t? . . ‘ F77? ‘ ‘ .- ' .'.>'““« , ‘1 " T~ “1'4. :)'~;-_; 115" in} ‘ 13.3% ngigi‘azffiwfit ”l;- - ”‘4'? é‘.‘ “*7" , ~“€t§z§‘g§§g‘wms?§:§§ ggfin magic»; .- 1.. lbw" (,4; ‘31 .;v 4; m . . , 5251:“ {‘ ’21“? ,2 “fizafié‘y-HL. ":3; ‘11, ii: 333%,?» l, .‘ 4v.- , .. ' N A .( 44 . 1w ‘q‘. v‘ u-u ‘1‘ A X ' ‘ "- ‘ 9+, , ‘L‘: .‘M Um. ‘ w. . . “a?h§§‘5$~‘2§’:fi‘“§'7‘3‘3m » ‘3 9% " 21.x Emsfiw xm‘fiwmy‘tw- “"| ~ -' H ‘ " "‘ ' \ *‘ : ' ,‘ ‘ \ ‘ k‘ 3‘s:- .. *~. "'N‘J‘Wm'» ‘ -?'3“‘v“-"*v v. wt “1» ‘ E w .WWXH‘L 3“. a“ ”Axum ‘ ‘r’ ‘flfil‘z‘hfi' L 49.34”wax-4.515615%,“ ‘q W“‘2:‘L“€I‘ 5‘5. “"1. hfi‘wa‘fifi? ‘E‘t'w - ' a, n . v in u A H“‘=‘§' ."t,'§"':“f‘ “*fi ~'~ ’v-‘Q‘JR‘ "L “ ‘. 1‘ suits: c‘tfixr-‘r ‘ “' ‘V‘ " -' a" ”Q Jihi§asvfii M556} "at f; 9N}- c 2% junxsugmmwa» v33: . in“: . , , N. J,‘ 1, ' {if «it fiuiw 1“!ng 3111‘ l' a ‘ §? .139 2 511‘: ‘51.»: mt: 153-»th ‘EQY‘wt‘ itfwfi‘fy “11%; t u "Hfi-‘JII' 4’i‘v‘f'Rkafib‘s‘fix‘fivfifii’kk it. EV. ‘fkfif‘kfi wbfi%"{fi§‘£fi§c ‘ h'fiéi'ivfiz’fi'fl‘fi (Mn -%::,~a<»§~m, My 4 ‘1“.~§~:’i~1u~a3-* % maggégi; .. .. . :-J I .1 g 1.1, u 431'“ {1.31, M “Ella-‘3'. , 1 g "1713?? 1‘1.‘~g"3‘e"‘1 _ , . _' . . a ILW‘QNLV LL: #811“ $.41.“ 3‘41), ‘ ”a!” $1 'firl'x‘ 13% “I ‘ A“ ms. 4 :1: “KN-4% m: m _¢3‘,.,'ad,v a}. ‘2‘9- a ,_ .w‘v. " "'“i‘mfii' ‘ ‘im’é‘zcéfi*ibfhfiffim.‘53%,:zzkéh Quafii‘fi‘: "~"s =5 L ".3 4" 1 in ELLLHQNH v ”"3 «in. 0‘ .‘t' ‘1.” ‘k" 5“: . . ‘ W ' ‘-‘ “18W«‘a‘zuux'xfinHMav .. w Lith‘ ;. ‘ W «Lidia-zai'z-es'taém»,:=‘:~‘:;§:='$, me?»- ~~ " a" ‘zn-uhte'g . -. "A g: 5. f «a. . :9 I ‘1 A ‘ . ‘ w "1‘1"?”Y‘ z'p! 1.. 'W‘.‘fi .1 \‘L 4“.“ ~44 w: w 2‘: 3:.“ ' W3,“ 'QA“. ‘ V a 9"é,~§n‘{'~u‘ll ' w. \£?'§g‘a?§¥§$ :‘ a 3“» M _ ‘ 3P0 \ r v WWW}?! 1» . v liar #1 “En v i "H me u. J’q .1 \ '5 ‘ 6,2 A“ r13 _ I} J ,x ‘1 31': {‘ifi' "1‘4; 3 . (1", ,_ . . ‘ I 71‘ ‘1' ~ t w x m‘wzaw M .. .- W‘Jq‘fligfimw; W «it? W“ . (3’3? iflfififififfifi’xhkawtfl 5325M ‘. ’iz‘$§¢’3 "WWW ,5 ',v ‘z. , ‘ 'r‘g': . , ‘1, J ‘3 I EFF {jiglm‘ffikvg' filégfi'rfi; 3&2?" Riggs". t, M . w my 2:: 3.31:9 A it‘ll lv‘l"? it?” u N! 'i— 1 "(4.132% V ' H's. :_ a A! a . r r - 'V .' V 1 ' ANCRQ‘RKV H lit-‘1')“: ,. '1 11:23:" W ' .m ‘1' k ‘ . .3 . ., ,a' H; 49:,» - 1-) "N33 fia‘fwxfiiéifi‘ikiz 3:} ' ';\}$J'{-.““w- W 1 . ‘ Mg‘mx“ ‘35 3 ‘figflvlmfxfi‘yfié‘ ‘qu 3gzk4‘éfv ‘ g" ragmw , a»: "* 1%??? q? .1. W 'i' 3: ~ v * $3118.“; Q U g " ‘.‘i “ g ’1‘; ’a~ u 2 1 \ , . ' ." ‘ My“ 'wii 431114“ “x: J «ma; Mm ! c $1” {4‘4 1}»: ngfiukfih “L’fl‘fi’fic‘xf. m, u,- ‘3fiifi'r‘fiv‘rhfl‘éit‘M‘i’"Qihgfimzwswam 4g V‘fi‘fli‘ifim ' a "“5? :h’fiéf«;€;:: Hm “ 4 Q n 3,,1“ "J" N h! 1 41K {3119' - t_, 1 1,, ‘ ”I V .47, “Eggfi‘ A‘, K 5311“; 31;». w: I ‘ Q) . 2% x. \u‘a’ .‘ ‘. . a. - A .; , ‘ ‘ .1 . H * ‘ - i ski} \. fl“- tfi" : A . . ’ A. 7—599” - 9‘qu ." 3 A _ "c. ‘ “" w- ”4-! .r‘ 2‘51..- . —. 'v a: g. < 3.. < ‘ V- .14.: ‘ *t .. a.“ 3‘: __=_ 45;. ‘_ v ._ M v .. ‘ _ _ ... ‘mufi . ’ ‘ h ‘ ‘ ' ‘ 3 ‘ ‘1 $5 ‘ “'5'; J“ '5‘": ’9. M - - A ‘ . ’ V "- z: ' "' “446; J"? wi- 2‘ , .‘ . . :., a 5‘.“ .W. . - “I .- n: .3; V , _ ‘ " '6. 1.4.. w x _ _ v_ . . ‘ _ l P ' V l“. 'r' 31.5%, v .‘1 {Eye . ‘ 1. t a wig?“ ”“3? 1kg" :7. 5!. HER-“R t‘k;1§1‘s“!,€‘fi,v\'{;‘~¥q 9.1 ‘ 3.9%. {my .u ,. a- *0“;- J91. R‘"! 4145!? _ ‘ "33” m’v‘h‘" 6‘??? 'aw 5: 35"} o ‘f ‘Jfix’fifl‘i’a‘xfi' “a", . hr "- .::“7M‘1. : ‘., .. . , w... 5“ 'hf J.“ . I , 4 "d , :2 4’; 1.4; 1‘34: .1 U I a? I". S «‘1‘? (3&4; h :8 lug; w “an A a We: {3‘ ~ '. a? .. A”: ‘ ‘ 4.34; ' 'k" a; 4E & r . ‘m. fi fi- ”1 439:9; fig‘a’ ‘ u \ ‘E.’ ,' ‘fi: f! .H Xi ' '7 .c i: ,. 3} '~' .L ‘ ‘ ~ «~, ~: , "A, 9‘» .g' aunt “1.3“" L'L My] ‘_ 3‘ ‘ “a! .v‘7l‘ ,_ . F. ”"1: I” ”31'" "@1313“ .311. E kif‘sé‘j'zkr‘j’ik MW“ . . w 1-1”. ,3}: U ‘3, ,- ‘. 1v h.“ r (r “a 1‘ ‘ ' .. ‘11. .11..»‘05 n ‘ '.L'§3‘*"T“{J~"§AZ;‘%.NE ,. . _ , ,. ”"313 thumh 1m ’1 lot-1.3.“ “'1ka \Mifi‘ ‘- . zqf {.ig" “43a fife Vz‘f‘fi; mrdfi'wifm ‘1’" wk}.- _ ’t :~ {3, «_ A313 #135: . ‘r , 2‘“... fl: E‘m‘ r . “m r "idfivfix‘ifim ‘ N * “5.115715%; _; ~ 2-; ‘~ n" {Vt ‘ W mam“ Michigan State LUH‘VGMW T“? This is to certify that the dissertation entitled PHYSICAL AND CATALYTIC PROPERTIES OF HYDROXY-METAL INTERLAYERED SMECTITE MINERALS presented by Steven Douglas Landau has been accepted towards fulfillment of the requirements for Ph.D. degreein Chemistry Thomas J. Pinnavaia Major professor Date 3/9/8l‘ MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 IIIIII I I IIII II I 4218 )V1ESI_J RETURNING MATERIALS: Place in book drop to LIBRAklfis remove this checkout from _-3_ your record. FINES Will be charged if book is ' Va.u_~ returned after the date stamped below. 3. fi’fifii? a g 0635216 6°93 PHYSICAL AND CATALYTIC PROPERTIES OF HYDROXY-METAL INTERLAYERED SMECTITE MINERALS BY Steven Douglas Landau A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1984 ABSTRACT PHYSICAL AND CATALYTIC PROPERTIES OF HYDROXY-METAL INTERLAYERED SMECTITE MINERALS BY Steven Douglas Landau Recent advances in the intercalation of robust poly- oxocations into smectite clay minerals has led to the for- mation of pillared clay catalystsl’z. The interlayer cations are converted to metal-oxide props or "pillars" which then impart high thermal stability, a large N2 BET surface area, and Bronsted acidity. The pillared clays typically exhibit pore sizes comparable to faujasitic zeolites. This dissertation will describe the synthesis of a new pillared clay interlayered with hydroxy-chromium polymersB. This new chromium clay is distinguished by a 27 A lattice expansion, as well as by excellent high temperature stabil- ity and novel catalytic activity. These new chromium interlayered clays possess both hydrogen transfer activity and a weak Bronsted acidity. Cy- clohexane dehydrogenation has been used as a model reaction for naphtha reforming whereas dealkylation of B-isopropyl- naphthalene has been used to probe Bronsted acidity. The results for the chromia pillared clay will be compared to those for a commercial Cr(III)/A1203 dehydrogenation cata- lyst. Aluminum pillared clay, the subject of several ..~... -.. --~u~. .—-. Steven Douglas Landau I studies , will also be discussed. Two types of poly- oxoaluminum solutions, base—hydrolyzed AlCl and aluminum 3 chlorhydrate (ACH) with OH/Al ratios of 2.00-2.42, and 2.50, respectively, have been used as reagents for the pillaring of Laponite-RD®, montmorillonite, and fluorohectorite clays. The physical prOperties of the intercalate depend to a greater degree upon the washing and drying condition than the pillaring solution employed. The difference in apparent pore size Openings and their dependence on drying conditions (air-drying vs. freeze- drying) can be explained by use of a clay flocculation model5 in which both lamellar (face-face) and delaminated (edge-face, edge-edge) associations of the layers can occur. B-isopropylnaphthalene dealkylation5 has been used to probe Bronsted acidity and further illustrate the effects of air-drying and freeze-drying upon montmorillonite. Gas- oil cracking6 was employed to elucidate pore structure facets of laponite, a delaminated catalyst, which were not obtainable from the dealkylation reactions. Steven Douglas Landau REFERENCES Vaughn, D.E.W. and Lussier, R.J.; Preprint, 5th Inter- national Conference on Zeolites, June 2-6, 1980, Naples, Italy. Pinnavaia, T.J., Science, 1983, 220, 365. Landau, S.D.; Tzou, M.S. and Pinnavaia, T.J., Abstracts, Clay Minerals Society Annual Meeting, September 30-Oct- ober 3, 1984, Baton Rouge, LA. Occelli, M.L. and Tindwa, R.M., Clays and Clay Minerals, 1983, 31(1), 22. Pinnavaia, T.J.; Tzou, M.S.; Landau, S.D. and Raythatha, R.H., J. Mol. Catal., 1984, 21, 195. Occelli, M.L.; Landau, S.D. and Pinnavaia, T.J., g; Catal., 1984, 0000, 0000. To My Family ii .‘ll... III... ‘4' ACKNOWLEDGEMENTS I would like to thank Dr. T.J. Pinnavaia for the tangible and intangible concepts he so skillfully imparted during my graduate studies. I hope my graduation will sig- nify the commencement of a long and fruitful exchange of ideas between us. I would also like to thank Michigan State University and the Department of Chemistry for giving me the oppor- tunity to undertake this study, along with thanks to the numerous support personnel associated with the chemistry department. I would like to express my gratitude to Dr. H.A. Eick for his guidance in editing this dissertation, as well as Dr. Tzou, Dr. Raythatha, and Dr. Atkinson for technical assistance supplied in various aspects of my work. Financial support from the Dow Chemical Company, USA, as a 1984 Summer Industrial Research Fellowship recipient is greatly appreciated. Lastly, I would like to thank my past and present re- search group members, and the Friday afternoon social club for the fond memories. iii TABLE OF CONTENTS CHAPTER PAGE LIST OF TABLES ........................................ Vi LIST OF FIGURES ..... . ................................. viii CHAPTER I - INTRODUCTION .............................. 1 Introduction... ................................... 1 Objectives of Dissertation Research ............... 27 CHAPTER II - EXPERIMENTAL METHODS ................ ..... 28 Clay Preparation..... ............................. 28 Pillaring/Delamination Reaction........... ...... .. 30 X-Ray Powder Diffraction (XRD) Measurement ........ 31 Chemical Analysis.................... ........ ..... 31 BET Surface Area Measurements..... ................ 32 Adsorption Uptake Isotherms ....................... 32 ESR Measurements ................................... 33 Catalysis....... ......... ......................... 33 CHAPTER III - CHROMIUM INTERLAYERED CLAY CATALYSTS.... 35 Introduction................ ...................... 35 Synthesis of Pillars and Intercalation ProcedureSOOOO00.0.0000...OOOOOOOOOOOOOOOO ........ 38 Physical Properties of Chromium Clay Catalysts I, II, and IIIOOOOOO0.0000......OOOOOOOOOOOOOOO0.0 40 iv CHAPTER PAGE Catalytic Properties of Chromium Catalysts I, II, and III..... ............. ...... ............ 52 Characterization of Cr-Laponite and Cr- Fluorohectorite Interlayered Clay ................ . 70 conCluSionSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO... 80 CHAPTER IV - ALUMINUM INTERLAYERED CLAY CATALYSTS ..... 82 Introduction ...................................... 82 Preparation of Pillaring Reagents.. ....... . ....... 84 Physical Properties of Products Obtained from Laponite and Smectites of Related Layer Charge and Morphology.... ...... . ...... ...... ............. 88 Physical Properties of Products Obtained from Montmorillonite, a Typical Smectite ............... 110 Physical Properties of Products Obtained with Fluorohectorite Possessing High Layer Charge and Large Particle Size .......... . ...... .. ........ 119 Catalytic Activity of Aluminum Interlayered Smectites. ................... ...... ............... 123 Conclusions.............. ........ ........ ......... 141 APPENDIX .............................................. 143 Washing Procedures ................................ 143 Drying Techniques ................................. 145 LIST OF REFERENCES................... ......... . ....... 147 LIST OF TABLES Table Page 1 Physical Properties of Some Commercially Important Zeolites ............................ 2 2 Classification of Some Molecular Sieves ....... 7 3 Partial Classification Scheme for Phyllo- silicates ..................................... 12 4 Unit Cell Formulas for 2:1 Layered Silicates ..................................... 13 5 Relative Rates of Esterification of C4 and C5 Carboxylic Acids ........................... 20 6 Physical Properties of Various Oxometal Pillared Clays ................................ 35 7 Physical Properties of Chromium Pillared Clays ......................................... 37 8 Unit Cell Compositions of Chromium Pillared Clay Prepared by Hydrolysis of Cr3+ with NaOH, Solid NaZCO3 and Solution Na2CO3 ........ 41 9 N2 BET Surface Areas for Air-Dried and Freeze-Dried Chromium Pillared Clays .......... 45 10 Hydrocarbon Adsorption Data for Solid and Solution Carbonate Hydrolyzed Chromium Pillared Clays ................................ 46 11 ESR 9 Values of Chromium Clays ................ 50 12 Effect of Alkali-Metal Promotion Upon Chromia Gel Cracking Activity ................. 54 13 Catalytic Parameters and Physical Charac- terization of Spent Chromium Catalysts ........ 72 vi Table Page 14 Characteristics of Natural and Synthetic Smectites Used for Chromium Interlaying ...... 73 15 Concentrations of Reactants Utilized for Chromium Cluster Growth.......... ............ 74 16 Physical Properties of Chromium Inter- layered Laponite, Fluorohectorite, and Montmorillonite .............................. 78 17 Chemical Composition of 50% w/w Chlorhy- drol Solution ..................... . .......... 85 18 Smectite Minerals According to Layer Electrical and Morphological Properties ...... 89 19 Physical Properties of Freeze-Dried Laponites Interlayered with Aluminum Chlorhydrate ............. . ........ . .......... 94 20 Physical Properties of Air-Dried Laponites Interlayered with Aluminum Chlorhydrate ...... 95 21 Physical Properties of Freeze-Dried Laponites Interlayered with Base-Hydrolyzed AlCl3 at r = 2.00...... ...................... 105 22 Physical PrOperties of (A) Saponite, and (B) 0.03u-Montmorillonite Interlayered with Aluminum Chlorhydrate ................... 108 23 Effect of Pillaring Reagent Upon Physical Properties of Pillared Montmorillonite a)....112 24 Effect of Washing Method on the Physical Properties of Pillared Montmorillonite a). .115 25 Effect of Drying Conditions Upon Physical Parameters of Pillared Montmorillonite.......118 26 Physical Properties of Fluorohectorite In— terlayered Clay Synthesized at Various r Values ........................ . .............. 122 27 Physical Properties of Smectite Inter— layered Clays Used for Catalytic Studies ..... 125 vii Figure 10 11 12 LIST OF FIGURES Framework Structure of Faujasitic Zeolite... Dependence of the Physical Prgperties of H-ZSM-S Upon Aluminum Content .............. Shape Selective Dehydration of n-C4OH and i-C4OH Utilizing Ca A (4.2 A) vs. the Non— Shape Selective Reaction Utilizing Ca X (7.5 23) ..................................... Clay Mineral Classification ................. Representative Structures of Some Related 2:1 Layered Silicates ....................... Model of [A11304 Metal Hydrolysis and Aggregation Pathways... XRD Patterns for Air-Dried and Freeze-Dried Cr3 s—Montmorillonite Purged at 25°C and 350°C for Two Hours in Argon ................ XRD Patterns for Air-Dried and Freeze-Dried Crl 24-Montmorillonite Purged at 25°C and 350°C for Two Hours in Argon ................ ESR Spectra of Purged Chromium Clay Cata- lysts ....................................... ESR Spectra of Reduced Chromium Clay catalystSOOOOOOOOOOOOO ..... 0.0... ........... Optimized Cyclohexane Dehydrogenation Activity for Cr1 88-Mont.(I), Cr3'53- Mont.(II) and Crl 24-Mont.(III) Catalysts... viii (7-x)+ (OH)24_X(H20)12] Ion... Page .. 4 .. 5 .. 9 .. ll .. 14 . 22 .. 26 .. 43 .. 44 .. 48 .. 49 .. 57 Figure l3 14 15 l6 17 18 19 20 21 22 23 24 25 26 Page Initial Cyclohexane Reactivity and Surface Areas After Two Hours on Stream as a Function of Hydrogen Pretreatment ........ 58 Cyclohexane Dehydrogenation Activity for Cr1 24—Mont.(III) After Fifteen Minutes and Seventy-Five Minutes on Stream ............ 60 Cyclohexane Contact Time Dependence Upon Reactivity for Crl 24-Mont.(III) Catalyst ..... 61 Cyclohexane WHSV Dependence Upon Reactivity for Crl 24-Mont.(III) Catalyst ................ 62 Cyclohexane Uptake by Spent Catalysts for Air—Dried Chromium Catalysts After Two Hours on Stream .................. . ............ 64 Cyclohexane Dehydrogenation Activity of Crl 24-Mont.(III) Catalyst with Time .......... 66 B-isopropylnaphthalene Dealkylation Activity for Cr .53-Mont.(II) Catalyst as a Function 0 WHSV ......................... 67 Dependence of Cyclohexane Dehydrogenation Activity on the Contact Time of Hydrogen Reduction for Cr -Mont.(III) and Cr 0 / 1.24 2 3 Al O ......................................... 69 2 3 Cyclohexane Dehydrogenation Activity for Cr1.88-Mont.(I), Cr3053—Mont.(II), Crl 24- Mont.(III) and Cr203/A1203.. .................. XRD Patterns for Air-Dried and Freeze-Dried Chromium Laponite Purged at 25°C and 350°C for Two Hours in Argon ................... ..... 76 71 XRD Patterns for Air-Dried Chromium Fluoro- hectorite Purged at 25°C, 110°C, and 350°C for Two Hours in Argon ................. ....... 77 27Al-NMR of Pillaring Solutions Employed for Interlaying of Smectite Minerals .......... 86 X-Ray Diffraction Patterns for Freeze-Dried Laponite Interlayered with Aluminum Chlorhy- drate, Purged for Two Hours at 25°C in Argon.. 90 X-Ray Diffraction Patterns for Freeze—Dried Laponites Interlayered with Aluminum Chlor- hydrate, Purged for Two Hours at 350°C in Argon .................. .. ..................... 91 ix TXFW' Figure Page 27 X-Ray Diffraction Patterns for Air-Dried Laponite Interlayered with Aluminum Chlorhydrate, Purged for Two Hours at 25°C in Argon ................................ 92 28 X-Ray Diffraction Patterns for Air-Dried Laponites Interlayered with Aluminum Chlorhydrate, Purged for Two Hours at 350°C in Argon ............................... 93 29 Physical Properties of Freeze-Dried Laponite Interlayered with Aluminum Chlorhydrate as a Function of Aluminum Ions Per Unit Cell ........................... 96 30 A Layer Aggregation Model for Air-Dried and Freeze-Dried Pillared Clays with a Particle Size <2u ................... . ........ 98 31 A Layer Aggregation Model for Air-Dried and Freeze-Dried Pillared Clays with a Particle Size <500 A ......................... 101 32 X-Ray Diffraction Patterns for Freeze- Dried Laponite Interlayered with Base- Hydrolyzed A1C13 at r = 2.00 and Purged for Two Hours at 25°C Under Argon. ........... 103 33 X-Ray Diffraction Patterns for Freeze- Dried Laponite Interlayered with Base- Hydrolyzed AlCl3 at r = 2.00 and Purged for Two Hours at 350°C in Argon.... .......... 104 34 X-Ray Diffraction Patterns for Freeze- Dried Saponite Interlayered with ACH and Purged at 25°C and 350°C for Two Hours in Argon.......... ......... . ................. 106 35 X-Ray Diffraction Patterns for Freeze- Dried 0.03u-Montmorillonite Interlayered with ACH and Purged at 25°C and 350°C for Two Hours in Argon ............... . ........... 107 36 X-Ray Diffraction Patterns for Air-Dried Montmorillonite Pillared Products (I) r= 2.40, (II) 0.80 g ACH/g Clay and (III) 2.6 g ACH/g Clay Purged for Two Hours at 350°C in Argon ..... ... ............... . ....... 111 37 X-Ray Diffraction Patterns for Montmoril- lonite Pillared Clays (r=2.50, Q=15.1) Sub- jected to Various Washing Conditions and Purged for Two Hours at'350°C in Argon ........ 113 X Figure 38 39 4O 41 42 43 44 45 46 47 48 49 X-Ray Diffraction Patterns for Air-Dried (Glass) and Freeze—Dried Montmorillonite Clay Purged for Two Hours at 350°C in Argon... 116 X-Ray Diffraction Patterns for Fluoro- hectorite Interlayered Clay Synthesized at Various r Values (Q=6.46) and Purged for Two Hours at 350°C in Argon ............... 120 B-isoprOpylnaphthalene Dealkylation Con- version Dependence as a Function of Reaction Contact Time ......................... 126 B-isopropylnaphthalene Dealkylation Con- version as a Function of WHSV ................. 127 Catalytic Dealkylation of B-isopropyl- naphthalene by Air-Dry (AD) and Freeze- Dry (FD) Forms of Montmorillonite (Samples XV-XVIII), and by Delaminated Laponite (Sample XIX) ............. . ..... . .............. 129 Gasoline (CS-Clz) Yields Obtained with Delaminated Clay (0), AAA-Alumina (I), and Zeolite-Promoted FCC (A) Catalysts ............ 132 (A) Light Cycle Gas Oil (LCGO) Yields and (B) Slurry Oil (80) Yields Obtained with Delaminated Clay (0), AAA-Alumina (In, and Zeolite-Promoted FCC (A) Catalysts ............ 133 (A) Propane and (B) Propylene Yields Obtained with Delaminated Clay (0), AAA- Alumina (I) , and Zeolite-Promoted FCC (A) Catalyst... ..................... .. ........ 134 (A) n-Butane, (B) i-Butane, and (C) Bu- tenes Yields Obtained with Delaminated Clay (0) , AAA-Alumina (I) . and Zeolite- Promoted FCC (A) Catalysts .................... 135 H2 Yields Obtained with Delaminated Clay (0) , AAA-Alumina (I) , and Zeolite-Pro- moted FCC (A) Catalysts..... ..... . ............ 136 Carbon Yields Obtained with Delaminated Clay (0) , AAA-Alumina (I) , and Zeolite- Promoted FCC (A) Catalysts.... ................ 137 (A) House-of-Cards Structure for a De- laminated Clay Catalyst. (B) The Long-Range Layer Stacking in a Well-Ordered Pillared Clay .......................................... 140 xi CHAPTER I INTRODUCTION Introduction The production of commercial chemicals involves exten- sive use of homogeneous and heterogeneous catalysts. Hetero- geneous catalysts were used for almost 60% of production capacity in 19831, estimated to be 499 x 109 lbsz. Zeolites, by far the largest volume commercial catalysts were employed at a rate of 7.5 x 106 metric tons per day as cracking cata- lysts in 19783. In 19604, zeolites were shown to possess good isomeri- zation ability. The high acidity and regular pore struc- ture of the zeolite family of catalysts enhanced gasoline production over that obtained with amorphous catalysts by increasing selectivity to gasoline along with lower gas and coke yields. Although a large number of zeolites have been synthe- sized, only the Faujasitic X and Y types as well as the other zeolites listed in Table 1 have found wide-spread uses. Faujasite type zeolites have the largest pores, correspond- 0 ing to 7.5 A. Modern cracking catalysts comprise between m N.m x m o.m mmnom m NNANoncNANoHacmmz muHcoHum m N.v mmmmo m NHHNoncNHANoHavNHmz « « o.N x m N.@ mmuom NH m.mmHNon.N.mANoHa.N.mmz mpHcmwuoz m m.m mmnom oH oaHNoncmANoHacmmz mlsz m m.N mmmmoummsm NH mmHANon.mmHNoH550°c ( \‘s ” \:Si’/ /Al// ) —————————> o’/ ‘o o \o o o N o o - o o o . \ . . / ( ::sr” ’Al:’ \\Sl’/ \:§l+ fi1\\ )+ H20 0 ‘o o 0 o’ \o o “o o 0 Through the selective substitution of a variety of ex- change cations, molecular sieve type compounds may be gen- erated. As early as 1932 it was known that the molecular volume of the sorbate had an influence upon the amount which could be absorbed by the zeolite8. Molecular sieving was first demonstrated in a series of papers and patents in the early 194059-11. Eventually zeolites were classified into five distinct types based upon sieving characteristics, as shown in Table 212. The sharp delineations in pore size and channel vol- umes shown in the table were recognized as a very desirable feature for applications of zeolites as Bronsted acid type heterogeneous catalysts. In 1960, Weisz, et al.13 published one of the first examples of shape-selective catalysis. :5. R. 0.0.. Eon: X2 05.... 2...: to. :5: 05m .09... ! 5 (il I I l ~09... 0:0: .39.: Xe. 0.6.0. 0.0:... 8.. ..E: 0~.m m 09C. 3343 20.. 32.: 2.5.0:.» 6:: o...o0~-am .0Hm HMHSUQHOE mEOm MO COHuMOHmemMHU .N GHQMB Their work made use of metal exchanged aluminosilicates and the conversion of decane to lower molecular weight olefins and paraffins. The first example of shape-selective catalysis with synthetic crystalline aluminosilicates was demonstrated in 1962 by Weisz and co-workers14 utilizing Ca A and Ca X (with 4.2 and 7.5 A pores) for the dehydration of primary and branched alcohols. These results are repre- sented in Figure 3. More sophisticated examples involve the alkylation of toluene to near quantitative yields of para-xylenels, and methanol dehydration to gasoline16 in one step. Shape selectivity can be manifested in several waysl7. Reactant selectivity is observed when reactants within a specific size range have access to the active site because of molecular sieving effects. Product selectivity occurs when only certain products generated within the pores have the correct molecular dimension to diffuse out of the open- ings leading to the cavities. Restricted transition state selectivity occurs when certain reactions are excluded be- cause the transition state necessary is unattainable due to steric limitations. Qualitative recognition of any of the selectivity prOp- erties mentioned above can be conveniently measured through the use of the constraint index18. The constraint index approximates the relative cracking rate constant for two hydrocarbons, hexane, and 3-methylpentane: Constraint Index = log10(Fraction of n-hexane remaining) log10(Fraction of 3-Methylpentane remaining) ‘-__~".__n -..—.‘- 100 -— X o\° c: . .3 m 50 - 3.4 CD > . 8 A U '4 0 * I 200 250 300 Temperature °C Figure 3. Shape Selective Dehydration of n-C OH and i-C OH o 4 4 Utilizing Ca A (4.2 A) vs. the Non-Shape Selec- O tive Reaction Utilizing Ca X (7.5 A). 10 A shape selective catalyst has a constraint index (C.I.) greater than unity. While the constraint index may 19 have some value, it has been criticized since the test reaction involves the simultaneous diffusion of two re- actants into the catalyst structure. Elaborate models20 have been formulated to describe the diffusional character- istics for ZSM—S in the constraint index reaction. Regardless of the precise mechanism of diffusion within these intracrystalline environments, the high acidity, large surface area and molecular sieving properties of zeo- lites make them well suited. Us perform as heterogeneous catalysts. Their only major disadvantage resides in their rather small largest pore opening, that being 7.5 A found in faujasitic type X and Y zeolites. Smectite clay minerals may be used effectively as molecular sieves and catalysts due to their ability to ex- change cations and to swell in the appropriate polar sol- vent. The relationship of various clay minerals is de- picted in Figure 4, along with a classification scheme pre- 21 sented in Table 3 and unit cell data in Table 4. The structure of montmorillonite and other related smectite minerals is shown in Figure 5. The smectite family is comprised of silicate sheets containing four sheets of oxygen atomszz. A center layer of octahedral holes is sandwiched between two layers of tetrahedral holes. The octahedral layer generally contains metal ions suitable for octahedral coordination, i.e. A13+ or Fe3+ and Mgz+. The tetrahedral layer usually contains Si4+. The upper ll CLAY MINERALS I 1 CRYSTALLINE AMORPHOUS eg. Allophanes I 1 TWO LAYER THREEILAYER MIXEDILAYER CHAINS eg. Kaolin eg. China Clay Chrysotile Asbestos I l EXPANDING LATTICE NON-EXPANDING LATTICE eg. Talc l UNLIMITED LAYER EXPANSION LIMITED LAYER EXPANSION ie. SMECTITES . ie. VERMICULITES I l DIOCTAHEDRAL TRIOCTAHEDRAL MONTMORILLONITE HECTORITE BEIDELITE SAPONITE NONTRONITE Figure 4. Clay Mineral Classification. silicate layer has its vertices pointing downward, and the lower sheet has its vertices pointing upward. These three sheets are combined so that the tips of the tetrahedrons of each silica layer and the hydroxyl oxygens of the octahedral layer form a common layer. When Al3+ occupies two out of three octahedral positions, the mineral such as montmoril— lonite is classified as a dioctahedral mineral. When all 2+ three octahedral positions are occupied with Mg , as in saponite, the mineral is said to be trioctahedral. 12 mpHCOHCHHU mMOHz mauuflum HmupmnMDUOHHB mpflummcmz move: mauuflnm Hmupmnmgooflo 00H: mauuflum o.m ouflmomoHnm mMOHS Hmupmnmvooflua mpfl>oomaz mMOHZ Hmuomnmu00Ho 00H: o.H GHHHSOHEH0> kuomnmuooflue myHHooHEHm> kumeMHUOHHB muHHDUHEHm> Hmupmnmpooflo mmpflasoHEHm> HmupmnmuOOHo muflasoNEHm> m.o:o.o mafiaoosmm mDHHovomm mmuflcommm .maflcommm Ho mmufluomEm HmuomLMDOONHE muflaouuaoz mpflcommm mpflaamoflmm muflcoHHHHOEpcoz ImpflquHHHoEucoz muflcoHHHHoEHGOZ Ho mmufluomEm HMHUoSMHOONQ Ho mufluomEm o.oumm.o came moama Hmnpmnmuoowua oamu mpflaamnmouwm mmuflaamcmouwm Hmnomzmgooflo Iogflaamnm0Hmm c Hum mHHDOm>H£O mmqflpcmmumm Hmuposmuooflue muflcfiaomx mmuflcflaomm HmuomSMDOOHo muflcflaomm o HHH peso masanom moflommm moonmnsm moouw mom mmumnu mmhe .mmDMOHHHmoaamnm How mfimzom GOHHMOHMHmmmHU Hmfluumm .m 0Hnme l3 4 .o.NHmo.0Hm. 00 01:0.ONOAxHHx Ho.omz. H..rHéHoiNo .xHHxlo.0mz. «.mo.om.o.wfl .mo.ONo .o.0mz.N: Hmomoa>nm -o.mHm. N .:\x o m» +:z muflcommm .0 me. N .:\x 0 mm +:z OHHHOnvam m..o.0az. UHMB H.Eo.ONo.o.NH.,Ho.0Hm.HONHEN: mgfl>oom52 HV.mo.ONo.xH. s :\x +: opflcoHHHHOEusoz H2.183205%..002. mpflfiahzmou>m nova: mmpfluomEm Gama I wuflaahsmonhm HMHUoQMDOOHHB HmuomSMDOOHQ msouw Hmumcflz .moDMOHHHm pmuowmq Hum Mom mmaofiuom Hamo #NGD .v OHQMB 14 BEIDELLITE NONTRONITE AI" m 5i“ Figure 5. Representative Structures of Some Related 2:1 Layered Silicates. 15 Existing between each layer are exchangeable cations. The number of exchangeable cations existing between each layer establishes the Cation-Exchange-Capacity (CEC) of the mineral. The CEC can be measured by saturating a sodium exchanged clay with ammonium or barium ions, and determining the quantity of ammonium or barium incorporated at pH 7. These exchangeable cations serve to balance the net negative charge distribution for a smectite silicate sheet. The charge distribution for pyrophyllite is as follows: 6 o’2 4 Si4+ Layer common to Oh' Td sheet 4 02—, 2(OH)_ 4 [Al3+] Layer common to Oh, Td sheets 4 02-, 2(OH)- 4 314+ 6 o”2 yielding the theoretical formula NaX(OH)4818(Al4)020.nH20 with an average distribution of 66.7% SiOZ, 28.3% A1203 and 5% H20. As can be seen from the charge distribution, no charge deficiency exists. Isomorphic substitution of A13+ by Mg2+ or Fe3+ in the octahedral layer will produce a net negative charge in that layer. It is this negative charge that attracts and holds interlayer cations giving rise to the structural integrity of a 2:1 layered silicate. Sub- stitutions within the lattices are also common in the tet- rahedral layer where aluminum easily replaces silicon. This substitution effect accounts for approximately 16 80% of the observable CEC. Broken bonds within the struc- ture account for approximately 20% of the observable CEC. The contribution of the hydroxyl groups to the CEC is minimal since these groups do not possess exchangeable pro- tons. With pyrophyllite or talc, where aluminum and magnesium respectively occupy the octahedral layer, there is no net negative charge upon the silicate sheets. These two clays have no intercalation properties. Smectite clays are characterized by a well ordered structure, exchangeable cations, and an expandable (002) interlayer spacing through the use of the apprOpriate polar solvent. These properties make smectites suitable for the intercalation of molecular props (pillars) between the layered sheets. Pillared clays can act as molecular sieves by limiting the size of the molecule accessible to the in- terlayer. To make these pillars catalytically active, we can heterogenize previously homogeneous catalysts in the smectite interlayer. There are advantages to heterogenizing a homogeneous catalyst, among them: (1) easy recovery of the catalyst from the reaction medium; (2) sometimes increased stability to high temperatures and hydrolysis; (3) no solubility limitations and (4) inducement of shape selectivity. Pinnavaia, et al.:x3have shown that heterogenized di- rhodium acetate complexes are useful for alkene and alkyne hydrogenations with total quenching of the isomerization of l-hexene to 2—hexene. This isomerization accounts for 35% 17 of the products in the homogeneous systems. Chang and Pinnavaia24, and Farzaneh and Pinnavaia24 have shown that asymmetric hydrogenations and hydroformylation reactions can be carried out on smectite silicates with equivalent or enhanced activity as compared with homogeneous cata— lysts. Thomas, et al.25 have shown that refluxing l-hexene, l-heptene, and 1-octene in a hexane solution with a cation exchanged smectite converted these alkenes to their cor- responding bis-sec-alkyl ethers. The exchanged cations were Cu2+, Fe2+, Fe3+, Cr3+, and Al3+. Intercalation of the cation is necessary for reaction, along with a swelled interlayer. The importance of this synthetic route is apparent when compared to the standard Williamson synthesis of ethers. In the Williamson synthesis R'-X + Na-OR-———)R'-OR + Nax primary alkyl halides work best, whereas secondary and tertiary halides yield the elimination product as well as the substitution product. The synthetic route of Thomas, et al. does not have this rearrangement problem; it allows for the easy production of such ethers as 1-hexyl-2-propyl ether and di-Z-hexyl ether. These examples illustrate that smectite silicates have been recognized as useful supports. In order to utilize these interlayered smectites in industrial processes, high temperature stability must be achieved to allow the catalyst to withstand regeneration procedures. There have been 18 several successful attempts to construct a swelling re- sistent, cross-linked smectite which could serve as a backbone for catalysis or act as a molecular sieve. Barrer and Macleod26, and Barrer and Reay27 in the 19505 replaced the charge balancing cations of montmoril- lonite with the tetraalkylammonium cations N(CH3)4+ and N(C2H5)4+. These pillaring reactions increased the in- tracrystalline porosity and yielded basal spacings of 13.5 A and 13.9 A for the tetramethyl and tetraethyl am- monium intercalates, respectively, as compared with 9.6 A for a Na+-montmorillonite. At 70 K and 90 K it was shown that the order of sorp— tion for some gases was 02 > N2 > Ar. At room temperature pentanes were sorbed in the following sequence, n-C5H12 > iso-C5H12 > neO_C5H12' This sorption was in the inverse order of cross-sectional molecular diameter. Clementz and Mortland28 similarly showed that tetra- alkyl ammonium cations can be intercalated into reduced charge montmorillonite. Charge reduction is achieved by _exchange with lithium cations before intercalation of the tetraalkyl ammonium cations, with internal surface areas as high as 200 m2/g being reported. Several tris-metal chelates have been intercalated 2+ into layered silicates, including Cu2+ and Fe 1,10- phenanthrolinezg; Fe2+, Cu2+ and Ru2+ tris-bipyridleO; Co3+ and Cr3+ tris ethylenediamine31; and others32. In- tercalation was shown to occur by two methods: (1) cation exchange up to the mineral CBC and (2) intersalation of 19 excess salt beyond the mineral CEC. Ion exchange proceeds through electrostatic interactions, while intersalation de- pends on the anion present. Intersalation tends to be favored by the anion's ability to ion pair (SO42— > Br- > Cl-); and the ability of the M(Chelate)3n+ to shield the anion from the electrostatic charge of the silicate sheet. Structural stability to 250°C is attainable for some of these tris-chelates. Basal spacings of 18 A were obtained through CBC-type exchange, and spacings of 29 A were re- ported for the intersalation product. Silica33 has been intercalated in smectites by two different mechanisms: (1) hydrolysis of Si(acac)3+ ex- changed smectite, or by (2) the in situ reaction of SiCl4 and H(acac) to yield Si(acac)3+, followed by hydrolysis of this product. These two methods yield basal spacings of 12.6 A, with surface areas from 40 to 240 m2/g and tempera- ture stability to 620°C. This system may have industrial importance for selective adsorption and as a catalyst sup- port. The protonated dication of l,4—diazabicyclo(2,2,2,) octane, also known as DABCO34’35, can be intercalated into smectites yielding 14.8 A basal spacings. This pillar introduced Bronsted acidity to the inter- layer, structural stability to 220°C and shape selectivity caused by an effective interlayer spacing of 5.3 A and a 0 calculated interpillar distance of 6 A. These two effects are illustrated in Table 5. 1! I l 3! \ 20 Table 5. Relative Rates of Esterification of C4 and C5 Carboxylic Acids* Acid H—Montmorillonite DABCO—M CH3 (CH2)2 COOH 1.65 1.00 (CH3)2 CH COOH 1.49 0.48 CH3 (CH2)3 COOH 2.01 1.24 (CH3)2 CH CH2 COOH 1.95 0.67 * reacted with n-butanol T = 190°C, LHSV = 6.0 hr‘1 As can be seen by the data in Table 5, DABCO-M has greater shape selective properties. As the kinetic diameter increases (normal XE' branched) the relative conversion rate decreases by a greater percentage for DABCO than for montmorillonite. This can be attributed to a rigid pore size for DABCO as opposed to the variable interlayer of a free-swelling smectite. Several inorganic supports have been synthesized which confer Bronsted acidity and shape selectivity to the smectite interlayer. These pillars can be generated by hydrolysis of a metal chloride solution over a range of OH:M ratios36a_e. The extent of cation oligomerization is controlled through selected experimental conditions (pH, age, and temperature). The mixing of hydrolyzed metal ion solutions with a suspension of fully dispersed sodium ex- changed smectite leads to insertion of the polyoxy cation between the negatively charged silicate sheets. The 21 flocculated product is then washed and air dried or freeze dried, followed by calcination which is believed to trans- form the metal complex into a metal oxide with a smaller hydration sphere36d. An Al—PILC system has received considerable attention; it is characterized by a d(OO£) basal spacing of over 18 A at room temperature, and at 600°C the basal spacing is O 16 A36a,b,c,d. 27 37 The Al pillared species has been probed by Al-NMR and appears to be a polyoxy aluminum cation of the family (7—x)+ 38 [A11304(OH) where x = 4. The structure 24-xH2012] (Figure 6) has one central aluminum atom tetrahedrally coordinated to four oxygen atoms, with three aluminum octa- hedra attached edge on to each vertice of the central aluminum tetrahedron. This ion is consistent with an ob- served interlayer spacing of approximately 9 A. Surface areas range as high as 500 m2/g with average pore sizes in the 20 A range. The mechanism of crosslinking to the silicate sheet has not been fully elucidated. It appears that the first stage of attachment is mainly an electro- static one. A plausible mechanism for the observed Bronsted acidity can be related to the degree of hydration of the complex. Mortland and Raman39 showed that the pK of hydrated ex- changeable cations decreases with dehydration. As the water content of the metal complex decreases (as caused by calcination), the polarization forces arising from the metal centers become more concentrated on the remaining 23 water molecules. This causes an increase in hydrolysis, with an increase in proton donating abilities. Several catalytic systems which proceed through Bronsted acidity have been investigated. Large molecule 40’41, esterificationszs, and catalytic cracking 42,43 cracking of gas-oil were the most illustrative reactions em- ployed. It was shown that for a pillared or a delaminated clay, the gas-oil cracking selectivity more nearly re- sembles a commercial zeolite promoted FCC catalyst than an amorphous aluminosilicate catalyst. In addition, the delaminated clay yields greater amounts of light cycle gas oil (b.p. 221-343°C) and lower amounts of slurry oil (b.p. 343-426°C) than the zeolite based catalyst. Brindley andYamanaka44 have synthesized a chromium hydroxide montmorillonite which possessed a 17 A lattice expansion at 200°C. Tzou45 recently reported the synthesis of a chromium interlayered clay which was characterized by a 27 A lattice expansion at room temperature, excellent thermal stability and novel catalytic activity46. Generally there are two distinct methods for the synthesis of an intercalate. The first method involves the modification of an intercalated metal complex to form a hydroxy-metal polymer. The second method involves the exchange of a previously synthesized hydroxy-metal polymer into the layered silicate. Method one is illustrated in Scheme 3 for the series of reactions utilized to synthesize silica pillared clay33. Materials generated in this fashion generally contain 24 Scheme 3. Si(acac)3+ + Na+ -———€> Si(acac)3+ + Na (1) Si(acac)3+ + H20 -———-> Si(OH)4 + H+ + 3(CH3CO)2CH2 (2) monolayer metal-hydroxide polymers. Eventually brucite or chlorite—like structures fill the clay interlayer. This method leads to low surface area materials which are not desirable due to the restricted interlayer. Method two, direct exchange, is the method of choice for producing well-ordered, high surface area pillared smectites. Cationic hydroxy—metal polymers may be synthe- sized by the base hydrolysis of almost any desired metal. If the size of the polyoxo—polymer can be regulated, the lattice expansion of the clay can be varied accordingly. As the charge density upon a metal increases, the polari- zation forces exerted upon the coordination sphere of solvent increases. In aqueous solution, the positive charge upon the metal will cause the loss of a water proton. Upon the addition of base, soluable polymeric species of high molecular weight will form. This is represented in Scheme 4: Scheme 4. y+1 [ I *I + . O - O\ (y-l) / \ on \ / / [M(H20)x_10H] -———+ (HZO)M\\ ’,M(H20)x-;]'——9 .rM.\ ,.M~. H <—— 1 25 This mechanism, termed olation, yields condensation poly— mers that appear as hydrous gels which are generally more than 99 volume percent water. If the hydrated metal is allowed to react with hydroxide at elevated temperature, or at an artificially adjusted pH, the hydroxy bridged polynuclear species may undergo oxolation reactions as de- picted in Scheme 5. Scheme 5. H I x+ (x-4)+ \M/O\M/O\M’ > \M/O\M/O\M/ + 411+ / \O/ \O/ x / \0/ \O/ \ I I n H H As the degree of aggregation increases, high molecular weight polymers will result. If the solution is allowed to age long enough, colloidal particles will result, with eventual precipitation of a metal-hydroxy polymer. A scheme for metal ion hydrolysis aggregation is presented in Figure 7. Through the appropriate choice of pH, temperature, and aging, the size of the polyoxocation grown may be con- trolled. In this manner, the synthesis of intercalated clay catalysts with pore openings larger than 7.5 A may be undertaken. 26 .mwm35umm coflummmummm cam MHmmaonpmm Hmumz 8.3.3: .222 33:02 Illllllllllllfi o o o 0 ‘ll _+x 6 n 8 ..— .........L$:..OV s. iguasfov m‘ 2:902 _ lac Unv o o o a ‘II N n n 33358 .IN L959 .2 Al 173.22 .2 820230; N“ N N 2:294 ....AII+A~-»N.AIQ 2111839 .2 3.8.2.8 coo-035:8 2:230 . . . ZIfifocz flzzolelu . . N 13-3.192113 3921;: 5.950 3202.952 £3.05»... .5 musmflm uouobanbv k fir 27 Objectives of Dissertation Research A variety of cations have been shown to permanently expand smectite-type layered silicates. The physico— chemical properties of clay pillared by polyoxocations are poorly characterized and understood. In general, certain properties of smectite clays are typically overlooked in pillared clay chemistry. The layer charge and partical size are quite heterogeneous. This charge heterogeneity influ- ences the interaction of pillaring species upon the clay surface. It might also serve to promote cluster rearrange- ment or degradation once the pillaring species is in close approach to the silicate layers. Since the properties of a heterogeneous catalyst are intimately related to its synthesis history, the effects of various washing, drying, and calcination methods are of interest. The history of the hydrolysis reaction itself upon cluster synthesis is also of paramount importance. In this dissertation, these questions will be addressed by probing the physical and catalytic properties of smec- tite clay minerals pillared and delaminated by polyoxo- cations of chromium and aluminum. CHAPTER II EXPERIMENTAL METHODS Clay Preparation The clay minerals Laponite-RD® (Laporte Ind. England) and fluorohectorite (Corning Glass, USA) were used as sup- plied from the manufacturer. Montmorillonite (Source Clay Minerals Repository) con- tains impurities which must be removed in order to simpli- fy and standardize the evaluation of clay catalysts. Solu- able salts and carbonates are removed from the clay in order to promote flocculation of the clay. Calcium carbon- ate may act as a cementing agent preventing intercalation of the desired metal oxide species. It may also interfere with CBC determination due to the equilibrium represented in Scheme 6. Scheme 6. 2.. Caco3——->Ca2+ + C03 The addition of sodium acetate (NaAC), buffered to pH=5 with acetic acid allows for the removal of carbon dioxide upon digestion at 70°C according to Scheme 7. The 28 29 Scheme 7. CaCO + 2H+——)Ca2+ + H0 + CO 3(5) A m 2(g) following procedure, based on a 5 g sample of clay mineral, was employed for the removal of soluable salts and carbon- ates: (1) Add 50 ml of l N sodium acetate buffered with NaAC and bring the clay into suspension. (2) Digest this suspension for 1 hour at 70°C with occasional stirring. (3) Centrifuge the solution, discard the super— natant. (4) Repeat the above three steps. Free (non-lattice) iron oxides which can also cement the clay are removed by treatment with a sodium citrate/ sodium bicarbonate solution and subsequent low tempera- ture (80°C) treatment with sodium thiosulfate. (1) Add 40 ml of 0.3 N Na-citrate and 5 ml of i N NaHCO . The citrate (C H O ) chelates with ferric iron and prevents precipitation of FeS. The bicarbonate maintains neutrality and fur- nishes hydroxide ion when hydrolyzed. (2) Warm the suspension to 75—80°C and slowly add 1 g of Na 8204. Do not_exceed 80°C or FeS may precipitafe. Digest for 15 minutes. This re- agent effects reduction of ferric ions to fer- rous ions which can then be washed away: 4 0H" + szo4z'——> 25032’ + 21120 + 2e" 2e' + 2Fe3+———) 2Fe2+ 2' + 2H 0 net: 4 OH‘ + s o 2" + 2Fe3+—-92Fe2+ + 2503 2 2 4 (3) The resulting solution is then cooled and con- centrated by centrifugation. The supernatant is discarded. 3O Hydrogen peroxide is then employed to digest organic matter present in the clay. At this stage the clay is sodium saturated by the addition of sodium chloride. The flocculated particles are then collected, washed until free of chloride ion (as tested with silver nitrate) and air dried. Pillaring/Delamination Reaction Two types of polyoxo—aluminum cation solutions were utilized for the aluminum interlayered clay work. One type of solution was a freshly hydrolyzed aluminum tri- chloride hexahydrate solution which contained between 0.0 and 2.50 mol of hydroxide per mol of aluminum. The other type of solution was the commercially available Chlorhydrol® solution (Reheis Chemical Company) of aluminum chlorohy- drate, freshly diluted with water to a concentration of 0.23 g. The amount of solution used in the exchange re- action was varied so that the ratio of mmol Al3+ to milli- equivalents of clay (Q), varied from 0.0 (no Al3+ present) to approximately 85. The pillaring solution utilized for the synthesis of chromium interlayered clay was a chromium nitrate solution freshly hydrolyzed with anhydrous sodium carbonate. This hydrolysis mixture was aged at 95°C for 36 hours prior to exchange with clay. The exchange reac- tion consisted of aqueous suspensions of clay ( 1 wt. %) slowly added at room temperature to the vigorously stirred pillaring solution. Aluminum reaction mixtures were al- lowed to age for two hours at room temperature, chromium reaction mixtures for 1.5 hours. The clay intercalates 31 were then washed free of excess electrolyte with deionized water. Final products were either air- or freeze-dried. Catalysts were dehydroxylated at 350°C in helium for two hours. Use of this procedure enabled clays to be stored in- definately in air without loss of catalytic activity. X-Ray Powder Diffraction (XRD) Measurements A Philips or Siemens X-Ray diffractometer with nickel filtered Cu KG radiation (kg = 1.5405 A) was used to measure basal spacings. For air-dried samples, 1 ml of 1 freshly prepared pillared clay was spread across a 1 inch by 1 inch glass slide and allowed to dry. These slides were then heated to specified temperatures for 2 hours in inert gas. Freeze-dried samples were heated for 2 hours in inert gas at specified temperatures, then slides were prepared before XRD measurements were made. The Bragg angle 20 peak positions were converted to d—spacings by use of a standard Cu KG radiation d-spacings 20 chart. Chemical Analysis Elemental analysis of Al clay samples were carried out at the inorganic laboratory of the Department of Toxicology, Michigan State University, with a Jarrell-Ash 955 Atom-Comp instrument. J.T. Baker instra—analyzed grade standards were used for the analysis of Si, Al, Fe, Mg, Na, and Ca. NBS plastic clay 98a served as a clay standard. Clay samples were prepared for analysis in my laboratory in the follow- ing fashion. All clay samples (0.05 g) were fused with lithium borate (0.3 g Gold Label, Aldrich) for 12 minutes 32 at 1000°C in preignited graphite fusion crucibles. The resultant glass was transferred to 30 ml of 3% HNO This 3. solution was mixed for 10 minutes, or until complete dis- solution of the glass was assured, then diluted to 100 ml with deionized water. Galbraith Analytical Laboratory, Inc. analyzed chromium clays for Cr and Si from which unit cell compositions were derived based on the assumption that the Si content of the clay remained constant throughout the pillaring reaction. Carbon analysis on spent catalysts was performed by Galbraith Analytical Laboratory. BET Surface Area Measurements Surface area measurements were determined on a Perkin- Elmer-Shell model 212B sorptometer at liquid nitrogen temperature with nitrogen as the adsorbate and helium as the carrier. Nitrogen adsorption capacities were obtained for three partial pressures at -196°C. Approximately 100 mg of sample was used for each measurement. All sam- ples were degassed for 2 hours at 350°C under flowing helium prior to nitrogen adsorption. Adsorption Uptake Isotherms The adsorption of organic molecules of various kinetic diameters was measured on a McBain balance equipped with quartz glass springs.and buckets. The samples were activ- ated at 350°C under dynamic vacuum for 2 hours prior to measurement. The probe molecules employed were benzene, 1,3,5-triethylbenzene and perfluorotributylamine (PFTBA) 0 whose kinetic diameters are 5.8, 9.2, and 10.2 A, 33 respectively. The probe molecules were stirred and equili- brated in a constant temperature bath maintained at 20tl°C, approximately 5°C below room temperature, to insure that capillary condensation did not take place within the sample and yield artificially high readings. ESR Measurements ESR spectra were obtained using X—band radiation with a Varian E-4 spectrometer. All spectra recorded at room temperature were obtained at 2000 G field set, and a 4000 G field sweep, as well as a 3600 G field set and a 4000 G field sweep with a 6.3 G modulation amplitude. Diphenyl- picyrl hydrazine (DPPH) was used as a reference. Clay samples were subjected to two temperature treatments: (1) They were purged at 350°C for 2 hours in helium as well as (2) heated for 1 hour in helium, 1 hour in hydrogen, and then sealed under helium. The first treatment was part of the standard synthesis procedure. The second temperature treatment represented catalytic pretreatment conditions. Catalysis Vertically mounted, fixed bed continuous flow micro- reactors constructed of 7 mm I.D. quartz tubing were opera- ted in the integral mode. The reactors contained up to 1.5 g of catalyst which was diluted with a-alumina (Norton Chemicals) to yield bed heights of 3 to 7 cm. The reactors were encased in a tube furnace powered by a three stage temperature controller (Eurotherm, model 919A) fitted with a chromel-alumel thermocouple. Reactant 34 was introduced with a syringe pump (Sage Instruments, model 341A) at the top of the reactor to a preheater zone. Helium, used as a vector gas, was purified with BASF R3—ll catalyst followed by 4 A molecular sieves. The helium flow rate was controlled through a two-stage regulator followed by a tri-flat flow meter. All catalysts used in B-isopropylnaphthalene (Alfa Inorganics, 99+%) dealkylations were pretreated at 480°C for 1 hour in purified helium. Dealkylation products were analyzed after they were trapped out of the effluent by an ice bath, and then diluted with approximately 25 ml of benzene. The samples were manually injected onto a 5 foot % inch stainless steel column composed of 3% SE-30 sup- ported on diatomite, operated at 160°C. Peak integrations were obtained on a Sargent SR recorder and analyzed with a compensating polar planimeter. Cyclohexane dehydrogenation catalysts were pretreated at 550°C for 1 hour in helium to remove adsorbed water. This was followed by an additional hydrogen pretreatment (1 hr. 1 atm pressure) which reduced the chromium to the catalytically active +2 oxidation state. Spectrophoto- metric grade cyclohexane (Fisher) was used as the reagent, products were trapped at the base of the reactor in a liquid nitrogen trap, allowed to warm to room temperature and manually injected onto a 6 foot % inch stainless steel column of 10% OV-l7 on Chromosorb W. Integrations were obtained with a Disc integrator (model 204) equipped re- corder. CHAPTER III CHROMIUM INTERLAYERED CLAY CATALYSTS Introduction The growth of an oxometal species suitable for inter- calation is limited only by the desired metal's hydrolysis chemistry. The properties of the resulting clay inter- calate are intimately related to the hydrolysis chemistry and synthesis history of the oxometal species. Table 6 lists the d -spacings and surface areas associated (00%) with pillared clays. Some metals, such as chromium (III) Table 6. Physical PrOperties of Various Oxometal Pillared Clays O d(002) A (a) Sugface Area Metal Ref 25°C. 500°C M /g, 500°C Ni 47 14.8 12.1 -- Zr 48 18.8 16.5 280 Bi 49 16.6 9.6 35 (a) d(OO£) spacings after being heated to the temperature indicated show extensive hydrolysis chemistry which is highly depend- ent upon reaction conditions. 35 36 Brindley and Yamanaka50 have produced a chromium interlayered clay characterized by a 16.8 A spacing and a 250 mZ/g surface area at 25°C. The pillaring cluster was generated from a OH/Cr mole ratio solution of 2.00 using sodium hydroxide as a source of base. This pillaring solution was allowed to age for ten days, was mixed with montmorillonite and then allowed to age three additional days. The 16.8 A spacing is greater than would be ex- pected for a chlorite type structure (14.2 A). This clay product exhibited low thermal stability in air and col- lapsed to 9.8 i at 400°C. Tzou and Pinnavaia51 recently synthesized a chromium pillared clay product at a OH/Cr ratio similar to the value used by Brindley for the 16.8 A clay, but more severe cluster growth conditions were employed. .The chromium solution (OH/Cr=2.00) was aged for 36 hours at 95°C. Solid anhydrous sodium carbonate served as the source of base. After high temperature aging, a clay sus- pension was added according to the procedure outlined in Chapter II. This method produced an expanded intercalate with a 27.6 A spacing and a 350 m2/g BET surface area. Results for these two Cr-clay intercalates are presented in Table 7. It is apparent from these results that the hydrolysis of chromium is very sensitive to the experi- mental conditions utilized to generate the high molecular weight oxocations. Indeed, the hydrolysis chemistry of chromium is very dependent upon the method of base addition. In 1908, Bjerrum52 postulated the existence of the 37 Table 7. Physical Properties of Chromium Pillared Clays. (a) ° mmol Cr3+ OH/ p——-9(ooz)§————1(b’ 25A wt. % Cr Clay MEQ Clay CR 25°C 350°C 550°C m /g(500) in product I (Brindley) 7.3 2.0' 16.8 14.2 13.7 15 13.88 II (Tzou) 62.5 2.0 27.6 23.1 19.6 350 18.44 3+ (a) Ratio of Cr to clay used in reaction, (b) Heated under argon. 6+ 6+ . 53 Cr6(OH)12 and Cr12(OH)30 ions. He later undertook a study of sodium hydroxide hydrolyzed chromium nitrate at 75°C. Products isolated contained OH/Cr ratios of one and two, probably representing the Cr2(OH)24+ and Cr6(OH)lg+ ions, respectively. Precipitated crystalline Crzo3 . xHZO was also observed. This study also showed that equilibrium was generally attained in approximately 110 hours at reflux temperatures. Laswick and Plane54 studied chromic per- chlorate solutions generated without addition of base under reflux conditions. They showed that the rate of olation is extremely slow under these conditions. The dimeric complex [Cr2(H20)4(OH)2]4+, attained a maximum concentra— tion within minutes of reaction, which accounted for 11% of the total chromium present. Approximately 74% of the chromium ions remain as unreacted Cr(H20)63+, even after 27 days. The addition of sodium hydroxide at an OH/Cr ratio=1.00 produced an undetermined colloidal polymer after only three hours of reaction that accounted for 13% of the available chromium. Thirty-eight percent of this base hydrolyzed solution remained as the hexaquochromium (III) 38 species. Several other chromium species, such as a [Cr2(H20)OH]S+ dimer55 and [Cr30(CH3c02)6(H20)3]+ trimer56 have recently been isolated. These results suggest that the method used to form chromium polymers (i.e. temperature, pH, and aging condi- tions) will have a pronounced effect upon the structure of the polyoxochromium intercalated clays. This section concerns itself with the generation of several chromium intercalated clays; including those produced by Brindley's and Tzou's procedure and their characterization with re- spect to their physical and catalytical properties. Synthesis of Pillars and Intercalation Procedures Three distinct types of chromium interlayered clays were used in this study. Those of Brindley (I) and Tzou (II) are summarized in Table 7. Synthesis of high surface area, large pore pillared interlayered clays (PILC) is best achieved by method II, that of Tzou and Pinnavaia. Tzou's method of hydrolysis was carried out in the following manner, based on 1 g of clay to be pillared. An- hydrous sodium carbonate (5.3 g) is added slowly at room temperature to a 500 ml solution of 0.1 M chromic nitrate. This addition is done slowly enough to prevent loss of solution due to release of carbon dioxide upon dissolution of the carbonate. Typically this addition is complete within five minutes. The resulting solution is then aged for 36 hours at 95°C. Upon completion of this aging period, a 1 wt.% slurry of montmorillonite is added dropwise to 39 the vigorously stirred pillaring solution, and then allowed to further age for 1.5 hours. The resulting solution is washed free of excess electrolyte by repeated centrifuga- tion and dispersion in distilled water, then it is air dried on glass plates. The addition of solid anhydrous sodium carbonate to a 0.1 M chromic nitrate solution at an OH/Cr ratio of 2.00 always produces some precipitate. The amount of pre- cipitate was dramatically decreased after the solution was allowed to age for 36 hours at 95°C. The remaining precipitate was removed by centrifugation prior to the pillaring reaction. The presence of precipitate after high temperature aging indicates that equilibrium was not attained. This observation strongly suggests that a variety of polymeric species are formed since equilibrium is not attained. A cluster capable of producing the observed Ad(00£) of 17 A as well as chromium hydroxide precipitates are among the products. The addition of any solid sodium carbonate at any rate invariably produces localized hy- droxide concentrations conducive to chromium hydroxide formation. This precipitation is disasterous to cluster synthesis for several reasons. The hydroxide and chromium concentrations are not easily controlled as chromium hy- droxide effectively removes reactant from the solution. Once formed, these hydroxide species can become large, and may facilitate the removal of large amounts of carbonate, and disturb attainment of equilibrium. 40 Synthesis method III was designed to alleviate the problems associated with solid carbonate addition through the use of dissolved sodium carbonate (solution) as a source of base. A solution of sodium carbonate (0.25 M) was added dropwise to a chromic nitrate solution (0.17 M) to achieve a final OH/Cr ratio of 2.0. Chromium ions were formally present at a 0.05 M concentration. This addition is done very slowly, generally at a rate of 10 ml carbon- ate per minute while the chromium solution is agitated vigorously. Upon conclusion of this addition, the re- sulting solution was aged at 95 to 100°C (reflux) for 36 hours. This freshly prepared pillaring solution is now utilized in the direct exchange method of clay inter- calation. All chromium clay catalysts (I, II, and III) are then either air-dried or freeze—dried, and purged at 350°C for two hours in either inert gas or vacuum. This procedure allows for long term storage without al- teration of physical or catalytic properties, probably due to conversion of the pillaring species to a chromium oxide cluster. Physical Properties of Chromium Interlayered Clay Catalysts I, II, and III. Typical unit cell compositions are presented in Table 8, along with chemical formulas derived from analytical re- sults. Variations in chromium hydrolysis conditions leads to a variety of intercalated species. The X-ray diffrac- tion patterns for air-dried and freeze-dried products sub- jected to various temperature treatments are shown in 41 Table 8. Unit Cell Compositions of Chromium Pillared Clay Prepared by Hydrolysis of Cr3+ with NaOH, Solid Na2C03 and Solution NazCO3. Na+ 1(a) II III Mont. (NaOH) (Solid NazCO3) (Solution Na2CO3) 5102 53.84 wt.% 45.4 wt.% 32.26 wt.% 35.64 wt.% A1203 18.86 16.2 10.71 13.28 Fe203 3.85 3.06 2.29 2.71 MgO 2.18 1.87 1.07 1.45 NaZO 2.13 0.00 0.02 0.01 Cr203 0.00 13.8 18.44 6.49 Chemical Formula: Na+-Montmorillonite Na Al 0.6[Al3.23Feo.42Mgo.47](517.88 0.13’020(OH)4 I (NaOH) 34(b) [cr(OH)2.66 11.88 [A13.14F°0.40M90.48](517.84A1 o.14’°20(OH’4 II (Solid Carbonate) .17 [cr(OH)2.83]3.53 [Al3.23Feo.42M90.47](517.88Al 0.13)°20(°H)4 III (Solution Carbonate) .48 - . [cr(°H)2.52]1.24 [Al3.23F°o.42M90.47](Sl7.88Alo.13)020(°H)4 (a) Taken from Brindley and Yamanaka, Amer. Miner., 1979, 64, 830. (b) THis number represents the average charge per Cr atom. 42 Figures 8 and 9, for catalyst II and III, respectively. Catalyst (I), synthesized by Brindley's procedure has a chromium content (1.88 Cr per unit cell) near that of catalyst (III) (1.24 Cr per cell) whose pillar was synthe— sized by hydrolysis with liquid carbonate. The Ad of (OOI) 7.3 A for product I is the smallest lattice expansion of any chromium pillared product. Catalyst II and III have similar Ad(OO£) spacings of m 16 to 17 A, but they differ significantly in their chromium content. Catalyst II and III exhibit strong differences in the X-ray intensities. Cata- lyst III (Figure 9) possesses much stronger and sharper OOR reflections, which could be related to a narrower pillar distribution produced by hydrolysis or selective inter- calation into the clays. It is worthwhile to note that all three catalysts (I, II, and III) generally do not exhibit more than two 002 reflections. This signifies that these catalysts possess good layer order, but that variability in layer spacing exists. As many as 6 orders of 00% re- flections are obtainable for some Al-PILC systems indica- tive of a much greater degree of layer ordering41 than is present in this system. Table 9 lists the surface area results obtained by utilizing the BET Theory of Multilayer Adsorption57. All samples were pretreated as described in Chapter II prior to measurement. The clay catalysts generated from high temperature cluster synthesis produce much larger surface area with solution carbonate hydrolysis yielding the highest value of 433 m2/g for air-dried products. 43 O FD (350°C) 11.78 A FD (25°C) AD (350°C) O 14.02 A 26.80 8 AD (25°C) II I I I I I I I I 20 18 16 14 12 10 8 6 4 2 Degrees 26 Figure 8. XRD Patterns for Air-Dried and Freeze-Dried Cr3 S-Montmorillonite Purged at 25°C and 350°C for Two Hours in Argon. 44 O 11.78 Al FD (350°C) FD (25°C) O 11.04 A AD (350°C) 26.80 A O 12.80 A AD (25°C) I I I I l l l l 18 16 14 12 10 8 6 4 2 Degrees 20 Figure 9. XRD Patterns for Air—Dried and Freeze-Dried Cr1 24-Montmorillonite Purged at 25°C and 350°C for Two Hours in Argon. {Do (150 45 Table 9. N2 BET Surface Areas for Air-Dried and Freeze- Dried Chromium Pillared Clays. SA (mz/g) Base (350°C) Catalyst Hydrolysis Cr/cell AD FD I (Brindley) NaOH 1.88 61 115 II (Tzou) Solid 3.53 353 368 III (Landau) Solution 1.24 433 421 Na2CO3 Hydrocarbon adsorption data for catalyst II and III are presented in Table 10. The molecules used have dif- ferent kinetic diameter558. Catalyst II has a apparent pore volume smaller than that of catalyst III as judged by adsorption data. Catalyst III adsorbs more hydrocarbon per gram of catalyst for all adsorbates employed. Even perfluorotributylamine with a 10.4 A kinetic diameter is readily adsorbed by both catalysts. As adsorbates increase in size, the freeze-dried clays show small but discernable increases in adsorption capacities. This effect is re— lated to differences in layer aggregation induced by air- and freeze-drying. Air-drying facilitates the maximiza- tion of long range basal plane-basal plane interactions of clay platelets due to surface tension forces acting upon the clay sheets. Freeze-drying inhibits layer reorgani- zation since the clay is frozen, and excess water is sub— limed off. This results in the generation of clays that are somewhat delaminated, a situation in which edge to basal plane and basal plane-basal plane interactions are .. i . I . x . 5h: .v» 1....) III) a .. .\.\v‘ .w ‘ ‘ .... .4 III flu ‘. Iv; . - II. . .-3 .. ., . .. x U L .. 4v 4 . -«.o p, . u. q . . \. ..., .a‘trt‘u a r-.lol| 1114111.. --x> in 46 .m\oo CH UmQHOmpm panoEm ucmmmummn mammsucmumm QH m0:am> Anv .Amm.ouom\mc Anon mm.o .6:A88 nasusnfiuuouosaugmm “Amo.ouom\mc Anon mm.o mamncmnuasnumfluuum.m.fl “Ash.o nom\mv Huou mm .mcmmcmm “m3oHHom mm 0H03 mmusmmmnm .Ooom Hm mwumnuomow cacmmuo mo pcmsmusmmma OD Hoflum muson N How omumsom>m oum3 mmHmEmm Ham Amv Amna.v Amma.v Amma.v omv. cam. wo.m Hmw pofluplonmwum «N.H HHH Avma.v AHvH.V AHNH.. mmv pofluonufia omv. omn. vo.m Avma.v Amma.v Amwa.v mmv. cam. om.H mom pmfluolmmmmum mm.m HH AmmH.v AnoH.v AQVANSH.. mmm pmfluonuflm mvv. omm. mm.H m v.oH m m.m m m.m Aooomm. Begum: Hfimo\uo oflmsmm ZMAmmwov mmmum m.m.H mamncwm m\NE maflmun poDHOmom m\HoEE Amv.m%mao UmHmHHHm ESNEOHSU Gamma Ionpmm mumconumu coausaom can Uflaom How MDMQ :ofi#mH0mpm sonumooupmm .oa magma 47 present. This delamination phenomena41 induces a degree of macroporosity which is available for adsorption above and beyond that attributed soley to the interlamellar region. The larger pore volume, as well as the lower chromium content of catalyst III implies that the pillaring oxo- cations are further apart than in catalyst II. This effect is related to the different products formed by solid and solution carbonate hydrolysis. The pillared product formed by solid carbonate hydrolysis is probably like Cr(OH)3 and the chromium must have a lower positive charge when more is present. The incorporation of colloidal dimers could also account for somewhat smaller pore volumes as well as the higher chromium content in II than in III. It is conceivable that Crl 24-mont (III) is a "cluster-like" cation, while Cr3.53-mont (II) is a "rod-like" or a two dimensional cation. X-Ray diffraction only tells us that at least one dimension, that which produced the 27 A spacing is similar. A preliminary ESR study of catalysts 1,211,111, and Cr3+-exchanged montmorillonite was undertaken. Instru- ment settings and sample preparation were described in Chapter 11. Figures 10 and 11 represent ESR spectra ob- tained on purged and reduced clay samples, respectively. Table 11 lists g values and peak-peak widths obtained for low (g1) and high (g2) field resonances. The low field resonances for calcined samples all have similar 91 values of 3.62 to 3.65. Cr3imontmorillonite exhibited the most 48 .___//\\h”’/r I Na+-Montmorillonite I Cr3+-Montmorillonite Crl 9-Montmorillonite Cr3.53-Montmcrillon1te -Montmorillonite H x 10 (Gauss) Figure 10. ESR Spectra of Purged Chromium Clay Catalysts. 49 Cr3+-Montmorillonite I_—i/I\A’T’flfifififflvfi——_AINKI\\\E:l 9-Montmorillonite _44\g L_Q | Cr3.53-Montmorillonite 7*— “7~—— T H0 x 103 (Gauss) Figure 11. ESR Spectra of Reduced Chromium Clay Catalysts. 50 Table 11. ESR g Values of Chromium Clays. Sample Cr3+ Cr -mont(1) 1.88 Cr -mont(II) 3.53 Cr -mont(III) 1.24 (A) Conditions g1(2100G) Purged 3.65 (383G) Reduced 3.57 (310) Purged 3.63 (251) Reduced 3.52 (226) Purged 3.63 (237) Reduced 3.54 (165) Purged 3.62 (237) Reduced 3.54 (167) (B) 92(4100G) 1.98 2.34 1.91 1.89 1.96 1.97 1.96 1.92 (599G) (275) (767) (964) (790) (1929) (766) (1679) (A) See text, Chapter II, ESR. (B) Resonance's peak—peak width. 51 intense signal at this low field. At higher field strength, the Cr3+ -clay exhibited the least intense resonance, with the narrowest peak separation. Crl 24-mont possessed the broadest peak separation along with the most intense signal. The g values are all between 1.91 and 1.98. When the samples were reduced as in a catalytic run, the low field resonance (g=3.52-3.57) exhibited a small de- crease in g value along with a reduction in peak to peak widths. The high field resonances (g=1.85-2.34) all de- creased in intensity and broadened considerably, except Cr3+-montmorillonite which increased in intensity. It should be noted that all reduced chromium pillared clayS' had a g value (1.85-1.92) upfield of DPPH, while Cr3+—ex- change clay had a 9 value (2.34) downfield of DPPH. Supported chromium has been characterized by three 59'60; Cr3+ions which are electron- 3+ distinct ESR resonances ions which are electroni- 5+ ) ically decoupled (6 phase), Cr cally coupled (8 phase), and oxidized (Cr4+ or Cr chromium (y phase). These resonances are characterized by a first derivative maximum of 1500 gauss, a broad resonance centered near 3400, and a very sharp resonance near 3400 gauss, respectively. The 0 phase resonance59 has been 3+ ions in a strong axial crystal field dis- assigned to Cr torted to lower symmetry. This phase (1500 gauss) was not observed in this investigation for any catalyst. This in— dicated that no electronically isolated chromium clumps were present, regardless of the hydrolysis method employed for cluster synthesis. 52 The absence of large crystal field distortions implies that the chromium polymers do not experience a large degree of internal strain, which would account for their excellent thermal stability. The y phase, as expected, was not ob- served for these clay catalysts. The 0 to 1600 gauss region was not shown in Figures 9 or 10. The resonance character- ized by g1 values in Table 11, centered near 2100 gauss, was also observed for Ngemontmorillonite. This resonance was caused by iron oxide impurities present in the clay. The resonances centered near 4100 gauss behave like 83 resonances characterized by Ellison6l. The g2 values for the chromium clays (g=1.89-1.97) are similar to those ob— tained by Ellison (g=l.97) if the Cr3+ clay (g=2.34) is not included. The peak to peak width of B3 (4100 gauss) varied widely (964 to 1929 gauss) for reduced clays, which was_also observed by Ellison. These EPR results show that cyclohexane dehydrogenation activity resides in the 83 phase. Hydrogen pretreatment 62,63 3+ produced a mixed-valence phase Cr2+/Cr cluster. This has also been prOposed by Groenveld, et al.64 and Wittgen, et al.62 as the active site for oxidative addition of an olefin to supported chromia on alumina. Catalytic Properties of Chromium Catalysts I! II, and III Chromium based catalysts hold a prominent place in catalytically promoted chemical reactions. Reforming of cracked and virgin naphthas is an important industrial pro- cess. This process generally involves several reactions which include the dehydrogenation of cyclohexanes, the 53 dehydroisomerization of alkylcyclo-pentanes, the isomeri- zation of butenes, and the hydrocracking of paraffins65. Chromium catalysts also accounted for 3.08 x 109 Kg of styrene produced in 197866 via ethylbenzene dehydrogenation. Two major types of catalyst can be utilized for the aforementioned reforming reactions: reduced metals and metal oxides. Other catalytic systems have been developed67 but have not achieved wide-spread industrial application. The reduced metals are of group VIIIB, typically nickel, platinum, or palladium, which are deposited on a support of silica or alumina. Metal oxides are typically chromia or molybdena. In catalytic reforming, the dehydrogenation of naph- thenes to aromatics is the chief octane upgrading reaction. In reduced metal catalysis, reaction conditions vary de- pending upon the catalyst. Supported Pt and Pd reactions take place typically around 250 to 300°C, whereas Ni re- quires temperatures of 350 to 400°C. Metal oxides operate in the 450 to 600°C temperature range66. The supported metals have a higher dehydrogenation activity than the metal oxide catalysts. To attain com- parable levels of activity for metal oxide promoted re- actions, one must operate the reaction at a higher temper- ature, a lower H2 partial pressure, and a lower reactant space velocity. Perhaps the most important commercial con- sequence of higher catalyst temperature is that it facili- 68 tates the rapid coking of these metal oxide systems and caused frequent regeneration periods. 54 The bifunctionality of transition metal oxide cata- lysts was illustrated by Clark69 who utilized several dif- ferent reactions. Table 12 depicts the hydrogenation and cracking of isobutylene dimers over a chromium oxide gel catalyst. Table 12. Effect of Alkali—Metal Promotion Upon Chromia Gel Cracking Activity. % Saturation Catalyst of C3 Product % Cracked Chromia gel 99.5 12-15 *NaOH + Chromia gel 98.5 1.9 Conditions: 390°C, 700 psig, 1.0 LHSV Hz/HC = 2. *100 ml chromia gel soaked in 300 ml 2.5% NaOH for 3 hours. Chromia gel has a hydrogenation and acidic functionality which do not occupy the same active site on the catalyst. This is inferred from the loss of cracking activity upon alkali treatment. The characterization of Cr—PILC catalyst activity utilized the dehydrogenation of cyclohexane: 0—70 I 3H2 and the cracking of B-isopropyl naphthalene: 03L —"* 00 *CH3‘°H=°”2 The reaction pathway of cyclohexane dehydrogenation over metal oxides such as chromia is schematically repre- sented in Scheme 870. 55 Scheme 8. D Benzene A H Cyclo-___.> -—-A Methylcyclo- —>Methylcy- hexaner—[CyCthexene] <——[ pentane ]\——clopentene H l A J D H D hydrocracked coke where: H,D - Reaction on a dehydrogenation-hydrogena- tion site. A - Reaction on an acid site. As Clark69 and Bridges, et al.71 have shown, alkali pro- moters quench the acid sites in the support and catalyst to inhibit the cyclohexene to methylcyclopentene transition and help reduce coke. Mechanistically the dehydrogenation of cyclohexane is lst order with respect to cyclohexane with the rate deter- mining step being the dissociative chemisorption of the hy- drocarbon upon the metal72, as shown in Scheme 9. Scheme 9. 02‘ CH" C H + OZZ—é Crm+ --—%> 02:——- Crm+ ______ C H n 2n l I n 2n-l 02' 02‘ The catalysts previously described were used as dehy- drogenation catalysts through the procedure outlined in Chapter II. The Weight Hourly Space Velocity (WHSV) reported for each run is reported with respect to the grams of reagent per gram of Cr203 present in the catalyst per hour. This reporting scheme was selected to enable the 56 catalysts to be compared more accurately since they con- tained various amounts of metal. Catalyst 1, Cr1.88-mont (I) produced by Brindley's method, showed negligible activity. Catalyst 11, Cr3.53- mont used as an air-dried catalyst, exhibited an initial conversion of sixty percent that decreased to thirty-four percent after two hours. Catalyst 111, Cr1.24-mont, main- tained high conversion levels after two hours, quantita- tively converting cyclohexane to benzene. These results are shown in Figure 12. Cyclohexane dehydrogenation is proposed to proceed via dissociative chemisorption of the hydrocarbon upon the metal. The +2 oxidation state, which appears to be the active oxi- dation state, is first generated by pretreatment of all catalysts in purified hydrogen. Hydrogen pretreatment is a necessary requirement for catalytic activity. This is carried out using hydrogen (P/PO = 1 atm) at a variety of contact times [volume catalyst bed (m1)/H2 flow rate (ml/ min)]. Figure 13 represents the relationship between initial conversion and post reaction surface areas as a function of hydrogen contact time. As the contact time of 3+ . is re- hydrogen increases, the ratio mmol HZ/mmol Cr duced. Catalyst III's initial reactivity is Optimized be- tween 0.90 and 1.6 sec. hydrogen contact time. At short contact times down to 0.90 sec, conversion is quantitative but the catalyst suffers a loss of surface area. The wt % carbon deposited upon the used catalyst also increases away from optimum conversion and maximum surface area % Conversion 57 90 ‘1 Cr1.24-Mont. (III) (3 WHSV) 80‘— 70 - 60'- I. 50 _ . Cr3.53-Mont.(11) (3 WHSV) 40'— 30 - 20.. 10 _ Crl 88-Mont.(1) (l WHSV) 0 NT Tn I I I T I 15 30 45 60 75 90 105 120 Minutes Figure 12. Optimized Cyclohexane Dehydrogenation Activity for Crl 88-Mont.(1), Cr3 53- Mont.(II) and Cr1 24-Mont.(III) Cata- lysts. 58 C. H12 Dehydrogen; Cru ' Mom- I I I I I I «500 c " 400 2 0 E 8 a 300 5,, 0 \ x a E a 200 ’E H -|00 27%C i L l l 1 L L 10 20 3D H.(&mmmtTMn:un. Figure 13. Initial Cyclohexane Reactivity and Surface Areas After Two Hours on Stream as a Function of Hydrogen Pretreatment. 59 retention. The loss of catalytic efficiency and surface area may be attributed to several causes. High contact times correspond to slow movement through the catalyst bed. The reduction of Cr3+ to Cr2+ liberates water as hydrogen is oxidized. This water could act to re-oxidize the cata- lyst. Also, steaming the pillaring cluster may cause structural rearrangements leading to loss of catalytic activity. At very short hydrogen contact times, reduction is so complete that structural modifications required upon change in oxidation state might produce a highly strained active site, and cause deactivation. If the active site is a Cr3+/Cr2+ composite, then over reduction to all Cr2+ would result in loss of activity. Figure 14 represents conversion over catalyst 111 after 15 and 75 minutes on stream for a variety of pretreat- ments. The general shape of the 75 minute conversion curve is similar to that of the initial activity curve. Figure 15 represents the variability of reactivity with contact time of reagent at a constant WHSV of 3 and a con- stant H2 pretreatment time of 1.6 sec. Short (2.7 sec) re- actant contact times produce a significant decrease in sur- face area. In addition, a higher wt. % carbon results than at a 6 sec contact time. Longer contact times (12.8 sec) produce smaller losses Of surface area, but greater carbon fouling. The intermediate contact time of 6 sec consistent- ly produced high conversions and low carbon results with- out any appreciable loss of surface area. Figure 16 illustrates the WHSV dependence (on a Cr203 60 C. H12 Dehydrog; CrL1 - Mont. I I T I I I '00- -( 80- a g t-Ehmn "-1 i3 60- - 0 > c 8 40- "75111611“ o\° 20- ‘ L l l l I L IO 20 30 H,Cawmn1mwms«; Figure 14. Cyclohexane Dehydrogenation Activity for Crl 24-Mont.(III) After Fifteen Minutes and Seventy-Five Minutes on Stream. % Conversion 61 5.9 sec 100‘- 2 402 m _ 2.75 sec 1.6% C/g 80 — ' .e* o 2.85 sec 6 306 m2/g ' 1.9% C C) 60'— a 347 mz/g 3.2% C 40... Temp: 550°C 3 WHSV 20‘— 0 I I I . I I I' 0 30 60 90 120 Figure 15. Cyclohexane Contact Time Dependence Upon Reactivity for Cr -Mont.(III) Catalyst. Minutes 1.24 62 fitCawufiaI 20- 40 80 |20 minutes Figure 16. Cyclohexane WHSV Dependence Upon Reactivity for Cr1 24-Mont.(III) Catalyst. 63 basis) for catalyst III at constant hydrogen and cyclo- hexane contact times. As expected, an increase in WHSV leads to greater carbon deposits, diminished surface areas, and reduced conversion levels. It should be kept in mind that all of the catalysts maintained a ~19.6 A spacing after reaction. Decreases in conversion at high WHSV values is probably due to the saturation of active sites by the re— actant, as well as by carbon deposition. The dependence of conversion upon pretreatment, WHSV, and reactant contact times for catalyst III are also ex- hibited by catalyst 11. The conversion dependence upon air- drying and freeze-drying was minimal. Freeze-dried cata- lysts exhibited similar dehydrogenation results, but they were harder to manipulate due to their powdery morphology. Differences among the three catalysts (Figure 12) may be explained in part by differences in pore structure. Upon completion of reaction at 550°C, catalyst I has a Ad of (002) 0 only 0.60 A, while catalysts II and III maintain large Ad (002) of 10 A. The pore openings of the latter catalysts is lar— ger than the kinetic diameter of cyclohexane (6.0 A) and allows for easy diffusion into the intracrystalline pores where reactivity is effected. This is illustrated by the diffusion characteristics of spent air-dried catalysts as shown in Figure 17. Freeze-dried catalysts exhibited simi- lar post reaction adsorption characteristics. Catalyst 1, which is collapsed after reaction, exhibits only external adsorption. The differences in the pore sizes of catalyst 11 and 111 are significant for these spent catalysts. 64 C, H.2 Uptake After Rxn. Cr“ (zoom L4 - I.2 - LO _ Cru(20.5M Figure 17. Cyclohexane Uptake by Spent Catalysts for Air—Dried Chromium Catalysts After Two Hours on Stream. 65 Catalyst III (Crl 24-mont) was subjected to an 8 hour C6H12 dehydrogenation run. This is shown in Figure 18. Initially the reactivity is constant for 3 hours, fol- lowed by diminished conversions with increasing time. Sur— face area is decreased by 70% accompanied by a large amount of carbon deposited. It appears that deactivation is due to pore blockage or active site coverage by physi-adsorbed carbon. Attempts to regenerate the catalysts by hydrogen treatment proved unsuccessful. Oxygen regeneration was also attempted, but the carbon layer could not be oxidized to CO2 at the temperatures employed (300°C) and collapse of the Cr interlayer clay to 10.2 A resulted. The acidic nature of these chromium catalysts and the dealkylation characteristics were probed using B-iSOpropyl- naphthalene as a substrate. This probe molecule was chosen for two reasons: (1) it is an analog to a-iSOprOpylbenzene (cumene) which is used extensively in the petroleum industry 73 characterization, and (II) it is much for Bronsted acidity larger than cumene and illustrates the cracking versatility of a large pore pillared clay versus a faujasitic zeolite. The results of two representative catalytic dealkylation runs are depicted in Figure 19. Dehydrogenation reaction conditions are most clearly approximated by the reaction at 2.2 WHSV. At a lower tem- perature of 400°C and a lower WHSV of 2.22 (vs. 550°C and a 3.0 WHSV for dehydrogenations) the catalyst dealkylates quite well. Deactivation is rapid, and attributed to the blockage of active sites as judged by the high wt. % C % Conversion 66 O 100 - 19.6 A 116 m2/g 7.45% C 80'- 60-— 40-— 20 -' Temp: 550°C 3 WHSV 6 sec CT ‘ 1.9 sec H2 0 ' l I I v 1 0 1 2 3 4 7 8 Hours Figure 18. Cyclohexane Dehydrogenation Activity of -Mont.(III) Catalyst with Time. cr1.24 67 '00 T I r I T T I T .- C735 - Mont. -: 400‘,6.0s CT 80 - - 8 "a 60 u- q 3 > S - . U 0 °\ 40 - ‘ WHsv- 2.2 20 ‘ n.4, u.o'/.c ‘ - fl 1 L TMm,mm. Figure 19. B-isopropylnaphthalene Dealkylation Activity for Cr3 53-Mont.(II) Catalyst as a Function of WHSV. 68 value of 11.45 %. If this reaction was run at identical dehydrogenation conditions, deactivation and carbon foul- ing would probably be more rapid and severe. Catalyst reaction at 1.00 WHSV illustrates the control- ling factor in carbon deposition. The relatively long 6 sec contact time produces the same large coke values (~11.00%) at both WHSV employed. This suggests that satur— ation of the surface occurs readily. Although Bronsted acidity-is present in the catalyst, no isomerization pro- ducts of C6H12 dehydrogenation were produced. It is con- ceivable that hydrogen pretreatment prior to C6H12 intro- duction quenches the acidity of the catalysts by reducing 3+ to a Cr3+/Cr2+ active site which only possessed the Cr very weak acidic character. Chromium catalysts II and III have been compared to commercially available Cr203 cur alumina catalyst obtained from Chemical Dynamics Corporation, USA. The commercial catalyst is approximately 19 wt. % Crzo3 with a surface area of 7.7 mz/g. Figure 20 compares the initial reactivity vs. hydrogen pretreatment conditions for catalyst III and the commercial catalyst, designated Cr203/A120 Two 3. features are worth noting: (l) the commercial catalyst re- quires almost three times as much hydrogen as catalyst III to achieve maximum activity, and (2) the commercial cata- lyst is not as effective as Cr1.24-mont (III) in dehydro- genating cyclohexane. Numerous hydrogen pretreatments over a wide range of conditions upon Cr203/A1203 were unsuccess- ful in increasing its overall conversion levels. Initial % Conversion 69 Temp: 550°C, 6 sec CT 100 - r1.24-Mont(III) (3 WHSV) 80 60 40 20 0 H2 Contact Time (Sec) Figure 20. Dependence of Cyclohexane Dehydro- genation Activity on the Contact Time of Hydrogen Reduction for Cr1 24- Mont.(III) and CrZOB/A1203. 70 ,The reactivity of Cr -mont (III) and chOB/Alzo 1.24 3 are shown for comparison in Figure 21. Analytical results on the spent catalysts and the dependence on reaction para- meters are given in Table 13. Catalyst Cr -mont (III) 1.24 is the most effective catalyst under the conditions em- ployed. This low Cr content clay has better order than catalyst Cr3.53-mont (II) as judged by XRD. This is pos- sible due to the existence of more bridging chromium atoms that have undergone oxolation reactions during cluster synthesis. The Cr3.53-mont (II) prepared by hydrolysis of Cr3+ with Na2C03(S) probably has a larger amount of chromium hydroxide like species incorporated into it. If indeed olation predominates over oxolation in solid carbonate synthesis, the greater amount of chromium in Cr3OS3-mont (II) would be distributed in pillars pro— 0 ducing Ad of 17 A, as well as large islands of mono- (OOR) mer and dimer sitting in the pores upon the basal plane or attached to the pillar. This would account for the ob- served reduction in surface area, the pore volume and the increased Cr content in catalyst II. The presence of these monomers and dimers would also facilitate coke formation by providing the acidic centers necessary for reaction. Characterization of Cr-Laponite and Cr-Fluorohectorite Inter- layered Clay The work presented in the last section for montmoril- lonite has been extended to two other smectite type minerals. Laponite-RD? which is a synthetic analog of hectorite, has a lath-shaped morphology and a manufacturer specified particle 71 550° Ce HI: W C. ”0+ 3”: r f u r I I I r ICC (" “ crLl / ” (WHSV- 3.0) ‘ 80 r- “ '- ' chOS/Alzos q (WHSV-LO) 60- 5r../ - 40t % Conversion (WHSV- 3.0) 20 )- " :- Gr" '1 (wnsv- (oh, L__¢ A A: Jeufiti A A .fi__4 0 3O 60 90 IZO MUM” Figure 21. Cyclohexane Dehydrogenation Activity for Crl 88-Mont.(I), Cr3 53-Mont.(II), Cr1 24- Mont.(III) and Cr203/A1203. 72 .mmsHm> :oauommulmum pcmmmummu mononucwumm :fl mquESZ ADV o AEUM H " om\mv usmEummuamum mm ESEHumo H50: H “mason N "Emmuuw so oEHu .oEHu yomucoo ocmxmnoHomo osooom xwm “Uoomm u damp QOHuommm “cofluommu mHOan pofluoluwm onB mumwamumo Ham Am. Ae.ec a.m a- AH.o o.H o.mH memaa\mo~no Ammec mme Am.mmc mo.om em.H o.m me.m AHHchcoe-e~ Huo Ammmv oefi Ao.mmc mm.om mH.m o.m mw.hH .Hchcoaumm muo lam. N.mm inexa.mfic N.oH .. o.H m.mH AHVucosnmm Huo m\ms a aonumo >mm3 no pmmamumo mm mooo w u: w DB Amvmumwamvmo ESHEOHSU ucmmm mo coaumNflanomumnu HMOHmmsm paw mumuwfimumm Uflumamumo .MH manna 73 size of a few hundred angstroms. It also possesses a low layer charge. Fluorohectorite possesses a very high layer charge and is characterized by a particle size much larger than Laponite-RDDor even montmorillonite. These clay char- acteristics are presented in Table 14, along with those of the montmorillonite used for generation of clay catalysts I, II, and III. Table 14. Characteristics of Natural and Synthetic Smec- tites Used for Chromium Interlaying. Layer Charge Particle Clay per 020(OH)4 unit Size Laponite-RD® 0.36 ~100 A Wyoming-Mont 0.86 ~2000 Fluorohectorite 1.6 >>2000 Synthesis of the pillaring species was carried out un- der identical conditions to those described in the previous work and utilized the solution carbonate synthesis method. The concentrations of reactants utilized for cluster growth are summarized in Table 15. The ratio of mmol Cr3+/MEQ clay was maintained at a value of 62.5 for all clay minerals regardless of their cation-exchange-capacity by adjusting the volume of pillaring solution used for each clay on a gram basis. Fluorohectorite was washed free of excess electrolyte by repeated suspension in deionized water, fol- lowed by centrifugation. The final product was air-dried. Laponite, because of its small particle size, could not be centrifuged without an appreciable loss of product. The 74 .>Ho>fluoommmu .m mm.o can a hH.o mnt Immoo paw +mHU mo soapmuucmocoo HmHuHcH I m.mm moHH.o moHH.o Ame mnHHOpoosouosHm m.mm ememo.o emo.o mm @omumuflcommq smao om: Aumnea\mmaoev Aumnfla\mmaos. Amooa\omzc swam +mno Hoes om_ummooiam2uom omfl+mmolam2uom omo .SDBOHQ HopmSHU EDHEOHSU Mom UwNHHfluD mucmuomom mo m:OHumupcmocou .mH manna 75 small particle leads to low sedimentation coefficients and prevents efficient collection. Dialysis was used to wash laponite, after which it was air-dried and freeze- dried. X-Ray diffractograms for Cr intercalated laponite and fluorohectorite are shown in Figures 22 and 23, respectively. Physical prOperties of the materials are presented in Table 16, along with those of montmorillonite intercalated clay prepared under analogous conditions. Chromium laponite shows atypical behavior as compared with other chromium interlayered clays. Dialysis of Cr- laponite prevented the removal of large hydroxy-metal poly- mers that were unable to diffuse through the dialysis tubing. The wash water was almost colorless during the entire wash procedure. This indicated that very little Cr3+ was washed from the clay during dialysis. This is confirmed by the high chromium content (40 wt. %). The d(00£) spacing of 21 to 22 A is lower than that generally obtained with montmorillonite (~28 A). This is attributed to hydrolysis of the pillaring agent. As hydrolysis pro- gresses, surface area and d spacing decreases. A1- (002) though the adsorption results are comparable to freeze- dried Al pillared montmorillonite (~0.3-0.4 mmol/g perfluoro- tributylamine adsorbed), Al pillared laponite is capable of almost twice the adsorption of Cr exchanged laponite. The XRD results from Cr-laponite are interesting. Low temperature treatments (25°C) produce discernable, but broad and poorly characterized d reflections. At elevated (002) '76 FD (350°C) m 22.64 A FD (25°C) AD (350°C) 0 22.07 A AD (25°C) I I ‘ I l l 7 T l 16 14 12 10 8 6 4 2 Degrees 20 Figure 22. XRD Patterns for Air-Dried and Freeze-Dried Chromium Laponite Purged at 25°C and 350°C for Two Hours in Argon. 77 15.78 3 350°C 110°C 10.30 i 1* 25°C . I I I I I I I I I 20 18 16 14 12 1o 8 6 4 2 Figure 23. XRD Patterns for Air-Dried Chromium Fluorohec- torite Purged at 25°C, 110°C, and 350°C for Two Hours in Argon. ‘-.-. ‘1 . 78 . mAHOHflOGHMQH Aaoov 6 sum pannnxm no: 6H6 Hmeumums mesa Am. eo.m mmv me.m m.mm m.om ma muflcoHHeuosuco: eem.o ewe mm.sH H.6H ~.mH om mufluonomaoeosam I- mom mm.o¢ I- m.mm om muflcommq I- com mm ow Amen- m Hm om m\aoaa .mca m\ms Ho ooomm comm 60:00: swao 0:0N20m am A .03 a «006 mcflmno 0 .mUHcoHHHHOEusOE can .mpfluouomsouooam .muflcommq communaumusH EDHEOHQU mo mmfiuummonm HMOHmhnm .mH GHQMB 79 temperatures, the clay intercalate appears to be delaminated. The amorphous x-ray pattern of the elevated temperature samples may be related to the high Cr content as well. Chromia oxide gel, formed as a condensation polymer of chromium hydroxide, is also x-ray amorphous. It is possible that this new chromium polymer, generated by high tempera- ture base hydrolysis, is also x-ray amorphous when not stabilized by the clay's silicate sheets. Cyclohexane dehydrogenation was attempted with air dried Cr-laponite, but the experiment proved to be unsuc- cessful: no benzene was produced. This might be the re- sult of dialysis facilitated hydrolysis of the pillaring agent. It is very reasonable to speculate that the pillaring agent present after dialysis is different from what would be obtained by removal of excess electrolyte. Hydrolysis rates are markedly different as the ionic strength of the electrolyte changes within the clay interlayer. Differ- ences in dialysis and centrifuge washing may account for the anomolous XRD patterns as well as diminished catalytic activity. Several interesting features are also exhibited by O Cr-fluorohectorite. The d of 19.2 A is considerably (002) 0 lower than the 27.6 A observed for montmorillonite. The chromium content is similar to solid carbonate-hydrolyzed Cr-montmorillonite, although it was produced by solution carbonate hydrolysis. The surface areas and adsorption capacities are also very low in comparison. The high 80 chromium content for chromium fluorohectorite is related to the high layer change associated with this clay. A larger number of cations are required to neutralize the higher negative layer charge on fluorohectorite than is re- 0 quired by montmorillonite. The small d (~8 A) and (00%) low adsorption capacities and surface areas indicate that the clay picks up an amount of oxocations, which are re- quired for electrical neutrality, then they are degraded by interlayer forces to yield the observed d spacing (00%) of 19.2 A. Analogous intercalation procedures using mont- morillonite, which possesses a lower layer charge, pro- duced the previously described Cr interlayered montmoril- lonite. Indeed, this cluster hydrolysis mechanism is Opera- tive for many metal hydroxide intercalates. Catalytic experiments were not attempted due to the low adsorption capacities obtained. Conclusions The preceding results have shown that a new class of porous chromium clay intercalates has been generated. This is accomplished by high temperature aging of chromium nitrate and sodium carbonate, followed by intercalation into natural and synthetic smectite minerals. The method of cluster synthesis plays an important role in determining the physical properties of the intercalated clay. Solution carbonate hydrolysis affords a material with lower Cr con- tent, larger pores, and enhanced catalytic activity as compared to that produced by solid carbonate hydrolysis. These new clay catalysts have been characterized as 81 traditional metal oxide catalysts which possessed dual catalytic activity affecting both hydrogen transfer and Bronsted acid type reactions. Dehydrogenation activity was shown to be dependent upon chromium content and its re- duction to an active site which contains a mixture of Cr3+ and Cr2+, the latter produced by a dry reductor. De- hydrogenation upon these clays also illustrated the first use of an intercalated metal-oxide pillar for carrying out catalytic conversions in a pillared clay. These catalysts' Bronsted acidity, which was characterized by B-iSOprOpyl- naphthalene dealkylation, is very good, but is not a fac- tor in catalytic reactions when reduced catalysts are used. The results obtained on the synthetic smectites, lapo— nite and fluorohectorite, suggested that the hydrated oxo- cations fill the interlayer and hydrolyze to achieve elec- trical neutrality. This accounts for low surface areas and low adsorption capacities as well as smaller than expected d(00£) spac1ngs. CHAPTER IV ALUMINUM INTERLAYERED CLAY CATALYSTS Introduction Zeolites have been used extensively as catalysts and selective adsorbents since the mid-19605. Their primary uses include catalytic cracking, hydroisomerization, and reforming reactions74. Although a large number of zeolites have been synthesized, only a limited number have been ex- tensively utilized. Faujasite zeolites have the largest pore openings corresponding to approximately 8 A. These pore openings are too small to allow the conversion of heavy oils, present in crude oil, to gasoline due to reactant selectivity. Consequently, it is of interest to synthesize ordered porous materials with pore openings larger than 8 A. As noted in the introduction, Barrer and Macleod26, and Barrer and Reay27 were the first to demonstrate that the charge balancing cations of montmorillonite could be re- placed with tetraalkylammonium cations. These cations served as molecular props between the clay sheets, imparting a 4 A free interlayer spacing. In an attempt to further in- crease the interlayer spacing, several types of tris-metal 82 83 chelates, including Cu- and Fe-tris 1,10—phenanthroline29, tris-bipyridyl30 and tris-ethylenediamine cations31 have been employed. These cations generally produce a 9 A Ad(00£) interlayer spacing, but exhibit low thermal stabil- ity (<400°C) due to carbon-carbon bond cleavage in the in- tercalant. Several inorganic intercalants have been synthe- sized36a-e. Besides those presented in Chapter III (of. Table 6), aluminum oxocations have received considerable attention due to their large pore openings and excellent thermal stability. This clay intercalant has been charac- terized through the microactivity test for cracking cata- lysts by using gas~oil42. The results demonstrate that the catalytic_activity is comparable to zeolite-promoted com- mercial catalysts containing 15 to 20% by weight faujasite type zeolite. Clay catalysts used in these experiments showed a progressive decrease in Ad spacing between (002) 540°C and 650°C with a corresponding decrease in surface area as well as catalytic activity. At present it appears that clay intercalates suffer from instability to the steam that is typically employed as part of the stripping operation of commercial catalytic cracking procedures75 At present the structure and morphology of aluminum interlayered clay is not fully elucidated. It has been postulated that the polyoxocations undergo dehydroxylation at elevated temperature to form small oxide aggregates which are stable on the interlayer surface76. Crosslinking by the pillars to adjacent silicate sheets is another distinct 84 possibility36b. It has been recognized that by controlling pillar size and distribution it is possible to form intra- crystalline materials with pore sizes larger than those of faujasitic zeolites. It is also known that smectite layers may aggregate in an ordered lamellar fashion (basal plane- basal plane interaction) or that they may adopt a delamin— ated or "house-of—cards" structure due to edge-basal plane interactions dependent upon mode of synthesis41 The objectives of this present study were to determine the relationship between the OH/Al ratio of the pillaring solution, as well as clay morphology, cation exchange capa- city, and history of the clay intercalate reaction with respect to the catalysts' physico—chemical and catalytic properties. Preparation of Pillaring Reagents Three different pillaring reagents were employed in this study. Solution #1 was synthesized by the base-hydrol- ysis of AlCl3-6H20 with NaOH to yield an OH/Al mole ratio of 2.00. Solution #2 was prepared in an analogous fashion to Solution #1, except the OH/Al mole ratio was 2.42. Solu- tion #3 was the active ingredient of anti-perspirant, 50% w/w chlorhydrol solution obtained from Reheis Chemical Com- pany. The chemical formula as reported by Reheis is A12(OH)5C1, its chemical composition is listed in Table 17. Chlorhydrol is synthesized by the reaction of aluminum tri- chloride with excess aluminum metal77 in an acidic environ- ment as shown by the following equation: 5A1 + AlCl3 + 15H20——-)3A12(0H)5+ + 7.5112 + 3c1" 85 Table 17. Chemical Composition of 50% w/w Chlorhydrol Solution. A1203 23.7% Cl 8.21% Al : Cl (mole/mole) 2.01 : 1 S04 <0.025% Pb <10 ppm Fe 42 ppm pH 4.20 Specific Gravity 1.337 The aluminum chlorhydrol (ACH) used in the present work was diluted from a concentration of 6.2 M (as supplied by the manufacturer) to a concentration of 0.23 g and then utilized in the same fashion as the other pillaring reagents. The chemical species actually present in solution were probed by 27Al-NMR by Dr. Ming-Shin Tzou45. The 27Al-NMR spectra are shown in Figure 24. It is apparent from the NMR results that the chemical compositions of these three solu- tions differ significantly. Solutions #1 and #2, those which are obtained from base-hydrolyzed AlCl3-6H20, are dis- tinguished by a sharp resonance at 62.8 ppm, along with one at 0.0 ppm. Solution #2 has a larger contribution of alumi- num giving rise to the 62.8 ppm resonance than does Solu- tion #1. Chlorhydrol (ACH) exhibits resonances at 62.8 ppm, as well as a broad resonance at 10.8 ppm and a sharper one at 0.0 ppm. 27Al-NMR spectra similar to those observed here have 86 A1C13 + NaOH (r = 2.00) \ W 9 r‘ A1013 + NaOH (r = 2.42) // 46 L“. AIIOHI; ACH OCIohedrol AI Monomers (1' = 2'50) Oclohedrol Al Palymen Figure 24. 27Al-NMR of Pillaring Solutions Emg%oyed for Interlayering of Smectite Minerals . 87 been reported by Akitt and Farthing77-79 and by Bottero, et al.37 for related base and metal hydrolyzed aluminum solutions. The sharp resonance at 0.0 ppm, based upon data presented by these workers, is attributed to monomeric A1(H20)63+ species. The sharp line at 62.8 ppm is assigned to an Al13 oligomer which possess Keggin ion-like structure38 as shown in Figure 6 (Chapter 1, p. 22). In this structure one aluminum occupies a central AlO4 tetrahedral position while the remaining 12 Al occupy equivalent octahedral posi- tions defined by oxygen, hydroxide, and water ligands. The symmetric AlO4 unit gives rise to the observed 27Al resonance at 62.8 ppm. The octahedral Al ions probably experience a strong electric field gradient due to quadrupolar relaxation upon the I = 5/2 nucleus and are NMR silent. These assign- ments are strengthened by the 27A1 resonance of the tetra- hedral Al in AlW120405- ion which occurs at low field (72 ppm). The octahedral A1 of AlMo60213- occurs at 0.0 ppm. The nature of the species giving rise to the broad resonance at 10.8 ppm in Solution #3 is unknown. It has been sug- gested by Akitt and Farthing79 that this resonance is due to aggregates of higher nuclearity than A113, although they were unable to offer any conclusive proof. Although com- plete identification of the species present in solution is unobtainable, it is clear that the three pillaring solutions differ significantly for OH/Al ratios in the range of 2.0 to 2.50. These three different pillaring solutions were utilized to intercalate a variety of smectite clay minerals which may 88 be classified into three distinct subgroups as outlined in Table 18. Physical Properties of Products Obtained from Laponite and Smectites of Related Layer Charge and Morphology Laponite-RD® was allowed to react with various amounts of aluminum chlorhydrate (ACH) in the range of 0.00 to 10.00 g of ACH per g of laponite. The synthetic procedure de- scribed in Chapter II was utilized. The ratio mmol Al3+ / MeQ clay shall be designated as Q. These final products were air—dried on glass plates, as well as freeze—dried. Figures 25 and 26 represent the x-ray diffraction patterns for laponite interlayered with ACH that was freeze-dried and heated at 25°C and 350°C, respectively, for 2 hours in argon. It should be noted that freeze-drying at low Q (0.279) yields a discernable 001 reflection. As the value of Q in— creases from 0.875 to 84.20, the intercalated minerals all give either no 001 x-ray reflections or very diffuse x-ray reflections. Heating these freeze-dried laponites to 350°C for 2 hours in inert gas produces little variation in the samples ranging from Q=0.875 to 84.20. The Q=0.279 sample shows a decrease and broadening of its 001 reflection. The freeze-dried samples are very different from the air-dried samples shown in Figures 27 and 28. Air-drying leads to materials with distinct 001 reflections in the range Q=0.00 to 10.95. Heating these samples to 350°C, as shown in Figure 28 produces broadened 001 reflections, but they re- mained more pronounced than their freeze-dried counterparts. Table 19 lists the physical prOperties for the series 89 elmocomol emomolmflm.lo.Hflqe.emz_e.HHq Ami mH.o mm.eam._mm.omzfie.ommmm.maaimm.omz Ha Apv exmocomoxmv.oaafim.eflmcHmH.omm~m.oaamo.mmzLmom.omz Ans vxmoc omOAoo.mHmvHom.oquw.mm2_wm.OHq Amv “ma Hohoa 050 m0 :oflpflmomEoo Haoo pass 038 AHV m ooomAA om.H 4 ooomz om.o m oomz oo.H m ooomz om.o m come mm.o opfluouooSOHOSHm Amy 0 duo AMOS so Apv u: .fiH. u z onouflcoHHHHOEucoz 0 asomm Anv u. m AmvoomImpflsommq omflm macapumm omHmH cam omumno HohmH swan muflm maoauumm pom omnmao mommH Havamhu oNHm oHOADHmm.HHmEm Ho\©sm omumno Hmme 30H oNHm 0H0Huuwm woumno momma AflcsmHo >00H0£QH02 pom omumso Momma HMOflmoaosmHoz pom afloauuooam momma on mcfipuooom .mofluuomonm mfimumcflz mufluomEm .mH manna MD .279 10.95 21.89 43.78 84.20 N 14 12 10 8 6 4 Degrees 20 Figure 25. X-Ray Diffraction Patterns for Freeze-Dried Laponite Interlayered with Aluminum Chlorhy- drate, Purged for Two Hours at 25°C in Argon. 91 27.59 A 13.38 P .279 .845 10.95 21.89 43.78 84.20 I *I I I I I 141210 8 6 4 2 Degrees 20 Figure 26. X-Ray Diffraction Patterns for Freeze—Dried Laponites Interlayered with Aluminum Chlorhy- drate,Purged for Two Hours at 350°C in Argon. 92 O 13.4 A O 15.8 A 2 ° 0 5.2 AL 19.6 A m (Na+-Laponite) O 21.5 A gh— N I I I I 20 18 16 14 12 10 8 6 Degrees 20 Figure 27. X-Ray Diffraction Patterns for Air-Dried Laponite Interlayered with Aluminum Chlorhy- drate,‘Purged for Two Hours at 25°C in Argon. 93 0.00 (Na+-Laponite) 10.95 I T7 I I I I I I I 20 18 16 14 12 10 8 6 - 4 2 Degrees 20 Figure 28. X-Ray Diffraction Patterns for Air-Dried Laponites Interlayered with Aluminum Chlorhy- drate, Purged for Two Hours at 350°C in Argon. 94 .elmocomoloo.mflmc16m.oeq66.m621 ae mumsma may to cofiuemOQEOU flame pee: was .6. mwo.o I I I U0HHDI0N0MWM mmH.o «mm H.vH mm.m o~.vm mmv.o Hmm ov.m om.m mh.m¢ cmm.o mmm oo.m oo.m mm.HN Nmm.o mmw mm.m om.H mm.oH mom.o mmv mo.m oom.o mh.o bmh.o Hem mv.H 5mm.o mn.m mmm.o mum Hq.o mooH.o mvm.o Nv¢.o mmm mvm.o wmmo.o mh~.o omm.o mom . I I Mmemmmwwmmm Emma Mom HOSE m\NE Amvaaou “Ha: mfimonuchm CH ouflcommq 0m: .pmnnompm amBmm comm 006musm Mom Hm poms mom madam Hm HOSE +m .mpoupmsuoHno Edsflesam QDH3 pouommHHoucH mopflcomoq pofluolouooum mo moflguomoum HMOfimwnm .mH manna 95 Q = 0.00 to 84.20 for freeze-dried laponites. As the value of Q increased, the aluminum content per unit cell increased as expected. The N2 BET surface areas and perfluorotributylamine (PFTBA) adsorption capacities both traverse a wide range of values. The PFTBA adsorp- tions and N2 surface areas are shown in Figure 29 as a function of the aluminum content per unit cell. Both adsorbates exhibit a maximum at approximately 3 Al per unit cell, then a decrease to values similar to sodium laponite (Q=0.00). Table 20 presents physical prOperties of some air-dried Table 20. Physical Properties of Air-Dried Laponites Inter- layered with Aluminum Chlorhydrate. Q 3+ # Al(a) PFTBA mmol Al per unit Surface Area Adsorbed MEQ Clay cell (mZ/g) (mmol/g) 0.00 -— 201 .210 2.75 1.43 425 .794 6.75 3.03 451 .825 10.95 3.62 363 .833 (a) The unit cell composition is [Mg5 64LiO 36]Si8010(OH)4. laponites. Their surface areas and PFTBA adsorption capaci- ties are consistent with those obtained for freeze-dried materials. The dependence of the apparent pore size of interca- lated chromium clays upon drying conditions was explained in terms of a delaminated model in Chapter III, p. 45. This same model, used to explain the differences in adsorption of 96 IO )- I I I r I I I a O ' o \ 0.8 P . PFTBA Adsorption, 25 C d .< m I— 0.6b .I t 0.4 1 '8 E 0.2 .. E 0.0 l 1 l l 1 1 L 0 2 4 6 8 IO l2 I4 r o 500 P- I I fi I I I q 3‘ N2 BET Surface Area E 6 fi .1 L. '4 o u o w- -I h a (D +- L L l I L 1 d 1 0 2 4 6 8 I0 I2 I4 Apparent-Al Per Unit Cell Physical Properties of Freeze-Dried Laponite Interlayered with Aluminum Chlorhydrate as a Function of Aluminum Ions Per Unit Cell. Figure 29. 97 hydrocarbons on air—dried and freeze-dried chromium pillared clays, may now be invoked to explain the observed similari- ties and differences in air—dried and freeze-dried laponite interlayered clays. The layer flocculation model previously presented41 differentiates three distinct modes of layer aggregation: face-to-face, edge-to-edge, and edge-to-face association of platelets. These arrangements are depicted schematically in Figure 30. The central panel represents a flocculated clay product. The stippled regions represent face-to-face pillared aggregates, whereas the unstippled regions repre- sent delaminated aggregates produced by edge-to-face asso- ciations of platelets. If the flocculated clay is freeze-dried, the mixed laminated/delaminated structure tends to be preserved as the solvating water is sublimed. As will be shown later for montmorillonite, this produces a "broad pore" type pil- lared clay product. If the flocculated clay is allowed to air-dry at ambient room conditions, then the surface ten- sion forces between clay platelets tend to aid in reorgani- zing the clay platelets as face-to-face aggregates. In this manner the surface tension forces aid in increasing or optimizing the face-to-face aggregation which produces a "narrow pore" type pillared clay. The behavior of laponite, a small particle, low layer charged synthetic smectite, may be reconciled in terms of the layer flocculation model presented above. Laponite has a lath-shaped morphology and a layer Figure 30. 98 m; g /___ / I/\ I /\/_\\ A L/ S I...’ Typical smectite clay smpemion with a particle size < 2n. Pillared struc I Delaminated structure Freeze-Dry Structure of the flocculated clay is largely preserved by freeze drying. Face-to—face layer stack- ilg occms over > 4 layers. Typical- ly two orders of 00,1 r-ray reflee- tion are observed. The delaminated fraction is responsible for aaorption of large molecules. Bulk density is low. Polyoxycation solution Combinirg reams in a flocullated clay with most platelets (7Ho%) involved in W tace—to—(ace aggregation. 'meremainderofclayhuadelaminated home-of-carch structure reuniting from eQe-to-etke and eQe-to-Iace a-ociation. Amen Surface temion forces of liquid water during airdryitg enhances the extent of tace-to-face layer ordering. Order- ing occurs over tens of layers and eacntiafly all layers are ordered (pil- lared). Pore structure is regular. Multiple orders of Mir-ray reflee tions are observed. Bulk density is high. A Layer Aggregation Model for Air-Dried and Freeze-Dried Pillared Clays with a Particle Size <2u. 99 I diameter of 10 to 100 A as specified by the manufacturer. The aspect ratio of laponite (ratio of platelet width to height) is small and does not vary appreciably. Clay layers with a large aspect ratio should tend to aggregate in a face-to-face manner, while a small aspect ratio clay, such as laponite, should facilitate the formation of edge—face and edge-edge interactions80. This hypothesis is supported by the XRD pattern of air—dried Na+-laponite purged at 350°C (cf. Figure 28). The peak present at ~4 degrees 20 represents the pillaring of sodium laponite through edge- face or edge-edge interactions as expected on the basis of its small particle size and low aspect ratio. Laponite possesses a low layer charge which also plays an important role in layer ordering. As an intercalated clay product air-dries, it progresses through a wide range of states; from sol state, to gel state, to a fully dried state. Vari- ations in charge distributions accompany these variations in states as described below. Initially the flocculated clay product is characterized by a watery dispersed state. This initial state of the polyoxoaluminum-clay complex is analo- gous to that of typical alkali metal smectites. The plate- lets carry a surface negative charge as well as a small positive charge on the edges due to disruption of the lat- tice. In dilute solutions the basal surface charges are much larger than the small edge charges and repulsion occurs. This yields a colloidal dispersion which looks and flows like water. This state is referred to as a sol. As the ionic strength of the solution increases (due to solution 100 evaporation) the surface negative charge is reduced due to association of the negative clay platelets and the cations in solution. The repulsion between platelets is reduced and the dominating force becomes the face-to-edge interactions, causing the formation of a gel. This is referred to as a "house-of—cards" structure. As laponite air-dries, clay platelet reorganization re- sulting from surface tension forces are Operative while the solution is still in the sol state. This helps to increase face-to-face interactions over face-to-edge interactions. Laponite undergoes the transformation to a gel state much faster than other large particle or high layer charge smec- tites. This rapid gel formation impedes the quantitative formation of face-to-face aggregates by essentially "lock- ing" the clay platelets into whatever random association they exhibit at the onset of the gel state. This "locking" phenomenom is schematically represented in Figure 31. A clay following this drying track would eventually exhibit 00% reflections which are very broad and poorly character- ized, representing a large array of pillaring interactions (cf. Figure 28). Laponite was also pillared with a freshly prepared base- hydrolyzed aluminum chloride solution at r = 2.00. These materials were processed in a similar manner to that of the ACH pillared materials. All samples at r = 2.00 were freeze- dried. As the following results will show, differences in pillaring solution lead to clay products with different A1 contents, but similar physical properties. 101 / \ . ' , ' //;_/ I a . . . 7 I \ I . ‘ e \ / ’\ / I 0 ' o Smectite clay suspen- Polyoxycation solution. sion with particles < 500 A. In the resultim flocculated clay eQe—to-{ace and ewe-Me layer usociation competes favorably with tace—to—face layer association. in. Freeze-dryingpraervesthestructm-e Surfacetensionforceseneotnteredinair of the flocculated clay. Pace-to-{aee drying enhance the extent of face-to-{ace layer association or pillaring is short mation at the em of delamination. rage (< 4 layers). 00,! x—ray reflec- but the structure is largely dehminated. tion is very diffuse to non-existent. Structure is extensively delaminated. Delamination reams in formation of macro- and mesa-pores. Figure 31. A Layer Aggregation Model for Air- Dried and Freeze- -Dried Pillared Clays with a Particle Size <500 A. 102 Figures 32 and 33 depict XRD patterns for r = 2.00 freeze-dried laponite products after being heated at 25°C and 350°C, respectiVely, for 2 hours in argon. These pat- terns are similar to those obtained for ACH-laponite. Dis- tinct 00R reflections are present only in Q = 0.283; the remaining series members exhibit very broad, weak, 00% re- flections. At 350°C, Q = 0.283 exhibits the "self—pillaring" 002 reflection that is also seen in Na+-laponite, as well as ACH-laponite, indicative of edge-plane and edge-edge inter— actions. Table 21 presents analytical data, N2 BET surface areas, and PFTBA adsorption capacities for r = 2.00 aluminum inter- layered laponite products. The surface areas and PFTBA up- takes are consistent with those values obtained for ACH- laponite. The aluminum content per unit cell is considera- bly lower with r = 2.00 than with r = 2.50 (ACH). This is due to the different aluminum species present in each solu— tion. At r = 2.00, tetrahedral aluminum centers (62.8 ppm resonance) as well as octahedral aluminum monomer centers (0.0 ppm resonance) are both present, with the slight excess of the aluminum present being tetrahedral in nature. In ACH (r = 2.5), there is a small tetrahedral aluminum contri- bution (62.8 ppm), a larger octahedral aluminum monomer con- tribution (0.0 ppm), and a detectable amount of octahedral aluminum polymer contribution (10.8 ppm) not present in the r = 2.00 solution. It is conceivable that the high molecu- lar weight aluminum polymer (10.8 ppm) is unable to pene- trate the dialysis tubing in which the freshly pillared 103 14.97 23’. 2 0.283 6.75 21.89 84.24 Figure 32. X-Ray Diffraction Patterns for Freeze— Dried Laponite Interlayered with Base- Hydrolyzed AlCl3 at r = 2.00 and Purged for Two Hours at 25°C Under Argon. v ...: ox ...- .1} H N H O (I) N 104 O O 12.62 A 29.43 A1 2 0.283 6.75 21.89 84.24 I I l I I 14 12 10 8 6 4 2 Degrees 20 Figure 33. X-Ray Diffraction Patterns for Freeze-Dried Laponite Interlayered with Base-Hydrolyzed AlCl3 at r = 2.00 and Purged for Two Hours at 350°C in Argon. 105 Table 21. Physical Properties of Freeze-Dried Laponites Interlayered with Base-Hydrolyzed AlCl3 at r=2.00. 0 3+ # Al‘a) PFTBA mmol Al Initial per unit Surface Area Adsorbed MEQ Clay Al Conc.. cell (m2/g) (mmol/g) 0.283 0.0494 g 0.17 245 .400 6.75 0.0655 2.18 404 .875 21.89 0.0784 2.72 278 .883 84.20 0.0784 3.16 334 .850 (a) Unit cell composition = [Mg5 64Li0 36]Si8020(0H)4. laponite was washed. This would account for the high alumi- nun content per unit cell for r = 2.50 (ACH), but not for r = 2.00. In an attempt to better define which factor, particle size or layer charge, is the dominant one operating in the delamination of laponite, two different smectites have been used. A natural saponite, which has a low layer charge (0.30) close to that of laponite, but normal particle size, and high layer charge 0.03u particle size montmorillonite (gift of Dr. R.H. Raythatha) were pillared by ACH, dialyzed, and freeze-dried. Figures 34 and 35 show XRD patterns for the synthetic saponite and small particle montmorillonite, respectively. Table 22 lists analytical data, as well as N2 BET surface areas and hydrocarbon adsorption capacities for the two pillared clays. As shown by XRD data for each pillared clay mineral, both possess discrete d reflections. The (009.) surface areas obtained are consistent with pillared mont- morillonite, much lower than the surface area of a 106 O e!18.79 A III I I T’ I I 14 12 10 8 6 4 2 Degrees 20 Figure 34. X-Ray Diffraction Patterns for Freeze-Dried Saponite Interlayered with ACH and Purged at_25°C and 350°C for Two Hours in Argon. 107 O 12.62 Aw 25°C I I I I I I 14 12 10 8 6 4 2 Degrees 20 Figure 35. X—Ray Diffraction Patterns for Freeze-Dried 0.03u-Montmorillonite Interlayered with ACH and purged at 25°C and 350°C for Two Hours in Argon. 108 Table 22. Physical Properties of (A) Saponite, and (B) 0.03u-Montmorillonite Interlayered with Aluminum Chlorhydrate. Saponite Unit Cell Composition: Al Na0.303[M95.03AL0.52F90.19]($7.51 0.49’020IOHI4 Al0.49’020IOHI4 [AIIOHI2.86]2.15[M95.03A10.52Feo.19](517.51 Surface Area (FD) = 222 m2/g at 350°C PFTBA Uptake = 0.320 mmol/g 0.03u-Montmorillonite Unit Cell Composition: 1 (OH)4 Na1.00IA12.99F30.41M90.51](517.77A 0.23’020 [AlIOH’2.83I5_99[A12.99Fe0.41M90.51](517.77A10.23)020(0H’4 Surface Area (FD) = 46 m2/g at 350°C Benzene Uptake = 0.281 mmol/g PFTBA Uptake = 0.031 mmol/g 109 delaminated clay prepared at the same Q value. The hydro- carbon adsorption capacities are also lower than that ob— tained with delaminated laponite. Although saponite has a low layer charge on the order of laponites, the larger particle size apparently forces layer organization upon drying as supported by XRD. Even the small particle montmorillonite possesses some layer order with two 002 reflections. The high layer charge of montmorillonite (1.00 5/020(0H)4 unit), along with the dialysis washing method, leads to a very large number (5.99) of intercalated aluminum ions per unit cell. It may be con- cluded from these results that a low layer charge along with a small particle size significantly affects the de- lamination process. To briefly summarize this section: (1) A complete or nearly complete delaminated clay struc- ture may be produced by the reaction of polyoxocations with clays possessing a very small particle size as well as a low layer charge. Air-drying tends to reduce the extent of layer delamination, but it does not completely convert the delaminated materials to pillared or laminated clays. (2) This new class of delaminated clays possesses a high degree of macrOporosity, along with a meso- and microporos- ity. This combination of porosities present imparts novel adsorption characteristics (along with novel catalytic properties to be discussed later) as compared with conven- tional pillared clays. 110 Physical Properties of Products Obtained from Montmoril- lonite, a Typical Smectite The previous section discussed the inherent nature of a smectite mineral, namely ways in which its layer charge and particle size dramatically influence the physico-chemical properties of the resulting interlayered clay. This section is concerned primarily with montmorillonite, an intermediate layer charged, normal particle size smectite. This section will show how synthesis parameters, including type and con— centration of pillaring solution, washing techniques, and drying techniques are intimately related to the physical properties of the pillared species. Figure 36 shows XRD patterns for three intercalated clays prepared with different pillaring solutions. Table 23 provides the interlayer composition, N2 BET surface areas, and PFTBA adsorption capacities for each material. All of the polyoxoaluminum clay products contain between 1.99 and 2.59 aluminum ions per unit cell, all are pillared with Ad(00£) expansions between 6 and 10 A, and all exhibit high surface areas. It should be noted here that all data were obtained for products which were collected and washed by centrifugation and then air-dried. The PFTBA adsorption capacity for samples II and III are similar. Sample I has a larger lattice expansion at 25°C (Ad(00£)s 9 A), but col— lapses at 350°C (Ad(00£) E 6 A). This smaller Ad(00£) spacing prevents adsorption of the bulky PFTBA (kinetic 0 diameter = 10.4 A) molecule into the clay interlayer. Figure 37 illustrates the XRD patterns obtained when lll 9 15.1 0 20.07 A O 9.61 A—> J\ ° 4.6(II) 9.82 A 15.1(III l I l l I 1 14 12 10 8 6 4 2 Degrees 26 Figure 36. X-Ray Diffraction Patterns for Air-Dried Mont- morillonite Pillared Products (I) r=2.40, (II) 0.80 g ACH/g Clay and (III) 2.6 g ACH/g Clay Purged for Two Hours at 350°C in Argon. 112 .pmflupluflm paw manuaupcmo >9 Umnmmz AQV .wxmovomoxmfi.oaamm.sflmvHom.omzflv.ommmm.maa_ u coflufimomsoo umsmq Am. . . . AH.mHuo .om.muuv mfiv o mom no em em m swan m\mom moo.m HHH . . . Aw.¢uo .om.muuc How o mmm ma as mm H swan m\mo¢ mom.o HH . . . AH.mHuo . ¢.Nuuv ooo o mmm ms ma mm m Anemomz\ Hoaa H Am\aosa .mcmc Am\ms. locommc Hamo use: ucmmmmm mamemm amend .4.m « Aaooco Had Ha mcflumaaflm onHGOHHHHOEucoz emumHHflm mo mmdpummoum HMUHmmsm m com: ucmmmmm mkflwmaaflm mo pommmm .mm mHQmB 113 18.39 fi’al 19.13 3 O . \ O 9.50 A n 19.62 A (v1) 1 ‘20.07 i (V) (IV) (111) t I I I I l 14 12 10 8 6 4 2 Degrees 26 Figure 37. X-Ray Diffraction Patterns for Montmorillonite Pillared Clays (r=2.50, Q=15.l) Subjected to Various Washing Conditions and Purged for Two Hours at 350°C in Argon. 114 a clay product is subjected to different washing procedures. Samples II and IV were washed by centrifugation until the wash water tested free of Cl— ions. Sample III was air— dried, while sample IV was allowed to remain as a sol for an additional three days prior to air-drying. It appeared that sample IV was ordered to a greater degree as supported by the 002 reflections. Samples V and VI, which were washed by dialysis exhibited two 002 reflections, and possess the same degree of layer ordering in their XRD patterns. Table 24 presents physical characteristics of montmoril- lonite samples III through VI. The unit cell composition ranges from 2.27 to 4.24 aluminum ions per unit cell. Sam- ples III and V clearly illiustrate that dialysis washing of the polyoxoaluminum clay product produces a greater amount of aluminum per unit cell than washing by centrifuge pro- duces. Sample VI was centrifuged 2 times after it was washed free of excess electrolyte. Approximately 0.5 alumi- num ions per unit cell were removed, representing ion pairs generated from ACH which were unable to diffuse through the tubing because of their larger size. This implies that all of the aluminum per unit cell was not required to neutralize the negative charge upon the layered sheets. This removal of excess ions dramatically increased the PFTBA adsorption capacity. Figure 38 illustrates the effect of air-drying versus the effect of freeze-drying upon the XRD pattern of mont- morillonite for clays prepared from ACH (r=2.50, Q=15.1), washed by centrifugation with 2.87 A1 ions per unit cell. 115 .mmmam co emflucuuflm 6cm AH.mHuo .om.muuc moa an“; emumaaflm mmHQEMm Hag lac wxmovomoxma.oaamm.nflmcHmm.omzfle.ommmm.maa_ ":ofluflm0dsoo ummmq Am. Axm .ucmov sam.o Hem mm.mH qu.m mammamflo H> . . . A: oeflc smo o mam ma ma am 4 mamaamfla > Amado m mom on omm.o mom me.mfi om.m omzoaam can 662m63. >H .ucmo mfie.o mom so.om hm.m .ucmo HHH .m\aoasc Am\me. Accommc Hamo pass conumz oHQEMm .mcm .amemm .a m a A 026 umm Ha mafiamms UonHHHm Mo mmfluummoum HMOHmwnm map .A vouHCOHHHHoEucoz so mogumz mafinmmz mo pommmm .em manna 116 I365 (00!) A. 2.87 [AHOHfégg] per cell OH/Al= 2.50 IIO° dehydration . amfi 3.13 K 3.7 5 7: 4’72 A ( 003) (006) (005) (904’ ' . ' ' I8.2A B. 2.87 [AHOHEfif] per cell OH/Al ' 2.50 [Freeze Drie—cfl IIO° dehydration Degrees 29 Figure 38. X-Ray Diffraction Patterns for Air-Dried (Glass) and Freeze-Dried Montmorillonite Clay Purged for Two Hours at 350°C in Argon. 117 These XRD patterns are a consequence of layer ordering, which is induced by surface tension forces as previously ex- plained by the proposed layer flocculation model. Table 25 lists physical parameters associated with pillared clays that have been dried in a variety of ways. Some samples were washed by centrifugation and dried either on glass (sample III), on plastic (sample VIII), or freeze-dried (sample VII). Other samples were washed by dialysis and either dried on glass (sample IX), on plastic (sample XI), or freeze-dried (sample X). When a polyoxoaluminum clay product is washed by centrifugation, the drying procedure does not influence the N2 BET surface area, or PFTBA ad- sorption capacity to any appreciable degree. Conversely, dialyzed clay products exhibit striking differences in their properties when dried. Montmorillonite air-dried on glass (sample IX) typically exhibits a PFTBA adsorption capacity in the range of 0.040 to 0.128 mmol/g, whereas plastic or air-dried montmorillonite (sample XI) quantitatively ex- cludes PFTBA. Since the clay sol wets the glass more ef- fectively than the plastic, it may be concluded that the contact angle of the sol is smaller on a glass (hydro- phillic) plate than on a plastic (hydrOphobic) sheet. The glass plate, composed primarily of SiO2 units, has a smoother, more regular surface than does the plastic (poly- ethylene) which is "rougher" on a molecular scale. The hydrophobicity and "roughness" of the polyethylene sheet precipitates the molecular sieving activity of sample XI with respect to PFTBA adsorption. .oaummam mamasnumsaom A0. .quocomoxmfl.o mm.sflm.Hom.omzfiq.ommmm.maal "managed nosed An. .H.mH u 0 an mom nufiz omuommu mmHmEmm Ham Adv Hfi 118 Aoflummflmv swede ooo.o mew mm.mH ¢~.q nude .mfimsHmHa Hx mnm.o mos No.H~ «m.q wmflmmwwmmmmm x mmH.o osH ms.mfi em.¢ uwmmmwwwmmwwmm xH mfiv.o mem No.mH em.m AoluwmmumMMMWAmmwwm HHH> va.o omm mo.mH hm.m wwwmmmmmmmwm HH> mHv.o mmm so.o~ nm.~ udfimmmMMMWHmwwwm HHH .Am\aoss. Am\mec locomm. Hamo been maoflufleaoo .mcmamsam wee amemm m m aoo 90d Ha mcflmuo a 6 An. 0 .muficoHHHHoaucoz pmHmHHflm mo mHmDmEMHmm HMOHmhgm com: mcofluflpcov mcflmuo mo uommmm .mm magma 119 The preceding results indicate that the pillaring solu- tion employed in the synthesis of polyoxoaluminum clays exerts a minimal effect upon the resulting clay. The only significant difference between base-hydrolyzed AlCl3 (r = 2.40) and ACH (r = 2.50) is the lower thermal stability of base-hydrolyzed AlCl3 pillared clays, and its implications for adsorption of large molecules. The washing procedure used for the removal of excess electrolyte exerts a greater influence over the final product obtained. Centrifugation affords a high surface area, and a large PFTBA adsorption capacity. Dialysis methods produce materials with approxi- mately 33% less N2 BET surface area and 40% less PFTBA ad- sorption capacity when dried in an analogous fashion as centrifuged products. The drying procedure, whether air- dried or freeze-dried, is also instrumental in determining adsorptive properties of the final product, as referred to in the layer flocculation model. Physical Properties of Products Obtained with Fluorohec- torite Possessing High Layer Charge and Large Particle Size Figure 39 presents the XRD patterns of fluorohectorite clay interlayered with ACH (samples XII and XIII) along with base-hydrolyzed AlCl (sample XIV) at r = 2.00. The largest 3 d(00£) spacing is exhibited by sample XII, which was washed by centrifugation. Dialysis washing produced sample XIII, which poSsessed a higher degree of layer ordering. Sample XIV, which was washed by centrifugation, produced the smal- lest d spacing, presumably due to the inherent thermal (OOl) stability of pillared clays prepared from base hydrolyzed 120 O 18.8 A 18.4 A O 9.40 A?» (XII) r=2.50 Q 9.40 A fl 0 15.5 A r=2.50 r=2.40 I I I I I 12 10 8 6 4 2 Degrees 26 Figure 39. X-Ray Diffraction Patterns for Fluorohec- torite Interlayered Clay Synthesized at Various r Values (Q=6.46) and Purged for Two Hours at 350°C in Argon. 121 solution of low r value (r = 2.00). Table 26 lists synthesis conditions and physical para- meters associated with samples XII through XIV. A larger amount of aluminum ions per unit cell were intercalated in dialysis washing (sample XIII) than in centrifuge washing (sample XII) for ACH pillared clays. The surface area (14.0 m2/g) and benzene adsorption capacity (0.411 mmol/g) are correspondingly lower for the dialyzed material as well. Sample XIV, produced at r = 2.00 and washed by centrifuga— tion had the least amount of aluminum ions per unit cell. Sample XIV also exhibited a lower surface area (66.8 mZ/g) and a diminished benzene adsorption capacity (1.00 mmol/g) than sample XII, which was also washed by centrifugation. The high layer charge of fluorohectorite (1.6 e/020F4 unit) facilitates the incorporation of a greater amount of the pillaring oxocation in sample XII than would be required by montmorillonite (cf. Table 23, sample III). Samples XII and XIII exhibited the effect of centrifugation and dialysis washing with respect to their aluminum contents. Sample XIII possessed a lower surface area and a lower ad— sorption capacity due to the extra 1.08 aluminum ions per unit cell which occupy internal surface area. The aforementioned physical prOperties show that as the smectites' layer charge increased, the clay required a great- er amount of aluminum ions per unit cell to satisfy the electrical charge of that cell. The increased number of aluminum oxocations per cell in fluorohectorite occupied more interlayer area, and consequently yielded products 122 .vhomOmHmAm.HHA¢.¢mEv "madanom nomad ADV .ee.e u o no emnommh neon uSim n 9 Inc uom.~ u H Ame eo.H w.ee em.~ e m.mH mesmeuucmo Anvmaoae\mo >Hx Hae.e o.eH me.e m e.mH memsamflo AEmoe HHHx mm.H mes ee.m a m.mH mesmeuuemo AEmoe HHx O Am\aoaac Am\~ec Heme peas Accommv musemoOHm 20H95H0m maesmm osmwcmm .m.m onuwm Hm # aoo mdflSmmg mdflumaaflm doapmHOmU¢ p .mmsam> H msoaum> um Umwflmmnpdmm >6HO Umum>MHHmch mpflHOpomnouosam mo mmfluummoum HMUHmmnm .om magma 123 with lower N2 BET surface areas and adsorption capacities than montmorillonite clays synthesized in an identical fashion. Catalytic Activity of Aluminum Interlayered Smectites Several workers have carried out catalytic studies on montmorillonite pillared by aluminum polyoxocations. Vaughan76 and his co-workers first demonstrated the zeo- litic properties of the materials using spray-dried and oven-dried products. Occelli42 has also made use of oven- dried pillared clay slurries for the selective cracking of 81'82 have utilized freeze- gas oil, while Shabtai, et a1. dried pillared clays for the cracking of molecules with kinetic diameters greater than 9.0 A. The results of these workers demonstrated that polyoxoaluminum pillared clays were effective Bronsted acid catalysts. It was of interest to show how the catalytic properties of an interlayered smectite clay were related to the synthesis parameters of the clay. Dealkylation reactions with cumene (isopropylbenzene) have been used extensively in the petroleum industry73 as a test for Bronsted acidity. This reaction proceeds in the reverse manner of a Friedel-Craft alkylation. Since pil- lared clays have pore dimensions larger than faujasitic zeolites, B-isopropylnaphthalene (kinetic diameter 10.2 A) was chosen to probe Bronsted acidity in pillared clays. Pillared montmorillonites, along with delaminated laponite, were examined as catalysts for the dealkylation of B-isopropylnaphthalene. The reaction was investigated 124 over a range of weight hourly space velocities (WHSV, 0.27-2.23), contact times (0.6-6.3 sec), and temperature (300-400°C). Coke formation was especially severe for the air-dried pillared clays at higher space velocities and longer contact times. Table 27 lists the physical characteristics of the catalysts employed in dealkylation (samples XV through XIX) and gas-oil cracking reactions (sample XIX). Figure 40 depicts the dealkylation activity of pillared montmorillonite (sample XVII) as a function of the reaction contact time. As the catalyst-substrate contact time de- creases, the activity increases, along with a decreased d(00£) spacing observed after reaction. This decreased spacing was caused by the degradation of the pillar by the free protons present in the reaction. The protonation of bridging oxide units present in the pillar and their mi- gration to the silicate sheets where they were condensed and removed as water, facilitated the observed d(002) spacing reduction. Catalytic activity depended on the WHSV as exhibited in Figure 41. As the WHSV decreased, the longevity of the catalyst increased. This behavior is reasonable to expect. As the WHSV is lowered, the grams of substrate per gram of catalyst per hour is decreased, which prevents the satura- tion of the active sites and allows more efficient cataly- sis. A reduction in the WHSV also lowered the % C deposited upon the used catalyst. The reduced % C levels are consis- tent with a decreased pore blockage mechanism that accounted 125 VON oo.m mom.OHAvm.mmza .eAmo or Hmc ".xch mamsmm .vAmOVONOANH.o 4mm.bflmvmwm.omZHv.ommmm .mafia "AHHH>xI>xV mmHQEmm "mummmq mo :OHpHmOQEOO H Adv .moe .om.m u h .n. .maoam powwaoupmn mmmn Eoum Umumumcmm .Nv.m u H Am. me.o mmm I: oo.m om moe mm.Hm xHx me.o emm e.ee em.m om woe H.mfl HHH>x oo.o mem m.mH em.m ac Anemoe H.mH HH>x mm.o cam o.eH mm.m om momz\maoae H.mH H>x oo.o .mmm m.mH mm.m he Anemomz\maoae H.ma >x .m\aoea. Am\~ev Ame demo nee: maceneecoo enema o mamsmm .mee .mmemm .¢.m eoo6 Accumd Ha mcesuo mcflumHHflm emmeosum oeumamuwu MOM pom: mmmau UmumhmHHmch mufluomam mo mmflummmoum aboammcm .nm mHQmB 126 .mEHB poopaoo coflpommm mo cofluocsm 6 mm mocmpcommo coflmpo>coo :OHDMwaHme mamamnunmmcammoum0mfllm .oe musmflm Aepsv we.p ow p 8 ow on Nb 3 O u A a u '05. U E... 5.0.: 8.~ 5.: e: 2.: 8s 8.2 8 $3 3.2 ---- em.m_ .3 «:8? 0 dam >2: 8a HH>x H m 0.. e8 1 e2 seven—aspauo ecu—czagaoepxaogeomeu m 127 .>mm3 mo coeuocsm 6 mm coflmpo>coo coflumememo mcmamzunmmsawmoum0mfllm Aisvmfip cw. ed cm on u a q q d 4 q q q 1 . can pc.c l 8 "MN.” 9 e3 HH>x mflmsmm 1 Au a m.n. . sm.c J L none.: 1 -.o 5:: 1 covuepxxmumc ecu—azucea:_xeogaomwam .Hv musmflm cm cc om om cop uogsaaAuog % 128 for increased activity as the WHSV is lowered. A useful comparison of dealkylation properties for all five catalysts listed in Table 27 could be obtained at a WHSV of 0.27, and a contact time of 0.61 sec. 41 illustrates the results obtained for the Figure 42 dealkylation of B-isopropylnaphthalene at 300°C over air- dried and freeze-dried pillared products formed from base hydrolyzed AlCl3 (Samples XV and XVI) and ACH (Samples XVII and XVIII) as pillaring reagents. Included in the figure are the results obtained for delaminated laponite (Sample XIX). Curves very similar to those in Figure 42 also were observed for the dealkylation reaction at 400°C. The choice of pillaring reagent has little effect on initial cracking activity, the conversions being in the range 90 to 100% in all cases. However, the method used to dry the products significantly affects catalyst longevity. The two air-dried products exhibit decreasing activity with increasing reaction time, but the freeze-dried pillared pro- ducts, along with delaminated laponite, are stable over the range of reaction time investigated. The catalytic instability of the air-dried pillared catalysts may be related to the formation of coke on the reactive surfaces. The wt. % values listed in Figure 42 give the carbon contents of the catalysts following the 105 min. reaction times. Somewhat more carbon is deposited on .the air-dried pillared clays than on the corresponding freeze-dried products. Relative to the pillared catalysts, delaminated laponite exhibits the least tendency to form 129 \NTQQC: / (XIX) — 0.70 ICXDI' eF===eh===dt===eF===EF===EF===€LC:; (XVI FD 0.96 _ |.09 “ (XVIII FD) o 80 r- A [.(XVII AD) fl C 2 - I84 E j ' .c (XV AD) 3 60 - A A 2! A g — A l.2l 40 *- ‘2C) Jr J l l l l .L 30 60 90 |20 Reaction time, min. Figure 42. Catalytic Dealkylation of B-isopropylnaphtha- lene by Air-Dry (AD) and Freeze-Dry (FD) Forms of Pillared Montmorillonite (Samples XV-XVIII), and by Delaminated Laponite (Sample XIX). 130 coke. The results illustrate that dealkylation reactions, due to the nature of the products obtained, do not ade- quately probe the structure sensitivity of pillared clays. Delaminated laponite, synthesized at Q = 21.89, was probed by the catalytic cracking of gas-oil43 in an attempt to elucidate details of its pore distribution not obtainable from B-iSOpropylnaphthalene dealkylation. The lack of long-range layer stacking in a delaminated clay is indicated by the absence of distinct 002 x-ray reflections41. The absence of such reflections, however, does not preclude the possibility of short-range layer or- dering over a few clay sheets. If short-range layer order- ing does occur, it should be reflected in zeolite-like cata- lytic selectivity for the delaminated clay. Catalyst testing was performed by Dr. M.L. Occelli at the Gulf Research and Development Company, Pittsburgh, PA. Catalyst evaluation was performed using microactivity tests (MAT) similar to the one described by Ciapetta and Anderson83 The following reaction parameters were used: space velocity, 15; contact time 80 sec.; temperature 515°C; catalyst-to-oil ratio, 2.5. The feed stock was a gas-oil with a 260-426°C boiling point range. The same gas-oil was used previously to evaluate the cracking activity of a montmorillonite pil- lared by aluminum chlorhydrate84. The deactivation of the delaminated clay catalyst was accomplished by passing dry air or a mixture of 95% steam - 5% N2 over the catalyst for 6 hours at temperatures in the BOO-600°C range. The 131 commercial catalysts used as references were treated with steam at a higher temperature or for longer periods of time to obtain the desired conversion level. The activity of our delaminated clay catalyst for gas oil cracking is similar to the activity of an amorphous AAA-alumina catalyst (78% SiO 22% A1203). However, the 2: selectivity of the delaminated clay catalyst more nearly re- sembles the selectivity of a commercial zeolite-promoted FCC catalyst. The zeolite-like activity is indicated by the gasoline yields for conversions in the 54 to 67% range. As shown in Figure 43, the CS-C12 yields for the delaminated catalyst are in the same range as the commercial FCC cata—f lyst and significantly greater than those obtained with amorphous AAA-alumina. As shown in Figure 44A, the yields of light cycle gas oil (LCGO) obtained with the delaminated clay are higher than those obtained with either the AAA-alumina or the com- mercial FCC catalyst. The enhanced LCGO yields are accom- panied by low slurry oil (SO) yields, as illustrated by the results presented in Figure 44B. Thus, the delaminated clay converts more of the heavier gas oil components to the more desirable LCGO fraction. The selectivity of the delaminated clay toward gasoline and LCGO occurs without increasing the yield of light gases. This result is illustrated by the yields of propane and propylene in Figure 45 and by the yields of nfbutane, i-butane and butenes in Figure 46. Hydrogen and carbon yields are provided in Figures 47 and 48, respectively. The hydrogen yields for the 132 GASOLINE (cs-0,2) YIELDS 65 P 60 - A/A’A 55 - n/ 45r- -/ .4() L 1 1, L, 1 1 50 55606570 758 CONVERSION (V’AFF) GASOLINE YIELDS (W. FF) Figure 43. Gasoline (CS-Clz) Yields Obtained with De- laminated Clay (0) , AAA-Alumina (I) , and Zeolite-Promoted FCC (A) Catalysts. 133 A. LCGO YIELDS I: 30r- u. “\i?‘ 2,. . I e /* ...J L A lens ). 9 (I S 50.5» ca A ': 40 50 60 7O 80 CONVERSION (V'loFF) B. I-BUTANE YIELDS i ‘BUTANE YIELD (V°/. FF) 0) r 40 5O 6O 7O 80 CONVERSION I V°/.F F) E c. BUTENES YIELDS .\>o I0 '- ..‘-.—___._ I; 8 - / .- 3 I 0V LI.) 6 '- }..A/,—-——A—A A A'— ; (>73 A m 4 )- A/ u.) 2 E 2- 8 O 1 1 1 4 4O 50 60 7O 80 CONVERSION (V’/. FF) Figure 46. (A) n-Butane, (B) i-Butane, and (C) Butenes Yields Obtained with Delaminated Clay (0), AAA-Alumina (I), and Zeolite—Promoted FCC (A) Catalysts. 136 A 0.20 _ H2 YIELDS & C E, 0.15 - ./-/ ‘2’ './ / LU C1“3" "” I A// E; III], ‘51,,I15Ar E 005 - A "'Z/4 >- ' 4 A I: 000 l 1 1 1 4O 5O 60 7O 80 CONVERSION (V'loFF) Figure 47. H Yields Obtained with Delaminated Clay (0) , AAA-Alumina (I) , and Zeolite-Pro— moted FCC (A) Catalysts. 137 6 CARBON YIELDS E u. .\° - P 4 E Z 8 2+ (r 4 L) (3 1 1 1 1_ 40 50 60 70 80 CONVERSION (V’/.FF) Figure 48. Carbon Yields Obtained with Delaminated Clay (0) , AAA-Alumina (I) , and Zeolite- Promoted FCC (A) Catalysts. 138 delaminated clay are in the range typically observed for both zeolitic and amorphous cracking catalysts. The carbon yields, however, are significantly larger than is typical for a zeolite-type catalyst. In fact, the carbon yields for the delaminated clay are even higher than those for the amorphous AAA-alumina catalyst. The tendency for the de- laminated clay to form coke may be related to the macro- porosity of this catalyst. Certainly, aromatic molecules can readily adsorb in the macroporous structure. The ine teractions of the adsorbed aromatics with Lewis acid sites on the dehydroxylated oxocations could facilitate polycon- densation of the aromatic centers and the formation of coke. The selectivity described above for the delaminated clay catalyst is very similar to the selectivity exhibited by a conventional pillared clay for the cracking of the same gas-oil sample under analogous reaction conditionse4. The pillared clay gave a somewhat lower gasoline yield and somewhat higher light gas and carbon yields than the de- laminated clay, but the relative LCGO and SO yields are very similar to those for the delaminated clay. The ten- dency for the pillared clay to give less gasoline, more light gas and more coke than the delaminated clay is probab- ly related in part to the iron contents of the two catalysts. The pillared clay used to prepare the previously reported 84 was a natural montmorillonite with an pillared catalyst appreciable structural iron content. In contrast, our de- laminated clay is prepared from a synthetic smectite which I I 41 I 0 contains only trace amounts of Iron . Structural Iron In 139 aluminosilicate catalysts can facilitate the cracking of heavier hydrocarbons to light gases and promote coke for- mation. The similar cracking selectivities for delaminated and pillared clay catalysts suggests that the delaminated clay possesses significant two-dimensional zeolitic character. Zeolite-like catalytic properties persist for the delami- nated clay despite the presence of macrOporosity and the lack of x-ray evidence for long-range layer stacking. The previously proposed "house-of—cards" model41 for a de- laminated clay is consistent with the presence of zeolite- like micrOporosity, as well as macroporosity. The "house- of—cards" model is shown in Figure 49. Included in the Figure is a model for the structure of a well-ordered pillared clay. In the delaminated clay model shown in Figure 49A, the extensive edge-to-basal plane interactions lead to the for- mation of macropores. In addition, some basal plane to basal plane layer stacking takes place, in accord with the zeolite-like cracking selectivity found in the present work. However, the layer stacking is short-range, occurring only over a few (2 to 3) layers (40-60 A), thus accounting for the absence of discrete OOR x-ray reflections. It should be noted, however, that the intercalated oxocations in these short-range stacked layers may not be as regularly spaced as in a well-ordered pillared clay. Evidence sup- porting a distribution of micropores in a delaminated clay has been provided previously by Pinnavaia and co-workers41 ZEOLITIC MICROPOROSITY MACROPOROSITY B Bra—E“ 'Ehéfil?=%5==fifil; °-::-."I-_.'::.’2-:———:::::::z-:-= 'TIE?3:E%=£IEFfi? |._._Il_._E——W—J Figure 49. (A) House-of—Cards Structure for a Delami- nated Clay Catalyst. (B) The Long-Range Layer Stacking in a Well-Ordered Pillared Clay. 141 on the basis of physical adsorption data for molecules of different kinetic diameters. In contrast to the short-range layer ordering which occurs for a delaminated clay, the layer stacking in a well- ordered pillared clay occurs over a much longer range (cf.' Figure 49B). Also, the spacing of pillars is quite regular. Thus, a pillared clay exhibits several 00% x-ray reflections, along with well-defined molecular sieving properties. How- 4— ever, well-ordered pillared clays lack the macroporosity and facile diffusion properties41 of a delaminated clay. Conclusions Two types of polyoxoaluminum solutions, base—hydro- lyzed AlCl3 (r=2.00-2.42), and ACH (r=2.50) have been used as reagents for pillaring Laponite-RD®, montmorillonite, and fluorohectorite. ACH imparts larger d spacings and (00%) greater thermal stability to these smectites than to pil- lared clays formed from base-hydrolyzed AlCl3. Selective adsorption studies demonstrate that the method used to wash the exchanged smectite sol, as well as to dry the floccu- lated clay layers, has a pronounced effect on the resulting pillared products' physico-chemical properties. For mont- morillonite and fluorohectorite, centrifuge washing of the reactant sol removes large intercalated species. This leads to final products with a lower # Al bound per unit cell, larger d spacings, larger surface areas, and larger (00%) PFTBA adsorption capacities than obtained for dialyzed materials, regardless of the drying conditions employed. 142 Laponite, which was dialyzed free of excess electrolyte, exhibited two distinct XRD patterns depending upon the dry— ing method employed. Air-dring afforded a material with discernable, sharp XRD patterns, whereas the freeze-dried materials possessed no d reflections. The surface (00%) areas and PFTBA adsorption capacities were similar for both materials. A clay flocculation model is presented in which both lamellar (face-to-face) and delaminated (edge-face, edge-edge) associations of the layers can occur, depending in part on the morphology and charge of the layers. Layer delamination also has important catalytic consequences, as demonstrated by differences in the sensitivity of air-dried and freeze-dried pillared clays to coke formation in B-iso- prOpylnaphthalene dealkylation. Gas-oil cracking was used to probe the layer orientation in laponite, and showed that short-range order (40-60 A) does exist. This is reflected in zeolite-like cracking selectivities. APPENDIX APPENDIX Washing Procedures As was shown by the data presented in Table 24 (cf. p. 115), the method used to wash a pillared clay sol free of excess electrolyte greatly influences the physical proper- ties of the resulting pillared clay. In an attempt to standardize this critical step of the synthetic procedure, a detailed description of dialysis and centrifuge washing, as well as air- and freeze-drying techniques employed in this study are outlined below. Dialysis washing is accomplished in the following manner. After the approPriate aging period is completed, the pillaring solution-clay sol is transferred to standard cellulose dialysis tubing. The dialysis tubing employed in the preceding studies was supplied by VWR Scientific, Inc., (45 mm x 100 ft.) with a 12,000-14,000 molecular weight cut-off. The filled dialysis tubing was placed in deionized water which was repeatedly changed approximately every two hours during the course of a normal working day. This procedure was continued until all excess chlorine was re- moved, as checked by the silver nitrate test for chloride. Centrifuge washing can be accomplished in several ways. 143 144 The procedure given below must be followed exactly if one wishes to reproduce any material presented in this disserta- tion. The centrifuge must be set to maintain its internal temperature at 25°C. The intercalated clay is first separ- ated from the bulk pillaring solution mixture by centrifuga- tion at 5500 RPM for 15 minutes. After the first 1500 ml aliquot is centrifuged, the mother liquid is decanted. Ad- ditional pillaring solution-clay sol is added to the tubes without redispersion of the previousl; sedimented clay. The tubes are again centrifuged for 15 minutes at‘5500 RPM. .This procedure was continued until all the interlayered clay was collected. The clay present in the centrifuge tubes was dispersed before removal from the tubes by the addition of 100 ml of distilled water, along with two small stir bars, which were used to break up the sedimented clay through agitation. It should be noted that the sedimented clay must not be re- suspended upon a stir plate since this action facilitates pillar hydrolysis causing structural rearrangement. After dispersion, stir bars were removed and the clay suspension was transferred to a 2.0 l Erlenmeyer flask. The clay was then shaken in a minimum of water to further break up any small clay aggregates remaining. After all the lumps were removed, the solution was diluted to the volume necessary to fill the centrifuge tubes (1500 ml), shaken quickly to produce a homogeneous dispersion, and then transferred to the centrifuge tubes and centrifuged for 15 minutes. 145 After the centrifuge stOpped, the mother liquid was decanted and the clay was again redispersed according to the previous directions. The sequence of centrifugation speeds employed are as follows: - 2x at 5500 RPM for 15 minutes - 1x at 5000 — 1x at 4500 - 1x at 4000 - nx at 3500 Once the 3500 RPM speed was reached, the clay was washed at this speed until the decanted liquid yielded a negative silver nitrate test for chloride. Flocculated clay particles appeared in aluminum systems after four washings, but still contained a considerbale ex- cess of chloride. Typically it took 8 to 10 washings before the silver nitrate was negative. The chromium system gener- ally takes 7 to 8 washings before flocculated particles are noticeable. Since no chlorine was used in chromium cluster synthesis, this system was washed in an identical fashion as the aluminum pillared clays, with the following exception: after flocculated chromium clay particles were present (~7 washings), the clay was washed an additional two times at 3500 RPM. Drying Techniques Clay samples were either air-dried on glass sheets or polyethylene film, or freeze-dried. Air-drying was accom- plished by dispersion of the washed clay in distilled water at approximately 1 wt. %, which was then spread on the desired drying surface. 146 Freeze-drying was accomplished in the following manner. The clay suspension (~1 wt. %, 200 ml) was placed in a freeze-drying flask and immersed in liquid nitrogen. Initi- ally, while the solution was cooling down, the flask was swirled to prevent settling of the clay. Once the solution started to freeze, the flask was removed from the liquid nitrogen and rotated at a 45° angle. This action coated the walls of the flask with a thin film of clay. This procedure of immersion and rotation was continued until the remaining clay was frozen. LI ST OF REFERENCES 8. 9. 10. 11. 12. 13. 14. 15. LIST OF REFERENCES Mross, M.-D., Catal. Rev.-Sci, Eng., 1983, 25(4), 591. Webber, D., CENEAR, 1984, 62(19), 7. Heinemann, H., Catal. Rev.-Sci. Eng., 1981, 23(1 and 2), 315. Rabo, J.A.; Pickert, P.E.; Stamires, D.N. and Boyle, J.E.,"Proceedings 2nd Int. Congress Catalysis," Ed. Tech., Paris, 1960, p. 2055. Breck, D.W.,"Zeolite Molecular Sieves," Wiley, New York (1974). Olson, D.H.; Haag, W.O. and Lago, R.M., Jour. Catal., 1980, 1, 390. Benesi, H.A. and Winquist, B.H.C., Advances in Cataly- sis, Vol. 27, Academic Press, New York (1978). Baba, T., Bull. Chem. Soc. Japan, 1930, 5, 190. Barrer, R.M., J. Soc. Chem. Ind., 1945, 4, 130. Barrer, R.M. and Belchetz, L., J. Soc. Chem. Ind., 1945, 64, 131. British Patent #548,905 (1942); U.S. Patent #2,306,610 (1945). Barrer, R.M.,"Zeolites and Clay Minerals as Sorbents and Molecular Sieves," Academic Press, London (1978). Weisz, P.B. and Frilette, V.J., J. Phys. Chem., 1960, 64, 382. Weisz, P.B.; Frilette, V.J.; Maatman, R.W. and Morver, E.B., Jour. Catal., 1962, l, 307. Kaeding, W.W.; Chu, C.; Young, L.B.; Weinstein, B. and Butter, S.A., Jour. Catal., 1981, 67, 159. 147 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 148 Meisel, S.L.; McCullough, J.P.; Lechthaler, C.H. and Weisz, P.B., Chem. Tech., 1976, g, 86. Derouane, E.G. in "Catalysis by Zeolites," Iemlik, B.; Naccache, C.; Ben Taarit, Y.; Verdrine, J.C.; Coudurier, G.; Praliavd, H. Editors, Elsevier Scientific Publi- shing Company, The Netherlands (1980). Frilette, V.J.; Haag, W.O. and Lago, R.M., Jour. Catal., 1981, 61, 218. Dewing, J., J. Molec. Catal., 1984, 0000, 0000. Derouane, E.G. and Gabelica, Z., Jour. Catal., 1980, 66, 486. Lowe, B.M.; Whan, D.A. and Spencer, M.S., Jour. Catal., 1981, 0, 273. - Theng, B.K.G.,"The Chemistry of Clay-Organic Reactions," Wiley, New York (1974). Grim, R.E.,"Clay Mineralogy," 2nd Edition, McGraw-Hill, New York, 1968, pp. 77-92, 185-195. Pinnavaia, T.J.; Raythatha, R.; Lee, J.G.S.; Holloran, L.J. and Hoffman, J.F., J. Am. Chem. Soc., 1979, 101, 6891. “‘ Pinnavaia, T.J., 182nd ACS Meeting, New York, Inorg. 45, 1981. ‘ Adams, J.M.; Ballantine, J.A.; Graham, S.H.; Laub, R.J.; Purnell, J.H.; Reid, P.I.; Shaman, W.Y. and Thomas, J.M., Jour. Catal., 1979, 66, 238. Barrer, R.M. and Macleod, D.M., Trans. Farad. Soc., 1955, 51, 1290. Barrer, R.M. and Reay, J.S., Trans. Farad. Soc., 1957, 66, 1293. Mortland, M.M. and Clementz, D.M., Clays and Clay Miner- als, 1974, 32, 223. Berkheiser, V. and Mortland, M.M., Clays and Clay Miner- als, 1977, 3E: 105. Traynor, M.F.; Mortland, M.M. and Pinnavaia, T.J., Clays and Clay Minerals, 1978, 66, 318. Koppelman, M.H. and Dillard, J.G., Clays and Clay Miner- als, 1980, 26, 211. Loeppert, R.H. and Mortland, M.M., Clays and Clay Min- erals, 1979, g1, 373. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 149 Endo, T.; Mortland, M.M. and Pinnavaia, T.J., Clays and Clay Minerals, 1980, 66, 105. Shabtai, J., International Conference on Polymers as Reagents, Milan, Italy (1977). Mortland, M.M. and Burkheiser, V., Clays and Clay Min- erals, 1976, 66, 60. (a) Brindley, G.W. and Semples, R.E., Clay Minerals, 1977, 66, 229. (b) Lahav, N.; Shani, U. and Shabtai, J., Clays and Clay Minerals, 1978, 66, 107. (C) U.S. Patent #4216188 (1980). (d) Lussier, R.J.; Magee, J.S. and Vaughan, D.E.W., 7th Canadian Symposium on Catalysis, Edmonton, Alberta (1980). (e) Raythatha, R., Tzou, M.S. and Pinnavaia, T.J., 183rd ACS Meeting, Las Vegas, Coll. 197 (1982). Bottero, J.Y.; Cases, J.M.; Fiessinger, F. and Poirier, J.E., J. Phys. Chem., 1980, 66, 2933. Johansson, G., Acta Chem. Scand., 1960, $6, 771. Mortland, M.M. and Raman, K.V., Clays and Clay Minerals, 1968, 16, 393. U.S. Patent #4238364 (1980). Pinnavaia, T.J.; Tzou, M.S.; Landau, S.D. and Raythatha, R.H., J. Molec. Catal., 1984, 0000, 0000. Occelli, M.L. and Tindwa, R.M., Clays and Clay Minerals, 1983, 31(1), 22. Occelli, M.L.; Landau, S.D. and Pinnavaia, T.J., Jour. Catal., 1984, 000, 000. Brindley, G.W. and Yamanaka, S., Amer. Mineral., 1979, 64, 830. Tzou, M.S., Ph.D. Dissertation, Michigan State Univer- sity, 1983, p. 78. Tzou, M.S.; Landau, S.D. and Pinnavaia, T.J.,"Catalytic Properties of Clays Pillared by Chromia and Alumina," Sixth Annual Spring Symposium, Michigan Catalysis Soci- ety, East Lansing, Michigan (1983). Yamanaka, S. and Brindley, G.W., Clays and Clay Miner- als, 1978, 26(1), 21. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 150 Yamanaka, S. and Brindley, G.W., Clays and Clay Miner- als, 1979, 27(2), 119. Yamanaka, S.; Yamashita, G. and Hattori, M., Clays and Clay Minerals, 1980, 28(4), 281. Brindley, G.W. and Yamanaka, 8., American Minerologist, 1979, 66, 830. Tzou, M.S. and Pinnavaia, T.J., Unpublished Results. Bjerrum, N., Ph.D. Dissertation, Copenhagen, Denmark (1908). Bjerrum, N. and Faurholt, E., Z. Physik. Chem., 1927, 130, 584. Laswick, J.A. and Plane, R.A., J. Am. Chem. Soc., 1959, 6;, 3564. Thompson, M. and Connick, R.E., Inorg. Chem., 1981, 66, 2279. Finholt, E.; Thompson, M.E. and Connick, R.E., Inorg. Chem., 1981, 36, 4151. Brunauer, S.; Emmett, P.H. and Teller, E., J. Am. Chem. Soc., 1938, _6, 5. Hirschfielder, J.O., Curtis, C.F. and Bird, R.B., "Molecular Theory of Gases and Liquids," Wiley and Sons, New York (1964). O'Reilly, D.E. and MacIvers, D.S., J. Phys. Chem., 1962, 66, 276. O'Reilly, D.E., Advances in Catalysis, Vol. 27., Aca— demic Press, New York (1960). Ellison, A., J. Chem. Soc. Faraday Trans. 1., 1984, 66, 2581. Wittgen, P.R.M.M.; Groenveld, C.; Janssens, J.H.G.J.; Wetzels, M.L.J.A. and Schuit, G.C.A., J. Catal., 1979, 59, 168. McAuliffe, C.A. and Ashmawy, F.M., J. Chem. Soc., Faraday Trans. 1, 1984, 66, 1083. Groenveld, C.; Wittgen, P.P.M.M.; van Kersbergen, A.M.; Mestrom, P.L.M.; Nuijten, C.E. and Schuit, G.C.A., 6. Catal., 1979, 62, 153. Ciapetta, F.G.; Dobies, R.M. and Baker, R.W. in "Cataly- sis," Emmett, P.H., Editor, Reinhold Publishing Company, New York (1958). 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 151 Pines, H.,"The Chemistry of Catalytic Hydrocarbon Con— versions," Academic Press, New York, 1981, pp. 208-209. Moss, R.L. in "Catalysis," Kemball, C., Editor, The Chemical Society, Specialist Periodical Reports, Vol. 1, Wright and Sons, Ltd., England (1977). Satterfield, C. N.,"Heterogeneous Catalysis in Practice," McGraw-Hill, New York, 1980, pp. 268- 271. Clark, A. Ind. Eng. Chem., 1953, 45, 1476. Mills, G.A., Heinemann, H., Milliken, T.H. and Oblad, A.G., Ind. Eng. Chem., 1953, 66, 136. Bridges, J.M.; Rymer, G.T. and MacIvers, D.S., J. Phys. Chem., 1962, 6, 871. Van Reijen, L.L.; Sachtler, W.M.H.; Cossee, P. and Brouwer, D.M.,"Third Internation Congress on Catalysis," Sachtler, W.M.H.; Schuit, G.C.A. and Swietering, P., Editors, North Holland Publishing Company, 1965, p. 829. Corma, A. and Wojciechowski, B.W., Catal. Rev.-Sci. Eng., 1982, 24(1), 1. Jacobs, P.A.,"Carboniogenic Activity of Zeolites," Elsevier Scientific Publishing Company, New York (1977). Thomas, C.L.,“Catalytic Processes and Proven Catalysts," Academic Press, New York, 1977, p. 28. Vaughan, D.E.W. and Lussier, R.J., Preprints, Fifth International Conference on Zeolites, Naples, Italy, June 2-6 (1980). Akitt, J.W. and Farthing, A., J. Mag. Res., 1978, 33, 345. Akitt, J.W. and Farthing, A., J. Chem. Soc., Dalton Trans., 1981, 1617. Akitt, J.W. and Farthing, A., J. Chem. Soc., Dalton Trans., 1981, 1624. Van Olphen, H.,"An Introduction to Clay Colloid Chemis- try," Wiley, New York, 2nd Ed., 1977, pp. 95-98. Shabtai, J.; Lazar, R. and Oblad, A.G., New Horizons in Catalysis, 1981, 1, 828. Shabtai, J.; Lazar, R. and Oblad, A.G., Proceedings of the 7th International Congress on Catalysis, Tokyo, Japan, July 1-4, 1980, paper B-8. 152 83. Ciapetta, F.G. and Andersen, D., Oil and Gas Journal, 1977, 66, 88. 84. Occelli, M.L., I & EC Proc. Res. Dev. Journal, 1983, 22, 553. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII