, the misfit is accommodated within the normal, planar 1:1 layer structure. In chrysotile, the misfitis partly overcome by the cylindrical curvature of the layer. In antigorite, the misfit is overcome by the curvature of the alternating wave modulation. This is one of the few examples in layer silicates in which three different structures occur in a material of the same or similar chemical composition.73 Since most of the work in this thesis is based on anti gorite, 26 1 I I I v C N choice. 10"! __L_ Figure 1.8 Pairing of tetrahedral oxygen atoms (stippled) at the base of an upper 1:1 layer with the octahedral OH groups (large double circles) on the upper surface of the layer below.65 d§wrflfim ...-...; . 27 Layered W W W a Lizardite Scroll Inverted Wave W vvvvvy,‘ v Antigorite Figure 1.9 Schematic [010] projections of the three species of serpentine rninerals.65b 28 it is appropriate to discuss the antigorite structure in detail. Antigorite Antigorite is generally some shade of medium to pale green, but can be pale buff or gray. Antigorite serpentines are usually tougher and less porous than lizardite or chrysotile serpentines. The color of rocks composed of antigorite, as well as other serpentine minerals, is determined by the distribution of accessory minerals, principally magnetite. Antigorite occurs as fine to coarse, interpenetrating anhedral blades that give the rock a toughness not found in other serpentines. Some of the coarse blades have been used for single crystal studies. The first single-crystal study of antigorite was done by Aruja (1945).74 The structure given in Figure 1.10 (From Kunze, 1956)75 is in the form of an alternating wave extending along the x axis, the super structure direction. Kunze proposed the superstructure A: 43.3 A and other cell dimensions, b=9.23 A C: 7.27 A, similar to the other serpentine minerals.75'78 The tetrahedral sheet is continuous through the structure, but reverses polarity at the midpoint where the wave changes its direction of curvature. The octahedral sheet is also continuous, but it too reverses at the midpoint and is bonded to different tetrahedral sheets in each half of the structure. The antigorite structure A is not fixed at 43.3 A, but varies among certain preferred values. The most commonly occurring values are A = 33.7-43.31 A (Zussman et a1., l957)79.80, A = 25.7-51.5 A (Kunze, 1961)28 and A = 32.8-51.4 A (Uehara and Shirozu).81 Kunzu defined the superstructure period in terms of In, the number of tetrahedra in the super period. When in is odd, the reversal in the tetrahedral sheet at PP' and RR' (Figure 1.11) occur in normal 6-fold tetrahedral rings, but the reversals at 29 Figure 1.10 [010] projection of the structure of antigorite; the structure reverses polarity at PP', QQ; and RR'.65 30 QQ' occur in unusual 8- and 4-fold rings. Under these conditions, the Mg- octahedra at all inversion points are only slightly distorted.23 There are 17 tetrahedra and 16 octahedra within one super period.65 I.c Early Studies A number of natural silicates yield a residue of hydrated silica when acted on by acid. Rinne82 concluded that the hydrated silica residue constituted a well ordered remnant of the biotite structure and suggested the name baueritization for the natural hydrolysis of biotite. A few years later (Rinne, 1924) he reported that "X-ray crystallograms are no longer obtainable for the residual silica of zeolite and mica.‘3.3 Wyart examined hydrolyzed single crystals of zeolite by the rotating crystal X-ray method and obtained only characteristics of amorphous silica.84 Mehmel85 made an extensive study of the hydrolysis of biotite by acids. In those cases in which he studied the product by the powder X-ray diffraction method; only amorphous silica was indicated. Pabst found that Gillespite, a tetragonal sheet silicate, BaFeSi4O1o is hydrolyzed by hydrochloric acid leaving flakes of hydrated silica. These flakes retained the main features of the sheet structure to X-ray precession patterns. Gradual distraction of this structure has seen upon thermal dehydration to an unoriented structure.83 In 1959 Lopez-Gonzalez and Cano-Ruiz reported that the surface area of the acid treated vermiculite increased with increasing concentration of acid. They noticed that the hydrolyzed structure lost its surface area upon calcination.86 Luce and coworkers reported that the diffusion coefficient of magnesium is greater than that of silicon for some magnesium silicates towards acid hydrolysis and hence the rate of extraction of magnesium was greater than that of silicon.87 According to Rice and Strong, the dissolution of Mg2+ is 31 6-ring Lizardite 8-ring . , modules modules modules 1111:3111,“ H A A U CS ' *1 ring rjeversed 8-ring reversed 6-ring reversed a a a a... r .7 ‘7 ' a a ... V W \ I r 7 v v - as a a A C U a (A Figure 1.11 [001] projection of the tetrahedral sheet of antigorite structure. 32 caused by the rupture of weak Mg-O bonds.88 Any acid treatment depletec some silicon, but the rates were very small. Girgis and Mourad89 in 1976 have reported the same observation as Lopez-Gonzalez and Cano-Ruiz wit} serpentine upon acid hydrolysis. Both groups described the higher surface area obtained as a result of the formation of some kind of silica called "free silica". Suquet and coworkers90 recently obtained a porous, less crystalline hydrated silica material with original platy morphology upon acid hydrolysis of vermiculite. Harkonen and Keiski studied the conditions for the acid extraction of cations from phlogophite and obtained a porous material comparable to the porosity of active coal.91 Corma and coworkers (1987: have reported that the hydrolysis order of Mg?+ > Fe3+ > Al3+, fo: palygorskite indicates that most of the octahedral positions at the edges are occupied by Mg2+, some by Fe3 +, and partially or none by Al3+ ir palygorskite.92 Corma and coworkers also reported in 1987 that the acic' hydrolysis of up to 60% of the octahedral cations in palygorskite minerals does not provoke any structural changes or arnorphization. Gonzalez and coworkers (1989) reported that the acid activation is much higher in the magnesium-rich palygorskite, which undergoes a greater dissolution of its octahedral sheet as well as a more extensive alteration of its structure and a greater increase of microporosity.93 They indicated that as the acid hydrolysis progressed, the octahedral sheet increasingly dissolved, creating microporosity between the tetrahedral silicate sheets. In a third stage. increasing acid concentration and/or reaction time condensed silanol groups, and the microporosity decreased. N o microporosity was observed for the acid activated sample after silica extraction. The fibrous morphology of the samples was preserved throughout the acid treatment, both before and after silica extraction.94 Corina and coworkers (1990) concluded that acid attack 33 of palygorskite does not obey a purely diffusional controlled process According to them diffusion is the controlling factor at low levels 0: hydrolysis, while at high levels the chemical reaction (hydrolysis) can be the rate determining step. The table below shows the activation energies for the extraction of three different octahedral cations as reported by Corma anc coworkers.95 The dissolution of Mg2+ from sepiolites using acids was firs attempted by Abdu-Latif and Weaver (1969) who found that the dissolutior of palygorskite and sepiolite in excess hydrochloric acid was a first order reaction in terms of concentration of A1, Fe and Mg.96 In several studies the Table 1.3 Activation energies of octahedral cations. MgZ-l- A13+ 1363+ AE (kcal/mol) 3.05 9.01 4.2 acid activation of sepiolite has been carried out to determine the structural changes in sepiolite [Shirnosaka et a1.(1973)97; Corma et al.,(1986)98], tc improve the surface area and porosity [Jimenez-Lopez et a1. (l978)99; Rodriguez-Reinoso et a1. (1981)]100: the amount of surface acidity [Bonilla et a1. (1981)]101 and to obtain good quality products to use as adsorbents [Gonzalez et a1. (1984)]102, filling materials [Gonzalez et a1. (l982)103. Acosta et al. (1984)104] and catalysts [Dandyn & Nadiye-Tabbiruka (1982)105, Corma et a1. (1984)106, Corma & Perez-pariente (l987)107]. Jimenez-Lopez et al. reported a maximum of 391 ng‘1 BET surface area for acid hydrolyzed sepiolite after a series of acid reactions under different 34 conditions.99 Gonzalez et al. reported a fibrous silica gel with an acceptable purity level and a high specific surface area of 505 m2g.'1 102 Cetisli ane Gedikbey (1990) concluded that the dissolution of Mg from sepiolite is i first order reaction in terms of acid concentration and Mg content of the solid.103 The reaction rate constants were proportional to the acie concentrations, and inversely proportional to the square-root of the initia particle radii of the sepiolite. In 1987 Mendioroz and coworkers acid treatee bentonite and revealed an evolution in the hysteresis loops of the corresponding nitrogen adsorption-desorption isotherms.109 They concluded that in the first stage of the attack, an opening of the bentonite lamella was produced by the abstraction of interlayer cations anc corresponding water molecules. Thus accessibility, otherwise impossible, 01 nitrogen to the internal surface of the samples with subsequent increase ir nitrogen adsorption occurred. A second stage was produced when the octahedral aluminum sheet was dissolved as the acid attack progressed. A: higher levels of acid concentrations, the partial distraction of the tetrahedra sheet leads to the changes in the structure. Jovanovic and Janackocvic (1991) observed a similar pore structure and adsorption properties upon acic hydrolysis of bentonite.110 Bremner et a1. (1984) calcined various kaolir samples from 500 to 900 0C and found only little difference in alumina solubility with 26% HCl.111 Lussier (1991) found that narrow calcinatior windows give acid-reactive calcined kaolin which develops high surface areas.112 He also found that little or no Al3+ is depleted from the solic during the hydrolysis process. 35 References 1. Thomas, J. M. Angew. Chem. Int. Ed. Engl. 1988, 27, 1673. 2. Holderich, W.; Hesse, M.; Naumann, F. Angew. Int. Ed. Engl. 1988, 27, 226. 3. Pinnavaia, T. J. Science 1983, 220 , 365. 4. Vaughan, D. E. W. ACS Sym. Ser. 1988, 368 , 308. 5. Ozin, G. A.; Kupennan, A.; Stein, A. Angew. Int. Ed. Engl. 1989, 101, 373. 6. Stucky, G. D.; MacDougell, J. E. Science 1990, 247, 669. 7. Bein, T.; Enzel, P.; Bauneu, F.; Zupperol, L. Adv. Chem.. Ser. 1990 226 433. 8. a) Mitchell, 1. V. (ed), "Pillared Lamellar Structures", Elsivier, New York, 1990. b) Ozin, G. A. Adv. Mater 1992 4 No. 10 9. Gregg, S. J.; Sing, K. S. W. "Adsorption, Surface Area and Porosity", 2nd Ed., Academic Press, New York, 1982 p. 25. 10. Meier, W. M.; Olson, D. H., (eds) "Atlas of Zeolite Structure Types", 3rd ed., Butterworth-Heinemann, London, 1992 p.96. 11. Davis, M. E.; Saldarriaga, C.; Garces, J .; Crowder, C. Nature 1988, 331, 698. 12. 13. 14. 15. l6. l7. l8. 19. 20. 21. 22. 36 Esterrnann, M.; McCusker, L. B.; Baerlocker, C.; Merrouche, A.; Kessler, H. Nature 1991,352, 320. a) Krwsge, C. T.; Leonowicz, M. E.; Roth, W. J .; Vartuli, J .C. U. S. Patent 1992, 5,098,684. b) Beck, J. S.; Vartuli, J. C.; Roth, W. J .; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T-W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc.I 1992 , 114 , 10834. c) Krwsge, C. T .; Leonowicz, M. E.; Roth, W. J.; Vartuli, J.C.; Beck, J. S. Nature 1992 359 710. Guptha, R. P. Phys. Rev. B Condensed Matter 1985, 32 , 8278. Besses, B.; Moreweck, B.: Smith, A. K.: Basset, T. M. J.Chem. Soc. Chem. Comm. 1980, 1569. Wells, A. F. "Structural Inorganic Chemistry" 4th Ed. 1975 Clarendon Press Oxfoed. lone, K. G.; Vostrikova, L. A.; Mastikhin, V. M. J. Mol. Catal. 1985, 31, 355. Dailey, J. S. Ph.D Dissertation, Michigan State University 1991. Gude, A. J .;, Sheppard, R. A. Amer. Mineral. 1972, 57, 1053. Benake, K.; Lagaly, G. Clay Miner. 1977,12 , 63. Benake, K.; Lagaly, G. Amer. Mineral. 1977, 62 , 763. Pinnavaia, T. J.; Tzou, M.; Landau, S. D. J. Am. Chem. Soc. 1985, 107, 4783. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37 Pinnavaia, T. J .; Tzou, M.; Landau, S. D.; Raythatha, R. H. J. Mol. Catal. 1984, 27, 195. Barrer, R. M.; MacLeod, D. M. Trans . Faraday Soc. 1955, 51, 1290. Barrer, R. M. Pure Appl. Chem. 1989, 61, 1903. Johansson, G. Acta. Chem. Scand. 1960, 14 , 771. Bottero, J. Y.; Cases, J. M.; Flessinger, F.; Poirier, J. E. J. Phys. Chem. 1980, 84, 2933. Burch, R. (ed) "Pillared Clays", Catal. Today 1988, 2, 185. Pope, M. T."'Heteropoly and isopoly Oxometalates", Spring-Verlag, New York 1983. Kwon, T.; Tsigdinos, G. A.; Pinnavaia, T. J. J. Am. Chem. Soc. 1988, 110, 3653. Drezdzon, M. A. Inorg. Chem. 1988, 27, 4628. Kwon, T.; Pinnavaia, T. J. Chem. Mater. 1989, I , 381. Anthony, R. G.; Dosch, R. G. Stud. Surf. Sci. Catal. 1991, 63 , 637. Clearfield, A. Comments Inorg. Chem. 1990, 10 , 89. Johnson, J. W.; Jacobson, A. J .; Butler, W. M.; Rosenthal, S. E.; Brody, J. F.; Lewandowski, J. T. J. Am. Chem. Soc. 1991, 111, 381. Tomlinson, A. A. G. "Pillared Layered Structres" Mitchell, 1. V. (ed) Elsevier New York 1990 p.91. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 38 Alberti, G.; Constantino, V.; Marrnottini, F.; Vivani, R.; Zappelli, P. "Pillared Layered Structres" Mitchell, 1. V. (ed) Elsevier New York 1990, p.119. Cao, G.; Mallouk, T. E. Inorg. Chem. 1991, 30 , 1434. Yamanaka, S.; Hattori, M. Inorg. Chem. 1981, 20 , 1929. Burwell, D. A.; Thompson, M. E. Chem .Mater. 1991, 3 , 730. Schwieger, W.; Heidemann, Bergk. Rev. Chim. Minerale 1985, 22 , 639. Treacy, M. M.; Rice, S. B.; Jacobson, A. J.; Lewandowski, J. T. Chem. Mater. 1990, 2 , 279. Dines, M. B.; Cooksey, R. E.; Griffith, P. C.; Lane, R. H. Inorg. Chem. 1983, 22, 1003. (a) Landis, M. E.; Aufdembrink, B. A.; Chu, P.; Johnson, I. D.; Kirker, G.W.; Rubin, M. K. J. Am. Chem. Soc. 1991, 113, 3189. (b) Dailey, J. S.; Pinnavaia, T. J. Chem. Mater. 1992 4, 855. Peters, F. A.; Johnson, P. W.; Kirby, R. C. BuMines RI 1963, 5 7, 6290. Peters, F. A.; Johnson, P. W.; Kirby, R. C. BuMines RI 1963,57, 6229. Johnson, P. W.; Peters, F. A.; Kirby, R. C. BuMines R1 1964, 27, 6431. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 39 Carroli, D.; Starkey, H. C. Clays Clay Miner. 1977, I9 , 323. Gunter, J. R.; Oswald, H. R. Bull. Inst. Chem. Res. 1975, 53(2) , 249. Kohlschutter, V. Z. Anorg. Allg. Chem. 1919, I 05 , 35. Feitknecht, W. F ortschr. Chem. Phys. Chem. Eucken, A. (ed) Berlin 1930, 21(2A) , 1. Kohlschutter, H. W.; Sprenger, L. Z. Angew. Chem. 1939, 52, 197. Lotgering, F. K. J. Inorg. Nucl. Chem. 1959, 9 , 113. Bemal, J. D. Schweizer Archiv. 1960, 26 , 69. Mackay, A. L. Proc. 4th Int. Symp. React. Solids de Boer, J. H. (ed) Elsevier Amsterdam 1960, 571. Dent-Glasser, L. S.; Glasser, F. P.; Taylor, H. F. W. Quart. Rev. 1962, 16 , 343. Bemal, J. D.; Mackey, A. L. Tschermak’s Miner. Petr. Mitt. 1965, 10 331. Shannon, R. D.; Rossi, R. C. Nature 1964, 202 , 1000. Kleber, W. Kristall u. Technik 1967, 2 , 5. Deschanvres, A.; Raveau, B. Rev. Chim. Miner. 1968, 5 , 201. Bonev, 1. Acta. Cryatallogr. 1972 , A28 , 508. Gunter, J. R.; Oswald, H. R. J. Appl. Crystallogr. 1969, 2 , 196. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. ’ 73. 74. 4o Gunter, J. R.; Oswald, H. R. J. Appl. Crystallogr. 1970 3 21. Oswald, H. R.; Gunter, J. R. 7iemeCongr. Int. Micr. Electr. Favard, P. (ed) Granoble 1970, 417. (a) Bailey, S. W. "Crystal Structure of Clay Minerals and their X -ray Identification", Brinley, G. W. and Brown, G. Eds. Mineralogical Society London 1985. (b) Liebau, F. ”S tructural Chemistry of silicates" Springer-Verlag Berlin Heidelberg 1985, 93. Farmer, V. C. "The Infra-red Spectra of Minerals", Farmer, V. C. Eds. Mineralogical Society London 1974, 331. Dixon, J. B. "Minerals in Soil Environments", Dinauer, R. C. Eds. 1977, 357. Giese, R. F. Jr. Clay Miner. 1975, 23, 165. Beson, G.; Mitsud, C. T.; Mering, J. Clays Clay Miner. 1974, 22, 339. Suquet, H.; Calle, de la C.; Pezerat, H. Clays Clay Miner. 1975, 23, 1. Jones, B. F.; Galan, E. "Reviews in Mineralogy", Bailey S. W. Eds. 1988, 19 , 631. Giese, R. F. Jr. ”Reviews in Mineralogy”, Bailey S. W. Eds. 1988 , I9, 29. Wicks, F. J .; O'Hanley, D. S. ”Reviews in Mineralogy”, Bailey S. W. Eds. 1988, 19 , 91. Aruja, E. Mineral Mag. 1945, 2 7, 65. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. ' 41 Kunze, G. Zeit. Krist. 1956, 108, 82. Kunze, G. Zeit. Krist. 1958, 110, 282. Kunze, G. Zeit. Krist. 1959, 111, 190. Kunze, G. Fortschr. Mineral. 1961, 39, 206. Zussman, J .; Brindley, G. W. Amer. Miner. 1957, 42, 666. Zussman, J.; Comer, J. J. Amer. Miner. 1957, 42, 133. Uhera, S.; thirozu, H. Mineral. J. (Japan) 1985, 12, 299. Rinne, F. Math-Phys. Kl 1911, 33, 441. Pabst, A. Amer. Miner. 1958, 43, 970. Wyart, J. Bull. Soc. Fran. Min. 193, 56, 81. Mehmel, M. Chemie der Erde 1937, 11, 307. Lopez-Gonzales, J. De D.; Cano-Ruiz, J. 6th Annual Conferance on Clays Clay Miner. 1959, 399. Luce, R. W.; Bartlett, R. W.; Parks, G. A. Geochimica et Cosmochimica Acta. 1972, 36 , 35. Rice, N. M.; Strong, L. W. Canadian Metaiiurgical Quarterly b 1974, 13, 485. Girgis, B. S.; Mourad, W. Chem. Biotechnol, 1976, 24, 9. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 42 Suquet, H.; Chevalier, S.; Marcilly, C.; Barthomeuf, D. Clay Miner. 1991, 26, 41 . Harkonen, M. A.; Keiski, R. L. Colloids and Surfaces 1984, 11, 323. Corma, A.; Mifsud, A.; Sanz, E. Clay Miner. 1987, 22, 225. Gonzalez, F.; Pesquera, C.; Blanito, C.; Benito, 1.; Mendioroz, S.; Pajares, J. A. Applied Clay Science 1989, 4 , 373. Gonzalez, F.; Pesquera, C.; Blanito, C.; Benito, 1.; Mendioroz, S.; Pajares, J. A. Clays Clay Miner. 1989, 37, 258. Conna, A.; Mifsud, A.; Sanz, E. Clay Miner. 1990, 25, 197. Abdul-Latif, N .; Weaver, E. C. Clays Clay Miner. 1969, 17, 169. Shirnosaka, K.; Kawano, M.; Tanguchi, T. J. Clay Sci. Soc. Japan 1973, 13, 113. Corrna, A.; Mifsud, A.; Perez-Pariente, J. Clay Miner. 1986, 31, 69. J imenez-Lopez, J .; Lopez-Gonzalez, J. D.; Ramirez-Saenz, Z.; Rodriguez-Reinoso, F.; Valenzuela-Calahorro, C.; Zurita-Herrera, L. Clay Miner. 1981, 16, 315. Rodriguez-Reinoso, F.; Ramirez-Saenz, A.; Lopez-Gonzalez, J. D.; Zurita-Herrera, L.Clay Miner. 1981, 16, 315 Bonilla, I. L.; Ibarra, R. L.; Rodriguez, D. A.; Moya, J. S.; Valle, F. J. Clay Miner. 1984, 19, 93. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 43 Gonzalez, L.; Ibarra, R. L.; Rodriguez, D. A.; Moya, J. S.; Valle, F. J. Clay Miner. 1984, 19, 93. Gonzalez, L.; Ibarra, R. L.; Rodriguez, D. A. D. Angew Makromol. Chemie. 1982,103, 51. Acosta, J. L.; Rocha, C. M.; Ojeda, M. C. D. Angew Makromol. Chemie. 1984, 126, 51. Dandy, A. J.; Nadiye-Tabbiruka, M. S. Clays Clay Miner. 1982, 30, 347. Corma, A.; Perez-Pariente, J.; Fornes, V.; Mifsud, A. Clays Clay Miner. 1984, I9, 673. Corma, A.; Perez-Pariente, J. Clay Miner. 1987, 22 , 423. Cetisli, H.; Gedikbey, T. Clay Miner. 1990, 25 , 207. Mendioro, S.; Pajares, J. A.; Benito, 1.;Pesquera, C.; Gonzalez, F .; Blanco, C. Langmuir 1987, 3 , 676. Javanovic, N .; J anackovic, J. Applied Clay Science 1991, 6, 59. Bremmer, P. R.; Hicks, L. J .; Bauer, D. J. The Metallurgical Society of AIMG Technical Paper No. A84-47, 1984. Lussier, R. J. Catal. 1991, 129, 225. Chapter II Acid Hydrolysis of 1:1 Layered Silicate Structure II.a Introduction The serpentine-kaolin subgroup of silicates is the only group of minerals under the category of 1:1 clays. On the basis of the earlier work summarized in section I.c it can be seen that the octahedral magnesium in 1:1 clay minerals is selectively dissolved under acid conditions at elevated temperatures. Aluminum in the octahedral sites of kaolinite is much more difficult to deplete under acid hydrolysis .13 Girgis and Mourad in 1976 reported obtaining a high surface area silica upon acid hydrolysis of octahedral magnesium in antigorite serpentine minerals.3 Results obtained for the acid hydrolysis of kaolinite and antigorite are discussed in this chapter with appropriate conclusions. 45 II.b Acid Hydrolysis of Serpentine II.b.l. Experimental Starting Material The starting material was a greenish black antigorite serpentine rock from Lancaster county Pennsylvania. The major mineralogical component was antigorite and an appreciable amount of a magnetic impurities (magnetite). Large rock particles (about 5 cm diameter) were manually crushed to smaller particles (1 mm inch diameter) and the magnetic impurities were separated by using a magnet. The light green material obtained after the above separation was mechanically ground and the desired mesh fractions were obtained by sieving. The purified antigorite was characterized by elemental analysis, XRD, TEM, 29Si MAS NMR and by FTIR. Acid Hydrolysis After mechanical grinding and separation through sieves the desired particle size fractions were obtained. Acid hydrolysis was carried out by placing the sample in contact with 20 wt% HCl at room temperature and then heating the mixture to the desired temperature (60 0C , 80 0C or 105 0C) with constant stirring. The liquid/solid ratio was 100 cm3/5 g. Moles of H+ per mole of octahedral M2+ (M: Mg, Fe) in antigorite were 5 times stoichiometric ratio. After the desired reaction time at above temperatures, the mixture was filtered, washed until chloride free, and air dried. 46 Physical Measurements The percentage of Mg depleted from the antigorite structure upon acid hydrolysis was determined base on chemical analysis. Chemical analysis was performed by the ICP (Inductively Coupled Plasma) method by using a lithium metaborate flux for preparing the solutions suitable for analysis.9 XRD patterns were obtained by using a Rigaku rotaflex diffractometer equipped with DMAXB software and Ni-filtered Cu K; X-ray radiation. 27Al and 29Si MAS NMR experiments were performed on a Varian 400 VXR solid state NMR spectrometer. A Bruker multinuclear MAS probe equipped with zirconia rotors was used for all the measurements. The 29Si spin-lattice relaxation times (T1) were determined by the inversion recovery method. A total of 12 scans were accumulated for each sample. The spinning rate was 4.2 kHz and the delay time was 600 sec which was 5 times as large as the largest T1 of 29Si MAS NMR signals. Cross polarization experiments were canied out with delay times of 10 sec and contact times of 1000 ms. The delay time used with 27Al MAS NMR was 10 seconds. Adsorption and desorption measurements were carried out on an Ominisorb 360cx Coulter instrument with nitrogen as the adsorbate at 77K except the surface area measurements reported in section II.b.4 which were carried out on a Quantasorb Jr. surface area analyzer . Surface areas were calculated according to the BET ( Brunauer, Emmett and Teller) method. t- Plots were constructed according to either the Boer, Lippens and Osingalo,11 ( for samples with CBET constant >70) or Lecloux and Pirard method12 ( for samples with CBET constant < 70). Because it has been demonstrated that the standard isotherm has to be chosen according to CBET constant, which 47 depends on the adsorbent-adsorbate interactions. Adsorption desorptior isotherms constructed by using the data obtained from the continuou: sorption measurements correspond to the selected p/po values. The sample: were outgassed at 150 0C under a vacuum. The adsorption of othe adsorbates was determined by using a McBain balance equipped with quartz glass springs and buckets. Pore size distribution curves were obtained fron the nitrogen adsorption isotherms by the BJH (Barrett, J oyer and Halenda method13 by using a Coulter Ominisorb Beta 0.05 Analysis software program. Derivative uptake plots were obtained with the same software program used for the pore size distribution curves. Pore diameters tha correspond to the log p/po values were obtained according to the equatior proposed by Kresge et a114 by using nitrogen adsorption data. Micropore volume values were determined by plotting the following equation or Dubinin and-RadusikevichlSI P0 2 Log V= Log Vmic - D( log F . the intercept of which will give the derived value of (V mic in mlg-l). FTIR spectra were obtained on an IBM 1R44 spectrometer by the KB: pressed pellet technique. TEM photographs were obtained on a JEOL 1000) II transmission electron microscope operated at 100kV. Samples were embedded in polymerized epoxy resin and sectioned to 90nm thick films. Temperature programmed desorption (TPD) of ammonia was carrier out as follows. A sample was outgassed at 300 0C in a flow of Ar gas for Z hours and then cooled to 100 OC. Anhydrous ammonia gas was passed over the sample for 30 minutes and then flushed with pure Ar gas for another 3( minutes. Then the sample was switched to an on line GC system and the 48 temperature ramped at a rate of 10 0C/minute up to 300 0C. TCD (them conductivity detector) was used to detect the amount of ammonia desorbe as a function of temperature. The in situ FT-IR spectroscopy of adsorbed pyridine was followed i order to characterize the acid strength and Lewis acid and Bronsted acid si: distributions of the samples. A thin film of sample was prepared by using 1 mg of the sample. The sample was heated under heliumin the IR cell to 35 0C over a period of 2 hours by using a programmable temperature controlle and then evacuated for 1 hour, and finally cooled to 200 0C. The referent spectra were recorded at this temperature on a Nicolet FT-IR 4 spectrometer. Pyridine then was introduced at its vapor pressure level a room temperature for 20 minutes. The pyridine adsorption spectra wel recorded after 10 hour evacuation at 200 0C. The difference between the tw spectra was used to determine the acid strength and Lewis-Bronsted acid sil distribution. Alumination Experiments Alumination reaction of acid hydrolyzed antigorite and startin antigorite were carried out according to the method presented by Fripiat an coworkers for alumination of sepiolite.16 A 5 g quantity of the sample we allowed to react at 90 0C for 6 hours with 0.5M NaAlOz in 6N N aOH i such a way that the Al/Si ratio was between 3 and 14. After the reaction th solid was separated by centrifugation, washed five times with 100 ml ( diluted NaOH (pH=l2.8), and rinsed with 200 ml of water. Finally, th product was stirred for 1 hour in 1.25 L of a 1N NH3 solution at roor temperature three times and rinsed again with water. The conversion of th product into the ammonium exchanged form was completed by rinsing wit 49 a 1M NH4+-acetate solution, the excess salt being removed by washing with deionized water. A blank experiment was done without any antigorite present. The cation exchange capacity (CEC) of the aluminated product was determined by the ammonia exchanged method.17 Determination of Crystallinity Crystallinity is represented as a crystallinity ratio with respect to the starting material. Two methods were used to determine the crystallinity ratio. Method I This method is an extension of the procedure of Wirns and coworkers for determining the crystallinity of polymers.18 We used the following equation to calculate the C% (percentage of crystallinity). C % = JIsdq-UIAdqllkl x100 lIsdq where 13 and IA are the scattering intensities of the crystalline sample and the amorphous derivative. The scaling factor k is the ratio of the intensities at a particular 29 value outside of the crystalline peak region. All the intensities were baseline corrected. The biggest advantage of this method is that the percent crystallinity is independent of the absolute areas of the sample or the amorphous diffraction patterns. Thus it is not necessary to be concerned about measurement and normalization of the incident X—ray intensity 1, sample thickness, size, shape or the electronic structure. A crystallinity index was obtained with respect to the starting material. 50 The scattering intensities for the amorphous reference were obtained by using a totally amorphous antigorite sample, which was obtained by hydrolysis with 20% HCl at reflux temperature for 24 hrs. Oriented film samples were used for all the XRD measurements. Method 11 g In this method phlogopite was used as an internal standard. A 1:1 by weight mixture of phlogopite and sample, each <325 mesh, was prepared by grinding with a motor and pestle. Powder XRD patterns were taken for three different mixtures for each sample to average random errors due to mixing. The average integrated area of the 001 peak of the three different mixtures was obtained with respect to the internal standard. A crystallinity index with respect to starting material was calculated based on this intensity. II.b.2. Results and Discussion Characterization of Starting Material The approximate structural formula calculated from the elemental analysis results was (Si2)[Mg2,9 Ecol] 05 (OH)4. XRD results obtained were very similar to the XRD results reported by Brindley and Santos (Figure 11.1).4 TEM photographs agree well with the early work (Figure 11.2).4'6 29Si MAS NMR showed a single peak at -94.1 ppm relative to TMS. The same results were reported by Lippama and coworkers for serpentine from Sharum, Norway (Figure 11.3).7 FTIR peaks perfectly match the IR peaks reported for antigorite, but not any other 51 001 Lfl 020 1 L3: 3? 07110720‘3‘0‘. U) C Q) ...I E 002 01 ‘ l 2, ' r —---*-.._b 0 10 20 30 Two Theta Figure 11.1 X-ray powder diffraction of antigorite. Inset: X-ray diffra of antigorite reported by Brindley and Santos, 1971. 52 Figure H.2 TEM pictures of antigorite. Top: From Brindley and Santos, 1971. Bottom: From present work. 53 -94.558 -50 -7O -90 -110 -130 [mm Figure 11.3 29Si MAS NMR of starting antigorite. 54 serpentine group minerals (Figure 1.4).8 The BET surface area of the nor porous antigorite was 6 m2 g'l. Highly Active Material from Acid Leaching of Antigorite Girgis and Mourad showed that an increase in HCl concentratio beyond 20 wt% does not increase the weight loss, surface area or por volume (Table 11.1)3 of silica prepared by acid hydrolysis of 1:1 layere silicates at 75 0C. On the basis of these results by Girgis and Mourad, 20‘. HCl acid was used in the present study of antigorite with a particle size <10 mesh. Variable reaction times and temperatures have been studied in thi work. The fraction of Mg depleted as a function of hydrolysis time 2 various temperatures is plotted in Figure 11.5. It is very clear that the rate c acid hydrolysis depends on the reaction temperature. 100% depletion of M was reached faster at 105 OC (refluxing temperature) than 80 0C and 60 0C In addition, the Mg depletion rate at 80 0C is higher than that at 69 0C a indicated by the slopes of the curves at lower reaction times. Tables 11.2-4 show that there is agreement between the crystallinit; ratios obtained for acid hydrolyzed antigorite by methods I and H. Eve: though the method I was developed for determining the crystallinity o polymers where there are no compositional changes of the polymer durin, the polymerization process, it is reasonable to use the same methodology tr obtain the crystallinity of acid hydrolyzed clays where there is a loss 0 octahedral metal atoms. The mechanism for depletion of octahedral cation by H+ will be discussed later in detail. ‘Qb liANSMHlANCE 100 80 1 ransnutance 8 20 55 coco 1000 3000 3‘00 320. 1.0 2.00 :00 Y r . 4 T .fi Y 1 IO > 4 4 60 > II J ‘ 3565 \ 3500 ‘ IO 1 4 2° _ .73,“ I 3650 ‘ :mrz 4 o 1 x I a 1 A 2 O 2 O 3 O 10 12 1 0 3 o p d ‘I- p q r- d P II! b d _ 3679 a _ '1 p - 4000 3800 3600 3400 3200 3000 Vszenunuxr Figure 11.4a Top: IR spectra of antigorite from 3000 cm'1 to 4000 cm-1 (Vander and Beutelspacher, 1976). Bottom: FTIR spectra of antigorite (present work). 56 1900.000 1)» 13.0 "00 .000 ’00 T 1074 1111111 # 987 A L 1360 1140 VVauenunux: 1580 3 920 Figure 11.4b Top: IR spectra of antigorite from 700 cm“1 to 1800 cm-1 (Vander and Beutelspacher, 1976). Bottom: FTIR spectra of antigorite (present work). 57 700 030 600 $50 $00 ‘50 q .- .4 .4 '1 -e . .1 111111111 Figure 11.4c Top: IR spectra of antigorite from 400 cm'1 to 700 cm-1 (Vander and Beutelspacher, 1976). Bottom: FTIR spectra of antigorite (present work). 58 120 '0 . 0) (>3 1004 A Q) . ’- A g 80-4 A D O 13 ‘ :1 o o .. U) 60 U o 3 143 o 5 n U 40-1 o O B A At 105°C ‘1' ‘0 D At 80°C 0 At 60°C 20 . , . , . I . a 0 20 40 60 8 Reaction Time (hrs.) Figure 11.5 Percent of depletion of magnesium from antigorite (>100 me verses the reaction time at different temperatures. The initial ] concentration was 20 wt%. The H+/Mg2+ was 5/1. Table 11.1 Effect of various HCl concentrations on the physical and 59 chemical characteristics of serpentine (acid treatment for 7 h at 75 0C) Acid concentration 10% 20% 25% 30% 35% 40% Weight loss (%) 36.80 46.10 46.80 49.50 50.6 - Surface area 124 199 236 196 227 229 (ng'll Pore volume 0.123 0.165 0.165 0.169 0.177 0.224 (mlg'l) Girgis, B. S.; Mourad, W. E. Chem. Biotechnol. 1976, 26, 9. Figure 11.6 shows the relationship between the amount of octahedral Mg depleted and the crystallinity ratio obtained according to method I. It is obvious that the change in the crystallinity with the removal of octahedral Mg at these three different temperatures is more or less the same. In other words, the pattern of crystallinity changes depends on the amount of Mg depleted. Up to 70% Mg removal the rate of change in crystallinity is very low compared to the change that occurs beyond 70% Mg depletion. The initial change in the crystallinity ratio below 70% Mg hydrolysis most likely reflects the decrease in the number of Mg atoms available for x-ray diffraction, but the oxygen frame work of the hydrolyzed structure remains the same as that of the starting clay. Relatively rapid loss in the crystallinity ratio above 70% Mg depletion can be explained in term of loss of framework oxygen. From this observation we can conclude that a well crystallized 60 layered silicate can be obtained with only 30% of the original octahedra Mg remaining in the octahedral layer. Table 11.2 Crystallinity and magnesium depletion with respect to reactio: time for antigorite samples hydrolyzed at 60 0C Reaction time Mg% depleted Crystallinity ratio (hf) Mtd 1 Mtd 2 1 27 .88 93 3 39 .88 94 9 41 .88 92 15 52 .86 .89 24 56 .84 .96 48 69 .81 .83 72 79 .74 .56 Figure 11.7 shows a typical X-ray powder diffraction pattern for aci. hydrolyzed antigorite at different levels of Mg depletion. Broadening of till 001 XRD peaks was not observed with these oriented peaks as one WOUll 61 Table 11.3 Crystallinity and magnesium depletion with respect to reactio time for antigorite samples hydrolyzed at 80 0C Reaction time Mg% depleted Crystallinity ratio Mtd 1 Mtd 2 1 43 .86 .92 3 52 .85 .85 6 57 .85 .76 9 68 .79 .74 15 75 .54 .55 24 94 .32 .18 48 97 .6 .1 72 96 .1 _ .ll expect with structural alterations (depletion of Mg). Figure 11.8 shows thl XRD peaks observed in the region of 50-70 29 for the transparent powde samples run at transmission mode (reflected beam of X-ray was detected) The in-plan 24 3 0, 0 6 0. and 0 6 1 reflections can be observed. With the increasing depletion of Mg there is little modification in these in-plan XRI peaks. Modifications to a reflections other than 001 peaks suggests that there 62 Table 11.4 Crystallinity and magnesium depletion with respect to reactio time for antigorite'samples hydrolyzed at 105 0C Reaction time Mg% depleted Crystallinity ratio (hr) ' Mtd 1 Mtd 2 .25 — - .92 1 52 .86 89 3 78 .45 .43 6 87 .38 .24 9 93 .4 .24 15 98 .05 .06 24 99 <.05 ~0 may be some rearrangement of layers. Figures 11.9, 11.10 and 11.11 show the relationships between 1.11! percentage of Mg depleted, BET surface area, and the reaction time at 64 0C, 80 0C and 105 0C respectively. In all these cases a continuous increasl in the amount of Mg hydrolyzed was observed. However, the surface are: showed a maximum at about 60-80% Mg depletion and then started te decrease as Mg depletion become more complete. Similar behavior wa; observed by Girgis and Mourad for acid hydrolysis of antigorite serpentine 0.9 o 0% 951 48 0.74 o J U .2 [:1 E 0.5- 3. A 3’ ”:1 0.3~ D 0 0.1“ D A A at 105°C A D at ao-c -03 ° elm. . - . 0 20 4o ' 60 80 r1001120 Percentage Mg Removed Figure 11.6 Relationship between the amount of magnesium depleted and the crystallinity, as determined by method I. Intensity 4.0000004 3.2000904 2.4000404 Loco-.04 0000 0 2000 2240 .5. 1000 E .5 l 120 500 o 50% Mg removed . . fi .. a .. . .. .. «AT . 0 lo 1 . . .. .. .. d . d . 1 1 o 10 20 30 40 26 96% Mg removed 1 1 1 1 1 20 so 40 29 Intensity lnlcmily 0000 6400 4000 3200 1600 2400 1920 1440 050 400 1 1 .l A 1*l 75% Mg removed 20 :0 40 28 l 1 1 1 A A ‘ 99%!Hgnummnfl to 20 30 no 28 Figure 11.7 Typical X-ray powder diffraction patterns of acid hydrolyZl antigorite at different levels of octahedral Mg depletion. 65 Mg % depleted I. 70 Relative Intensity 26 Figure 11.8 24 3 0; 0 6 0 and 0 6 l diffraction peaks of antigorite and Mg- depleted antigorite. 66 .360 '8 C 6 . ‘7" p E -270 «5” 3 : .5; or . 2 _ g '8 I a > 0 8 _ 8 3 ”180 't 5' l 3 o r m ‘ r O r a. i 0 Mg removed I 20 I - . . . I f .- Surface area‘90 0 25 Y . 50 75 Reaction Time (hrs.) Figure 11.9 Percentage of Mg depleted from antigorite and the BET surface areas of hydrolyzed products plotted versus reaction time at 60 0C. 67 100 325 C G 5 '° ‘ L g '275 0 r E 80- . e b a. b 2 J " ~225 B b ¢ P .53” 60- I . I: i 8 -i75 3 b a- b 0 Mg removed ; 4O , I Surface area 125 o ' 170 ' 2'0 ' 3? 4‘0 ' 5‘0 ' 6'0 7‘0 80 Reaction Time (hrs.) Figure 11.10 Percentage of Mg depleted from antigorite and the BET surface areas of hydrolyzed products plotted versus reaction time at 80 0C. 68 125 L400 0 Mg removed l I Surface area: g :340 a “’5‘ t ’7‘ U3 E : N e d :280 é o L g i 85‘ C a 3* 1220 8 a 4 P D g b ‘1': § 655-4 l 3 °‘ E160 . n v T v '100 450 5 15 25 Reaction Time(Hrs.) Figure 11.11 Percentage of Mg depleted from antigorite and the BET surface areas of hydrolde products plotted versus reaction time at 105 0C. 69 (See Table 11.5). A maximum of BET surface area of 326 m2g'1 has bee observed for Mg depleted antigorite serpentine. But the authors were unabl to comment further on the origin of the maximum.3 It is very surprising I notice that from our results the maximum surface area values (312 - 35 m2g'1) obtained at all three reaction temperatures correspond to the Table 115 Effect of various duration of treatment on the physio: characteristics of serpentine (acid treatment at 75 0C)a Duration of treatment 3h 5h 7h 12h g Surface area (mZg-l) 186 219 236 211 Pore volume (mlg-l) 0.146 0.159 0.185 0.216 g a Data obtained from Girgis, B. S.; Mourad, W. E. Chem. Biotechnol 197 26 9 condition in which about 70% of the Mg is hydrolyzed. This observatio: agrees perfectly with the effect we observed for the crystallinity change with Mg depletion. This proves again that hydrolysis of more than 70% 0 the octahedral Mg leads to the almost complete collapse of the crystallin structure. . The decrease in the BET surface area beyond ~70% Mg depleti01 suggests that there is a formation of less active material such as an inactivl form of silica. The crystallinity ratios clearly indicate that the materia formed at the early stages of Mg depletion rapidly undergoes . transformation to an amorphous material with further Mg removal. Th: 70 highest BET surface area we obtained was 353 m2 g'1 (see Table 11.6)- W also obtained BET surface area of 300 m2 g'1 or more for all thre hydrolysis temperatures, compared Table II.6 BET (m2/g) surface area and MAS NMR curve fitting result with respect to Mg depletion for antigorite samples hydrolyzed at 60 0C Mg% BET (m2/g) Relative intensities of 2981 MAS NMR depleted peaks Q3 Q3 HOSiO3 Q4 (-92.4 ppm) (-103 ppm) (-112.0 pme 27 97 47 6 47 39 i 1 17 30 15 55 41 181 27 18 55 52 195 17 16 67 56 217 1 1 33 56 69 353 12 19 69 79 339 1 24 75 J to 6 m2 g-1 for the starting antigorite (see Tables II.7 and 8). These values above 300 m2 g'l, agree with the highest BET surface area value reported 31 71 far for acid hydrolyzed antigorite. Girgis and Mourad reported a BE’ surface area of 326 m2 g-1 for acid hydrolyzed antigorite.3 Table 11.7 BET (mZ/g) surface area and MAS NMR curve fitting result with respect to Mg depletion for antigorite samples hydrolyzed at 80 0C Mg% BET (mZ/g) Relative intensities of 29Si MAS NMR depleted peaks Q3 Q3 HOSiO3 Q4 (-92.4 ppm) (-103 ppm) (-1 12.0 ppm) , 43 161 21 19 60 52 I 205 22 19 59 57 270 10 30 60 68 307 14 27 59 75 312 5 23 72 94 300 — 17 83 97 225 - 15 85 96 200 — 20 80 Even under the mildest conditions used to hydrolyze the octahedra Mg from antigorite (20% HCl, 1 hr. and 60 OC - see Table 11.2) thre different 29Si MAS NMR signals were observed in contrast to a single pea] 72 observed at -94.0 ppm for the starting antigorite. The peaks observed fc hydrolyzed samples occurred near -92.4 ppm, -103.0 ppm and -112.0 ppm. Table 11.8 BET (m2/g) surface area and MAS N MR curve fitting result with respect to Mg depletion for antigorite samples hydrolyzed at 105 0C Mg% BET (mZ/g) Relative intensities of 29Si MAS NMR depleted peaks Q3 Q3 HOSiO3 Q4 (-92.4 ppm) (-103 ppm) {-92.4 pme - 207 28 19 53 52 300 20 13 67 78 320 5 35 6O 87 307 - 34 66 93 277 - 32 68 98 230 - 32 68 99 127 - 20 80 73 The peak near -92.4 ppm can be assigned to Q3 tetrahedral Si still bonded ‘ the remaining octahedral Mg through Si-O-Mg type bonds. The 2 pp: down field shift of this peak relative to the Q3 resonance parent antigori may be due to the initial changes in the octahedral sites such as weakenir of O-Mg bonds caused by the acid attack. The results of curve fitting of tl 29Si MAS NMR spectra are shown in Tables H. 6, 7 and 8 of the peak ne: -94.2 ppm (peak 1) fit well with the above assignment. Once more than 94‘ of octahedral Mg is hydrolyzed out, there is no detectable peak near -92. ppm. Figure 11.12 shows changes in 29Si NMR peaks for sample hydrolyzed at 80 0C. 29Si CP/MAS NMR of hydrolyzed samples shows 2 enhancement of the peak near -92.4 ppm (Figure 11.13). This enhanceme1 may be due to the Q3 Si nuclei like Sia indicated in Figure H.14. The 29! MAS NMR peak near -103 ppm (peak 2) can be assigned to Q3 sites whic are more accurately specified as HOSi(OSi)3. The enhancement of the pea near -103 ppm in 29SiCP/1VIAS spectra confirms this assignment. The pea near -112 ppm (peak 3) can be assigned to Q4 sites. Peaks near -103 pp] (peak 2) and -112 ppm (peak 3) may respectively be due to Q3 Sib and C Sic type nuclei as shown in Figure 11.14. This Q4 site formation can 0001 within layers. According to Engelhardt and Michel Q4 peaks around -11 ppm correspond to an Si-O-Si angle of 151 0 (Table 11.9).19 This chemicz Shift value is bigger than most of the commonly known silica types such 2 quartz (-107.4 ppm), crystobalite (-109.9 ppm) and silica gel (-109.3 ppm This higher Si-O-Si bond angle supports Sic type arrangement. The broa Q4 band suggests that there may be several overlapping peaks with th highest intensity at -112 ppm, possibly a result of a range of Si-O-Si bon angles. Rxn Time 72hrs 481m 24hrs 151118 121m: 9115 6hxs 31115 1111' YITTY TV.*I]..TT]'TTTITITTI -80 -90 -100 ~110 -:20 :cm Figure II.12 29Si MAS NMR spectra of acid hydrolyzed antigorite samples with different reaction time at 80 0C. Mg depletion 96% 97% 94% 75% 68% 57% 52% 43% 75 H 9 SiO-Si-OSi O Si No cross Polarization 'memmmmmwmmmmmmm A. 50% Mg depletion L -85 - :05 ' 30m / \J 1H cross polarized No cross polarization Vll’llilllillllliTlIllllIlllTlTfifi ‘80 '90 '100 -110 -120 DDm Figure 11.13 29Si MAS NMR and 29Si CP/MAS NMR spectra for Mg depleted antigorite at 80 0C. A. 50% Mg depletion, B. 96% Mg depletion. 3 ’ " s i r“ .. « C / 2” ‘7 .' C n n n a- n 0 § ‘ \ ‘ § ‘ O O C O 0 Silicon a Hydroxyl group . Magnesium 0 Oxygen Figure II.14 Scheme for the Mg hydrolysis of octahedral Mg in 1:1 layered silicate lattice. 77 Table H. 9 29Si chemical shifts, 8(ppm), relative intensities, nSia a; calculated mean SiOSi bond angles, a, of highly dealuminated zeolite ZS -5 -563pm) nSia (1(0)b ‘ —5(ppm) nSia 04(0)b 109.87 1 146.1 113.79 1 152.2 111.76 1 148.9 113.97 5 152.6 112.04 1 149.3 114.32 1 153.2 112.55 1 150.1 114.54 1 153.6 112.70 1 150.4 114.83 1 154.2 112.83 1 150.6 115.16 1 154.8 113.44 3 151.6 115.90 1 156.3 113.55 2 151.8 115.99 1 156.5 117.00 1 158.8 a Relative to a total of 24 Si atoms. b Calculated from the shifts according to the equation, = -247.0 cosa/(cos OH) + 2.19 Engelhardt, G.; Michel, D. "High-Resolution Solid-State NMR of Silicate and Zeolites" John Wiley & Sons 1987, 304. 78 This can be explained by the presence of both Q3 and Q4 sites in the : silicate layer as shown in Figure 11.14. With this kind of a layer (a l which contains both Q3 and Q4 with variable Si-O-Si bond angles), possible to keep the initial templates parallel to each other and then retain long range order. But it is also possible for the resulting hydr silicon tetrahedra to become distorted, resulting in poor crystallinity Figure II.15). It is very unrealistic to conclude that all the newly formec sites are edge shared as shown in the Figure 11.14. as there is only silicate structure with edge shared tetrahedra reported in the literatun Upon 70% depletion of the Mg there might be a delamination of structure that gives a compacted arrangement without much void space. Figure II.16 shows the FTIR bands in the region 2000 to 400 cm-1 acid hydrolyzed antigorite samples with different amounts of Mg depleti According to Fanner21 the IR bands near 1076 cm'l, 980 cm'1 and 448 CI for the starting clay are assigned to perpendicular vibrations of the pseu hexagonal layer, and Si-O stretching and bending vibrations, respective The bands near 625 cm"1 and 565 cm"1 involve Mg-O vibrations. The El band near 1076 cm"1 gradually shifts to a shoulder near 1200 cm'1 w increasing amounts of octahedral Mg depletion. Similarly, the band 111 988 cm-1 shifts to a band at 1095 cm'1 and the band near 448 cm“1 shifts a band near 467 cmrl. The Mg-O associated bands near 621 cm'1 and 5 cm-1 diminished as the percentage of Mg depletion increased. Comparal modifications in the FTIR spectra of acid hydrolyzed palygorskites w: reported by Gonzalez and coworkers in 1989.22 Figure 11.17 shows the FTIR bands between 4000 cm-1 and 2500 on for acid hydrolyzed antigorite samples with different amounts of l depletion. If the guide lines given by Ledoux and White23 are followed 79 iii:188811111283 x vvvv ‘ A A A 311111lilllllilllil-llllil ::y/’ ilillllllillillil + AA mmmmmm ,mmu Distorted tetrahedra x = Leached out octahedra Note : Clay layers are drawn straight Figure 11.15 Schematic representation of antigorite showing disordered structure 4Q on acid hydrolysis 1096 1200 467 96% 75% 50% Absorbance 25% 988 448 1076 621 ’56 l o 2200 1800 1400 1000 600 200 Wave numbers (cm") Figure II.16 FTIR bands in the region 2000 cm'1-400 era-1 for antigorite and acid hydrolyzed antigorite at different levels of Mg depletion. 81 Mg removed Q, U C O ‘9 O U) .0 < 3443 i O 0% 4050 f 5850‘ ' '32'50' ' 2850' ' 2450 Wavenumbers cm“ Figure II.17 FTIR bands of antigorite and acid hydrolyzed antigorite at different levels of Mg depletion between 4000 cm'1 and 2500 cm’l. 82 classify -OH groups, the following assignments can be made for the starting clay. The strong sharp peak near 3679 cm'1 is assigned to inner-surface hydroxyls located at the surface of the octahedral sheet opposite of the tetrahedral oxygens of the adjacent clay layer and is indicated with an X in Figure 11.14. The broad band near 3443 cm"1 is assigned to inner -OH groups ( indicated with a Y in Figure II.14). The very weak FTIR bands between the above two -OH bands can be assigned to outer -OH groups at the surface, including both -OH at broken edges and at the octahedral sheet found at upper surface. Upon acid hydrolysis of octahedral Mg all -OH bands observed for the starting clay gradually are replaced by a single broad SiO-H band with its maximum near 3432 cm'l. These changes in hydroxyl group vibrations clearly agree with the structural hydroxyl changes shown in Figure II.14. Similar changes in hydroxyl bands have been observed for acid hydrolyzed palygorskites.21 Micro and MesoPorous Materials Obtained from Acid Hydrolysis of Antigorite. Earlier in previous section it was discussed that a crystallinity index of 0.70 could be retained, upon 70% removal of magnesium, when the particle size of the starting material is <100 mesh. Table II.10 Shows the relationship between the crystallinity index and the percentage of magnesium depleted with different reaction times where the starting material was <325mesh (45 microns). These data indicate that the 0.70 crystallinity index limit can be reached with 50% of the magnesium depleted. This observation suggests that there may be a particle size dependence of crystallinity upon hydrolysis could be a diffusional controlled process. 83 However, the results recorded in Table II.11 give us a very surprising fact about the initial hydrolysis phenomenon. For three different particle Table II.10 Crystallinity, BET surface area, MAS NMR and percentage of magnesium depletion with respect to hydrolysis time for 325 mesh antigorite samples hydrolyzed at 80 0C. (NOTE: Different batches of starting clay were used in samples with reaction time 3, 4 and 6-12 hours) Reaction Mg% Crystallinity CBET BET Total pore time depleted ratio surface volume (hourS) (ng'l) (mlg'1)* 3 39 .99 135 135 .078 4 41 .99 87 130 .096 6 50 .76 43 140 .096 9 61 .64 50 169 .130 12 70 - 46 178 .160 * Total pore volumes estimated at P/Po= 0.95 from nitrogen adsorption isotherms. sizes, namely particles with <150 -75 um, <75 -45 um, <45 urn there is no appreciable difference in BET surface area, crystallinity ratio or percentage of magnesium depleted after acid hydrolysis for 3 hrs (or when 40% of the magnesium is depleted out). This observation clearly suggests that the initial acid hydrolysis mechanism is not diffusional control. 84 Considering both facts we can suggest a two step hydrolysis mechanism, namely, an initial non diffusional control process and a secondary diffusional control process. Mechanistic possibilities for the two processes will be discussed in further detail later. Figure II.18 shows the 29Si MAS NMR of samples with different amount of Mg depletion. The same peaks as reported previously were observed. The Q3 peak observed at -92.4 ppm has a higher intensity compared to other peaks especially for samples with less than 50% Mg depletion. This observation suggests minimal structural alteration of the starting material. Table 11.1] Dependence of the surface area, crystallinity and the percentage of magnesium depleted on the particle size for acid hydrolyzed antigorite. Particle Sizecu) % Mg depleted Surface Crystallinity Area(m2g'1) Index <150-75 41 119 .98 <75-45 42 134 .90 <45 39 130 .99 Figure II.19 shows t-plots obtained according to the Boar, Lippens and Osingaloa11 method for samples with 39%, 50% and 60% Mg depletion. Since the t-curves for samples a and b are overlapped the total 85 Peak 1 = -92.4 ppm -112 Peak 2 = -103.0 ppm Peak 3 = -113.0 ppm Antigorite single peak at -94.7 ppm % Mg depleted 92.4 70 Q3 Q4 41 \’_/ 39 llllIlllllllllllllllllllllllllllllllfll'm -60 -80 -100 -120 ppm Figure 11.18 29Si MAS NMR spectra of acid hydrolyzed antigorite (325 mesh). 86 60 .50 '- .w 40 u c \efio“ a 39070 Mg 0 p _ _ >5 1 slapc 3 lg 30 -1 U) 1: < slope 2 a. o E 20 " .2 ° 4 > slope 10 - ‘ Total surface area = 130 ngl O I I I I I I ' I I I I I 0 1 2 3 4 5 6 7 8 9 10 ill a. 39% Mg depleted b. 50% Mg depleted Figure II.19 (parts a and b) t-plots of antigorite hydrolyzed at 80 0C (325 mesh). 87 Volume adsorbed (mlg-1) oufi I ' I ' I I I I I ' I ' I ' I ' I 0 1 2 3 4 5 6 7 8 9 10 t (A) Figure II.19 (part c) t-plots of antigorite hydrolyzed at 80 0C (325 mesh). 88 surface area of samples with 39% and 50% Mg depletion can be obtained from the slope l drawn in Figure 1119 (139 ng'l) . This total surface area agrees with the BET surface areas of both samples (135 m2g-1 and 140 mZg- 1). The difference between these two samples, See Table 11.10, is in their CBET constants and total pore volume. This suggests the presence of some structural differences in pore volumes. And also we see a difference between slope 2 and slope 3 in Figure II.19 which are the second Slopes drawn for sample a and b respectively. In other words these two slopes give different microporous surface areas. The microporous surface area obtained from slope 2 for sample a is 131 m2g'1 and that of sample b from slope 3 is 65 ng'l. This difference in microporousity and total pore volume hint at the difference in the Mg depleting process. This new process can be correlated with the secondary diffusional controlled mechanism already discussed in the earlier part of this section. Part c of Figure 11.19 shows the t-plot for the sample with 60% Mg depletion . It is very clear that the choice of reference (aluminum oxide) used in this t-plot was not suitable. The same phenomenon was observed with all other samples with higher percentages of magnesium depletion. The CBET values in Table II.10 Show that they decrease with increasing percentage of magnesium depleted. Figure 11.20 shows the t-plot for sample c according to the Pirard and Lecloux method with the same volume adsorbed as that used in Figure II.19. Numerical values of the standard data 11: v/vm for 100 Z CBET 2 40 reported by Pirard and Lecloux were used to match the CBET constants reported in Table II.10. These standard values have been correctly fitted ( within 5% up to p/po=0.6) by the n-layers of the BET equation. Slope in Figure 11.20 gives the total surface area of 172 ng'l. The BET surface area of 169 m2g-1 of the sample c agrees well with the total surface area obtained. Even though 89 80 Adsorbed Volume (ml/g) 0 “ v I I f I T I ' I 0 1 2 3 4 5 ' 6 Lecloux-Pirad Thickness (A) Figure ILZO Lecloux-Pirad plot of the nitrogen adsorption data for acid hydrolyzed antigorite with 60% Mg depletion, illustrating the absence of microporosity in the sample. 90 there is a negative deviation in the t-plot, the partial pressure correspond to the deviation point is too high to consider the system to be microporr No upward deviation due to capillary condensation was noted either. Figure H.21 shows the pore size distribution for samples with differ amounts of Mg depletion according to the BJH (Barrett, Joyer and Halen method13 obtained by using desorbed nitrogen volumes. The sample v 39% Mg depletion has a very small amount of pores with radii around 19 but the contribution of these pores to the surface area is negligible. T result confmns that the total surface area of this sample mostly comes fr micropores below 15 A diameter. The micropore volume calculated plotting the Dubinin-Radusikevich equation15 for the same sample is 0.( mlg-1 (Figure 11.22). This value shows very good agreement with the tr pore volume of 0.078 mlg-1 reported in Table II. 10. Two kinds of po were found in samples with higher amount of Mg depletion. Pore r2 corresponding to these values agree with the literature values of 14 A-22 reported by Girgis and Mourad3 for a serpentine compost that predominantly antigorite. Figure 11.23 shows the nitrogen adsorpt desorption isotherms of samples with Mg depletions of 39% and 60%. '1 sample with 39% Mg depletion has only one hysteresis loop whereas that the sample with 60% Mg depletion has two distinguishable hysteresis 100 This observation agrees well with the pore size distribution curv Moreover, the isotherm for 39% Mg depletion shows the basic typ isotherm related to its microporosity. The isotherm for 60% Mg deplet shows nonmicroporous behavior at low relative pressures. Figure H.24 shows the derivative of the nitrogen uptake obtained using omnisorb software. Pore diameters obtained according to the metl 91 G °§ '9“ Mg depletion 8 «4 g 70% g 4.. '1: d.) O ' m 60% “~ 39% «1b 30 12 i4 A IS A 18 36 225—“'1 Pore Radius (A) Figure H.21 Pore size distribution of acid hydrolyzed antigorite sam obtained by BJH method.13 92 840 {a}; AAAAAAA 1 rrrrrrrrr r 111111111 r 111111111 r 111111111 1 ,,,,,,,,, «t J 805 ‘1 A 0 DD 4 > 1 E . v V I o 1 _ u' 1 o g 770 . 3 a 0 3a a ‘ o :1 0E ‘ O 32 . o 3 .2” 735 3 o ..l j o o o 4 OO . 0000303336 4 1 700 YYYYYYYYY I """"" I ''''''' I vvvvvvvvv 1 vvvvvvvv fi,,,YVYYTW 0 10 20 30 40 50 6 [RT.ln (PdP)]2.103[Goules/mole) 1 Figure H. 22 Dubinin-Radusikerich plot of hydrolyzed antigorite with 2 Mg depletion. 93 120 ‘ a. 39% Mg depleted too .1 b. 60% Mg depleted :5 1 .5. aa- 0 E 1 2 g 5° ‘ r j 'a o r . e . ’ O (n 'a < o ' I 4 I fl I T I I 0.0 0.2 0.4 0.6 0.8 * p/pO Figure 11.23 Nitrogen adsorption-desorption isotherms for acid hydrolyzed antigorite with 39% and 60% Mg depletion 94 . 39A 204 60% Mg depletion 15' 111-1 5-1 'r'ro'7 rut!“3 r'ro‘5 i'rrr‘ 81113 1'10’2 r-ro'l 1 Derivative uptake CD 39% Mg depletion . / 5 \'// 25A mo" moi 1'10‘5 1'10" 1'10'3 irrrr2 1'10'1 1 mo CD Figure II. 24 (parts a and b)Derivative of nitrogen adsorption versus log partial pressure. 95 99% Mg depletion f 50- / / 04 J 4/ R107 1'05 1'05 mo'4 1310-3 1-10-2 lrm-i 1 Derivative Uptake a 351 70% Mg depletion I 15-1 11H 1'10'7 mmS 1*10‘5 mo" r183 1"10‘2 1"10'1 1 Pi?!) Figure 11. 24 (parts 0 and d)Derivative of nitrogen adsorption versus log partial pressure. 96 of Kresge and coworkers15 are reported in the same figure. These values suggest a micropore diameter of 8 A and a mesopore diameter of 39 A. When the system goes from the microporous to the mesoporous region there is an intermediate pore diameter of about 25 A can be seen in the derivative uptake curves which was not observed with the t-plots or pore size distribution curves. As discussed above, the derivative of the desorption curves gives the mesopore diameter of 35 A and derivative of the uptake curves gives the mesopore diameter about 39 A. The Similar results shown in adsorption and desorption modes suggest that the nanopores formed are regular. Derivative data also explain the hysteresis loops observed in the adsorption-desorption isotherms. Figure H. 24 gives the derivative uptake vs p/po plot for an amorphous sample at 99% Mg depletion in which we do not observe any porosity. absence of porosity suggests that the mesoporous structure formed at 60% Mg depletion disappears with further magnesium depletion. Table 11.12 shows the physical adsorption volumes of some adsorbates with different effective kinetic diameters at saturated vapor pressures for samples with different amounts of Mg depletion. The kinetic diameters and the calculated cross sectional areas of the adsorbates used are given in Table 11.13. The amounts of ethanol and benzene adsorbed for all the samples were very similar. However, the adsorption of nitrogen was always highest. The high nitrogen adsorption capacity may be because the small cross sectional area of nitrogen allows a higher packing density. But the amounts of 1,3,5 trimethylbenzene and perfluorotributylamine adsorbed were much lower compared to those of the rest of the adsorbates used because of the larger kinetic diameter of the adsorbates. The mesopore radii 97 obtained from the distribution curves (Figure H.21) and the amount of any of these adsorbates adsorbed, (with the kinetic diameters given in Table H.13), agree very well. Table II. 12 Adsorption of different molecules by acid hydrolyzed antigorite (mmoles g'l). Adsorbate (Kinetic diameter) Mg% Nitrogen Ethanol Benzene 1, 3, 5 Perfluoro- Trimethyl- tributyl- depleted (3.64 A) (—) (5.85 A) benzege amine (7.50 ) (10.2 A) 39 3.05 1.60 1.62 0.77 0.35 41 4.97 1.62 1.45 0.85 0.37 61 5.97 3.60 2.63 1.77 0.78 70 6.63 4.30 4.35 1.90 0.91 To propose a better explanation for this behavior the unique Structural characteristics of antigorite must be considered. Figure 1.10 shows an edge- on view inverted array of tetrahedral and octahedral sheets of the antigorite layer, whereas Figure I. 11 shows the 001 projection of the layer. At every other inverting point in the structure there are eight-membered rings and 98 Table 11.13 Molecular parameters of the adsorbates used Molecule Kinetic Diameter (nm)9 Cross sectional Nitrogen 0.364 162 Ethanol - 0.230 Benzene 0.585 0.323 1,2,3-trimethylbenzene 0.750 0.442 Perfluorotributyl-amine 1.02 - four-membered oxygen rings instead of only hexagonal rings as in most other clay structures. Earlier we discussed the lack of a particle size dependence for the initial acid attack up to 40% magnesium depletion (-see Table II.10). At this point we can conclude that the initial acid attack of the Mg octahedral sheet is through the z direction (hence no particle Size dependence), specifically via the so-called eight-membered rings. It also has been known for several decades that any acid will selectively deplete by edge attack the, octahedral cations (especially magnesium).3r22 It is reasonable to expect a very rapid Mg depletion process at the 8-membered ring inversion points in the octahedral sheet along the z direction perpendicular to the antigorite layer. Inverted octahedral sheets with the six 99 and the four-membered ring inversion points, as in palygorskite could not be attacked by the acid through the z direction. Thus the mechanism for acid hydrolysis of clay is mostly reported as edge hydrolysis or diffusion control.3r25 Because of the depletion through the 8-membered ring channels, a structural topochemical magnesium removal can take place in antigorite. At about 40% of Mg depletion a completely nanoporous structure is generated. The higher the depletion the lower the microporosity (Figure 11.19). However, a new kind of mesopore formation (Figure H.21 and 23) and an increase in the BET surface area were observed. These mesopores may be generated because of lateral hydrolysis starting from already hydrolyzed octahedral sites. The lateral hydrolysis process which generates smaller mesopores around 17 A radii supports the secondary diffusional control process. Growth of this lateral hydrolysis may propagate from outside to inside of the particle. At the same time a slow edge hydrolysis process (compared to the above-discussed process) can be expected. The cartoon in Figure 11.25 below depicts the zones expected in the process. Hydrolysis of layered silicates also removes some of the silicon atoms from the tetrahedral sites of clay minerals, but the quantities are very small.24r 25, 26 According to the our proposed mechanism for hydrolysis of antigorite most accessible silicon for hydrolysis is that around the eight- membered ring outside the clay particle. This hydrolysis of tetrahedral silicon will tend to make the pores bigger and hence facilitate the lateral hydrolysis. This argument explains the adsorption of nitrogen, ethanol and benzene in reasonable amounts. If the pore openings were just the same size as those of the eight membered rings of the silica sheet there would not be any physical adsorption of the above adsorbates simply because the 100 Initial Mg depletion through 8 member ring channels at octahedral sheet inversion Amophous pornts silica Edge _ hydrolysis Edge hydrolysis Partial Mg depletion with retention of layered structure Figure 11.25 Edge view of acid hydrolysis of the octahedral sheets of an antigorite clay particle 101 adsorption would be controlled by the narrowest opening of the tetrahedral sheet. Therefore, there should be a rearrangement of Si tetrahedra in such a way that the original layered oxygen framework structure is preserved. A more detailed illustration of the hydrolysis steps are shown in Figure H.26. Electron Micrographic Evidence for Acid Hydrolysis of Antigorite Severe experimental difficulties were encountered in electron microscopy studies of serpentine minerals due to the poor stability of the minerals under the electron beam (Mellini, et al 1986, l987).27r28 We also experienced difficulty in obtaining good quality TEM images. Figure H.27 shown a HRTEM picture of antigorite reported by Mellini.28 In this picture twin planes (010) appear due to stacking disorders and run from top to bottom. Figure H.28 illustrates the origin of the disorders.29 Figure 11.29 explains the so called lattice fringes.3O In figure 11.27 the lattice fringes (100) run from left to right. Figure II.30 shows a transmission electron micrograph of the starting antigorite clay used in the present work and Figure 11.31 shows a acid hydrolyzed sample with 39% octahedral Mg depletion (Table II.10). Resolution and the magnification of these images were good enough to see twin (001) planes. Spinnler systematically analyzed effects of specimen thickness, defocusing (Figure 11.32) and tilting (Figure 11.33) on calculated HRTEM images.29 Since the quality of the images of antigorite depend on these factors as shown by Spinnler, the poor HRTEM images were obtained in the present work. The acid hydrolyzed antigorite showed no appreciable structural alteration compared to the starting material. This also supports the most of the observations discussed above. Moreover we 102 Top View Antigorite ................................................................................................................................................... .......................................................................................................................................... ............................................................................................................................................................................................ ............................................................................................................................................................... .................................................................................................................................................. . .3. . . .. .. . .. . . . . . .. ..3 3. ... ..3.3.3.3.3.3. ,.3.3.3.3.3. .... . ... . . 3.3- ............................................................ , . .. Side View 8-ring cavity 6-ring cavity Primary acid hydrolysis , _ ._._ q ............................................................................................................................................... ............................................. .................................................................................................................................................................................. -3._._.3.3.3.3.3.3.3.3.3_3.3. ._.3.3. 3 3,3.3._.3.3.3~3r3.3~3n3._.3.3-3.3-3. . .3._.3.3.3-343 3 33.3... ...3._.3. . .3.3 .3u3r3r3e3.3.3-3.3r3~3¢3n3n3n‘-3-.o.n_o.5.3.3.3.313....o3..:3..~ v- ._.3. .3.33.3. 33333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333 ................................................................................................... .. .3.3.3.'.3.' ‘ .‘3. 3.3.3.3 .3'. .z. ........................................ ..................................................... Secondary acid hydrolysis Tetrahedral silica sheet Octahedral silica sheet Figure 11.26 Stepwise acid hydrolysis of antigorite. 103 Figure II.27 HRTEM of antigorite (Mellini, reference 28). Twin planes (001) run from top to bottom, lattice fringes (100) run from left to right. 104 Figure 11.28 Schematic drawing of (001) faults. The curved solid line represents the tetrahedral and octahedral layers, and the 6 and 8 refer to the types of reversals. In (a), a mirror perpendicular to c*, and in (b), two-fold axes parallel to b, the types of reversal that do not change across the fault. In (c) an a-glide perpendicular to c* produces a change in the type of reversal across the fault. 105 6-ring D U . I ‘ o u... ”I. .t. ...... ... ... .... ......1... ‘ ....Eézés ......ahé. E“aha”§a§_u~_.. gs Ziggzsgséés . ....Efiazgssas= _ Meagan”? .. ......_.....a ...a. . .. ...... . ....- r. “:57: a 5.5.... ... .. .a. a...— .... Figure 11.29 Lattice fringes of antigorite. Figure 11.30 A TEM image of the starting antigorite used in this work. Figure 11.31 A TEM picture of an acid hydrolyzed sample a (acid hydrolyzed for 3 hours at 80 0C). Juhmruulil «r .. , . . . .,$.5..waw.mrldi.rrn mww , . , ... . . Figure 11.32 Calculated HRTEM images of antigorite as viewed down (010). 108 THICKNESS oerocus -750 A -850 It ~9so A -roso A -rrso X -1250 A 4350 It 4450 K E—fié—‘I‘i' -rsso A ~1650 A I :=-\ .. m m m -1750 X um um m m m: m 4850 K m __—r 109 [001] A 2.1° [601] 1 7% K...— k w “w "‘5... , M —‘— h M guy-i 3 : >[100] .. Mg%; :--u—_"‘.'f""‘:='- :___. -;=. Figure 11.33 Calculated HRTEM images of antigorite as a function of crystal tilt. 110 observe the same morphology as in the starting materials upon acid hydrolysis of up to 40% of the octahedral magnesium. Moya and coworkers (1984) reported the same results as this work with TEM studies of acid hydrolyzed sepiolite.29 Figure 11.34 and Figure 11.35 show TEM images of antigorite with 70% Mg depletion (Section II.b.l). In Figure 11.34 the layered structure can be seen in areas labeled L. Light color bands running from top to bottom are so called lattice fringers. Areas labeled A could be either improperly focused or thicker regions. Figure 11.36 is an image of an antigorite clay from Spinnler given for comparison with the same magnification as the other images. The existence of lattice fringes with white colored paths at this higher magnesium depleted state supports the hydrolysis through the eight- membered inverted points discussed in Section II.b 4 (compare with Figure H.27). Figure 11.35 shows a 001 view of an acid hydrolyzed antigorite. Light colored areas may represent the pores in the clay particle that exist because of depletion of octahedral magnesium. Acidity of Acid Hydrolyzed Antigorite Neither Bronsted nor Lewis acidity was observed for acid-hydrolyzed antigorite by FT-IR spectroscopic studies of the pyridine adsorption. No pyridine adsorption was observed at the 150 °C outgassing temperature, only physically adsorbed pyridine. A peak in the temperature programmed desorption (TPD) of adsorbed ammonia was observed at 150 0C (Figure 11.37) for samples with 39%, 60% and 70% Mg depletion. This peak also may be due to ammonia physically adsorbed in the pores. Furthermore, since the peak was observed at fairly low temperature it could not be due to Figure 11.34 A TEM image of acid hydrolyzed sample with 70% Mg depletion (acid hydrolyzed 12 hours at 80 °C). “383}: .. .. . . , .... A m Figure 11.35 A 001 TEM view of antigorite with 70% Mg depletion (acid hydrolyzed 12 hours at 80 °C). 200A 113 \ ,..~.\:\\\\\\ $ .xx . \\.\ . . \\.\.\\\. § ..s. s.” Mm , 1985. nnler from Spi lgorite fant ICtLll'e O . \ . . h“: ~ . a ‘\\\ \ . it o l \e . \\\ . \\ sxs ..s§fi.:.. .....MQ \ s \\x . . A: .. . \ x \ \ ,\\. . . ......” n .vs.\\.s. .\.\\.s\\.\\\. . . s... s x a: x \s \.\ \\\\\\\‘ . R .xx An HRTEM p .36 H Figure 114 l l l -6- 70 % Mg depleted -B- 60% Mg depleted .9. 39% Mg depleted Relative Intensity 100 120 140 160 180 200 Temperature (°C) Figure 11.37 Temperature Programmed Desorption of ammonia for acid hydrolyzed 325 mesh antigorite with different levels of Mg depletion. Outgassing temperature was 400 °C. 115 any acid sites. The amount of ammonia released increases with the total pore volume ( Table II.10). This fact confirms that the peaks shown in Figure 11.37 are in fact due to ammonia trapped in the pores. Thermal Stability of Acid Hydrolyzed Antigorite Table H. 14 shows the effects of outgassing temperature on the BET surface area of acid hydrolyzed samples with different amounts of Mg depleted. For samples with 60% Mg and 70% Mg depletion there is little temperature effect on the BET surface area. In fact, there is a slight increase in BET surface area with degassing temperature for samples at higher Mg depletion. This confirms that the mesoporous structure discussed above is stable towards thermal effects and differ from the data that Girgis and Mourad.3 Antigorite samples hydrolyzed with 25% HCl for 7 hours show thermal instability in the BET surface area. (Table II.15). But on the other hand samples treated with 40% HCl for 7 hours show a BET surface area increment when the degassing temperature is changed from 100 0C to 200 OC ( Table H.15 ). According to the BET surface area values reported for the nanoporous samples at 39% and 41% Mg depletion there is a thermal instability of the material that depends upon the degassing temperature. The hydrolysis mechanism and the structures of samples with lower Mg depletion are different from those of highly Mg depleted samples as discussed in previous sections. Alumination of Acid Hydrolyzed Antigorite In all the alumination experiments including the blank experiment a white fine powder was obtained. Alumination was done by reacting samples with NaAlOz/NaOH solution at 90 °C according to Fripiat and coworker.16 116 Table II.14 Effect of the outgassing temperature on the BET surface area (mZg-l) of acid hydrolyzed Antigorite. Mg% depletion Degassed at Degassed at Degassed at 150 0C 250 0C 350 0C 39 135 82 80 41 130 81 81 60 l 10 120 128 75 178 184 195 Table II.15 Variation of BET surface area of antigorite hydrolyzed with 25 and 40% HCl, according to Girgis and Mourad3. Temperature (°C) BET Surface area (ng'l) 25 % HCl (7h) 40% HCl (24h) 100 236 263 200 210 326 280 168 244 117 The blank experiment was an NaAlOz/NaOH solution cooked without any antigorite present. Figure 11.38 shows 29Si MAS NMR of aluminated antigorite, 39% Mg depleted antigorite, and 70% Mg depleted antigorite. The aluminated acid hydrolyzed samples give a single peak at about -87.6 ppm. Figure 11.39 shows the 29Si MAS NMR of acid hydrolyzed samples with 39% Mg depletion and 7 0% Mg depletion. According to the prior assignments given in Section II. b.3 acid hydrolyzed samples had Q3 (-92.4 ppm), Q3 with -OH groups (-103.0 ppm) and Q4 sites (-112.0 ppm). There is no appearance of Q4 or Q3 with Si-OH groups for the aluminated sample shown in Figure 11.38. This could be due to the dissolution of silicon in the strongly basic media used. But the resonance for the aluminated sample was shifted about 5 ppm downward from Q3 sites reported for acid hydrolyzed samples (shifted from -92.4 ppm to -87.6 ppm). This shift could be easily explained by the tetrahedral substitution of silicon by aluminum. Each A104 tetrahedron that is connected to a SiO4 tetrahedron shifts the 29Si signal by about 5-6 ppm to low field.31 This is further supported by the fact that the 27Al MAS NMR of aluminated resonance of the 39% and 70% Mg depleted antigorite (Figure H.40) shows only tetrahedral aluminum. However, the blank experiment also showed the tetrahedral alumina. The resulting solid may be a mixture of tetrahedrally aluminum substituted antigorite and the aluminum salt as in the blank experiment. For aluminated antigorite that has not been subject to acid hydrolysis, a single peak at -92.86 ppm was observed. This is a noticeable shift from pure antigorite which gives a single peak at «94.1 ppm (Figure 11.3). In the 27A1 MAS NMR only octahedral aluminum is present in the aluminated antigorite. The shift in the 118 406 70% Mg depleted 39% Mg depleted \¥V/-- Amdghm ITTTIITITTTTTITITTTITTTTWIIIITIITIIIITIII[XIII] ~60 -80 -100 -120 DDm Figure 11.38 293i MAS NMR of aluminated antigorite and acid hydrolyzed samples. 119 -113 70% Mg depleted 39% Mg depleted lufiyfiu IITTTIITIIIIITIIIIWIIIITTIIIITTTTIIIITWT -70 -90 -110 DDm Figure 11.39 29Si MAS NMR of acid hydrolyzed samples. (4) (6) fl -3 3/ J Blank IITIIUIIIIIHIIIIIIIIHI‘IIIIIH"ll”IIIIUIHHIIIIIIIHIIIIIIIIIIIIHIIIIIII‘III Hill] 100 80 60 40 20 0 --40 ppm Figure 11.40 27Al MAS NMR of, 1. Blank, 2. Aluminated antigorite, 3. Aluminated sample with 39% Mg depletion, 4. Aluminated sample 70% Mg depletion. 121 29Si NMR peak could also be due to aluminum substitution for Mg in octahedral sites, because the value of -92.8 ppm lies between the values for the 29Si NMR shift for antigorite, -94.1 ppm (trioctahedral clay) and kaolinite, -91.5 ppm (dioctahedral clay). Figure II.4l shows the X-ray diffractograms of the blank, the aluminated antigorite and the aluminated acid hydrolyzed samples. Both aluminated versions of the acid hydrolyzed samples show relevant peaks for antigorite (Figure 11.1). There also exists in the antigorite sample a new phase whose peaks exactly match the XRD peaks of the blank. This clearly shows that there are two phases present in the aluminated acid hydrolyzed samples. The two different phases are tetrahedrally aluminum substituted antigorite and an unidentified aluminum phase which is different from the starting sodium aluminated, gibbcite [Al(OH)3] or alumina (A1203). The XRD of aluminated antigorite exhibits only the antigorite phase. This agrees well with the MAS NMR results and prove that only one reaction has taken place in the system, namely substitution of Mg by A1 at octahedral positions of antigorite. Table H. 16 shows the cation exchange capacity (CBC) and the BET surface area of the two aluminated acid hydrolyzed samples. Both acid hydrolyzed samples show a very high CEC after alumination. This confirms the tetrahedral substitution of aluminum suggested by 298i MAS NMR. Low surface areas can be explained by the dissolution of amorphous silica containing Q3 Si-OH groups and Q4 Si groups in the strong basic medium used for the reaction. For future studies it would be worthwhile to do the same reaction in either neutral or a slightly acidic medium in order to maintain the high surface area. 122 YI‘ITIVIITIVIUIr l ....I....,....,....I.. I70% Mg depleteJl A I. 3 i 3‘ '8 fl .9. g 39?: Mg depletejim J; E d) m Antigorite 4 V L #4 A; ,__.~.._____, Jim 1 a- W 111111.11111111111111111111111111111 4111 o 5 10 15 20 25 so 35 4o 29 Figure 11.41 X-Ray Diffraction pattem of aluminated samples. 123 Table II.16 Cation Exchange Capacities and BET surface Area of aluminated acid hydrolyzed samples and antigorite. Sample Cation Exchange BET Surface Capacity(meq/100g)a Area (ng'l) Before After alumination surface area alumination before alumination 39% Mg depleted 0.42 46 10 antigorite 70% Mg depleted 0.39 53 7 antigorite Antigorite 0.30 b 6 l i a The CEC were measured by the ammonia desorption method17 b Not measured 124 II.c Acid Hydrolysis of Kaolinite II.c.l Experimental Acid hydrolysis of kaolin was carried out by placing the sample in contact with 50 wt% HNO3 at room temperature and then heating the mixture to reflux with constant stirring. The liquid/solid ratio was 80 cm3/5g or 5 moles H+ per mole of A1. After the desired reaction time the mixture was filtered, washed until chloride free, and air dried. The percentage of A1 depleted was calculated based on the results obtained from chemical analysis. Chemical analysis was performed by the ICP method by using the lithium metaborate flux method for preparing solutions suitable for analysis.9 The BET surface area measurements were carried out on a Quantasorb J r. surface area analyzer with nitrogen as the adsorbate at 77 K. II.c.2 Results and Discussion Table II.17 shows the percent of aluminum depleted, the Si/Al ratio and the BET surface area of the resulting residues after the acid hydrolysis reaction. No hydrolysis of octahedral aluminum was observed upon acid hydrolysis at elevated temperatures. The stability of aluminum in clay minerals toward acids has been observed before.32'33 Present work with out precalcination we find very low surface area. 125 Table 1117 Comparison of percentage A1 depleted and BET surface area to reaction time of kaolinite. Reaction time (hours) 2 4 Al% depleted 2.5 2.7 Si/Al ratio 1.0 1.1 BET surface area (ng'l) 18.27 19.51 II.d Conclusions 1. Based on the changes in the crystallinity index and the BET surface areas with the percentage of octahedral Mg depleted by hydrolysis with 20% HCl acid, it can be concluded that up to 70% of the octahedral Mg can be depleted without dramatically changing the structure of 100 mesh antigorite. The Si Q4 SiO4 sites formed upon the acid hydrolysis process help maintain the layered structure during the acid attack. A BET Surface area of more than 300 m2 g'1 was obtained. This is a very large increase compared to the 6 m2 g'1 for the Starting clay. The prolonged attack (removal of octahedral Mg beyond 75%) leads to the disappearance of the layered structure and the formation of an amorphous, low surface area silica. 2. The mechanism for acid hydrolysis of antigorite is considered to have three main Steps: 126 i. Initial acid attack of the octahedral Mg sheet through the eight membered rings located at the inversion points in the basal plane of the silica sheet. ii. Secondary lateral hydrolysis of the octahedral Mg Sheet starting from the already hydrolyzed 001 planes that correspond to the eight membered rings. iii. A relatively slow, compared to the steps i and ii, edge hydrolysis process. A microporous material with an average pore diameter ~8 A can be obtained from the selective removal of 40% of the octahedral magnesium from antigorite. The hydrolysis of magnesium beyond 40% leads to the secondary hydrolysis discussed above with the formation of a mesoporous material with an average pore diameter of 39 A. Mg2+ is depleted from the wave structure by H‘t/Mg2+ migration through the 8-membered ring of the basal surface. A regular nanoporous magnesium silicate was synthesized by topochemical hydrolysis of Antigorite. The nanopore can have two forms one with a diameter of ~ 8 A and the other with ~ 39 A form depending on the Mg2+ depletion and the rearrangement of the Si02 sheet. 3. The TEM images Show clear evidence for the topochemical acid hydrolysis of antigorite and evidence for rapid acid hydrolysis through the eight membered ring channels. 4. No acid sites were found by pyridine chemisorption in acid hydrolyzed antigorite. 127 5. Aluminum could be tetrahedrally substituted in acid hydrolyzed samples to obtain material with a higher cation exchange capacity. Also aluminum could be substituted for octahedral Mg in antigorite. 6. Owing to the eight-membered openings in the inverted silica sheet of antigorite, the acid depletion of octahedral Mg could be carried out to provide resulting materials with good physical properties. Topotactic hydrolysis has not been observed with 1:1 and 1:2 clays bearing only hexagonal cavities in the 001 plane. But as mentioned in the section 1.0 clays with ribbon structures (palygorskite and sepiolite) and hence bigger cavities in the silica Sheet have shown special behavior towards mineral acids.32r34'48 Silicates with cavities bigger than the hexagonal cavities in the 001 plane of the tetrahedral layer would be the perfect candidates for similar studies. Silicates such as manganpyrosmalite, Mn8 [Si601 5](OH,C1)1049, bementite, Mn7[Si(,015](OH)35O and apophyllite, KCad,[Si4O1()]2(F,OH).8H2051’52 would be good candidates in this aspect (Figure 11.42). Carlosturanite53 and Greenalite54 (Figure H.43) also Should be especially suitable candidates to be examined. Both of these minerals are as serpentine group minerals. Even though, for simplicity they are not included in clay mineral classification given in Table 1.2. Figure II.44 shows two other layered Silicates which have modulated tetrahedral sheets with continuos octahedral Sheets as do the other examples given above, namely zussmanite and stilpnomelane.55 7. Kaolinite with aluminum in the octahedral sites would not be a good candidates for acid hydrolysis reactions without thermal pretreatment. 128 Figure II.42 Sructures I Suggested for future leaching experiments. a. Manganpyrosmalite, b. Bementite, c. apophyllite, d. Hypothetical structure. 129 ‘ .45‘ 45‘ .454? .45 .45‘ iéaneag