'Mlmimrfl' 4 .- -ua4rhm W > «1' 3-27.31 {“11" ~' rt? 1193‘; 1 3' ' £1)" ¥y ‘Il'i “I V ‘- .m‘tz'} .p ‘ .A‘IM‘H-bt' .{I'm 1." J-P1h'""" I J ~'I' \ Hg 5*? "[11"in 1 -'1' 1'1 .'. 5‘: J... Ihlrs l . ‘I. ' 1 3:31:13? 1". 4.” "\"JL' ‘IVI-V.‘ "V3.31" ‘. #3114}- " (a. 4:" {"1“} 1‘1"“. . ‘14:"..1'.‘ V‘:\ 3",}: “Pisa J“)\u £1?“ "v‘PM' .11 {Cf . 1 ,f-HVV. 3.1 .8 wh\i\'k‘“u- ‘H'l 1' ;,,' "I“ '1 “1;; 1111‘ v a .41)"; . *3 I . .1“... 'I. m ‘2'. Rh *3ng :11: - 41.3111? ‘: 2;:i‘quV... a. <11- 1.- "‘5’, 41 . W . $1" VH-"ZJJV: 35"“ I .5gi 1.. “tnéfi \.:'“ '11'..'~"'»". '5‘"? 1'1?“ I'I'u‘\\“u 11.1 ‘0' 5'5. 3; .1 {Ida‘m‘fib ‘6'. "M .In fiflfi {‘:'” fig" ,1. "‘V"" c I‘M‘V ‘JxVIIJI‘. "mitt 1.1'1‘ . in)“. “1‘10‘1‘ u 1-IfouuU1H4I. o ’1‘: “pk 'W at (“£11 531.. ‘17.“: .1 :43?”‘n"4‘ iHESl‘“ LIBRARY Michigan State University This is to certify that the dissertation entitled Intercalation of Transition Metal Complex Salts in Layered Silicate Clays: Applications as Phase Transfer Catalysts presented by Abbas Kadkhodayan has been accepted towards fulfillment of the requirements for Ph.D. degree in _Chemisj;ny_ 5 J. Pinnavaia ' f ”/ff Dr. Th Date/20”“ 23/, // MS U is an Affirmative Action /Equal Opportunity Institution 0-12771 MSU LIBRARIES .—c-_. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. INTERCALATION OF TRANSITION METAL COMPLEX SALTS IN LAYERED SILICATE CLAYS: APPLICATIONS AS PHASE TRANSFER CATALYSTS BY Abbas Kadkhodayan 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 INTERCALATION OF TRANSITION METAL COMPLEX SALTS IN LAYERED SILICATE CLAYS: APPLICATIONS AS PHASE TRANSFER CATALYSTS BY Abbas Kadkhodayan Several chelated transition metal complex salts as well as some organic salts have been intercalated in smectite clays (hectorite and montmorillonite). After the character- ization of these systems by a variety of physical and chemical measurements, their utility as phase transfer catalysts was examined. Among the complexes used in the present investigation tris-orthophenanthroline metal complexes of the type M(phen)3xo4 (M-Ni,Fe,Zn) and (X04 =MoOi-, Croz’, woi’, 502’) and organic salts such as Nile Blue were capable of binding to smectite clay in excess of its cation exchange capacity (Intersalation reaction, see Chap. 1). The adsorp- tion isotherms demonstrate that the complexes have a marked affinity for the clay surfaces in aqueous suspension. Chemi- cal analysis of M(phen)§+/XOi--hectorite indicates the presence of one equivalent of M(phen)§+ to balance the layer charge of the clay and one equivalent as M(phen)§+/Xoi- ion pairs. The solid state of these intersalated phases yield a d(001) spacing as high as ~30 A which is consistent with the presence of an anion layer sandwiched between two layers Kadkhodayan, Abbas of the complex cation within the interlayer of the smectite. Intercalation of various complex cations and anions showed that the intersalation reactions are dependent on the nature of the complex cation as well as counter-anion. Several intersalation phases (e.g. Fe(phen)§+/SOi-- hectorite) have been tried for the desorption of complex salt from the interlayer, when they were equilibrated in aqueous solution as well as organic solvent. The results indicate that the binding of these ion pairs to the clay interlayers are quite strong. For example, the coefficient for desorption of Fe(phen)§+/Soi- ion pairs from the two-CBC hectorite intersalate in water at 20°C is 8.2 ><10_4 M. However, among the exchange ions Ni(phen)§+/Cr0 --hectorite by M00 of MoOi-, essentially all of the intercalated CrO investigated, acetate has the lowest affinity for replacing intercalated Croi- anion. Kadkhodayan, Abbas X-ray diffraction measurements were used to study the swelling preperties of Ni(phen)§+/Soi--hectorite system in two solvents used in the catalytic studies (toluene and water). Virtually no swelling occurred with toluene, but in the case of water an expansion oflvs A was observed. Zn(phen)§+/SOi-—hectorite intersalate (2 equivalent CEC) was synthesized and a nitroxide spin probe [peroxyl- amine disulfonate (PADS)] was doped into the anion inter- layers at about 2% level of exchange. The spin probe was used to estimate the interlamellar mobility at different relative humidity. For example, under fully wetted condi- tions the tumbling mobility of the intercalated 2-(PADS) was Tc 5 3><10-10 sec. Therefore, the interlayer takes on appreciable solution-like properties. An indication of the retention of chemical and structural constitution of the complexes upon intercalation was observed by IR and electronic spectroscopy. The ease with which xoi‘ anions can be replaced in M(chel)§+/X0i--hectorite intersalates suggested that these compounds may be effective catalysts in a triphase reaction system involving anionic nucleophiles. In this respect several types of layered silicate intersalates were synthe- sized and used for halogen exchange reactions on different alkyl halides as well as carboxylate and cyanide ion displacements. It was found that Ni(phen)§+/SOi--hectorite is 140 times more selective for the reaction of l- bromopropane than the largest substrate (l-bromo hexadecane) tried. The intersalation compounds were also examined for Kadkhodayan, Abbas shape selectivity. However, despite the great size selectivity of the intersalates, no dramatic shape selecti— vity was observed. All catalysts except the cetylpridinium clay intersalate, exhibit a pseudo first-order dependency on the substrate concentration and the amount of catalyst. THE SUCCESS OF THIS WORK IS DEDICATED TO MY GRANDMOTHER; ANA. FOR WHOM NO ACKNOWLEDGMENT WOULD BE SUFFICIENT. I HAVE ALWAYS WISHED THAT I COULD DO SOMETHING IN RETURN FOR ALL SHE HAS DONE FOR ME...BUT NEVER COULD! -11- ACKNOWLEDGMENTS Credit must be extended primarily to mankind's development of the Scientific Method and, thereby, the only means we have of separating fictitious thought and fantasy from scientific realities of this universe. This method has and certainly will continue to change the world. For one reason or another I never seriously worked hard on my Ph.D. until near the end. For Professor Thomas J. Pinnavaia, my research advisor, who was very patient and tolerant, I am very appreciative. I sincerely thank him for all that he has done for me. Special thanks are also due Professor Carl H. Brubaker, who served as my second reader and also to Professor Max M. Mortland for providing special research facilities and helpful discussions. I also wish to thank Professors Ashraf.M.El-Bayoumi. and Richard Schwendeman, who also served on my guidance committee. Special acknowledgment is due the members of the academic and technical staff of the Chemistry Department and also to colleagues of my research group whose cooperation and help has been invaluable. -iii- Gratitude is expressed to Ms. Margy Lynch for her great assistance in typing this dissertation and help in many occasions. Also, I would like to remember here Martin Rabb and Emmanuel Giannelis, who have become dear to me as friends. Many thanks also go to all the pe0ple whose work has been administered in the following ways: a scholarship from the University of Teacher Education, Tehran, Iran; a Teaching Assistantship from the Department of Chemistry, Michigan State University; a Research Assistantship from the National Science Foundation, Michigan State University; a Sage Foundation Award from the Sage Foundation, Michigan State University. Finally, I am deeply grateful to Nahid, with whom I share my life, for all her help, encouragement and friendship throughout the course of this study. _iv_ so: 2. Ev for 9.858 a... 8 8.55:8 2.. 53:8 .63 £8» a :2. co 3 2%» V8133..." .2: .2828. . 4:9; 62. ceases... E 3333...: TABLE OF CONTENTS CHAPTER PAGE LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . ix LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . . .xi LIST OF SYMBOLS AND ABBREVIATIONS. . . . . . . . . . . . .xv CHAPTER I INTRODUCTION. . . . . . . . . . . . . . . . . 1 A. Structure and Properties of Layered Silicates. . . . . . . . . . . . . . . . .1 B. Intercalation of Transition Metal Complex Salts in Smectite ("INTERSALATION). . . . . . . . . . . . .12 C. The Principles of Phase Transfer Catalysis. . . . . . . . . . . . . . . . 19 D. DevelOpment of Supported Phase Transfer Catalysis. . . . . . . . . . . . . . . . 32 1. Concept of Triphase Catalysis (TC). . . . . . . . . . . . . . . . .32 2. Relationship of Triphase to Phase-Transfer Catalysis. . . . . . .33 3. Polymer-supported Phase Transfer Catalysts. . . . . . . . . . . . . 36 4. Polymer-supported 'Solvents' and 'Cosolvents' . . . . . . . . . .43 5. Inorganic-based catalysts. . . . . . 44 E. Anion Exchange and Catalytic Properties of Smectite Intersalates. . . 48 CHAPTER II EXPERIMENTAL. . . . . . . . . . . . . . . . .52 A. Materials. . . . . . . . . . . . . . . . 52 1. Natural Hectorite. . . . . . . . . . 52 2. Sodium Montmorillonite (Wymoning). . . . . . . . . . . . . 52 3. Synthetic Hectorite (Laponite- RD). . . . . . . . . . . . . . . . . 53 4. Fluoro-Hectorite. . . . . . . . . . 53 5. Reagents and Solvents. . . . . . . . 53 B. Synthesis. . . . . . . . . 54 l. M(phen)3SOa and M(bp)3SOu Complexes (m= -Ni, Fe, Zn). . . . . . 54 2. Ni(phen)3x0a (XOL==Moou, Croa, W0“). . . . . . . . . . . . . . . . 55 -vi- CHAPTER CHAPTER III C. D. E. F. G. 3. Fe(L)3SOu, [(L==l,lO-orthophen- anthroline)3,4,7,8-tetramethy1- l,10-phenanthroline, and 4-7- diphenyl- -l ,10-phenanthroline (Bathopheanthroline)]. . . . . . 4. Fe(TPTZ)2(C10a)2,4H20,Fe(TPTZ)2- 504. . . . . . . . . . . . . . . 5. Ni(en)3SOa and Ni(NH3)GSOa. . . 6. F6(bP)3(ClOu)2 and F6(bP)3SOu. . 7. K3CI(OX)3'3H20. . . . . . . . . 8. Elemental Analysis and Storage. Preparation of Intercalates. . . . . Preparation of Intersalates. . . . . Desorption Coefficient Measurement. Ion Exchange-Adsorption Isotherms. . Anion Exchange Adsorption Isotherm Studies. . . . . . . . . . . . . . Kinetics of Biphase and Triphase- Catalyzed Displacement Reactions. Reaction of Benzyl Chloride With Salts of the Acetate Ion. . . . . . Modification of the Intersalate With Silane Coupling Agent. . . . . . . . Physical Methods. . . . . . . . . . 1. X-ray Diffraction Studies. . . . 2. ESR Studies. . . . . . . . . . . 3. UV-Visible Spectroscopy. . . . . 4. Gas Chromatography. . . . . . . 5. Infrared Spectra. . . . . . . . RESULTS AND DISCUSSION. . . . . . . . . A. Adsorption of Complex Salts on Smectite-(Intersalation Reaction). . 1. Adsorption of Metal Complex Salts. . . . . . . . . . . 2. X-ray Diffraction of M(che1)2+/ X02 -Hectorite. . . . . 3. An1on Effects on M(chel)2+ Adsorption. . . . . . 4. Effects of Ligand on M(chel)2+ Adsorption. . . . . . . . . . . 5. Electronic Spectra. . . . . . . 6. Adsorption of Organic Salts. . . 7. Desorption Coefficient Measure- ment. . . . . . . . . . . . . . 8. Anion Exchange Properties of Smectite Intersalate. . . . . . 9. Swelling of the Intersalation Phase. . . . . . . . . . . . . . 10. ESR Studies. . . . . . . . . . . -vii- PAGE .69 CHAPTER APPENDIX. B. Catalytic Properties of the Intersalates. . . . . . . . . . . 1. Kinetics of the Haolgen Exchange. . . . . . . . . . . 2. Substrate Size and Shape Selectivity. . . . . . . . . . Structural Modification. . . . Particle Size Effect. . . . . Carboxylate Ion Displacement. Supported Cetyl Pridinium Bromide for TC. . . . . . . . 7. Halogen Exchange Using Organic Salt Intersalates. . . . . . . 8. Kinetics of Cyanide Displace- ment on l-Bromooctane. . . . . aims-w LIST OF REFERENCES. . . . . . . . . . . . . . . . . -viii- TABLE 10 11 LIST OF TABLES PAGE X-ray Basal Spacing, d(001) of Na+-hectorite Exchanged with Different Complexes (Total Loading is 73 meg/100 gram clay). . . . . . . . . .71 X-ray Basal Spacing d(001) of Na+-Hectorite Exchanged with Different Complex Salts With Loading of 2 Equivalent CEC. . . . . . . . . . . . 74 Cation Exchange Capacity and X—ray Basal Spacing, d(001) of Different Clay Exchanged with Ni(phen)3SOu. . . . . . . . . . . . 35 Effect of Ligand, Counteranion and Loading on the d(001) Spacing of Intersalated HectoriteSo O O O O O O O O O O O O O I O I O O O . 88 X-ray Basal Spacing, d(001) for Na+- Hectorite Exchanged with Different Organic Molecules (at l and 2 CBC Exchange Level). . . . . 93 Desorption Coefficient for 2CEC Hectorite Intersalation Complexes. . . . . . . . . . . . . .105 Basal d(001) Spacing for (PADS)2'-Exchanged Hectorite Intersalates of Different Relative Humidity. . O O C C O C . C O C C O C O O 117 Calculated Rotational Correlation Times (nsec) of PADSZ' Doped into Zn(phen)3SOu- Intersalate of Hectorite Clay. . . . . . . . . . .119 Phase-Transfer Catalyzed Halide Displace- ment Reactions at 90°C with Ni(phen)§+/SO§'- Laponite Intersalates as Catalysts. . . . . . . . 125 Dependence of kobs on the Amount of Sodium Chloride Present. . . . . . . . . . . . . . . . . 127 Phase Transfer Catalyzed Halogen Exchange Reactions. . . . . . . . . . . . . . . . . . . . .130 -ix- TABLE 12 13 14 15 16 17 18 Kinetics Data for the Reaction of Butyl Bromide and Aqueous NaCl at 90°C. . . . . . . Pseudo First-Order Rate Constants (lozkobs (hr'1)) for the Reaction of Aryl and Alkyl Bromides with NaCl at 90°C. . . . . Pseudo First-Order Rate Constants (102kobs (hr'1)) for the Reaction of Alkyl Bromides with NaCl at 90°C. . . . . . . . . . Effect of Particle Size on Pseudo First- Order Rate Constants (lozkobs (hr’1)) for the Reaction of Alkyl Bromides with NaCl at 90°C. 0 O O O O I O O O I O O O O O O O O O 0 Reaction of Acetate Ion with Organic Substrate. . . . . . . . . . . . . . . . . . Phase Transfer Catalyzed Halogen Exchange Reactions Using Cetyl Pyridinium Hectorite Complex (72 hr at 90°C). . . . . . . . . . . Kinetics Data for the Reaction of Butyl and Octyl Bromide with Aqueous NaCl at 90°C. . . -X- PAGE 134 138 142 FIGURE LIST OF FIGURES PAGE Layer structure of smectite (hectorite) illustrating the intracrystalline space which contains the exchangeable cations. ox gens Si +, occasionally Al3+ Al3+, M92+, Fe3+ Hydroxyls. . . . . . . . . . . . . . . . . . . .3 The conceptual structure of 2:1 layer silicate mineral. The trioctahedral structure of talc. . . 4 The dioctahedral structure of pyrophyllite (2:1 layer silicate mineral). Both Figs. 2 and 3 adoped from Reference [8]. . . . . . . . . . 5 Schematic representation of homoionic smectite containin 1 CBC equivalent of M(phen)32? the 18 g intersalated phase which contains ca. 0.2 CEC equivalents of excess M(phen)3§5u salt; and the ~30 A intersalate which contains EE- 1 CBC eguivalents of excess M(bp)350a salt. x ’ represents the SORZ‘ ion. . . . . . . . . . . 17 Schematic representation of a triphase system. 0 O O O O O O O O O O O O O O O O O O O O 34 Reaction flask used to minimize creeping of finely divided intersalated catalyst: (A) Teflon-lined screw cap, (B) reaction chamber 4(diameter)><2 cm., (C) magnetic stirring bar. . . . . . . . . . . . . . . . . . . 62 Schematic representation of ethylene diamine, phenanthroline and some of its derivatives and 2,4,6-tripyridyl-l,3,S-triazine ligands. . . 72 001 X-ray reflections (Cu-kc) for an oriented film sample of Ni(phen)§+/MoO§’-hectorite intersalate; d-spacings are given in Angstrom units. . . . . . . . . . . . . . . . . . . . . . .76 ..xi_ FIGURE 9 10 11 12 13 14 15 16 17 18 19 PAGE 001 X-ray reflections (Cu-kc) for an oriented film sample of Ni(phen)§*/SOfi'-hectorite intersalate: d-spacings are given in Angstrom units. . . . . . . . . . . . . . . . . . . . . .78 The crystal structure projected along the b axis. Complex ions drawn with heavy lines are lying relatively in a higher part of the unit cell than those drawn with light lines. . . . . . . . . . . . . . . . . . .79 Adsorption isotherms (25°C) for [Ni(phen)3] SOH, Ni(phen)3]CrOa, and Nile Blue on hectorite in aqueous suspension. . . . . . . . .82 X-ray diffraction pattern of Fe(TPTz)§YSOE'- hectorite prepared by addition of 15 equiva- lent CEC of the complex to Na*-hectorite. . . . 90 X-ray diffraction patterns of 10,5,2, and 1 CBC equivalents of Ni(phen)§+/Cl;- hectorite. O O O O O O O O O O O O O O O O O O O 91 Possible orientations of the Fe(TPTZ)§+ ion on the surface of Na*-hectorite: A) The complex cation oriented perpendicular to the silicate layers. B) Tilted orienta- tion. C) The complex cation oriented parallel to the silicate layers. . . . . . . . .93 UV-visible spectra of [Fe(phen)3]SO, [Fe(3,4,7,8,-Mea-phen)]SOa, [Fe(4,7-diphenyl- phen)3]SOa in solution and NaI-hectorite intercalated with these compounds. . . . . . . .96 X-ray diffraction patterns of 2 CBC equi- valents of Nile blue and 3,3-(4,4'- biphenylene)-bis-(2,5-diphenyl-2H- tetrazolium chloride)-hectorite. . . . . . . . 101 Anion exchange isotherms for Ni(phen)§+/ CrOfi'-hectorite intersalate at 25°C. . . . . . 108 X-ray diffraction pattern for Ni(phen)§+/ SOfi’-hectorite intersalate solvated by water. . . . . . . . . . . . . . . . . . . . . 110 ESR spectrum at 20°C of the nitroxide spin probe, (PADS)2', in DMSO solvent. The vertical line on this spectrum represents the position of g==2.00. . . . . . . . . . . . 113 -xii- FIGURE 20 21 22 23 24 25 26 27 28 Rigid-limit x-band ESR spectrum of PADS; the field increases to the right. . . . . . . . .114 ESR spectra at 20°C of the nitroxide spin probe, (PADS)2' doped at the 2% exchange level into Zn(phen)2+/SO§'-hectorite at different relative gumid1ty. A free electron signal at g==2.0023 is shown. . . . . . 115 Schematic representation of the three- phase system, using smectite intersalate as a solid phase. . . . . . . . . . . . . . . . .122 Experimental set-up for triphase catalytic reactions. Corning N° 9826 Culture tube (20 Xl50 mm) containing organic, aqueous, and solid phase; a teflon coated magnetic stirring bar (lo/15 xs/js in. octagonal bar with pivot ring) is shown. . . , , , . . , , 123 Plot of ln (unreacted) butyl bromide in the organic phase as a function of time for the reaction of 2 mL of 0.5 M n-butyl bromide in toluene with 3.0 mL of an aqueous 3.3 M NaCl catalyzed by 0.069 meq of Ni(phen)§+/Sofi-ion pair in Laponite at 90°C. . .124 Plot of 102 kobs as a function of the amount of catalyst used. The catalyst and reaction conditions were similar to those described in Figure 24. . . . . . . . . . . . . . . . . . .131 Plot of ln(unreacted) l,8-dibromooctane in the organic phase as a function of time for the reaction of 2 mL of 0.5 M 1,8- dibromooctane in toluene with 3.0 mL of an aqueous 3.3 M NaCl catalyzed by 0.069 meq of Ni(phen)2+/SOZ’ ion pair in natural hectorite aE 90°C. . . . . . . . . . . . . . . . 133 Selectivity induced by the hectorite inter- salate catalyst on the pseudo first-order rate constants of Br-Cl exchange for normal and branched alkyl bromides with aqueous NaCl as a function of the number of carbon atoms in the alkyl chain. For the reaction conditions see Table 13, footnote a. . . . . . . 140 Schematic representation, showing edge view of the intersalated hectorite catalyst and the triphase catalyzed reaction zone. . . . . . 144 -xiii- FIGURE 29 30 PAGE Plot of ln(unreacted) l-bromooctane in the organic phase as a function of time for the reaction of 2 mL of 0.5 M 1-bromooctane in toluene with 10 mmol of sodium cyanide dissolved in 1 mL of water catalyzed by 0.069 meq of tricapryl methyl ammonium chloride ion pair in hectorite at 90°C. . . . . 153 Plot of 2h/d (d is basal spacing) versus n (order of reflection), using least square fitting technique. . . . . . . . . . . . . . . .157 -xiv- LIST OF SYMBOLS AND ABBREVIATIONS acac AEC aq Bathophenanthroline bp Bu CEC chel DMSO Et en ESR GOLOC. hr IR intercalate intersalate Angstrom acetylacetone anion exchange capacity aqueous phase 4,7-diphenyl-l,10- phenanthroline 2,2'-bipyridyl Butyl cation exchange capacity chelating ligand clay mineral for catalyst support dimethylsulfoxide ethyl group ethylene diammine electron spin resonance Gauss gas liquid chromotography hour infrared a clay sample which has been intercalated with a cation a clay sample which has been intercalated with a complex salt. intersalation obs Me pm or u mEq or meq mmol M M(chel)§+/XOi--hectorite The term intersalation has been suggested for the inter- calation of ionic salts in clays with neutral layer charge, e.g., kaolinite [41]. The term has been used in this dissertation to empha- size the presence of cation- anion pairs in smectite- clay interlayers. observed rate constant ligand general metal center methyl group milliliter micron milliequivalent millimoles molar concentration designation for a catalyst supported on hectorite n- normal OEt ethoxy group -OH hydroxyl group org organic phase ox oxalate P polymer for catalyst support (PADS)2- peroxyl amine disulfonate anion P'I‘C phase transfer catalysis Phen ortho-phenanthroline or 1,10-phenanthroline -xvi- r.h. 0+ or Q(+) TC TPTZ Vac. wt x,§ and y, y phenyl group alkyl- or aryl-group relative humidity cationic complex as catalyst triphase catalysis 2,4,6-Tripyridyl-5- Triazine rotational correlation (tumbling) time ultraviolet vacuum weight general anion to be displaced in a triphase catalyzed reaction any tetrahedral—ate anion (e.g. MoO4') -xvii- OBJECTIVES Layered silicate clays are ubiquitous constituents of the natural environment. These minerals are known to interact with numerous organic, inorganic and organometallic compounds to form complexes with a wide range of properties. The intercalation properties of the layered silicates provide numerous opportunities to modify or enhance the selectivity of a catalytic species. Because of the crystalline nature of layered silicate clays, one might expect to make use of these materials in a phase transfer catalysis (PTC). PTC is a relatively new field. As in classical phase transfer catalysis, reactions are carried out in an aqueous-organic two-phase system, but a third, immiscible solid phase (triphase catalysis) is also present which may be recovered at the end of the catalytic reaction by simple filtration. Here lies the potential importance of such a technique in industry, since both discontinuous processes with dispersed catalyst and continuous processes on a fixed-bed catalyst are feasible. Earlier studies have shown that intercalation compounds of hectorite, montmorillonite, and other smectite clay minerals can be tailored to function as selective liquid- solid phase transfer catalysts, particularly when the -xviii- spatial requirements of oriented reaction intermediates on the intracrystal surfaces differ for different substrates. Surface chemical equilibria can also be controlled and used to enhance the selectivity of intercalated clay catalysts. Since clay layers are negatively charged, the rational design of clays as phase transfer catalysts is generally restricted to intercalates in which the host structure is interlayered with catalytically active cations such as transition metal complexes. Anion intercalation is normally precluded on electrostatic grounds. Therefore, it could be of considerable interest to synthesize a new type of complex clay with anion intercalation that can then be used for triphase catalysis (TC). In the preSent investigation, several catalyst systems based on intercalation compounds of smectite clays have been made. For example, intercalation of complex salts 2- 4 M==Fe,Ni,Zn, produce well ordered systems in which the such as M(Phen)§+so , where Phen==1,10-Phenanthroline and pore sizes can be adjusted to adsorb substrates in a limited size range. Therefore, catalytic selectivity based on the size or selective adsorption of the substrate can be anticipated. -xix— CHAPTER I INTRODUCTION A. Structure and Properties of Layer Silicates The layered silicates, hectorite and montmorillonite described in this dissertation are smectite minerals, a class of naturally occurring clay minerals. The term "clay mineral" refers to a specific silicate with particle size less than 2 u and with a definite stoichiometry and crystalline structure. Smectites are composed of units made up of two silica tetrahedral sheets and a central octahedral sheet. All the tips of the tetrahedrons point in the same direction and toward the center of the unit. The tetrahedral and octahedral sheets are combined so that the tips of the tetrahedrons of each silica sheet and the hydroxyl groups of the octahedral sheet form a common layer. The layers are continuous in the a and b directions and are stacked one above the other in the c direction. In the stacking of the layer units, oxygen planes of each unit are adjacent to oxygen planes of the neighboring units, with the consequence that there is a very weak van der Waals interaction between them. The outstanding feature of -1- -2- the smectite structure is that water and other polar molecules, such as certain organic molecules, can enter between the unit layers, causing the lattice to expand in the c direction. The c-axis dimension of smectite is, therefore, not fixed but varies from about 9.6 A to substantially complete separation of the individual layers in some cases. Figure (l) is a diagrammatic sketch of the structure of smectite. In the 2:1 trimorphic or three sheet mineral of smectite there are two broadly divided structures: dioctahedral [aluminum silicate (Figure 2)] and trioctahedral [magnesium silicate (Figure 3)] minerals. These can be subdivided into groups in which the layer charge arises predominantly from isomorphous substitution in the octahedral layer and those with charge arising from tetrahedral layer substitution. Further subdivisions can be made on the basis of layer charge density. Thus the different minerals of the smectite group form a continuous multidimensional series. They tend to fall into discrete populations differentiated mainly by the source and density of their layer charge [1,2]. In the case of montmorillonite, isomorphous cation substitutions (commonly aluminum for silicon in the tetrahedral layer and magnesium and iron for aluminum in the octahedral layer) occur during formation and subsequent alteration. This leads to an imbalance of electrical charge which is compensated for by the presence Exchange Cations Silicate Sheet 9.65 Exchange Cations Figure 1 Layer structure of smectite (hectorite) illustrating the intracrystalline space which contains the exchangeable cations. O oxygens 0 814+, occasionally Al. O 1A13+p M92+' Fe3+ . Hydroxyls 3+ lnteriayer Region W ‘Hllflg 1D Si 0° ©OH Figure 2 The conceptual structure of 2:1 layer silicate mineral. The trioctahedral structure of talc. lnterlayer Region W Am 00 ass @OH Figure 3 The dioctahedral structure of pyrophyllite (2:1 layer silicate mineral). Both Figs. 2 and 3 adopted from Reference [8]. -6- of easily exchangeable cations on the mineral surface. A delicate balance exists between interlamellar attractive forces and the repulsive forces generated when the crystals are suspended in polar liquids. Often the repul- sive forces are sufficient to cause separation of the individual sheets, exposing the entire surface to the liquid. Interlamellar spacings in general depend upon the species of exchangeable cation present, the nature of the solvent, and whether or not any electrolytes are present. In aqueous suspensions interlamellar spacings as large as several hundred Angstroms have been observed. In fact, the silicate layers of Na+-montmorillonite in dilute aqueous suspension are completely dispersed (delaminated). As the concentration of the dispersion is increased, gelation occurs. The gelation phenomenon which occurs at concentration as low as 2 wt% clay, is believed to result from layer edge to face interactions which generate a "house-of-cards" structure [3]. An ideal montmorillonite may be defined as having the unit cell composition shown below [2] in which the superscripts (IV) and (VI) in the formula denote the respective tetrahedral and octahedral layer cations, and (M+) represents a univalent or equivalent compensating cation. . IV. VI. + [(518) (A13.33Mgoo67) 020(OH)4]0.67 M -7- The isomorphous substitution consists predominantly of Mg-for-Al in the octahedral layer, resulting in a net anionic charge of 0.67 units per unit cell. The Mg-for-Al substitution has been shown to result in incomplete neutralization of the negative charge on the apical oxygens and hydroxide groups coordinated to the magnesiums. The anionic charge of the aluminosilicate layers is neutralized by the intercalation of compensating cations and their coordinated water molecules. The montmoril- lonite crystal structure thus consists of superimposed aluminosilicate layers, each of which is interleaved with a "layer" of hydrated, exchangeable compensating cations. These cations can alternatively be described as occupying the interlamellar spaces or the regions between the basal surfaces of opposing silicate layers. Although the compensating cations are normally located in the regions adjacent to the points of anionic charge on the basal surfaces, small anhydrous cations, principally Li+ or H+ ions, are capable of migrating through the basal surface oxygen sheet to the neighborhood of the isomorphous substitution sites. The protons appear to associate with the octahedral hydroxyl groups, instead of forming hydroxyl groups by reacting with the incompletely neutralized oxygens [4]. Hectorite is the trioctahedral analogue of montmoril- lonite and contains predominant octahedral Li-for-Mg substitution. In hectorite, which typically exhibits a -3- layer charge of 0.67 e/SiBOZO' the cation exchange capacity on an anhydrous basis is 87 meq/100 9., about one-fifth the exchange capacity of sulfonated styrene- divinylbenzene resins. Since the average distance between exchange equivalents in the mineral is ~8.3 A [5], cations with cross sectional diameters greater than this value can saturate the interlamellar surfaces before 100% exchange is reached. Thus although the interlamellar surface is very large (~750 mz/g), the size of the exchanging ion can be a limiting factor in determining ion loading. . . . IV. . VI Hectorite. [(818) (Mgs.33L10.67) '020(OH)4]0.67 M+ The commercial filler, Laponitec)is a synthetic low-charge hectorite which, unlike the natural mineral, can be obtained with a negligible iron content. Another grade of LaponiteG>consists of a fluorohectorite in which the octahedral lattice hydroxyl groups have been replaced with fluoride ions. Hectorite lacks the potentially acidic, exposed aluminum ions that are partly responsible for the acidic character of the surfaces of the montmorillonites. However, both hectorite and montmorillonite contain other acidic species adsorbed on the basal surfaces and in the interlamellar spaces. Mortland et al. have reported [6] that residual water molecules remaining under vacuum in a base-saturated -9- montmorillonite are able to convert NH3 into NHZ. The extent of the transformation suggests that these residual water molecules have a degree of dissociation higher than usual. The NMR studies performed by D.E. Woessner and 3.8. Snowder [7] also support the proposed dissociation mechanism and show that the dissociation of interlayer water is 106-107 times higher than in liquid water. The surface acidity of dried minerals may exceed that of concentrated sulfuric acid [8]. This can result in the catalysis of many desirable, undesirable, or unexpected reactions when these materials are incorporated in organic media. This acidity is related in origin, and is often similar in strength, to that of the acidic alumina-silica and zeolite cracking catalysts used in the petrochemical industry [9]. The surfaces of the silicate minerals may contain both Lewis and Bronsted acidic species, although the Bronsted acidity is probably the more significant in catalysis of reactions by the minerals. The compensating cations of the montmorillonites, hectorites and related minerals can be exchanged with other hydrated inorganic cations, with metal complex 2+ 3)! variety of organic cations. Therefore a large active cations (for example, Fe(Phen) and with a wide surface area (700—800 mz/g) allows an enormous range of guest molecules to be intercalated. In the intercalation processes several binding mechanisms may operate [9-11]. -10— Due to the high negative surface charge (cation exchange capacity, CBC) and layer morphology, these minerals also serve as nature's own model for a Platy hydrophobic colloid of the "constant charge" type. The layer charge, indicated by a given structural formula, should only be regarded as an average over the whole crystal because this charge may vary (between certain limits) from layer to layer. The experimental evidence for this view and its implications for the colloid stability of montmoril- lonite suspensions have been described by Lagaly and co-workers [12,13]. Smectite clays possess a combination of cation exchange, intercalation and swelling properties which makes them unique. Their capacity as cation exchangers is fundamental to their intercalation and swelling properties. This distinguishes smectites from the micas and perphyllite/talc groups of minerals, which are not ion exchangers [14]. Because of the ability of the minerals to imbibe a variety of cations and neutral molecules, an almost limitless number of intercalates are possible. Because of the small particle size (<2 u) of smectite clays and their unusual intercalation properties, they afford an appreciable surface area for the adsorption and catalysis of organic molecules. Recently, intercalated clay catalysts have been expertly reviewed by T.J. Pinnavaia [5]. Indeed, the probable catalytic role of smectites has been recognized in several "natural" processes [15], including petroleum forming reactions [16,17], -11- chemical transformations in soils [10,18,19], and reactions related to chemical evolution [20-25]. Recent advances in the intercalation chemistry of smectite clays has rekindled interest in these minerals as catalysts or catalyst supports. By mediating the chemical and physical forces acting on interlayer reactants, one often can improve catalytic specificity relative to homogenous solution. A second new class of smectite intercalation compounds makes use of robust cations as molecular props or pillars between the silicate layers. Recently it has been found that silicon could be introduced into the interlamellar regions by ion exchange of triacetylacetonato silicon (IV) cations (Si(acac);) by in situ reaction of acetylacetone-solvated clays with SiCl4, or by formation of polychlorosiloxanes (-SiOC12-—)n [26]. The pillaring phenomenon leads to the formation of porous networks analogous to zeolites. Since pillared clays can have pore sizes larger than those of zeolites they offer a promising new means of facilitating the reactions of large molecules. By the intercalation of polynuclear hydroxy cations of some transition metals in smectite, new pillared clay catalysts with pore size larger than 10 A have been synthesized [27]. As the smectite interlayers are swollen beyond the dimensions of the coordination sphere of the aquated metal ion, the complex becomes solvent-separated from -12- the surface and begins to tumble rapidly. Also, under multilayer salvation conditions only the first one or two layers adjacent to the silicate surface exhibit significantly restricted motions [28]. Complementary ESR and NMR experiments [29-32], along with quasi-elastic neutron scattering studies [33], have provided incisive information on the interlayer dynamics of clay intercalates. A general picture of the interlayer environment has emerged from these studies. At low degrees of interlayer salvation (e.g., 1-3 water layers) the solvated exchange cations adopt oriented positions on the interlamellar surfaces. Though oriented, the solvated cations are in a dynamic state and undergo anisotropic rotations about specific molecular axes. Uncoordinated water molecules between the solvated cations are capable of translational diffusion within and between the "cages" defined by the solvated cations and the silicate layers [33]. Restricted motions and preferred orientations also have been observed for intercalated organic species [30,34]. B. Intercalation of Transition Metal Complex Salts in Smectite ("INTERSALATION") In contrast to the vast literature on clay-organic reactions [8,9,35], only a few references have been found describing interactions of chelated transition metal complexes on smectites. -13- Properties of the layered silicates change when different ions or chelated ions are adsorbed. Adsorption of ions or chelated metal ions by smectites may create changes in the chemical and physical properties of the clay. Adsorption of gases, surface area measurements, x-ray basal spacings, oxidation-reduction potentials as well as catalytic properties may be altered by exchange of complex metal ions onto the silicate surfaces. It was found [36] that related tris-ethylenediamine complexes Cr(en)§+, Co(en)§+, and Cu(en)§+ afford appreciable surface areas when they occupy the exchange sites of montmorillonite. The metal complex clays showed promise as chromatographic supports for separation of light hydrocarbons and nitrogen oxides. Mortland and Berkheiser [37] have demonstrated that Cu(Phen)§+ and Fe(phen)§+ have marked affinity for exchange on to the Na+-hectorite surface and the homoionic exchange forms of the mineral have an N2 surface area in the range 200-300 mz/g. This surface area is available for the adsorption of polar (H20) and larger nonpolar (benzene) molecules. In addition to finding some unusual redox properties of the mineral bound ions, they also found that Fe(bp)§+— and Cu(bp)§+-hectorite are capable of binding metal complex in excess of two times the cation exchange capacity of the mineral through intercalation of the bromide or sulfate salts. Similar observation by Uylterhaeven et a1. [38] showed that the -14- Cl- uptake on hectorite is accompanied with the adsorption of an excess [Ru(2,2'-bipy)3]2+ up to 1.55 meq/g. Adsorption of M(chelate)§+ by ion exchange and intersalation mechanisms has been studied in considerable detail [39,40]. Recently Pinnavaia et a1. [39] have shown that the binding of tris-bipyridyl metal complexes of the type M(bp)§+ (M==Fe2+, Cu2+, Ru2+) to hectorite surfaces to occur by two mechanisms: (1) replacement of Na+ ions in the native mineral by cation exchange up to its cation exchange capacity and (2) binding of metal complex beyond the CEC of the mineral through "intersalation" of metal complex salt. (The term intersalation has been suggested for the intercalation of ionic salts in clays with neutral layer charge, e.g., Kaolinite [41]. The term will be used in this dissertation to emphasize the presence of cation-anion pairs in smectite-clay interlayers.) The binding of excess salt is believed to result from a screening of the electrostatic charge of the clay by the complex cation, permitting the penetration of more of the complex cation and its anion into the interlamellar regions of the layered silicate. In this case apparently the large ligands are responsible for this screening since simple inorganic salts do not behave in this fashion. Complexes of the types M(bp)§+ or M(Phen)g+ are of special interest, in part, because they have a spherical shape and are among the more thermally stable complexes known. These complexes, due to their spherical shape, -15.. should be capable of propping the interlayers of the mineral to a greater extent, therefore one now expects the regions between the exchange ions to be available for adsorption or surface catalyzed reactions. In fact several attempts have been made [37,39,40,42-44] to intercalate relatively large cations to act as molecular props to keep the layers apart. The initial result by Mortland et a1. [37] demonstrated that Fe(Phen)§+ and Cu(Phen)§+ do in fact act as ~8 A molecular props in hectorite. They also concluded that van der Waals interactions appear to be responsible for the adsorption of the complex cations in excess of the exchange capacity. These van der Waals forces may also comprise an important driving force which causes a complete interlayer to be saturated before adsorption occurs in succeeding interlayers. The intersalation reactions are very dependent on the nature of the counter-anion with the binding decreasing in the 2.. order SO4 ,Br >ClO4 has a cross-sectional thickness of approximately 8 A 1>Cl'. A complex cation like M(Phen)§+ along the C3 axis [37] while its associated counter-anion 2- 2- 2- 2- 4 (S04 , Moo4 , CrO4 , etc.) has a cross sectional diameter of approximately 4 A. Thus, considering of the type XO the thickness of the silicate layer (9.6 A), the d(001) reflection of an air dried intersalated sample is expected to be'~30 A. The size of the pockets created by the cations is large enough to accommodate the anions and still retain enough space for free rotation in a highly swelled Figure 4 -15- Schematic representation of homoionic smectite containing 1 CBC equivalent of M(phen)§+; the 18 A intersalated phase which contains 93. 0.2 CEC equivalents of excess M(phen)3504 salt; and the ~30 A.intersalate which contains 22* 1 CBC equivalents of excess M(Phen)3SO4 salt. X2- represents the 80:- ion and L represents the Phenanthraline ligand. -17... Homoionic, I8A° [_ _ _ _¥ l— — — ——j Monoloyer Intersalation, l8A° — — —J 4» ML)? NHL): NHL)? NHL)? NHL)? MiL)? lb‘l — — —J Biloyer Intersalation.~30A° -18.. system. In the present work peroxylamine disulfonate radical (PADS) spin probe is doped into the anion interlayers to estimate the interlamellar mobility. The result showed that PADSZ- molecules have enough room to tumble fairly rapidly (i.e., 3 ><10-10 sec), regarding the size of molecule, therefore, take an appreciable solution-like properties (see Chapter III). The homoionic M(Phen)§:hectorites, which exhibit rational 18 A x—ray reflections, have been characterized [37] with regard to surface area, H20 absorption, types of water present, and orientation of the complex ion in the interlayer regions. Two solid state phases have been isolated with 18 i and ~30 3 spacings [40]. The 18 A phase results from the presence of a monolayer of complex cation, whereas the ~30 A phase contains two layers of complex cation. Schematic representations of the homoionic and intersalated phases are shown in Figure (4). In general, the surface areas of the new intersalated phases are low (~30 mZ/g), because nearly all of the internal surface area is covered by complex cation. The homoionic and intersalated phases are thermally stable to ~350°C. Above this temperature the 18 A homoionic and intersalated phases collapse to ~12 A, due to loss of Phen or bp ligands, however, the ~30 A intersalated phases behave quite differently. Thermolysis of a ~30 A intersalate containing M(Phen)§+ or M(bp)§+ and 80:- ions affords anew high-temperature phase with an -19- (001) spacing near 18 A. These high-temperature phases are thermally stable to at least 550°C, and they exhibit N2 surface areas up to 360 mZ/g. Chemical and spectroscopic studies indicate that the new high temperature phases consist of silicate sheets separated by two layers of carbon. Quantitative kinetics studies [45] have shown, rather remarkably, that these "graphitic clays" are so stable that the silicate layers can be dehydroxylated (640°C) without collapse of the interlayers. The rates of dehydroxylation are appreciably larger than those for ordinary smectite suggesting that the expanded interlayers are sufficiently porous to facilitate migration of water out of the structure. C. The Principles of Phase Transfer Catalysis Reaction between two substances located in different phases of a mixture is often inhibited because of the inability of reagents to come together. The classical solution to this problem, and by far the one most frequently used in the laboratory, is simply the use of a solvent which can dissolve both reagents. Use of solvents is not always convenient, and on an industrial scale it frequently is expensive. The technique of "phase transfer catalysis" provides a method which avoids the use of solvents [46-49]. Basically, all phase transfer catalyzed reactions involve at least two steps: (1) transfer of one reagent from its "normal" phase into the -20- second phase; and (2) reaction of the transferred reagent with the nontransferred reagent. Jarrause [50], as early as 1951, observed that the quaternary ammonium salt benzyltriethylammonium chloride markedly accelerated two—phase reaction of benzyl chloride with cyclohexanol (Reaction 1) and the two phase alkylation of phenylacetonitrile with benzyl chloride or ethyl chloride (Reaction 2). [::IIOH C6H5CH2N+Et3C1-(aq) [::]r'O-CH2-C6H5 C6H5-CH2C1 + + NaOH <+ org. org. + NaCl + H20 (1) fzns CHZCN CH-CN + _ C6H5CH2N Et3Cl (aq) + c2H5c1-FNaaH s; -+Nac1-+HZO (2) org. In addition to this work, a number of other publications and patents appeared during the period 1950-1965 in which quaternary ammonium or phosphonium salts were used as catalysts for two-phase reactions [51-66], although in these instances either the general nature of phase transfer catalysis was apparently missed, or the catalytic activity was believed to involve only surfactant properties of the quaternary salts. Makasza and ca-workers [67] later -21- reexamined the two-phase alkylation technique in great detail and published their findings in a number of papers, greatly expanding the understanding and utility of this alkylation method. Gibson [68] in two-phase permanganate oxidations, Makasza [67] with alkylation reactions, Hennis [69] with carboxylate displacement reactions, Bréndstrdm [70] with alkylation reactions, and Starks [47] with a variety of reactions, each recognized many of the elements of phase transfer catalysis by quaternary ammonium salts. The name "phase transfer catalysis" was first applied to the technique in patents [47] and in the journals [46], after which detailed evidence for the mechanistic pathway illustrated in (Scheme I) [48] was adduced. The recognition of crown ethers as phase transfer catalysts, in both liquid-liquid and liquid-solid reactions, was first published by Liotta and co-workers [71]. Since these earlier patents and papers, the literature on phase transfer catalysis has grown rapidly, so that now there are more than 1000 publications. Phase transfer catalysts are onium salts, usually quaternary ammonium or phosphonium species, crown ethers, cryptands, or linear polyethers. The mechanism of catalysis is thought to vary to some extent with the type of system involved [72], but the simplest case arises with SN2 displacement reactions, typified by the reaction of a nucleophile, Y, in an aqueous solution with an alkyl halide, RX, in an organic phase [46] (Scheme I). +- +- QY + Rx ———————+ RY + QX (organic) ——H ________ H————— +- +- +- +- QY + MX .e——————- MY + QX (aqueous) (Scheme I) +— +- MY is the alkali metal salt of the nucleophile and QX is the onium salt catalyst. The latter becomes involved in a fast equilibrium in the aqueous phase to generate 5; ion pairs and because of the relatively lipophilic character of O. the anion, Y, is effectively 'ferried' into the organic phase. Here intimate contact with the alkyl halide is established and displacement readily occurs. Finally, the displaced anion, A, is 'ferried' back to the aqueous phase as the ion pair, OR, and the cyclic series of events is completed. The most common example for which a large amount of data are available is simple cyanide displacement an alkyl chlorides or bromides (Reaction 3) R-Cl + NaCN 4 R-CN + NaCl (3) org, aq. org. aq. Simply heating and stirring a two-phase mixture of l-chloroctane with aqueous sodium cyanide leads to essentially zero yield of l-cyanooctane even after several days of reaction time. However, if a small amount of an appropriate quaternary ammonium salt is -23.. added, then rapid formation of l-cyanooctane is observed in essentially 100% yield after 1 or 2 hr. Quaternary cation, Q+==R4N+, selected for its high compatibility with the organic phase, transfers cyanide ion into the organic phase as Q+CN-, which then undergoes reaction with chlorooctane to produce cyanooctane. Coproduced Q+Cl- is rapidly reconverted to Q+CN-, either in the aqueous phase or at the aqueous-organic interface, by anion exchange with sodium cyanide from the aqueous phase. The concept of phase transfer catalysis is not limited to anion transfer systems, but is much more general, so that in principle one could also transfer cations, free radicals, whole molecules, or even energy (in a chemical form). Very little work has been reported on such systems, although we may certainly look forward to it in the future. PTC is not only an academic curiosity. To date there exist about 20 processes which produce polymers, pharmaceuticals, and intermediates for dyestuffs and agrochemicals, and their outputs are 50-500 tons/y each. Polycarbonates: In 1958, Bayer made polycarbonates by PTC-before PTC was "officially” discovered [73]. The reaction is between carbonyl chloride and the sodium salt of bisphenol -24- CH - CH — 3 3 I l ° + - - + N Na 0 —C O Na + COCl2 —* -r- O —C 0C"- CH3 “- CH3 in + 2NaCl (4) Penicillin esterification: The original commercial penicillin was benzylpenicillin: IL-—:;I:T,-;]:E: with R= -C6H5CH2CO-— chemically modified penicillins, developed in the 1960's, had different R side chains; in particular, Beecham developed the successful ampicillin (R= CGH 5CH(NH2)CO--). Penicillins are acids and are usually marketed as their salts. Recently, two penicillins based on ampicillin in which the carboxyl group is esterified have been commercialized [74,75]. The idea is that the penicillin will no longer appear as a zwitterion when applied. Instead, the new lipophilic form will diffuse into the tissues, penetrate pus, and will then hydrolyze to ampicillin at a wide range of sites. By use of a compound that hydrolyzes slowly, blood plasma levels of the antibiotic can be maintained with fewer doses per day. -25- To achieve these aims, a labile ester is required that will not hydrolyze too quickly but will be less stable than the B-lactam ring in penicillin. Natural penicillins and the precursors of semisyn- thetic penicillins are manufactured by fermentation of a carbohydrate substrate with penicillium chrysogenum in the presence of phenylacetic acid (for penicillin G) or phenoxyacetic acid (for penicillin V) [76]. The penicillin is harvested by extraction with butyl acetate, but there are losses of at least 10% in waste streams. Further extractions with butyl acetate are uneconomic because unreacted phenylacetic acid or phenoxyacetic acid are simultaneously extracted at an unacceptable level. The penicillin can, however, be recovered by ion-pair extraction [77]. This is neither a chemical reaction nor a catalytic process, but it uses phase- transfer catalysts and the phase-transfer technique. Additionally, phase transfer catalysis may function not only through liquid-liquid systems, but also with liquid-gas, liquid-solid, solid-gas, and presumably solid-solid systems. For example, in the liquid-gas combination, hydrogenation of olefins with the Wilkinson catalyst [78] is a system in which a gas phase reagent is transferred into the liquid phase and activated for reaction (¢3P)3Ru(CO) R-CH---=CH2 +112 4. RCHZ-CH3 (5) org gas with the organic reagent. -26- Indeed, perhaps the oldest of all phase transfer reactions is the transfer and activation of oxygen by hemoglobin from the air in our lungs into the blood and throughout our bodies to where energy production is necessary. All phase transfer catalysts do not behave equally well for all kinds of reactions; some are often sensitive to such factors as: (i) the chain length and types of alkyl groups attached to the quaternary nitrogen or phosphorus atom; (ii) the presence of strong base or acid in the aqueous phase; (iii) the concentration of the inorganic reagent in the aqueous phase; and (iv) the presence of certain ions, such as iodide, perchlorate, or toluene sulfonate. Most anions prefer to reside in an aqueous phase rather than an organic phase, even a highly polar one, because of the favorable thermodynamic effect afforded by anion hydration. This effect results from spreading the electronic charge over the greater volume of the hydrated species, and is therefore dependent on the charge to volume ratio of the anion. To be highly effective as phase transfer catalysts for two-phase displacement reactions, the catalyst cation-anion pair needs to be strongly partitioned into the organic phase. The optimum distribution of catalyst cation in anion transfer catalysis is that which allows the rate of the organic phase reaction to equal the rate of catalyst regeneration. For example, in the cyanide displacement reaction sequence: -27- k R-Cl-+Q+CN ——9—+ R-CN-+Q+Cl organic phase reaction (6) k Q+Cl -+NaCN ——£-+ Q+CN -+NaCl catalyst regeneration (7) Reactions (6) and (7) will have maximum rate when kO(RCl)(QCN) = kr(QCl)(NaCN) (8) If we assume that catalyst regeneration takes place only in the aqueous phase, then a definite aqueous phase concentration of the catalyst cation would be required (i.e., in contrast to the situation where all the catalyst cation is in the organic phase and catalyst regeneration occurs by anion transfer across the phase boundary). Values of kr typical for anion exchange reactions in aqueous media are 5-10 orders of magnitude greater than k0, so that even if catalyst regeneration occurred only in the aqueous phase, only a few hundredths of a percent of the catalyst need be in the aqueous phase to satisfy or exceed the criterion represented by equation (8). Thus, for most phase transfer catalyzed reactions involving anion transfer from aqueous to organic phases we wish to select a catalyst which will be soluble in the organic phase. In fact, the distribution of catalyst cations between organic and aqueous phases depends not only on the organic structure of the cation, but also on the nature -23- of the anion associated with the cation, the polarity of the organic phase, the concentration of inorganic salt in the aqueous phase, the presence of solvating compounds, and possibly other factors. The kind of anion associated with the catalyst cation has enormous influence on the extent to which a given cation-anion pair is extracted from the aqueous to the organic phase. This means, for example, that a catalyst cation which easily transfers iodide anions from the aqueous to the organic phase might be totally inadequate for the transfer of chloride ions. Two principle characteristics of the anion most influence its tendency to increase or decrease the ability of a catalyst cation to transfer. First, anions are hydrated to different extents, depending mostly on the charge-to-volume ratio of the anion, and the more the anion is hydrated, the more strongly it will be attracted to the aqueous phase and the more difficult it will be to transfer. This water of hydration may or may not accompany the anion when it is transferred into the organic phase, although most measurements [48,79,80] indicate that the water is transferred. Increasing the concentration of inorganic salts in the aqueous phase tends to salt out inorganic salts, pushing them into the organic phase. Increasing the inorganic salts concentration also ties up additional water of hydration, reducing the amount of water -29- available for anion hydration, providing in some cases for easier transfer of the anion into the organic phase. Usually, the best phase transfer catalysis conditions are realized when the aqueous phase is saturated with the inorganic reagent. One also expects factors such as polarity of the organic phase and structure of the catalyst cation to influence selectivity of anion partitioning into the organic phase. In reactions where the phase transfer step, rather than the organic phase reaction step, is rate limiting, the concentration of inorganic reagent in the aqueous phase not only affects the degree of anion hydration, as noted above, but will also directly affect the rate of the phase transfer step and therefore the rate of the overall catalysis sequence. This effect may be positive or negative, however, depending on the principal site of anion transfer for the particular system being used. When the concentrations of anions are not maintained constant, or approximately so, then the rate equation must take these variations into account. We must first consider the mechanism by which the anion is transferred into the organic phase. The most obvious possibilities appear to be: (a) Simple ion exchange across the interface (liquid ion exchange) [81] -30- k (Oc1)-+CN” e=§e (QCN)-+Cl- (9) org aq org aq from which it may be shown that k Q [(CN-) ] = a 0 as [(QCN)or ] - - (10) g ka[(CN )aq]-+[(c1 )aq] (b) Transfer of 0+ back and forth across the inter- face with anion exchange in the aqueous phase: + +CN’ #1:,— CN (11) Qaq aq (Q )aq kTT _—___\ (OCN)aq e——— (QCNlorg (12) .112.» (OC1)aq e——— (OC1)org (13) + -+c1' sfiée ( c1) (14) Qaq aq Q aq From this sequence of equilibria it may be shown that pkkarQo(CN )a [(OCN) ] = - - org E;ky-+(l+6kp)ky(€l )-+(l+pkfl)kx(CN_7aq (15) where p==Vb (volume of organic phase)/VA (volume of the aqueous phase). (a) Transfer of the inorganic salts into the organic phase for exchange: -31- (NaCN)aq ¢‘(NaCN)Org (16) (NaCN)Org-+(QC1)org =2i(QCN)org-t-(NaCl)org (l7) (NaCl)org $‘(NaC1)aq (18) (d) Transfer of anion at the organic-crystalline solid interface: (QCl)org+(NaCN)solid as (OCN)org + (NaC1)solid' (e) Formation of micelles in the aqueous phase and transfer of anion across the micelle interface: +CN’ s (QCN) +c1 (19) (QCl)micelle aq micelle aq (QCN) =‘(QCN)Or (20) micelle 9 Little is presently known about how important each of these mechanisms is to anion transfer, but it is likely that each will be favored in particular experimental circumstances. Mechanism (a) would be expected to be favored by using a very large catalyst cation so that essentially none of it is partitioned into the aqueous phase. Mechanism (b) would be favored by use of a catalyst cation that partitioned into both phases. Tetrabutylammonium salts are a common example. Mechanism (c) will be favored by use of an organic solvent in which the inorganic salt has appreciable solubility, such -32- as methanol or acetonitrile, along with an aqueous phase saturated with the inorganic salt (or even dry inorganic salt). An anion having appreciable organic structure will also facilitate this route to anion exchange. Mechanism (d) will be expected when the catalyst can interact directly on the surface of the solid, such as with crown ethers or cryptates, which can exchange not only anions, but also entire salt molecules. Mechanism (e) will be favored by use of catalyst cations that are good micelle-forming agents, such as RN+(CH3)3 salts where R is a long chain such as C16H33-, and by the use of dilute inorganic solutions. Mechanism (e) represents a transition zone between true phase transfer catalysis and reactions that are catalyzed by micelles [82]. Most reported work deals with kinetics that probably have either mechanism (a) or (b) as the predominant mode of anion transfer, and the kinetics equations to be developed are based on these. Kinetics for situations in which mechanisms (c)-(e) are likely to predominate have not been well studied at this time and therefore will be left for the future. D. Development of Supported Phase Transfer Catalysts 1. Concept of triphase catalysis (TC) Triphase catalysis (TC) has recently been introduced [83] as a unique form of heterogeneous catalysis in which -33- the catalyst and each of a pair of reactants are located in separate phases. Based on this concept, new synthetic methods have been developed for aqueous phase-organic phase reactions using a solid phase catalyst. Although it is only at an early stage of development, TC shows considerable potential for practical use. Regen has shown [84] that the development of a technique which used insoluble catalysts to accelerate aqueous-organic phase reactions would not only be an interesting possibility but could also provide the basis for synthetic methods competitive with or even superior to existing ones. Furthermore, recently he critically reviewed the triphase catalysis. Figure (5) illustrates the general features of a triphase catalyst system. Two immediate advantages would be (1) simplified product work-up and (2) easy and quantitative catalyst recovery. From the standpoint of industrial applications, this concept is a very attractive one due to low energy and capital requirements for processing. In addition, such a technique would be ideally suited for continuous flow methods. 2. Relationship of triphase to phase-transfer catalysis The term, "phase-transfer catalysis", was originally introduced by Starks to characterize a process in which: "reaction is brought about by the use of small quantities of an agent which transfers one reactant across the -34- >m xm >m + ..x xx + ..> 030m I X ..> r 0.2495 mDOwDO< L m_m>.._<._.vermiculite > biotite:>pyrophyllite. This observation further argues for the reactivity of the edge surface towards polyanions since such minerals as biotite and pyrophyllite have few, -50- if any, exchangeable cations with which the functional groups of the polymer may interact. The failure of synthetic polyanions to form interlayer complexes with expanding 2:1 type layer silicates has led to the widely accepted view that the active sites on clay surfaces, in general, are mainly located at the crystal edges and identifiable with Al-OH or its protonated form, Al-Oflg. Although such sites clearly play an important part in the adsorption by kaolinite-type minerals they may not be of great significance to polyanion uptake by 2:1 type layer silicates. In the present work, an attempt is made to demonstrate some of the unique properties of anion intercalation within the interlamellar regions of smectite clay minerals. As previously described in Section B, a novel class of intercalates in which two layers of M(chelate);+ and a layer of counteranions are bound in the interlamellar region of the layered silicate. The anions present between interlayers can actually be replaced by other anions, in fact this is a first indication that layered silicates can act as anion exchangers with anion exchange capacity (AEC)'~50 meq/100 g of intersalate (see Chapter III for detail). The anion exchange takes place without significant loss of the complex cation, which remains immobilized on the surface. It is expected that layered silicate -51.. intersalates to be of considerable importance in part, because they convert a mineral which is normally a cation exchanger into an anion exchanger which might be utilized as sinks for the deactivation of hazardous and pollutants anion. In addition, they provide an opportunity to investigate the effects of intercalation on the catalytic activity and selectivity of anions. The positioning of the anions between two layers of metal complex cations can influence their salvation properties, therefore the interlayers may be sufficiently lipophilic to allow penetration Of the interlayers by organic reagents while at the same time allowing the anions to be replaced by exchange with salt from an aqueous phase. Therefore one of the objectives of this research is to elucidate the catalytic properties of intersalates, particularly application of the intersalates in a triphase catalyst system. CHAPTER II EXPERIMENTAL A. Materials Unless stated otherwise, all reagents were obtained commercially and used without further purification. 1. Natural Hectorite Naturally occurring sodium hectorite (Bl-26) with a particle size <2 um was obtained from the Baroid Division of NL Industries in the precentrifuged and spray-dried form. The idealized anhydrous unit-cell formula of hectorite is Na [L ](Si 0.66 10.66M95.34 8.00)0 the experimentally determined cation-exchange capacity is 73 meq/100 g of air-dried clay [138]. 2. Sodium Montmorrillonite (Wyoming1 This mineral was obtained from the Source Clay Mineral Repository. The mineral was allowed to sediment and then was saturated with Na+ ions by adding an excess of sodium chloride. The clay was centrifuged and dialyzed until free of excess sodium chloride and then freeze dried. Ion exchange of Na+ in the native mineral with 1.0 M Cu(NO3)2 and subsequent analysis of the minerals -52- -53- for copper indicated the cation exchange capacity (CEC) of 80 meq/100 g. 3. Synthetic Hectorite (Laponite-RD) A synthetic hectorite with the structural formula of Na0.22LiO.l4[MgS.64Li0.36](518.00)020(OH)4 was obtained in powder form from Laporte Company. The cation exchange capacity of this hectorite is 55 meq/100 g and the size of average particle is ~10 A thick, with an average diameter of 200 A. 4. Fluoro-hectorite This synthetic clay was donated by Corning Glass Works. The compound was provided in (11.2 g of fluoro- hectorite/100 ml suspension). The structural formula was given as Li1.4[Mg3.4Lil.6]($18.00)020(F)4. The particle size is >>2 um and the CEC is 193 meq/100 g. All of the hectorites and montmorillonite were freeze dried and stored in a desiccator over anhydrous CaClZ. 5. Reagents and Solvents Ethylenediamine (en); 2,2'-bipyridyl(bP);l,10- phenanthroline monohydrate (Phen); 4,7-diphenyl-l,10- phenanthroline or Bathophenanthroline (BathPhen); 3,4,7,8-tetramethy1-l,lO-phenanthroline (TM-Phen), 2,4,6-tri-(Z-pyridyl-s-triazine (TPTZ) were purchased from Aldrich Chemical Company and G. Fredrick Smith Chemical Company, respectively. TPTZ was purified by recrystallization from petroleum ether. -54- All alkyl and aryl halides as well as internal standards and solvents used in this study were purchased from Aldrich Chemical Company except l-bromo-3,5,5- trimethylhexane which was obtained from K & K Laboratories. Tricaprylmethyl-ammonium chloride (poly Sciences), nile blue (MCB Manufacturing Chemists, Inc.), 3,3- (4,4'-biphenylene)-bis-[2,5-diphenyl-2H-tetrazolium chloride] (Morton Thiokol, Inc., Alfa) cetylpyridinium chloride and polybrene CH3 CH3 '+ '+ .— [-—N -(CH2)6-N-(CH2)3-]2Br X CH3 CH3 (Aldrich Chemical Co.) were stored in a vacuum dry desiccator over CaClz. Dowex exchange resin 2 X3, 100-200 mesh, Cl-form was a gift from the Dow Chemical Company. Silane coupling agent Z-6076 [ClCH ] was purchased CH CHZSi(OCH 2 2 3’3 from Dow Corning Corporation. The Spin probe peroxylamine disulfonate, potasium salt, (PADS) (Aldrich Chemical Company) was kept under N2 atmosphere to prevent possible decomposition. All solvents and inorganic reagents used were ACS reagent grade. B. Synthesis 1. M(Phen)3§9 and M(bP)3§_9_4 Complexes (M==Ni,Fe,Zn) 4 -55- Metal tris(1,10-phenanthroline) and tris(2,2'- bipyridyl) complexes were prepared as sulfate salts by reaction of the free ligand and an aqueous solution of the metal sulfate according to literature methods [139, 140]. The crystals were extracted with hot benzene to remove any adsorbed free ligand. 2. Ni(Phen)3£)_ (XOA =MoO,,CrO4,WO,,)_ 4 An anion exchange method was adopted for the preparation of these salts from the chloride salt. An amount of anion exchange resin (Dowex 2><3, 100-200 mesh, Cl-form) was saturated by diluted HCl solution and then washed several times with deionized water. After filtration the resin was transferred to a chromatographic column where the chlorine was exchanged for so:— by treatment with concentrated (l M) Na SO solution. The 2 4 eluants were checked with silver nitrate for the absence of C1-. The direct exchange between the Cl--form of the 2- 2- resin and MoO4 , CrO4 or W02- anions was not practical; because the exchange process is slow and requires a large excess of XOi-. However, the SOi--form of the resin could be readily exchanged with the appropriate X04- anion. BaCl2 was used to check for the absence of sulfate. A 0.05 M solution of tris-(l,lO-phenanthroline)-nickel (II) chloride heptahydrate prepared by literature method [140] was passed through the exchange column. The flow rate was about 1 mL/min and the ratio of resin to salt was 2:1 (meq:meq). The X02- complex salts were -55- recrystallized from hot deionized water. Despite the tediousness of this process the yield was almost quantita- tive. 3. Fe(L)3§Q4, [L==1,lO-orthophenanthroline;3,4,7,8- tetramethyl-l,10:phenanthroline, and 4-7-d1pheny1- 1,10-phenanthroline (Bathopheanthroline)] Fe(L)3SO4 complexes were prepared according to previously reported procedures [141-144]. The absorption spectra of the complexes were in agreement with those reported in the literature. 4 . Fe (TPTZ) 2 (C10412 , 4H20,Fe (TPTZ) 2&4_ Bis 2,4,6-tri(2-pyridyl)-s-triazine) iron (II) perchlorate tetrahydrate and sulfate were synthesized by the method of Watton et a1. without modification [145-146]. Care must be taken in handling perchlorate complex salts as these salts are potentially explosive 2 The electronic spectra agreed well with literature values [141]. 5. Ni(en)3§9 and Ni(NH316§94_ 4 The tris(ethylenediamine) nickel (II) sulfate, was synthesized according to the procedure of George and Wendlandt [147]. Hexammine nickel (II) sulfate was prepared by the addition of an excess of concentrated aqueous NH3 to a saturated solution of NiSO4-xH20 The precipitation of small violet-colored crystals was accomplished by the addition of 95 percent ethanol. The -57- hexammine complex was washed with ethanol and ether and then air dried. 6. Fe(bP)3(ClO412 and Fe(bP)3§Q4_ Fe(bP)3(ClO4)2 was prepared according to the methods of Burstall and Nyholm [148]. Tris(2,2'-bipyridyl) 4-7H20 was prepared by adding an excess of the amine to a solution iron (II) sulfate heptahydrate, Fe(bP)BSO of FeSO4-7H20. The solution was heated one hour at 90°C. The solution was evaporated to a low volume and the dark red paste extracted several times with hot benzene. The remaining material was dissolved in water and crystallized from aqueous solution. Crystals were collected in a bfichner funnel and dried over P40 . 10 29 Potassium trioxalatochrominate (III) trihydrate was 7. M3Cr(OX)3-3H synthesized by a known procedure [149]. 8. Elemental Analysis and Storage Elemental analyses were performed either by Gailbraith Laboratories, Inc., Knoxville, TN or Spang Microanalytical Laboratory, Eagle Harbor, MI. All complexes and catalysts were stored in a desiccator over anhydrous CaCl2 or P4010. C. Preparation of Intercalates Intercalation reaction of all complex cations were performed under similar conditions, since all complexes -53- were quite stable and also soluble in water. In a typical experiment a 1.0 wt% aqueous suspension (e.g. 0.1 g of the mineral in 10 m1 H20) was added to a vigorously stirred aqueous solution of the complex cation containing ~1.2 meq/meq of clay. The reaction mixture was stirred over an hour. After an equilibration time of 24 h the clay was repeatedly washed with deionized water and then collected by centrifugation. The resulting homoionic metal-complex exchange forms of the minerals were either air-dried or freeze-dried, as desired, and stored in a desiccator over anhydrous CaCl . 2 D. Preparation of Intersalates Identical procedures were used for the intersalation reaction of all M(Chel):+ (x==2,3) and organic salts. In a typical intersalation reaction 0.10 g of freeze-dried hectorite with a CEC of 73 meq/100 g was slurried overnight in 10 mL deionized water (1.0 wt%) and sonified for about half a minute before being added to a vigorously stirred aqueous solution of the complex salt (0.146 meq) at room temperature. Typically, the concentration of the salt solution was 2 mg/mL (~0.002 M). After an equilibration time of 24 hours, the products were collected by centrifugation and washed with a minimum amount of deionized water. The -59.. intersalates then were either freeze dried or spread on a glass surface for air drying. Air dried samples are more convenient for weighing and transferring to a reaction flask. (For film sample preparation see Section 2 of Physical Methods.) The intersalation of the complex salts in smectite is highly dependent on the initial state of dispersion of the clay. To ensure optimum dispersion of the platelets, the 1.0 wt% mineral suspension was sonified by use of a cell disruptor model W185 from Ultrasonic Inc. immediately prior to its addition to the aqueous complex salt solution. B. Desorption Coefficient Measurement Procedures similar to that described for the Fe(Phen)§+/SOi--hectorite intersalate were followed for all desorption coefficient measurements. A 0.04 g sample of Fe(Phen)§+/SOi--hectorite containing one cation exchange equivalent of Fe(Phen);+ and one equivalent of Fe(Phen)§+/SOZ- ion pairs was stirred in 10.00 mL of an appropriate solvent (ethanol or water) at 20°C in a thermostat. After 48 hours of equilibration time the suspensions were filtered by use of a milipour filter (0.02 u) and the filtrates were analyzed by atomic absorption (Perkin Elmer Atomic Absorption Spectrometer 560). The instrument was calibrated against Fisher standards. -60- F. Ion Exchange-Adsorption Isotherms Ion exchange isotherms were developed for the adsorption of Ni(Phen)§+SO4, Ni(Phen)3CrO and Nile blue 4 on Na-hectorite by adding the apprOpriate amount of complex salts aqueous standard solution to a known weight (0.04 g) of Na-hectorite and bringing the suspensions volume to 25.00 ml. The suspensions were allowed to equilibrate for 48 hours and then centrifuged. The supernatant solution was analyzed for the concentration of complex salts by UV- VIS spectrophotometery. Standard curve of absorbance versus concentration was constructed with a known concentration of the complex salts. The absorption maximum was obtained at 520, 369, and 633 nm for Ni(Phen)3SO , Ni(Phen)3CrO4 and 4 Nile blue respectively. Adsorption isotherms were reproducible within experimental error as determined by checking certain points on the isotherms. G. Anion Exchange Adsorption Isotherm Studies Anion exchange isotherms were developed for the 3- 2- 2- 4 ' M°°4 4 and MeCOO on Ni(Phen)§+/SOi--hectorite. An appropriate adsorption of PO , Cr(OX)§-, wo , ocoO, CI amount of salt solution was added to a known weight (400 mg) of complex-hectorite and the suspension was diluted to a volume of 50.00 mL. The suspensions were allowed to equilibrate for 48 hours and then centrifuged. 2- The supernatant was analyzed for the concentration of CrO4 -61.. anion by spectrophotometry. The adsorption maximum was obtained at 369 nm for CrOi- anion and the amount of anion adsorbed by the complex-hectorite was determined by CrOi- desorption from the intersalates. Selected points on the isotherms were checked for reproducibility. In general, the results were reproduced to within 110%. H. Kinetics of Biphase and Triphase-Catalyzed Displacement Reactions Procedures similar to that described for the conversion of alkyl bromides to alkyl chlorides were followed for all displacement reactions described in this dissertation. The reactions of alkyl bromides with NaCl under triphase. conditions were carried out as follows. Air dried Ni(phen)§+/soi'-hectorite (0.1566 g 0.069 meq of Ni(Phen)3SO4 as catalyst) was dispersed in 3.0 mL aqueous 3.3 M NaCl in a 15>(150 mm Pyrex culture tube fitted with a teflon-lined screw cap and magnetic stirring bar. A specially designed flat-bottomed reaction flask (see Fig. 6) was sometimes substituted for the culture tube. This design was especially effective in minimizing "creeping" of the finely divided mineral-supported catalyst up the walls fo the flask during the course of the reaction. The mixture was stirred for a few hours until a homogeneous suspension was obtained. To the suspension was added 1.0 mmol of the appropriate alkyl bromide in 2 mL toluene. The tubes were sealed with a -62... Figure 6 Reaction flask used to minimize creeping of finely divided intersalated catalyst: (A) Teflon-lined screw cap, (B) reaction chamber 4(diameter)><2 cm., (C) magnetic stirring bar. -63- teflon-lined screw cap and vigorously shaken for a minute. The kinetics experiment was then started by placing the tube in an oil bath maintained at the desired temperature. The reaction was followed by withdrawing l-uL samples of the organic phase at different times (no less than 30 minute intervals) and monitoring the disappearance of the reactant by GLC. For sampling, the tube was removed from the oil bath, shaken, quickly cooled to nearly room temperature, opened, resealed and returned to the bath. The overall process took less than 1 min. The temperature of the oil bath used for the kinetic experiments was controlled to i0.5°C with the aid of a temperature controller (4831 automatic temperature controller, model 4831, Parr Instrument Co.) attached to an Iron-Constantan thermocouple. Product mixtures were analyzed by GLC on a Hewlett-Packard model 5880A with a flame ionization detector and a capillary 12.5 M XO.2 mm crosslinked dimethyl silicone column. First-order rate constants were determined by fitting the experimental data to the best linear curve by a least-squares method. Similar methodology was utilized in studying the reaction under biphase (liquid-liquid) conditions except that appropriate amounts of complex salts were used in place of the hectorite intersalates. -54- I. Reaction of Benzyl Chloride with Salts of the Acetate Ion Benzyl chloride was purified by distillation under reduced pressure before use. Air dried Ni(Phen)§+/SOi-- hectorite (0.1566 g, 0.069 meq) was carefully placed in the bottom of a 50-mL culture table. Sodium acetate (1.1 mmol) in 3 mL deionized water was slowly added to the tube, and then 2.0 mL toluene was added. A sample of benzyl chloride (1 mmol) was then added to the organic layer via syringe. The tube was then sealed with a screw cap, shaken vigorously for a minute, immersed in an oil bath maintained at 90°C for 48 hr, withdrawn, and cooled to room temperature. The concentrations ratios were measured by GLC analysis. The product benzylacetate was characterized by GLC retention time. Only two peaks (benzyl chloride and benzylacetate) were detected. J. Modification of the Intersalate with Silane Coupling Agent A 1 mL sample of chloropropyltrimethoxysilane (26076, Dow Corning) was shaken vigorously with 100 mL deionized water in a 250 ml Erlenmayer flask, until the solution became clear with no haze. The pH of the solution was adjusted to 3.0 with HCl and 10 mL of this solution was then added to a paste of Ni(Phen)2+/SOi-- hectorite intersalate (~0.2 g). The mixture was stirred for a few hours and then centrifuged. The clay was -65- washed with a very small amount of water, spread on a glass slide, and stored in a desiccator upon drying. K. Physical Methods 1. X-ray Diffraction Studies A11 d(001) basal spacings were determined with a Philips x-ray diffractometer or Siemens crystallaflex-4, both equipped with Ni filtered Cu kc radiation. The film samples were prepared by allowing an aqueous suspen- sion of the mineral to evaporate on a microscOpe glass slide and monitoring the diffraction over a range of 20 values (2-40°). In some cases the self-supported films were supported on a glass slide by the aid of double surface tape. Highly ordered self-supporting oriented films samples of intercalates or intersalates were obtained by allowing a ~l% water suspension of the complex mineral to evaporate at room temperature on a flat polyethylene surface and then peeling the dried films away. Basal spacings of fully solvated and swelled samples were obtained by first forming a thin film of the mineral on a 1.XlV, porous ceramic tile in place of microscope slide. This sample was then suspended under solvent and allowed to equilibrate for 30 min. The solvent adsorbed in the pores of refactory material, allowing the tile to function as a solvent reservoir and prevented the film from drying during the x-ray -66— diffraction measurements. This is a quite convenient method for determining the basal spacings of minerals wetted with a solvent which is very volatile. In a blank experiment, no diffraction patterns from the microscope slide or the ceramic tile was observed in the scanning range of interest. The peak positions in the angle 20 were converted to d basal spacing with a standard chart, (Cu Ka' A==l.5405 A). 2. ESR Studies Electron spin resonance (ESR) spectra were recorded at x-band frequency with a Varian E-4 ESR spectrometer, using quartz tubes containing thin films of the doped hectorite intersalates. Highly ordered self-supporting films of Zn(phen)§+/SOi--hectorite intersalates containing' PADS - as a spin probe at the 2% exchange level were prepared by evaporating at room temperature an aqueous suspension of the intersalate on a flat polyethylene or teflon surface and then peeling the films away. The self-supported film samples of the intersalates were placed in chambers of 52% or 98% relative humidity (r.h.) or vacuum dried for a few days before the measurements were made. Since the crystallites are oriented with their silicate layers parallel to the film surface, narrow strips of film (ca. 3 X12 mm) placed vertically in a quartz glass ESR tube (I.D. 4 mm) or on a teflon holder could be positioned in the cavity of an -57- ESR spectrometer with the silicate layers at a known angle to the external magnetic field. In the case of fully wetted samples, a suspension of the clay sample was centrifuged, and a glass capillary tube was immersed into the paste. Then approximately 1 cm of the tip of the capillary which was filled with the paste was cut and placed in a quartz glass ESR tube for the measurement. Standard pitch served as a standard for which 9 =2.0023. 3. UV-Visible Spectroscopy Electronic spectra were recorded on a Varian Associates Cary Model-l7 spectrophotometer. Absorption spectra of complex solutions were obtained using matched 1 cm path-length quartz cells. In the case of clay- complexes, the samples were prepared by mulling in mineral oil and placing the mull between silica disks. A mull sample of native clay was placed in the reference beam to reduce the effect of scattering. 4. Gas Chromatography All product mixtures along with solvent and internal standards were analyzed by gas-liquid chromatography (GLC) on a Varian Associates Model 920 single-column chromatograph equipped with a thermal conductivity detector. The column was a 5'><0.25" 5% SE-30 on chromosorb W. Also, a Hewlett-Packard Model 5880A gas chromatograph system with flame ionization detector -68- (minimum detectable level: 5810-.12 g/sec. of carbon) was used in some studies. The output of the detector was recorded on 5880A high speed printer/plotter with key stroke programming for data handling system. The triphase catalytic products were separated on a high speed metal capillary column (12.5 M><0.2 mm cross- linked dimethyl silicone) and were identified by comparison of GLC retention times with those of an authentic sample. The percentage of products was determined by integration of the chromatographic peaks. Integrations were carried out by the dual channel integration and computation, cartridge tape units, and BASIC pragramming. 5. Infrared Spectra Infrared spectra were recorded by use of, or with a Perkin-Elmer Model 457 grating spectrophotometer. The samples were prepared by using a KBr matrix or mulling the sample in fluorolube (Hooker Chemical Company) and placing the mulls between CsI disks. A wire mesh screen served as an attenuator in the reference beam of the spectrometer. CHAPTER III RESULTS AND DISCUSSION A. Adsorption of Complex Salts on Smectite (Intersalation Reaction) 1. Adsorption of Metal Complex Salts Metal tris chelates of the type M(chel)§+ (x==2,3), where M==Fe, Ni, Zn and che1==phen and its derivative, bipy, TPTZ, en, rapidly displace the surface Na+ ions from aqueous dispersions of hectorite. The exchange reaction may be schmatically represented by Equation (39), wherein the heavy lines represent the negatively charged silicate layers with a thickness of ~9.6 A: No" y Na+ ~.\. / 2+ M (chel): ( 3 9 ) The exchange reaction, which is quantitative up to the cation exchange capacity (CEC) of the clay and independent of the -69- -7o- counter-anion affords homoionic clay intercalates (Table l). The values given in the Table are in good agreement with the value expected for a monolayer of intercalated complex. A schematic representation of the ligands used are also given in Fig. (7). Analogous reactions can be observed with other smectite clays such as montmorillonite [40]. The sulfate salts of M(chel))2(+ complexes will also react with Na+-hectorite according to Equation (39), provided that the amount of salt initially present does not exceed the CEC of the mineral. However, if the sulfate salts are present in excess of the CEC, then some of the metal complex will bind to the clay interlayers as M(chel)§+/Soi- ion pairs. The intersalation process may be represented schematically by Equation (40). y\ hf"<} kaNNQPSQf. + (excess) V/m (40) Analogous intersalates were obtained using other X02- salts, specifically, CrOi-, MoOi—, WOi-, in place of 80%-. Chemical analysis of Ni(phen)§+/Soi- and Maoi— hectorite indicates the presence of one equivalent of Ni(phen)§+ to balance the layer charge of the clay and one equivalent as Ni(phen)§+/xoi_ ion pairs. The basal -71- .ouaamouousw on» can muomoa ommmaaoo on» How Aaoovo somzuon mocOHOMMAo on» ma “Hoovoq a“ m.mt ow muomoa ouoowawm on» How mmocxownu mamm3 moo cm> one a .musuauomfiou soon on oowuo moamsmn how one mmcwommmlo ones m.oH H.o~ v.6 o.aa H.o n.5a e.oH o.o~ a.o m.ms e.m o.mH TN-a- .-Hoo-o< « .-Hoo-o c. mcwommm Ham-om +~Hm-cocouamcocosous.a-ong m iv +~m “some mzum.o.c.m-ORL m +Nm Aconmvmmu m +~H -Nsms-Omu +~Hmlco-szL +NHo-mnz-nzL xoamsoo mo omNH .Aamao scum ooa\va mp ma mswomoq Houoav moxoamfiou ucouommwo sues oomsmnoxm ouwuouooni+mz mo Aaoovo .osflommm Human hautx .H canoe -72- ), IO- Phenanthroline 4,7 - diphenyl- phenanthroline 3,4,7, 8 - tetramethyl - phenanton ine 9 010" 2.4.6‘ TfiPY'idY' ' |.3.5- triazine o ”o N N / \ ethylene diamine Figure 7 Schematic representation of ethylene diamine, phenanthroline and some of its derivatives and 2,4,6-tripyridyl-1,3,5-triazine ligands. -73- spacing of these intersalates (Table 2) also suggests the presence of two molecular layers of metal complex cation. Figures (8) and (9) illustrate the 001 X-ray reflections of an oriented film sample of Ni(phen)§+/XOi-- hectorite. These results are in accordance with the findings of Berkheiser and Mortland [37,40], and Pinnavaia et a1. [39,151] where several complexes with different chelated ligands and counter-anions were studied. The results for M(chel)3SO4 intersalates suggest that the stacking of complex cation and $04 ions in the interlayers is similar to stacking patterns of hydrated M(chel)3SO4 salts in the crystalline state, wherein layers of complex cation are separated by layers of 4 ions [152,153]. Tanaka et al. [152], in their crystal structure determination of tris-(2,2'-bipyridy1) hydrated SO nickel(II) sulfate hydrate, have shown that the complex cations and hydrated sulfate anions crystallize in alternating sheets. The arrangement of complex ions, sulfate and water of crystallization are shown in Fig. (10). They concluded that the pseudo three-fold axis of the complex is nearly vertical to the (001) plane and each optical isomer (A and A) is arranged along the a axis in a separate layer occupying almost all the half-cell. In the study of crystal structure of tris- (o-phenanthroline)iron(III) perchlorate hydrate White et al. [153], also concluded that the disposition of the perchlorate anions and tris(o-phenanthroline)iron(III) -74- Table 2. X-ray Basal Spacing d(001) of Na+-Hectorite Exchanged with Different Complex Salts With Loading of 2 Equivalent CEC. Basal Spacinga Ad(001). (Alb Type of Complex Salt d(001) . (A) Ni(phen)3SO4 29.1 19.5 Ni(phen)3MoO4 29.4 19.8 Ni(phen)3WO4 26.0 16.4 Ni(phen)3CrO4 28.5 18.9 Fe(phen)3804 28.0 18.4 Zn(phen)3SO4 27.5 17.9 Fe(3,4,7,8-Me4-phen)3504 28.5 18.9 Fe(4,7-diphenyl-phen)3804 29.4 19.8 aThe d-spacings are for air dried samples at room temperature. The van der Waals thickness for the silicate layers is ~9.6 A; Ad(001) is the difference between d(001) for the collapse layers and the intercalate. Figure 8 -75- 001 X-ray reflections (Cu-kc) for an oriented film sample of Ni(phen)§+/MoOi--hectorite intersalate; d-spacings are given in Angstrom units. -75- Figure 9 -77- 001 x-ray reflections (Cu-kn) for an oriented film sample of Ni(phen)§+/SOi--hectorite intersalate; d-spacings are given in Angstrom units. -78- O— m— ON muquuO ON mu On an Oéu BK— 0‘0 656. d and no.“ -79- Ni(bp) so - 75 H20 ain't“ 9a :f/ //tm nr ‘5/ / 1‘) tuft/gA Qtséit,’ [:4 WOdO , Sakabe . Tanaka ([976). Figure 10 The crystal structure projected along the b axis. Complex ions drawn with heavy lines are lying relatively in a higher part of the unit cell than those drawn with light lines. -30- cations within the cell is generally in the form of alternating sheets parallel to the a,b plane. However, unlike the pure salts, the interlayer of the intersalates can be swelled by solvent, and the anions are exchangeable (vide infra). The adsorption isotherms in Figure (11) demonstrate 2-, Croi for the surfaces of Na+-hectorite in aqueous suspensions. that Ni(phen)3xo4 (XO4=SO -) have a marked affinity It can be seen that the amount of the metal complex adsorbed exceeds the cation exchange capacity of the mineral (73 meq/100 g). In order to maintain electrical neutrality, the complex ions bound in excess of the exchange capacity must be accompanied by intercalated counterions. The isotherms show several interesting features. In the region between 0 and 146 meq of bound complex per 100 g of hectorite, the slope is vertical, indicating the exchange equilibrium Na+-hectorite+Ni(phen)§+ ¢=Ni(phen)§+-hectorite-+Na (41) lies 100% to the right. x-ray diffraction of air dried samples with loadings of 73 meq/100 g (1 CBC) gives ~18 A basal spacing independent of drying temperature over the range 20-150°C. Space filled models of the complexes show that the cations are approximately 8 A thick along the C3 symmetry axis. This observed spacing is consistent -81- Figure 11 Adsorption isotherms (25) for [Ni(phen)3]SO4, [Ni(phen)3]CrO4, and Nile blue on hectorite in aqueous suspension. -32- 89? «concnonwm NN - 1 lie 5 2:0. .zquEzm-ozoo seems-pom d mad 3.2 o .65 .-c2a-_z a .om.-§-a-.z o m. S O. J con AV'IO 600) masaosav NO) XB‘IdWOO 03w -33- ‘with the value ~8 A expected for monolyers of complex ions oriented on the surface with their three-fold axis perpendicular to the silicate sheets [39]. In the region between 73 and'~240 meq/100 g, excess salt begins to penetrate the interlayer regions. However, complete double layer intersalation should form at 146 meq/100 9, since the samples with loading of 146 meq/100 g and beyond give ~30 A basal spacings Fig. (9) due to excess complex salt accommodation in an ordered phase of intersalation. Additional complex salt is also intersalated at loadings beyond 146 meq/100 g (2 CEC) and this excess complex salt may be accommodated in the surface regions between the coulombically bound monolayers of exchange cations. This extra adsorption (beyond 2 CEC) is dependent on the size of complex and perhaps preferential ion pairing in solution at high concentration of complex salt as well as the salvation properties of the complex. The adsorption isotherm for Nile Blue (Fig. 11) showed results similar to that obtained for Ni(phen)3X04.except that the amount .of complex capable of being intersalated beyond the CEC is slightly more than for the Ni(phen)3x04 systems. The adsorption isotherms were reproducible within experimental error and the complexes were stable over the course of the investigation with no observable change in the molar absorptivities of the complexes in solution. Moreover, the intersalates exhibited no diffraction lines -34- corresponding to the free salt. Therefore all of the bound salt is intersalated. Despite the highly ordered intersalated phases obtained by adsorption of Ni(phen)3xo4, the X-ray diffraction patterns of all Ni(phen)§+/Xoi--hectorite [see Figs.(8 and 9)] indicate that the interlayers are interstratified. That is, some layers have spacings which are larger or smaller than the value indicated by the first order reflection. The interstratification can be caused by a nonuniform charge distribution in the silicate sheets or by the irregular distribution of complex salts within the interlamellar space of the smectite. Although the charge heterogeneity was not determined independently in this study, such inhomogeneous charge distribution has been observed previously in smectites [12]. It was observed that the intercalation of excess complex salt was highly dependent on the exchange cation, conditions of salvation, and the swelling of the clay at the time of addition of excess salt [40]. In addition, large quantities of the complex salt were intercalated into a second layer in the interlamellar region of the clay only in the highly expanded smectite, like Na+-hectorite or montmorillonite. 2. X-ray Diffraction of M(chel)§+/XO?--Hectorite The effects of the complex salt on the d(001) spacings of Na+-hectorite are shown in Table (2). One CEC equivalent complex of Ni(phen)3XO4 resulted in ~18 A basal -35- spacing, while a ~30 A basal spacing is found for two equivalent CEC of the complex. Intermediate loadings (between 1.4-2 CEC) often resulted in randomly inter- stratified systems with d(001) values between 18 and 30 A. Considering the size of M(phen)§+ complexes, one may approximate 102 milliequivalents of complex cation per 100 g of clay for a complete monolayer in the interlayer of an expanding clay. Since the cation exchange capacities of Baroid hectorite (73 meq/100 g) and Laponite (55 meq/100 g) are less than the concentration of complex cation needed for monolayer coverage, the 18 A spacing is sufficient to accommodate all the exchanged complex cations. However, the cation exchange capacity of fluoro-hectorite (193 meq/100 g) exceeds the concentra- tion of cation needed for monolayer coverage; therefore, we would expect at least a partial two-layer coverage, resulting in higher spacings, if all the exchange sites are occupied by complex cation. X-ray diffraction analysis of Ni(phen)§+/SOi--clay with loading of 1 CBC does show a one-layer interlamellar complex, d(001) 2 18 A, in natural hectorite and Laponite, but a two-layer complex in fluoro-hectorite, Table (3). At a loadings higher than 2 CBC of the complex salt the basal spacings of clays dried in air at room temperature did not exceed ~30 A, which is indicative of a two-layer interlamellar complex made up of the exchanged complex cations and excess complex salt. Therefore, quantities -35- .mmaao Ham on poops mcz onoEoo on» ma muwommao omcasoxo cowuco mo usoao>flooo ac o.mH w.hm m.hH m.mH -m- .-Hoo-o .ocsonom Hanan ow Amcfleowsv ouflsoaawuosucos Hausumz mmH Aoufluouooniouooauv ouauouoo: owuosucmm mm Aouflsomaqv ouwuouoos afluonucam ms Aowouamv ouwnouoon Hausuaz m ooawwoe (NWHU mo omNE cacao mo mafia-momma masonoxm cowuao as.eonm-soso-sz cons oooscsoxm wean ucOHOMMHo mo .Hoo-o .mcfiomow Hmmmm mmuix one huwomocu omsmcoxm sofluao .m OHQMB -37- of complex salt in excess of that required to produce the two-layer complex was evidently expelled from the interlayers of Na+-hectorite upon air drying. Presumably, the excess salt coated the external surfaces of the clay particles. Also, any unintercalated salt used in preparation of the clay complex would exist on the external surfaces of the clay particles upon air drying. 3. Anion Effects on M(chel)}2{+ Adsorption The differences in the basal spacing between Fe(TPTZ)2(ClO Fe(TPTZ)ZSO4 and between Ni(phen)3C12, 4)2. Ni(phen)3SO4 shown in Table (4) prompted an examination of anion effects on the intersalation of Fe(TPTZ)§+ and Ni(phen)§+ salts. The results of Table (4) suggest that the additional binding of a complex cation through the intercalation of salt, is dramatically dependent on the nature of the anion. At all loadings beyond the cation exchange capacity, the Fe(TPTZ)§+/(C10;)2-hectorite system exhibits monolayer spacings of 20 A. Presumably, all of the free salt is expelled from the interlayers upon drying the Fe(TPTZ)2(C104)2 system at high loadings. Considering the size of Fe(TPTZ)§+ complex (longitudinally almost twice as large as Fe(phen)§+ complex), one expects that even at loading of 1 equivalent CEC the concentration Of the complex cation would exceed the monolayer coverage. Therefore, partial two-layer coverage, with a higher d(001) spacing would be expected. Fe(TPTZ)§+ shows no tendency towards intersalation with hectorite clay when C104 was -33- Table 4. Effect of Ligand, Counteranion and Loading on the d(001) Spacing of Intersalated Hectorites. a Basal Spagingb ) E . qulvalents of Complex d(001), ( Type Of Complex: Salt/Equivalents of Clay Fe(TPTZ)2(ClO4)2 0.5 20.0 Fe(TPTZ)2(C104)2 1 20.0 Fe(TPTZ)2(ClO4)2 2 20.0 Fe(TPTZ)2(ClO4)2 15 20.0 Fe(TPTZ)ZSO4 1 20.0 Fe (TPTZ)ZSO4 2 20.0 Fe (TPTZ)ZSO4 15 29.4 Ni(phen)3Cl2 1 18.0 Ni(phen)3Cl2 2 20.0 Ni(phen)3C12 5 23.2 Ni(phen)3Cl2 10 27.6 Ni(phen)3504 1 18.0 Ni(phen)3SO4 2 29.1 Ni(phen)3SO4 5 29.3 10 29.2 Ni(phen)3SO4 aConcentration of the salt for reaction with 1 equivalent M(chel)3 salt is l.5><10'_3M_(e.g. 0.5 equivalent of M(chel)3 salt is 0.75)<10 M and so on). All samples were air dried oriented films. -39- the counterion. However, when the soi‘ salt was used, the tendency toward intersalation improved and finally at 15 equivalent CEC of the salt a higher d spacing of 29.4 was detected. Nevertheless, the degree of intersalation still was low since very high amounts of salt (15 equivalent CEC and beyond) were needed to detect any change in the d spacing. It is to be noted that Fe(TPTZ)ZSO4 at complete two-layer coverage of complex cation should give a d Spacing of 34 A. Therefore, the observed inter- stratified broad peaks at the highest loading of 15 equivalent CEC would be indicative of partial two-layer coverage with poor ordering (see Fig. 12). The diffraction pattern for the Ni(phen)§+ complex show interesting features. When the C1- salt was used the d spacing gradually increased as the amount of salt was increased and, eventually, at 10 equivalent CEC two molecular layers of the complex cation were obtained (Fig. (13)). In contrast, the $04- anion is much more effective in ordering two molecular layers of the complex cation. The Ni(phen)3804 intersalate exhibits 10 orders of reflection with d(001) 3 30 A, at a reaction stoichiometry of only 2 equivalent CEC. Similar results were obtained when MoOi_, W03- and Croi- was used in place of $02- counterion. The qualitative order for intercalation of the anions correlates well with the anion binding selectivity reported by Pinnavaia et a1. [39]. -90- [Fe(TPTZ),]SO‘ -Hectorite l4.0 I5 CEC. Equivalents l ”.2 I 29.4 l4 IO DEGREES 29 Figure 12 X-ray diffraction pattern of Fe(TPTZ)§+/SOZ-- 4 hectorite prepared by addition of 15 equivalent CEC of the complex to Na+-hectorite. -91- 27.61? 950 N I5.5A° 20.0A' 92A“ ' l scsc 'fOA' 920M I ‘ 2csc 92;)” 1686 l I I3 IO 5 2 DEGREES (28) Figure 13 X-ray diffraction patterns of 10,5,2, and 1 CEC equivalents of Ni(phen)§+/Cl;- hectorite. -92- 4. Effects of Ligand on M(chel)§+ Adsorption Complex cations of different size and shape were prepared by using various ligands. Basal spacings of the homoionic intercalated complexes were determined. The results are shown in Table (1). The dependence of the basal spacing on the size of the complex cation indicates that the intercalation of larger cations gives rise to a higher basal spacing. Hence it is possible to monitor the basal spacing by selection of an appropriate ligand in a given complex. Fe(TPTZ)§+ is a large cation due to the size of ligand. The intercalation of this iron complex in hectorite does not produce the expected 001 basal spacing, since the longer dimension of the complex (estimated by space filling models) is approximately 18 A, it is possible that the complex is oriented parallel to the silicate layers. Attempts at explaining the basal spacings by rearrangement of the cations are qualitatively illustrated in Fig. (14). In addition, Fe(TPTZ)ZSO and Fe(phen)3804 are 4 different only in the type of ligand. Due to the larger size of TPTZ, a more efficient layer screening effect is expected. However, the experimental data show that the Fe(TPTZ)ZSO4 complex has little tendency toward inter- salation. very poor ordering is observed when Fe(TPTZ)ZSO4 is intercalated in hectorite (see Table 4 and Fig. 12), while Fe(phen)3SO4 readily forms a well-ordered intersala- tion (see Fig. 9). Perhaps the dominant factor here is -93- Slllcote Layer ./ 9.6A' Fe ] Ad~IsA’ 4" IOA'< Ad< I8A° Jr \- Figure 14 Possible orientations of the Fe(TPTZ)§+ ion on the surface of Na+-hectorite: A) The complex cation oriented perpendicular to the silicate layers. B) Tilted orientation. C) The complex cation oriented parallel to the silicate layers. -94.. the effective cationic charge. In the case of Fe(TPTZ)§+ the positive charge is spread over a larger molecule in comparison with Fe(phen)§+. These results show that the intersalation process depends upon not only the nature of the anion but also on the type of the chelating ligand. As previously mentioned, it is believed that the screening of the electrostatic charge of the clay by the complex cation permits the penetration of more complex cation and its anion into the interlamellar regions. However, the screening of the electrostatic charge of the silicate layer seems to be very much dependent upon the size, shape and the charge of the cationic complex. Therefore ligands play an important role in the intersalation reaction, as they are responsible for this screening effect. Simple inorganic salts or a complex like Ni(NH3)GSO4 do not behave in this manner. Phenanthroline derivatives 3,4,7,8-tetramethyl phenanthroline and 4,7-diphenylphenanthroline both are larger than the phenanthroline molecule. Basal spacings for the homoionic intercalation compounds are 19.0 and 20.1 A respectively (see Table 1). In most cases the 2 CEC intersalation complexes of the sulfate salts give more or less the same basal spacings as the Fe(phen)BSO4 intersalate (~30 A). The reason may be explained by the charge distribution over a bigger molecule which may -95... lower the interlayer ordering and therefore lowers the d spacings. 5. Electronic Spectra . vAn indication of the retention of chemical and struc- tural constitution of all chelated complexes used in triphase catalysis (see Section B) upon intercalation were provided by an electronic spectral study. The spectrum of Ni(phen)§+/SOi--hectorite could not be obtained because of law absorption. From the extensive investigations [154-157] of the spectra of the ortho-phenanthroline complexes formed by iron, it has been concluded that the intense absorption bands of these compounds in the visible region (Fig. 15) are due to the transfer of electronic charge between the d-orbitals of the metal ion and the h-orbitals of the ligands. The intensity of the visible absorptions implies that charge-transfer configurations make an important contribution to the electronic ground state of these complexes [158,159]. That is, there is appreciable h- electron exchange between the cubic tzg metal d-orbitals and the n-orbitals of the ligands. The small shift toward lower energy for Am between the pure complex in ax aqueous solution and the intercalated complexes can be attributed to the change in the silicate layer environment. Thus, the spectral studies indicate that the chelated complexes retain their constitution in the intercalated state. -95- I II ABSORBANCE (7 ABSORBANCE I> l l A l 340 500 700 320 500 700 (1) (nm) (A) (hm) I ABSORBANCE 320 500 700 (k) (nm) Figure 15 UV-visible spectra of A) [Fe(phen)3]SO4, B)[Fe(3,4,7,8-Me4-phen)]SO4, and C) [Fe(4,7- diphenyl-phen)3]SO4- in I-aqueous solution and II-—Na+-hectorite intercalated with these complexes. -97- 6. Adsorption of Organic Salts Compounds 1-5 were studied with the aim of elucidating the behavior of intercalated organic salts in smectite as (g l ((3 c1’ triphase catalysis. 3 (1) + — 933 EH3 - (C8H17)3CH3N c1 IE- A-(CH2)°_—AR-(CH2)3-]23r x (2) (393 (1 3 (3) (4) (5) The above organic salts were intercalated at the l and 2 equivalent CEC level. For compounds 1-3 X-ray diffraction patterns showed one strong (001) peak, presumably first order, but the higher order reflections were very weak. Moreover, increasing the loading from 1 CEC to 2 CEC organic salts resulted in small change in d spacings (Table 5). In the case of compounds 4 and 5, Ad was increased two-fold and higher order of diffraction .musofiousmmos any whommn ousumuomfiou Econ no poauo awn can madam madam a co mEHflm MC Show on» cw who; mOHmEom Hams .1. 9 - .w.hN b.mH m o.m~ ES 1. or: an: m «.2 Tom N com com H -a- coo m now -Hoo-o -m- use H now -Hoo-o nonesz oesomsoo "mAHo>oq omcmsoxm umu m can A Day moaoomaoz oasamuo acouommao saws oomsmnoxm oufluouoomI+cz Ham Aaoovo .msfloaow Hammm woutx .m canoe -99- were pronounced. In fact, these organic salts in smectite clays, when intercalated behave much like the chelated metal complexes discussed earlier. Although no information was obtained about the orientation of the organic molecules in the clay, the two-fold increase in Ad may be indicative of an ordered double layer of cation/anion pairs. For example, Fig. (16) illustrates the basal spacing for 1 and 2 equivalents CEC intercalation of compound 4 and 5. The X-ray diffraction pattern of compound 5 at 2 CEC loading is well ordered, while compound 4 shows less ordering. The ordering was not improved when the intersalation reaction of compound 4 was repeated in the presence of excess sulfate anion. Perhaps the charges cannot be adjusted to the surface charge patterns in the hectorite, as with 5, when SO4- anion is introduced. The basal spacings of 1 CEC and 2 CEC hectorite cetylpyridinium complex (1) were determined and in both cases a value of 20.6 A was observed, [see Table (5)]. It is difficult to interpret these spacings in terms of the orientation of the molecule at the surface. For example, for monolayer coverage of compound 1, the molecule should give d(001) =14 A, as suggested by Greenland et al. [160]. These workers also concluded that higher spacings like 21 A for 1 CBC intercalates perhaps is due to interstratified systems containing varying -100- Figure 16 X-ray diffraction patterns of 1 and 2 CEC equivalents of A) Nile bue and B) 3,3-(4-4'- biphenylene-bis-(2,5-diphenyl-2H- tetrazolium chloride)-hectorite. -101- J 25 20 IS IO 5 2 DEGREES (2 9) l l 25 20 l5 )0 5 DEGREESQG) -102- proportions of the one layer (d(001)a=14 A), two layer (d(001)==18 A) and three layer (d(001)==22 A) complexes. The basal spacing of 20.6 A corresponds to an interla- mellar separation of 11.0 A and an interlamellar volume of approximately 8.2x1025 A? per 100 g (the surface area of Na+-hectorite==750 mz/g). This volume can accommodate approximately 286 meq of cetylpyridinium ions (volume of one cetylpyridinium ion taken to be 470 A3). It is to be expected, therefore, that complexes containing less than this quantity of the cetylpyridinium compound should show no change in d spacing. Intercalation of compound 2 also showed no signifi- cant change in basal spacing at loading of 73 and 146 meq /100 9 (see Table 5). Also, the organic salt was 4 anion, to examine the effect of counter-anion. No change was intercalated in the presence of excess SO observed in the d spacing, suggesting that perhaps the adsorption here is due primarily to van der Waals forces. Therefore, the intersalation reaction is not dependent on the nature of the counter-anion. The high spacing observed for this organic molecule may be indicative of a more or less vertical orientation of the molecule in the interlayer surfaces, which is not uncommon in molecules with aliphatic chain. In an attempt to minimize the desorption of salt from the clay interlayers, the intercalation of a chain- 1ike macromolecule (polybrene) was examined. Since the -103- intersalates are used for catalytic reactions, desorption of the catalyst from the interlayer is undesirable (see Sections A-5 and B-l). When polybrene (3) was intercalated at a loading of l and 2 CEC level, a difference of 2.7 A in 001 spacing was observed (see Table 5). It is interesting to note that when 2 CEC phase was equilibrated in de-ionized water for several hours no changes in d spacing was detected. Lagaly et al. [161] concluded that the ionene [-N@(CH2)X]n chains are adsorbed in the interlayers completely in trains. The rate of desorption should be very low and the probability that all of the adsorbed segments can be simultaneously detached from the surface is small. In addition, the translational entropy gained by the system provides the driving force for the strong attachment of the polymer to the surface. However, due to the complexity of the system more study is required to explore the details of anion adsorption. All organic salts (1-5) were checked by IR spectra- scopy for retention of structural constitution after intercalation with hectorite clay. The spectra showed absorption bands characteristic of the authentic molecules. 7. Desogption Coefficient Measurement Several intersalated phases have been examined for the desorption of complex salt from the interlayer when they were equilibrated in ethanol and water as a solvent. -104- After equilibration for 48 hours no significant loss of the complex salts was observed. The desorption coefficient (Table 6) shows that the binding of M(che1)§+/X0i- ion pairs to the clay interlayers is quite strong. For 2- example, the coefficient for desorption of Fe(phen)§+/SO4 from the two-CEC hectorite intersalate in water at 20°C 10 2 is 8.2 Xl0- molesz/liter . The desorption coefficients for Fe (phen) 3804 and Fe (3,4 ,7,8-Me -phen) 3SO are even 4 smaller in nonaqueous media (see Table 6). 4 Fe(4,7-diphenyl-phen)BSO has a higher desorption 4 coefficient in ethanol and this is because the complex is more soluble in ethanol than water. However, a very low coefficient for desorption of this complex was 13 2 observed in aqueous media, i.e., 5.2 X10- molesz/liter , at 25°C. Thus intercalated complex salts are about as soluble as barium sulfate, silver chloride, or calcium carbonate which they have solubility product of 10 10 9 2 1.08 x10” , 1.56 x10’ , and 8.7 x10' molesZ/liter at 25°C respectively. This result is somewhat misleading, however, because the desorption coefficient depends strongly on the M(chel)§+/Xoi- loading of the intersalate. As the loading decreases, the desorption coefficient increases. Consequently, the intersalate can be converted to homoionic M(chel)§+-hectorite by washing the compound a finite number of times with water. -105- oHIon H.s mHIon.~.m oHIon ~.e oHIon o.o HHIon s.m oHIon ~.o noufia\~moaoz uncoommwooo coflumuomoo moum moum N O m uco>aom vommAsonmIahaosmHoIn.¢vom eonm-cocolascocoaoas.e-on comm-socoueosl.m.s.e.m-oo sown-cocoleozuo.s.o.m-om comm-coca-om aohm-accept onmEoU mo omNB .moxoageoo cofiuaacmuoucH oufiuouoom omu N Ham usofiowwmooo cowumuomoo .w canoe -106- 8. Anion Exchange Properties of Smectite Intersalates Significantly, the intercalated X04 ions in the smectite intersalates can be readily exchanged at room 4 part, with singly charged anions such as halide and same temperature with a variety of other XO ions and, in organic anions. The adsorption isotherms were obtained for the exchange of CrOi- in the parent intercalation compounds by inorganic, organic, and complex anions of differing size, shape and symmetry. Figure (17) illustrates 2- the exchange isotherms for replacement of CrO4 in Ni(phen)§+/CrOi--hectorite by MoOi-, WOi-, P02-, Cr(OX)§-, Cl-, CH3CO§ and C6H5-c00' ions. The adsorption isotherms 2- were measured by quantitative analysis of CrO4 ion replaced by the incoming anion. Almost all of the 2- 4 replaced by anions with a charge greater than unity at an equilibrium concentration of 6 ><1o"4 M. p03" almost entirely perhaps due to the higher intercalated CrO (~43 meq/100 g of intersalate) is anion 4 charge of the phosphate anion, however, Cr(OX)§— anion 2- 4 size. The difference between Mooi- and W02- ion in the isotherm is not understood, since both have the same replaces CrO replaces CrO at a lower level, maybe due to its larger charge and the size difference is negligible. Though the 4 is favored over chloride and benzoate, a significant fraction of the Croi- can be replaced by binding of CrO these latter ions at relatively low equilibrium concen- trations. Among the exchange ions investigated, acetate -107- 2- 4 O The adsorption isotherms [see Fig. (17)] demonstrate that has the lowest affinity for replacing intercalated CrO hectorite intersalates in fact do act as anion exchangers and in many cases the amounts of anion replaced is very close to the calculated value for the anion exchange capacity (ABC) of the clay intersalates. As can be seen from the isotherms, the greatest tendency for anion exchange reactions occurs for anions with higher charge. The adsorption isotherms for anions with charges of -2 and -3 indicate that in the region between zero and 45 meq of adsorbed anion per 100 g of complex-clay the slope is very steep. A steep slope is indicative of an equilibrium displaced far to the right. 2- [M(che1)§+/Y2-]-hectorite-+X ¢=[M(chel)§-/X2-]-hectorite + Yz- . (42) X-ray diffraction patterns and d(001) basal spacings are almost unchanged after the anion exchange reaction, indicating that the intersalated phase remains intact during the anion exchange process. As described earlier, the anion exchange takes place without significant loss of the complex cation, which remains immobilized on the surface. This is important, since the negligible desorption, over long periods of time, is a desirable property, for catalytic activity of the layered silicate intersalates. -108- T CH5 ”0 ‘6 0 lap- 15 o “C "e lg lg 3 2 O to: 7...» e “ 3 6 o Equil. cone. of anion (x IO"molor) O O h o 504 aIDIDsIaIuI boou / paqlosqo uquo 'baw Figure 17 Anion exchange isotherms for Ni(phen)§+/CrOi-- hectorite intersalate at 25°C. -109- 9. Swelling of the Intersalation Phase To obtain an indication of the swelling properties of M(chel)§+/SOi--hectorite, X-ray powder diffraction measurements were carried out under conditions where the interlayers were solvated by the two solvents used in the catalytic studies (see Chapter III). The observed reflections [Fig. (18)] provide a qualitative indication of the extent of interlayer swelling. The position of the first order reflections were as follows: 33 A (water), 28 A (toluene). Virtually no swelling occurs with toluene, because the same reflection is observed when no solvent occupies the interlayer regions. Water, however, swells the interlayer region. It can be seen from the figure that the interlayers are interstratified. That is, some layers have spacings which are larger or smaller than the value indicated by the first order reflection. The interstratification can be caused by nonuniform charge distribution among silicate sheets and by the partial segregation of the complex salt. Neverthe- less, the observed reflections provide a qualitative indica- tion of the extent of interlayer swelling. In triphase catalysis, one of the liquid phases is water. Since no organic solvent can compete with water in swelling of the clays, the role of substrate in a possible swelling is not significant (see Chapter III-B). -110- 20 IS IO 5 2 Figure 18 X-ray diffraction pattern for Ni(phen)§+/SOi-- hectorite intersalates solvated by (I) water and (II) air-dried film. -111- 10. ESR Studies Paramagnetic cations such as Mn2+ and Cu2+ have been utilized as spin probes to characterize the environment of exchangeable cations adsorbed on smectite surfaces [162,163]. A nonrigid, solution like interlayer was observed under certain conditions. This observation led to the supporting of homogeneous rhodium phosphine complexes in the intercrystal environment, with retention of catalytic activity [131,164]. Also, attempts have been made to relate [165] the mobility of organic exchange cations such as protonated 4-amino-2,2,6,6-tetramethy1 piperidine N-oxide on smectites to the extent of dehydration of the smectite. In the present study, electron spin resonance (ESR) investigations of the orientation and mobility of clay intercalated anions under different degrees of relative humidity have been carried out. The anion of Fremy's salt, peroxyl amine disulfonate (PADS) was used as a nitroxide spin probe to examine the rate of tumbling of the exchange anion to determine the diffusion of organic molecules into the intracrystal environment of the intersalated system. The nitroxide spin probes are observed to tumble rapidly enough in aqueous solution at 25°C to average completely anisotropies in g and hyperfine splitting (A) values. The properties of dilute aqueous solution of K2 (PADS) have been studied and rapid rotational 12 correlation times, TRHRB Xl0_ sec, have been found [166-169]. -112- Unlike other commonly used nitroxides, [PADS]2- has the advantage that its slow motional and rigid spectra are not inhomogeneously broadened by unresolved intramolecular proton dipolar interactions. Also, contributions from intermolecular electron-nuclear dipolar interactions are negligible since the linewidths are the same for H O and D20 [166]. 2 Zn(phen)§+/SOi--hectorite intersalate (2 equivalent CEC) was synthesized and the nitroxide spin probe is doped into the anion interlayers at about 2% level of exchange, and used to estimate the interlamellar mobility. It has been found that the doping level affects the spectrum of probe adsorbed on clay [30]. It is likely that spin exchange is enhanced by the concentration of probe ions in certain interlamellar regions,a phenomenon of ion segregation (demixing) that is not uncommon. Very low loading levels on the exchange sites (less than 2% of ABC) are therefore preferable to avoid unnecessary line broadening and spin exchange. The asymmetrical three-line spectrum of the nitroxide spin probe [PADS]2- in DMSO solution and, also, its rigid- limit spectrum are shown in Figs. (19) and (20). The spectra of Fig. (21) for PADS 2--doped hectorite intersalate demonstrate some loss of rotational mobility on the clay at 98% and 52% relative humidity and on the vacuum dry clay. The anistropic behavior becomes more pro- nounced as the relative humidity of equilibration is reduced. -113- -5 ESR spectrum of 1.10 M solution of [($03)N0] 2'in DMSO solvent 3 2 .3 P.__... V F H k Figure 19 ESR spectrum at 20°C of the nitroxide spin probe, (PADS)2-, in DMSO solvent. The vertical line on this spectrum represents the position of g = 2.00. -114- 20 GAUSS I—IH -—-——>— Figure 20 Rigid-limit x-band ESR spectrum of PADS; the field increases to the right. -115- fully wetted .L 98%".h. fiu A l ..l. 52%rh. a Figure 21 ESR spgctra at 20°C of the nitroxide spin probe, (PADS) ' doped at the 2% exchange level into Zn(phen) +/SOZ"-hectorite at different relative humidity. A free electron signal at gI=2.0023 is shown. -116- The ESR spectra of adsorbed probe reported here do not necessarily reflect interlayer solvent viscosities, since rotational correlation times are modified by specific surface interactions in addition to the properties of the solvent. For example, by using Equations (43) and (44) to estimate To for adsorbed [PADS]2- under wetted conditions, values of about 3.85><10"'10 sec. are obtained [see Fig. (21)]. These values are'-100 times lower than the value observed in the aqueous state [166-168]. However, larger values for Tc are found for the 1 orienta- tion of the clay films in the magnetic field compared to the H orientation. This is a result of some anisotropic rotation of the probe in the interlayer. In addition, the variationinrc values for samples with different relative humidities can be related to the interlamellar space limitation. The resulting basal spacing suggest that increasing the relative humidity gives the anion more room to move much faster (Table 7). Under fully wetted conditions, the X-ray basal spacing is 33 A which may allow enough space for the [PADS]2- molecules to tumble quite rapidly (i.e.'~3 > one .mouwm omcmnoxo on» so IN-momm- wm).:u«3 ooooo mums moumaamuoucw ouauouoosivomm-coom-GNM m.oH m.m~ muo .Oo> o.oH o.n~ mm a.a~ m.am mm «.mm o.mm pounds sH-sm -«- -Hoo-o nN3: .-Hoo-o< monsoon dunno assesses ossunaom -n- a..>uaoassm o>sunaom Deonoooao mo moumacmuoucH oufluouoom oomcmnomeIN-modm- Ham mafiocmm AHoo-o Hmmcm .A Dance -118- Tel = -2211W0R-/Ho (43) TCZ = 0.65W0(R+-2) (44) - t 1 8 Ri - [(ho/h+1) J - [(ho/h_1) ] (45) If the difference and sum of (ho/h_1)* and (ho/n+1)* are taken, the above independent equations for the determination of Tc (correlation time) can be obtained -—one containing the ol term and the other containing the o2 term. These two estimates of TC are not generally identical [169]. Tel and Tcz are correlation (tumbling) times in nanoseconds, W0 is the linewidth of the central peak (6), and H is the magnetic field (G) corresponding 0 to the central resonance line, where ho, +1, -1 = the height of the middle, low, and high field peaks, respectively. The correlation time, Tc, was taken as the average of Tel and T02, a procedure which seemed to reduce random error. The equations used to obtain Tc are strictly valid only for isotropic rotation, however, they are useful for comparative purposes in studying anisotropic rotation at surfaces [165]. The calculated values of Tc are given in Table 8 for the hectorite intersalates at different relative humidity. The values of Tc are in the range of 1-6 ><10-9 sec, meaning that the tumbling mobility of the [PADS]2- is reduced by a factor of 400-1000 when compared with the probe in aqueous solution. But this is relatively rapid motion of spin probes at the surfaces [171]. -119- Table 8. Calculated Rotational Correlation Times (nsec) of PADSZ- Doped into Zn(phen)3SO4-Intersa1ate of Hectorite Clay. a,b fit Correlation Time of [(503)2N012' (x 109 sec) Relative Humitidy TC TC of Samples (%r.h.) l 2 II 1" II 1 98 1.20 1.39 2.92 3.31 52 1.23 5.15 4.81 11.35 ..7 “Amount of PADSZ- doped into the intersalate were % 2 of the A.E.C. bSelf supported film of the intersalate (3 Xl2 mm) were made. c" and 1 represent the orientation of the hectorite films to the magnetic field, H. -120- B. Catalytic Properties of the Intersalates A new type of heterogeneous catalysis termed ”triphase catalysis" has recently been introduced [83]. The underlying feature which distinguishes this from other forms of heterogeneous catalysis is that the catalyst and each of the reactants are located in separate phases. This principle has been successfully applied to certain aqueous phase-organic phase reactions employing a solid phase catalyst. The ease with which XOi- anions can be replaced in M(chel)§-/XOi--hectorite intersalates suggested that these compounds may be effective catalysts in a triphase reaction system involving anionic nucleophiles. Moreover, since the intersalates are well ordered, it was felt that substrate size or shape selectivity might be observed. In the present work the kinetics of one such triphase catalyzed process was examined in detail. Anions such as halides and carboxylates as well as cyanide function well under triphase catalysis conditions utilizing smectite intersalates for the catalysis. They were found to be efficiently transported to the organic phase by the phase transfer agent and, once there, behave as potent nucleophilic species in a variety of displacement reactions. Several systems were identified in which anions were involved in the displacement reaction of -121- organic substrate in the presence of different types of layered silicate intersalates. An illustration of the three-phase system is presented in Fig. (22). 1. Kinetics of the Halogen Exchangg Among all available procedures for exchanging halogen in organic halides, only a few have proven to be useful for converting alkyl bromides to alkyl chlorides [172,173]. It was found that the triphase catalysis technique using smectite intersalates furnishes a convenient method for carrying out such transformations. Halogen ion displace- ment on different n-butyl halides were conducted in 50-mL culture tubes [(Fig. (23)] using procedures described in the Experimental Section. Rates of reactions were monitored by following the disappearance of the starting n-butyl halide from the organic phase. Clean pseudo first-order kinetics were observed and in spite of the inherent complexity of these systems, the reproducibility of observed rate constants, kobs' were good. Only one type of reaction product was detected. Figure (24) illustrates typical kinetics data. Examples showing the utility of triphase catalyzed halogen exchange are provided in Table (9). In the classical mechanism of phase-transfer catalysis, the nucleophilic substitution reaction [Equation (46)] occurs in the organic phase and is the rate-determining step. -122- INTERFACE V SOLID PHASE ._ v- , RXWOY A09) catalystjp RYWW'. xm, Figure 22 Schematic representation of the three-phase system, using smectite intersalate as a solid phase. -123- Teflon- lined screw cap l ' ~ 9 ' .e'Q O 0 ‘ I...... . . ...a l' O O .0.... 0 e 0 . e organic phase — solid phase suspended in 9.53353 the aqueous phase =1, ‘— mognetic stir bar Figure 23 Experimental set-up for triphase catalytic reactions. Corning N° 9826 Culture table (20 Xl50 mm) containing organic, aqueous, and solid phase; a teflon coated magnetic stirring bar (lo/35 x5/15 in. octagonal bar with pivot ring) is shown. -124- -In [Bu-BE], / [Bu-Br]o Time (hours) Figure 24 Plot of ln(unreacted) butyl bromide in the organic phase as a function of time for the reaction of 2 mL of 0.5 M n-butyl bromide in toluene with 3.0 mL of an aqueous 3.3 M NaCl catalyzed by 0.069 meq of Ni(phen)§+/SO4 ion pair in Laponite at 90°C. -125- .ummamumo ouwcomchIvom\+m-conm-Hz «a was mmo.o .0 wxaa moss H "cannaocoo :ofiuocomc as m 5 mode: season Hose 3 .8533)... a sh coins H mm hm.o Hm./\)(\ H0 /&)(\ mm.a H /\)(\ HU./\)(\ mm.H Ho /\)(\ um /&)(\ w.NN H /\)(\ Hm,/\)(\ AHIHn- mnoxmoa uosooum ucmuomom ”wanna-sumo mm moucaamuoucH ouwcomquIwom\+m-sosm-flz nuns Doom um mcowuocom ucoEoomHomwo oowamm oowuamuau Hommcmuanmcsm .m Danae -126- + Rx(org)-+§(org) §§L2=le Ry(org)-+x(org) (46) — — k[Q+] — — x(org)-+y(aq) ;.====e x(aq)-+y(org) (47) For most entering and leaving groups, catalyst regeneration by exchange between the aqueous (aq) and organic (org) phases [Eq.(47)] is so rapid that it has no effect on reaction rates. Undoubtedly, in the case of intersalation compounds equation more complex than (46) and (47) must be considered, taking into account any processes of adsorption and/or diffusion, k Rx(org)-+Icla§|§'a=2e Rx' claily afi=2 Ry k ~|c1a§1§ {.._—1A Ry(org) +|c1a§13£ (48) cla grid; k Wag) + Lil—2:1; e=ad §- k c1a§|§ ¢—_—__f—> §(aq) + [cj 1:12]? (49) where Icla§|x and Iclaily represent the supported complex clay], clail represent Rx, Ry, § and salt in the x and y forms, respectively; Rx- ciail, y- clay] and £- x in the environment of the third solid phase. Processes Ry- b, c, d, and f are controlled by adsorption (through their respective constants kb-f)' by diffusion, or by both. Process e, however, is controlled by the reactivity -127- of the nucleophile at the surfaces of the intersalate. In reality, the situation may be even more complex than (48) and (49) the continuity of the two liquid phases up to the catalytic center is not guaranteed. Here, the concentration of nucleophile in the aqueous phase, does not affect the reaction rate (Table 10) as dramatically as it would if the Table 10. Dependence of kobs on the Amount of Sodium Chloride Present.‘z 2 -1 NaCl (mmol) 10 kobs (hr ) 12 2.05 10 1.92 8 ’ 1.83 1.96 aReaction of 1 mmol n-butyl bromide in 2 ml of toluene with the indicated amount of sodium chloride dissolved in 3 ml of water catalyzed by 0.69 meq of Ni(phen)2+/SOZ- Laponite intersalate at 90° C. diffusion of the nucleophile toward the clay surfaces was the step controlling the regeneration of catalytic centers. The assumption is that the anion exchange equilibria whereby anions are transferred from aqueous to solid phase are very fast relative to the rate of the solid phase displacement reaction. This assumption will, of course, not be correct if one greatly slows the rate of phase mixing. Therefore in all the reactions carried out, the reaction mixtures were always stirred vigorously. -128- The result shows that the coefficient for desorption of the ion pairs in aqueous media is extremely low (i.e.«~10-10 molez/L2 at 25°C), it is even smaller in nonaqueous media. In order to study the significance of complex desorption and also to ensure that the displacement reactions were being catalyzed by the solid phase, the reaction of chloride ion with l-bromobutane in the presence of Ni(phen)§+/Soi’-hectorite was repeated, but stopped after a 20% yield of l-chlorobutane was obtained. Then a portion of both the aqueous phase and the organic phase was filtered and transferred to a second tube, which, along with the original tube, was heated for an additional period of time at 90°C. Analysis of the product mixture in the tube containing the clay catalyst showed an increased yield of l-chlorobutane. In the absence of the clay catalyst, however, the yield of l-chlorobutane remained unchanged. Under triphase reaction conditions with the intersalate as catalyst, some intercalated 80:- ions are replaced by halogen from the aqueous phase and this intercalated halogen is accessible for reaction with adsorbed alkyl halide. X-ray diffraction measurements show the basal d(001) spacing of the solid phase is retained after the displace- ment reaction. Therefore, the intersalation phase remains intact throughout the reaction. -129- It is interesting to note that despite some limita- tions, the clay intersalates are quite competitive with polymer supported systems (see Table 11). for displacement of iodide ion bs (aqueous phase) in l-bromobutane (organic phase) as a A plot of k0 function of the catalyst amount yielded a straight line, [Fig. (25)]. These data indicate that the catalyst efficiency remains constant. Thus the diffusion of reactants or of products across the various liquid- liquid and liquid-solid is not rate limiting. Since [1-] >>[l-bromobutane] the complete kinetic equation can be written in the following form: -d[l-bromobutane]/dt==kobs[l-bromobutane] (50) An examination of the dependency of the observed pseudo first- order rate constant (kobs) for the reaction of Cl- with l-bromobutane (organic phase) on the chloride concentration revealed that a two-fold increase in the amount of sodium chloride produced no change in the observed rate of reaction (Table 10). Therefore, the Cl- concentration at the surface of the intersalate is constant and anion exchange at the intersalate is not rate limiting. The kinetics features here bear a resemblance to that observed for the displacement of cyanide ion on 1- bromooctane using well known "phase-transfer catalysis" technique as well as recently developed triphase -l30- .cofluauflumnsm and“ “NH “HoxmmsoucvmimmoVZNmo u m m Gamma mcmuhummaom om hm wwm m.vH umwamumu «0 vmxxm .gm O .Annaav mum .Ne ..swno .ouo .n .ammmm .q.mn .cofluoom HmucmEHuwaxm on» Ca cmnHuommp mum mums poms msoauwocouu a s 3H 8 H /\/\ um /\/\ o “2.38 +2 m e m cm ma H./\)(\ Hm./\}(\ muwGOQMAIINOm\+NACm£mvHZ loco Aunt Away Axmv ummamumo .QEmB mafia unsooum usmuomwm Nxmcowuommm mmcmnoxm ammonm pmnkamumo Hmwmcmua mmmzm .HH magma -13l- 240 200 I60 C) <2 C) G) C) Q' J l J 1 J o .n o .n o no 0 N N "' - ,,(SJnou)"°x ,0: Figure 25 Plot of 102 kobs as a function of the amount of catalyst used. The catalyst and reaction conditions were similar to those described in Figure 24. (m9) -132- catalysis [103,48]. In all these systems,the rate of nucleophilic displacement exhibits a pseudo first-order dependency on the alkyl halide concentration and is also linearly dependent on the amount of catalyst used. For the phase-transfer reaction it has been proposed that the organic-soluble catalyst acts by repeatedly bringing cyanide ions located in an aqueous phase into the bulk organic phase where the displacement occurs. It has also been suggested that micellar catalysis in which micelles could bring small amounts of aqueous sodium cyanide into the organic phase in a form suitable for displacement reaction is of negligible importance. The effect of substrate polarity on reaction rate was examined using l,8-dibromooctane and l-bromooctane as substrates. A pseudo first-order plot for replacement of Br by C1 in l,8-dibromooctane is shown in Fig. (26). Although the carbon number is the same as l-bromooctane, the rate of displacement reaction is enhanced almost by 4 times. l,8-Dibromooctane is a more polar substrate and perhaps it has a higher affinity toward the surface of the catalyst than l-bromooctane. Therefore, the rate of reaction can be influenced by the polarity of the substrate as well as its size. Table 12 provides the pseudo first-order rate constants, k for the reaction of butyl bromide in various organic obs' solvents with aqueous NaCl in the presence of Ni(phen)§+ /SOi--hectorite as catalyst. Included in the table, for -l33- o m o [s o co 0A K) 0’ L. :3 <3 05 ‘r 0’ .§ or- m o N 9 l l L l 0. 5- 04 «2 V. to. o o o o o o[auopoowo:q gp -9‘|] Taumaoowmqu -9‘El ul - Figure 26 Plot of ln(unreacted) 1,8-dibromooctane in the organic phase as a function of time for the reaction of 2 mL of 0.5 M l,8-dibromo- octane in toluene with 3.0 mL of an aqueous 3.3 M NaCl catalyzed by 0.069 meq of Ni(phen)§+/Soi- ion pair in natural hectorite at 90°C. -134- Table 12. Kinetics Data for the Reaction of Butyl Bromide and Aqueous NaCl at 90°C“. Catalyst Organic Phase 102kobs (hr-l) Ni(phen)2+/SOZ--hectorite C H CH 1.9 3 4 6 5 3 C6H6 2.0 1,2-C2H4C12 1.9 1,2-C6H4C12 2.1 Ni(phen)2+-hectorite C H CH 0.09 3 6 5 3 Na+-hectorite C6HSCH3 0.24 Ni(phen)3804 C6H5CH3 <0.05 aReaction conditions: 1 mmol butylbromide in 2 mL organic solvent, 10 mmol NaCl in 3 mL H20, 0.069 meq. catalyst. -135- comparison, are the kobs values for the same reaction using homoionic Ni(phen)§+-hectorite and Na+-hectorite as catalysts and Ni(phen)3804 as a liquid-liquid biphase catalyst. Little activity is observed with Ni(phen)3SO4 under biphase conditions, yet when the salt is intercalated in hectorite, significant catalytic activity is observed. Also, it is noteworthy that the activity of Ni(phen)§+/SOi-- hectorite is insensitive to the nature of the organic medium. However, phosphonium-based polymers as triphase catalysts exhibited a modest dependence on the nature of the organic solvent employed. It was suggested [174] that the organic solvent can control the extent of swelling which dilates the mesh of the resin which, in turn, allows a greater accessibility of the reagents to the catalytic sites [124]. Since the clay catalyst is swelled much better by H20 than by any organic solvent, the effect of organic solvent on reaction rate is insignificant. The organic solvent could influence the adsorption equilibria for reactants on the surfaces. Also the organic solvent could influence the nature of the microenvironment at the active sites and, therefore, affect the free energy of activation. However, such factors do not seem to be important for the clay catalyst system. The low activity of Ni(phen)3SO4 under biphase conditions presumably arises from the low solubility of Ni(phen)§+/C1; ion pairs in the organic phase. In marked contrast to the intersalated catalyst, homoionic -136- Ni(phen)§+-hectorite and Na+-hectorite are relatively poor phase transfer catalysts for halide replacement in butyl bromide. These results are similar to those given in earlier reports of the phase transfer properties of homoionic smectite clays in displacement reactions involving anionic nucleophiles [134,128]. In all of these cases, however, the catalytic activity undoubtedly results from the small anion exchange capacity (<5 meq/100 g) of smectite clays which is believed to result from the replacement of some structural hydroxyl groups at the edges of the clay platelets [175]. Some adsorption of cation- anion pairs may also contribute to the observed catalytic activity, but unlike the intersalated complexes, these ion pairs would exist only in low concentration at external surface sites. When natural hectorite complex was adopted for the catalysis in some cases the reproducibility of the kinetic reactions from batch to batch was poor. However, for the synthetic hectorites, as expected, this was not the case. For this reason all reaction done with natural hectorite in a series were performed by using the same batch in order to obtain consistent results. 2. Substrate Size and Shape Selectivity The possibility of substrate size and shape selectivity with Ni(phen)§+/SOi--hectorite as a phase transfer catalyst, was investigated for the conversion of alkyl bromides to alkyl chlorides. A range of hydrocarbons from C to C 3 16 -137- were examined. Table 13 provides the values of kobs for the intersalated catalyst under triphase conditions. Indicated in the table are the rate constant obtained for the same reaction using tricapryl methyl ammonium chloride as a catalyst under biphase reaction conditions. It can be seen that the reaction rates under biphase conditions are very similar, as expected. However, the intersalated catalyst is more than 140 times more reactive toward the smallest substrate (l-bromopropane) than the largest substrate (l-bromo-hexadecane) in the series. It is possible that the apparent size selectivity of the interSalated catalyst is related to the ability of the substrate to adsorb in the interlayer region occupied by C1-. The degree to which the substrate penetrates the interlayer, however, probably is limited to a distance of a few molecular diameters, because adsorption of alkyl bromides to the intersalate is undetectably low. This latter observation is consistent with the fact that the 001 spacing of the intersalate is little changed by the adsorption of water or alkyl bromide. There is an alternative explanation for the apparent size selectivity of the intersalate based on selective adsorption at the edge sites of the intersalate. Since under triphase conditions the intersalate prefers to be wetted by the aqueous phase rather than the organic phase, the extent of edge site adsorption may decrease with increasing hydrophobicity and size of the alkyl chain. .umsamumo awe moo.o .ONm A8 m ca Humz HOSE ca .mcwsaou 45 N cw mumuumnzm owcmmuo «0 H085 H "mcofiuflwcoo cowuommmc -138- In Ho.ovv ummmmmao an mHo.o ummmmmao ma mmo.o unmammouomfi an ma.o Hmmflmao ON mH.o nmaammo I: m~.o ummamoouomfl In mm.o ummammonn om mm.o umHHmmUIOmH mm mo.o umaammouc I: mn.o Hmumlmmovue om H.H Hmmmvu mm v.a umhmmo Amflmwmwumo mmmnmmmv .mflmwamumo mmmnmflua. uao+zxmmovmmmmo>1~movg mufluouommnuwom\+mAcmnmvfiz mumuumnsm H..uoom um Homz saws mocHEoum ahxad oxmoav mucmumcou mumm HmUHOIumHflm 0650mm .MH manna mo cofluomwm may “Omfiaalunv mn ~139- The possibility of shape selectivity was examined for the series of normal and branched alkyl bromides. The pseudo first-order rate constants are plotted as a function of the alkyl carbon number for the Ni(phen)§+/SOi-- hectorite catalyst system in Fig. (27). The results show that the rate constants (kobs) decrease as the alkyl Carbon number increases for both normal and branched alkyl bromides. The value of k0 for the branched bs alkyl bromides was lower than for normal alkyl bromides with the same carbon number. However, the observed rate differences for the anion substitution reaction of normal and branched-chain alkyl halides are small. Despite the apparent size over the C3-C16 range, a dramatic shape selectivity is not observed for isomers of the same carbon number. Moreover, the kobs for the reaction of l-bromo-3- phenylpropane with NaCl is relatively high, considering the attachment of the large phenyl group as well as the carbon number in the chain. This latter observation is somewhat contrary to what is expected on the basis of shape discrimination. On the other hand this result is consistent with the result shown in Fig. (27). It is possible that the adsorption of the substrates on the surface of the intersalated catalyst is dependent more on polarity than on size or shape. Nonetheless, shape selectivity clearly is observed for two substrates with the same carbon number. This shape selectivity is a -l40- Figure 27 ‘2 .3 l L. 00 q. k) m 9! ”I a g. L. k) + 49 L 00 L: G I” 1(D / l I ,3 3‘ 5 mo 2 Z a: <1 <3 (I 2! 00 «v J J J l l l l (V 00 )(I1:- “‘r’//// V\,u~»unfi‘hur Figure 28 Schematic representation, showing edge view of the intersalated hectorite catalyst and the triphase catalyzed reaction zone. -l45- .me>HuowmmmH >MHU o oOH\me mmH paw mn.mm mun auHuouomnnouosHm was ouHuouown UHonmm .muHcomaH Ham huHoamao mmcmnoxm GOHumo .ummHmumo was mmo.o .on HE m :H Humz HOEE oH .mcmsHou HE N cH mumuumnam HOEE H "coHqucoo coHuomwmd a muHuouowcuonosHm .oHuonuchm mo.o NAA mpHuouoonllmom\+mA:mnmvHz Hm./\)(\/\)(\ manouomclouosHm .oHumnucmm . I v m «m 0 «AA muHuouomn I om\+~Acmnmsz um /\)(\ . Awmlev cHoumm mH.o Nv oUHHouomnllvOm\+mAcmnmvHz Hm./\)(\/\)(\ AoNIHmv UHoumm H.H Nv wuHHouomnllvOm\+macmnmvHz um./\)(\ muHcommq .oHuocucwm vm.o mo.o oDHuouomnlleom\+mAcmnmsz Hm./\)(\/\)(\ ouHcommq .OHumnucwm «a .H mo .o 33382....oner 28%.: Hz um /\/\ mno a Any coma muHHouomm uthmumu mumuumnsm A xv on» no oNHm mHOHunmm no.uoom um Homz nuHs mooHEoum meHd mo coHuommm on» Ham any mnox OHV mucmumsou mumm umpuolumHHm ovsmmm so muHm mHOHuumm mo pummmm .mH mHnme 2? m -146- particle size. Thus, increasing the edge surface area would favor catalytic reactivity. 5. Carboxylate Ion Displacements Liolta et al. [176] reported that acetate solubilized as the potassium salt in benzene containing 18 crown—6 because sufficiently nucleophilic to react smoothly and quantitatively, even at room temperature, with a wide variety of organic substrates under phase transfer catalysis (PTC) conditions. Using quaternary ammonium salt under PTC conditions Starks [49] obtained high yields of a variety of esters from simple alkyl halides and sodium carboxylates. It was of interest in the present work to examine the properties of clay intersalates for catalytic displacement reaction involving carboxylate ion as a nucleophile. The displacement reaction between benzyl chloride and acetate ion was carried out under triphase conditions by using hectorite intersalate as a catalyst. + _ + Q L CGHSCHZCI +CH3COO M catalyst v CSHSCHZOCOCH3 +MC1 (51) The yield after a reaction time of 48 hours at 90°C is shown in Table 16. When Na+-hectorite was used as a solid phase catalyst, a negligible amount of benzyl acetate was formed. Under identical reaction conditions butyl bromide also reacts with acetate, but the yield of ester is lower than that observed for the benzyl chloride system. This -l47- Table 16. Reaction of Acetate Ion with Organic Substrate‘z Substrate Product (%) Yield Benzyl chloride Benzyl acetate 61 Benzyl chloride Benzyl acetate 61) n-C4H9Br n-C4H90Ac 44 “Reaction conditions: acetate salt, 1.1 mmol; benzyl chloride, 1.0 mmol; eq Rx/eq Q(+)==28; toluene, 2 ml; water, 3 ml; reaction time, 48 hr at 90°C. bPlain clay was used as a solid phase. is because benzyl chloride is a more reactive substrate than n-butyl bromide. 6. Supported Cetyl Pyridinium Bromide for TC It is well known that the physical properties of clay can be modified considerably by ion-exchange of quaternary ammonium salts and related compounds [177,178]. On the other hand the adsorption of the cetyl pyridinium ion beyond the exchange capacity of Na and Ca-montmorillonite has also been shown [160]. In the present investigation 2 CBC cetyl pyridinium bromide-hectorite complex has been utilized for triphase catalysis. Cetyl pyridinium bromide was supported on the hectorite clay by the method indicated in the experimental section. The cetyl pyridinium-hectorite system exhibits 20.6 A diffraction peaks at loadings between 73 and 146 meq/100 g. The result is consistent with the data reported by Quirk et a1. According to their results [160], the observed maximum in the adsorption isotherm for cetyl -148- pyridinium bromide is 300 meq per 100 g montmorillonite. Therefore, at loading of 146 meq/100 g hectorite there is plenty of room for small anions to be accommodated within the interlayer. Examples illustrating the utility of triphase catalysis for some halogen exchange reactions are provided in Table 17. Here the results are given in terms of % yields. Pseudo first-order rate constant could not be determined, due to micell formation by the catalysis. The cetyl pyridinium salt itself acts as a catalyst in a biphase system. However the yield is lower than the corresponding triphase system (see Table 17). 7. Halogen Exchange Usinngrganic Salt Intersalates Several different organic salts (compound l-4, see Table 5) were intersalated and the resulting phases were used as triphase catalysis. The organic salts were supported on hectorite by the method described in the Experimental Section. X-ray diffraction measurement of the basal d(001) spacing confirmed that the compounds were indeed supported (Table 5), and in some cases the X-ray diffraction pattern was very similar to what is found for the metal chelated complex intersalation [see Fig. (16)]. These organic intersalates were used for halide ion displacement in l-bromobutane and 1-bromooctane. The experiments were conducted in the same manner as described previously in Section B—l. The rate data for the triphase catalyzed -149- Table 17. Phase Transfer Catalyzed Halogen Exchange Reactions Using Cetyl Pridinium Hectorite Complex (72 hr at 90°C).a Reactant Product (%) Yield Form of Catalyst ’AV/\‘Br ’AV/\~I 98 2CEC hectorite intersalation ’AV/\‘Cl ’AV/\‘I 59 2CEC hectorite intersalation ’A\/\‘Br ’A\/\‘Cl 55 2CEC hectorite intersalation /A\/\‘Br ’A\/\‘Cl 9 lCEC intercalation (homoionic) /A\/\‘Br /AV/\-Cl 39 Aqueous organic two-phase system using cetyl pridinium bromide “Reaction condition: hectorite clay, 1 mmol of substrate and 6 mmol of appropriate salt were taken. Toluene used as solvent, 0.1 gr The amount of quaternary ammonium salt is the same as any other intersalation system. -150- halogen displacement reactions are provided in Table 18. For all reactions pseudo first-order kinetics were observed. It is worthy to mention here that the difference in the reaction rate going from biphase to triphase systems for all these compounds is not dramatic. In addition, considering factors such as the molecular structure, the degree of layer ordering, the amount of anion present, hydrophobic interactions, ion pairing and the solvation properties of these molecules, it would be very difficult to rationalize the results. From the X-ray diffraction pattern of the Nile blue intersalate it seems that this compound is well ordered and similar to the M(chel)§+/XOi—- hectorite systems. However, the behavior of the Nile blue intersalate as a phase transfer catalyst was somewhat surprising. The rate of the exchange reactions under triphase conditions is only twice as large as the rate under biphase conditions. These organic salt-clay systems, like the organo- metallic—clay system mentioned earlier, show almost the same selectivity toward the longer size alkyl halide. The clay supported organic salts have advantages over the clay supported organo-metallic salts discussed earlier. The utility of transition metal complexes for a wide range of catalytic reaction is limited due to their reactivity with nucleophiles, but organic salts are less sensitive to nucleOphiles. -151- .ouHuouomcn mz nqu uHmm omHMOHvaH Gnu mo ucme>Havw umo o.~ mo coHuommH an pwummwum mums mummHmumo mmwcmHHB a .umxHaumo was moo.o .ONm HE m cH Homz HOEE OH .ocmsHou HE N CH moHEoun Hausa HOEE H uncoHqucoo coHuommma . F ..1 £5 ~m.o 85 $6 flammHumlmmovlzmlfwmon-M_ . _ . mmo mmo « a _ J 26 £6 :5 36 com ' m~.H «H.H Hm.o mm.H ow mm o>.m mm uHo+zmmomHnHmmou mpHEoum mcHEoum opHEoum mpHEoum umNHmumo Havoc Hanan Hmuoo Hanan m 0 m 0 taunt n x NOH Hulunc n x NOH wwmemumo mamanm mHmemumo mmmannB a u .003 um Homz msomsqd £DH3 mpHEoum quoo was H>uzm mo coHuomwm man How sumo moHumcHx .mH mHnma -152- 8. Kinetics of Cyanide Displacement on l—Bromooctane Phase transfer catalysis, as well as triphase catalysis procedures have been successfully utilized by several workers [48,83a] in cyanide displacement reactions involving simple alkyl halides. In the present work, tricapryl methyl ammonium chloride-hectorite complex (0.069 meq) was used as the solid phase along with 3 m1 aqueous potassium cyanide (3.33 M) plus 3 ml toluene solution of l-bromooctane (0.33 M). It was found that heating such heterogeneous mixtures at 90°C resulted in the production of l-cyanooc- tane in high yield. The typical procedure is outlined in Chapter II. Despite the complexities of the system, it was found that the above displacement reaction obeyed simple first-order kinetics, i.e., the rate of reaction showed a first-order dependency on the l-bromooctane concentration. Typical kinetics data for triphase systems as well as biphase are illustrated in Fig. (29). In this respect the kinetic reaction here shows a resemblance to that observed for the displacement of cyanide ion on l-bromooctane using the well develoPed "phase transfer catalysis" technique [48]. However, the details of the mechanism for this specific triphase-catalyzed system remains to be established. As was mentioned earlier, the utility of the transi— tion metal chelates intersalation compound for displace- ment reactions involving nucleophiles such as cyanide is O 0.2 O '3“ C O ‘6 8 0.4 E O .3 I..-1—_: \ '3‘ 0.6 C 2 g .D CEJ CH3 C T |.O Figure 29 -153- _. ° AMABr --+/VVV\CN o Kobsm) - |.09 (hr' ') O Kws(B)' L92 “1".” 0 0 b B A l J l 0.25 0.5 0.75 Time(hours) Plot of ln(unreacted) l-bromooctane in the organic phase as a function of time for the reaction of 2 mL of 0.5 M l-bromooctane in toluene with 10 mmol of sodium cyanide dissolved in 1 mL of water catalyzed by 0.069 meq of tricapryl methyl ammonium chloride ion pair in hectorite at 90°C. -154- limited due to the reactivity of these complexes. Herein lies the potential of organic salt intersalates. APPENDIX APPENDIX A. Calculation of d(001) Basal Spacing Using Bragg Law The following procedures were used to determine d(001) basal spacing and to assign the orders of reflection obtained. Equation (52) is the Bragg equation for a cubic system [179]. nA = [ZaO/(h2+k2+12)%] sine (52) 2+12)35 in Equation (52) is simply The factor ao/(h2+k the interplanar spacing d for the plane (hkl). The Bragg equation in its general form is then written n). = 2d sine (53) Multiplying the equation by 2n and dividing by d one can obtain ZunA/d = 411 sine or 21T/d = 41r sin 8/n1. By inserting the value for d we have 2w/d = (Zn/ao)n. By defining 2n/d = q, and 21t/ao = m, therefore, the equation can be simplified to q mn or 2n/d = mn. A plot of q versus n can be tried by least square technique to get the best fit of data point for different -155- -156- order of reflections. Here m is the slope of the line. Thus, the d value when n==l or d(001) basal spacing would be equal to Zn/m. For example, application of the above procdures using Braggs law for x-ray diffraction pattern of Ni(phen)§+/SOi--hectorite system indicate that the true d(001) to be 29.7 (see Fig. 30). B. Sample Calculation for the Amount of Catalyst The amount of Ni(phen)3SO complex needed in order 4 to prepare 2 CBC of 0.1 g of natural hectorite (with 2 CBC 4, 7 H20 a F.W. = 947.3 g/mole or 473.7 g/eq. 73 meq X0.1 g of 73 meq/100 g) is as follows: Ni(phen)3SO hectorite/100 g = 0.073 meq of the complex needed for 1 equivalent CEC of 0.1 g hectorite or (7.3'x10”5 eq). 7.3><10'5 eq><2 (factor for 2 CEC)><473.7 g/eq = 0.0692 9 711(1 needed for 2 equivalent CEC of 4’ 2 0.1 g hectorite. Now, 7.3 x10-5 eq of 503' subtracted from the total complex (i.e. 0.0692 g) needed of Ni(phen)3SO should be for 2 equivalent CEC. Hence 7.3><10-5 eq><(48 g/eq i”) 4’ that will wash out of the clay and 0.0692 g-0.0035 g = 0.0657 g of complex cation and so = 0.0035 9 so ion pair in 0.1 g of hectorite. Therefore 0.1 g hectorite + 0.0657 g = 0.1657 gr of Ni(phen)§+/SOi-- hectorite intersalate would contain 0.073 meq of ion pair which is taken to function as a catalyst. In other 2.. words, we need 146 meq of Ni(phen)§+ and 73 meq of SO4 2.5 2.4 2.3 2.2 2.( 2.0 (.9 (.8 (.7 I5 (.5 (.4 q (.3 (.2 LI (.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.( -157- (- (- (- . 27 /m=d:29.0A° (— - Ni(Phen)3 M0 04 - hectorite system L 1 J 1 1 l 1 l 1 O I 23456789l lll2|3l4l5l6l7l8 n Figure 30 Plot of ZW/d (d is basal spacing) versus n (order of reflection), using least square fitting technique. -158- to prepare 0.1657 g of Ni(phen)§+/SOi--hectorite intersalate. LIST OF REFERENCES 10. ll. 12. 13. LIST OF REFERENCES G. Brown, Ed., X-Ray Identification and Crystal Structure of Clay Minerals, Mineralogical Society, London, 1961. G.W. Brindly and G. Brown, Eds., Crystal Structures of Clay Minerals and Their X-Ray Identification, Mineralogical Society, London, 1980. H. Van Olphin, "An Introduction to Clay Colloid Chemistry", 2nd ed., John Wiley, New York, 1977, chapter 7. J.D. Russell and A.R. Frazer, Clays and Clay Minerals 12, 55 (1971). Thomas J. Pinnavaia, Intercalated Clay Catalysts, Science 220, 365-371 (1983). M.M. Mortland, J.J. Fripiat, J. Chaussidon, and Uytterhoeven, J. Phys. Chem. 61, 248 (1963). B.S. Snowder and Woessner, J. Colloid. Interface Sci. 30, 54 (1969). D.H. Solomon and D.G. Hawthorne, Chemistry of Pigments and Fillers, John Wiley & Sons, New York, (1983). B.K.G. Theng, "The Chemistry of Clay-Organic Reactions", Johy Wiley & Sons, New York, (1974). M.M. Mortland, Adv. Agron. 22, 75 (1970). M.M. Mortland, Trans. 9th Intern. Congr. Soil Sci. 1, 691 (1968). G. Lagaly and A. Weiss, The Layer Charge of Smectite Layer Silicates., Proceedings of the International Clay Conference, Mexico City, pp. 157—172 (1975). G. Lagaly, G. Schon and A. Weiss, Kolloid Zeitschrift und Zeitschrift ffir Polymere 250, 667-674 (1972). -159- 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. -160- H. Suquet, C. delaCalle, H. Pezerat, Clays Clay Miner. 23, 1 (1975). J.J. Fripiat and M.I. Cruz-Cumplido, Ann. Rev. Earth Plant Sci. 2, 239 (1974). A. Shimoyama and W.D. Johns, Nature Phys. Sci. 232, 140 (1971). W.D. Johns and A. Shimoyama, Amer. Ass. Petrol. Geol. Bull. 56, 2160 (1972). G.W. Bailey and J.L. White, Residue Rev. 32, 29 (1970). M. Schnetzer and H. Kodama in ”Minerals in Soil Environments", J.B. Dixon and S.B. Weeds, Eds., Soil Science Society of America, Madison, WI, 1977, Chapter 21. J.E. Bernal, "The Physical Basis of Life", Rutledge and Kegan Paul, London, 1951. M. Paecht-Horowitz, J. Berger, A. Katchalsky, Nature 228, 6361 (1970). M. Paecht-Horowitz, Angew. Chem. Int. Ed. Engl. 12, 349 (1973). J.J. Fripiat and G. Poncelet, Adv. Org. Geochem. Proc. 6th Intern. Meet., 875 (1974). N. Lahav, D. White, S. Chang, Science 201, 67 (1978). D.H. White and J.C. Erickson, J. Mol. Evol. 16, 279 (1980). (a) T. Endo, M.M. Mortland and T.J. Pinnavaia, Clays and Clay Minerals 28, 105 (1980); (b) T. Endo, M.M. Mortland and T.J. PIHnavaia, Clays and Clay Minerals 22, 153 (1981). Ming-Shin Tzou, Ph.D. Dissertation, Chemistry Department, Michigan State University, 1983. D.E. Woessner, J. Mag. Res. 32, 297 (1980). T.J. Pinnavaia in "Advanced Techniques for Clay Mineral Analysis", J.J. Fripiat, Ed., Elsevier, New York, 1981, pp. 139-161. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. -161- M.B. McBride in "Advanced Chemical Methods for Soil and Clay Mineral Research", J.W. Stucki and W.L. Banwart, Eds., Reidel, Holland, 1980, pp. 423-450. W.E.E. Stone in "Advanced Techniques for Clay Mineral Analysis", J.J. Fripiat, Ed., Elsevier, New York, 1981, pp. 71-112. J.J. Fripiat, "Advanced-Techniques for Clay Mineral Analysis" Elsevier, Amsterdam (1981). P.L. Hall in "Advanced Techniques for Clay Mineral Analysis", J.J. Fripiat, Ed., Elsevier, New York, pp. 51-75 (1981). W.E.E. Stone, in "Advanced Techniques for Clay Mineral Analysis", J.J. Fripiat, Ed., Elsevier, New York, pp. 71-112 (1981). B.K.G. Theng, "Formation and Properties of Clay- Polymer Complexes", Elsevier Scientific Publishing Company, New York (1979). V.J. Thielman and J.L. McAtee Jr., Clays & Clay Minerals 23, 173 (1975). V.E. Berkheiser and M.M. Mortland, Clays and Clay Minerals 25, 105-112 (1977). R.A. Schoonheydt, J. Pelgrims, Y. Heroes and Jan. B. Uytterhoeven, Clay Minerals 13, 435 (1978). M.F. Traynor, M.M. Mortland, and T.J. Pinnavaia, Clays and Clay Minerals 26, 318 (1978). R.H. Loeppert, M.M. Mortland, and T.J. Pinnavaia, Clays and Clay Minerals 21, 201 (1979). H. Van Olphen, "An Introduction to Clay Colloid Chemistry", 2nd ed., Wiley-Interscience, New York, p. 70 (1977). D.M. Clementz and M.M. Mortland, Clays & Clay Miner. 22, 223 (1974). M.M. Mortland and V.E. Berkheiser, Clays and Clay Miner. 24, 60 (1976). R.H. Loeppert, R. Raythatha, M.M. Mortland, and T.J. Pinnavaia, Preprints, Catalysis Soc. 6th North American Meeting, Chicago, (1979). -162- 45. R.H. Loeppert and M.M. Mortland, Clays and Clay Minerals.3l, 373 (1979). 46. C.M. Starks, J. Amer. Chem. Soc. 33, 195 (1971). 47. C.M. Starks and D.R. Napier (to Continental Oil Company) U.S. patent 3,992,432 (1976); British patent 1,227,144 (197 ); French patent 1,573,164 (1969); Australian patent 439,286 (196 ), Netherlands patent 6,804,687 (1968). 48. C.M. Starks and R.M. Owens, J. Amer. Chem. Soc. 33, 3613 (1973). 49. C.M. Starks and C. Liotta, "Phase Transfer Catalysis", Principles and Techniques, Academic Press, New York, (1978). 50. J. Jarrouse, C.R. Hebd. Seances Acad. Sci., Ser. C 232, 1424 (1951). 51. DuPont, British patent 632,346 (1949). 52. P. Edwards (to American Cyanamid), U.S. Patent 2,537,981 (1951). 53. Pest Control. Ltd., British patent 692,774 (1953). 54. R. Kohler and H. Pietsch (to Henkle), German patent 944,995 (1956). 55. H. Rath and U. Einsele, Melliand Textilber 39, 526 (1959); C.A. 33, 16531 (1957). 56. H.B. Copelin and G.B. Crance (to DuPont), U.S. patent 2,779,781 (1957). 57. Farbenfabriken Bayer, German patent 959,497 (1957); C.A. 33, 13665 (1959). 58. B. Graham (to Ethyl Corp.), U.S. patent 2,866,802 (1958). 59. G. Maercker, J.F. Carmichael, and W.S. Port, J. Org. Chem. 33, 2681 (1961). 60. Gavaert Photo-Produce N.V., Belgian patent 602,793 (1961); C.A. 33, 11491 (1963). 61. B.E. Jennings (to ICI), British patent 907,647 (1962). 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. -l63- M.A. Iskenderov, V.V. Korshak, and S.V. Vinogradova, vysokomol, Soedin 3, 637 (1962); C.A. 33, 9239 (1962). W.S. Port, British patent 912,104 (1962). R.W. Kay (to Distillers, Ltd.), British patent 916,772 (1963). F. Nerdel, British patent 1,052,047 (1966). B.C. Oxenrider and R.M. Hetterly (to Allied Chem.), U.S. patent 3,297,634 (1967). M. Makosza and B. Serafinowa, Rocz. Chem. 33, 1223 (1965). N.A. Gibson and J.W. Hosking, Aust. J. Chem. 33, 123 (1965). H.B. Hennis, L.R. Thompson, and J.P. Long, Ind. Eng. Chem. Prod. Res. Dev. 3, 96 (1968). A Brfindstrfim and U. Junggren, Acta Chem. Scand. 33, 2204 (1969). C.L. Liotta and H.P. Harris, J. Amer. Chem. Soc. 33, 2250 (1974). E.V. Dehmlow, Angew. Chem. Inst. Edn. 33, 493 (1977). H. Schnell, Angew. Chem. 33, 633 (1956). N.O. Bodin, et a1. Antimicrob. Agents Chemother. 3, 518 (1965). Beecham Group Ltd., British patent 1,364,672 (19 ). Chem. Eng. Prog. Symp. 33, 66 (1970). A. Molin, Dissertation (Ph.D.) Department of Chemical Technology, Royal Institute of Technology, 5-100 44 Stockholm, Sweden. P.S. Hallman, B.R. McGarvey, and G. Wilkinson, J. Chem. Soc. A.P. 3143 (1968). D.S. Allam and W.H. Lee, J. Chem. Soc. A.P. 426 (1966). V.L. Kheifets, N.A. Yakovleva, and B. Yakrasilschik, Zh. Prikl. Khim. (Lenningrad). J.B. Gordon and R.E. Kutina, J. Amer. Chem. Soc. 33, 3903 (1977). 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. -164- E.J. Fendler and J.H. Fendler, Adv. Phys. Org. Chem. 3, 271 (1970). (a) S.L. Regen, J. Amer. Chem. Soc. 97, 5956 (1975); (b) S.L. Regen, J. Amer. Chem. Soc. 3, 6270 (1976); (c) S.L. Regen, J. Org. Chem. 33, 875 (1977). S.L. Regen, Angew. Chem. 33, 421-429 (1979). D. Landini, A. Maia, F. Montanari, J. Amer. Chem. Soc. 100, 2796 (1978). Euchen Conference on Phase-Transfer Catalysis and Related Topics, Gargano (Italy), 1978. M. Makosza, E. Bialecka, Tetrahedron Lett., 183 (1977). (a) W.P. Weber, G.W. Gokel: Phase-Transfer Catalysis in Organic Synthesis, Springer, Berlin (1977); (b) E.V. Dehmlow, Angew Chem. 89, 521 (1977); (c) E.V. Dehmlow, Angew Chem. Ifit. Ed. Engl. 33, 493 (1977); (d) A. Brandstram, Adv. Phys. Org. Chem. 33, 267 (1977). E.H. Cordes, R.B. Dunlap, Acc. Chem. Res. 3, 329 (1969). C.A. Bunton, Prog. Solid State Chem. 3, 239 (1973). F.M. Menger, J. Amer. Chem. Soc. 33, 5965 (1970). J.M. Brown, and J.A. Jenkins, Chem. Comm. 1976, 458 (1976). M. Cinquini, S. Colonna, F. Montanari, and P. Tundo, Chem. Comm. 1976, 392 (1976). H. Molinari, F. Montanari, and P. Tundo, Chem. Comm. 1977, 639 (1977). P. Tundo, Synthesis 1978, 315 (1978). M. Tomoi, O. Abe, M. Ikeda, K. Kihara, and H. Kakiuchi, Tetrahedron Lett. 1978, 3031 (1978). H. Serita, N. Ohtani, and C. Kimura, Kobunshi Ronbunshu 33, 203 (1978). H.J.M. Dou, R. Gallo, P. Hassanaly, and J. Metder, J. Org. Chem. 33, 4275 (1977). P. Tundo, Chem. Comm. 1977, 641 (1977). 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. -165- J. Smid, S. Shah, L. Wong, and J. Hurley, J. Amer. Chem. Soc. 31, 5932 (1975). S. Shah and J. Smid, J. Amer. Chem. Soc. 100, 1426 (1978). E. Blasius and P.G. Maurer, Makromol. Chem. 178, 649 (1977). S.L. Regen, in "Catalysis in Organic Synthesis" (G.V. Smith, Ed.), Academic Press, New York, 1977, p. 119. H. Komeili-Zadah, H.J.M. Dou, and J. Metzqer, J. Org. Chem. 33, 156 (1978). E. Chiellini and R. Saloro, Chem. Comm. 1977, 231 (1977). E. Chiellini, R. Saloro, and S.D. Antone, Makromol. Chem. 178, 3165 (1977). S. Colonna, R. Fornasier, and U. Pfeiffer, J. Chem. Soc. Perkin I, 1978, 8 (1978). T. Yamashita, H. Bull. Chem. Soc. Yasueda, N. Nakatani and N. Nakamura, Japan 33, 1183 (1978). T. Yamashita, H. Yasueda, and N. Nakamura, Bull. Chem. Soc. Japan 33, 1247 (1978). (a) S. Ohashi and S. Inoue, Macromol. Chem. 150, 105 (1971); (b) S. Ohashi and S. Inoue, Macromol. Chem. 160, 69 (1972). S. Inoue, S. Ohashi, and Y. Unno, Polymer J. 3, 611 (197 ). S.L. Regen and L. Dulak, J. Amer. Chem. Soc. 33, 623 (1977). S.L. Regen, J. Amer. Chem. Soc. 33, 3838 (1977). M. Tomoi, T. Takubo, M. Ikeda, and H. Kakiuchi, Chem. Lett. 1976, 473 (1976). H. Normant, Angew. Chem. Int. Edn. 3, 1046 (1967). J.B. Shaw, D.Y. Hsia, G.S. Parries, and T.K. Sawyer, J. Org. Chem. 33, 1017 (1978). F. Rolla, W. Roth, L. Horner, Naturwissenschaften 33, 377 (1977). 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. -166- (a) C.G. Armistead, A.J. Tiler, E.H. Hambleton, S.A. Mitchel and J.A. Hochey, J. Phys. Chem. 33, 3947 (1969); (b) W.A. Aue and C.R. Hastings, J. Chromotogr. 42, 319 (1969); (c) H. Colin and G. Gviochon, ibId 141, 289 (1977); (d) P. Roumeliotis and K.K. Unger, ibid—333, 211 (1978). J.F. Fritz and J.N. King, Anal. Chem. 33, 570 (1976). I. Haller, J. Amer. Chem. Soc. 100, 8050 (1978). A.C. Zettlemoyer and H.H. Hsing, J. Colloid Interface Si. 33, 263 (1977). Amer. Chem. P. Tundo and Paolo Venturello, J. Soc. 101, 6606 (1979). S.L. Regen and C. Koteel, J. Amer. Chem. Soc. 33, 3837 (1977). I. Halasz, and P. Vogtel, Angew. Chem. Int. Ed. Engl. 33, 24-28 (1980). R.K. Iler, "The Chemistry of Silica", Wiley, New York, chapters 5 and 6, (1979). P. Tundo, Paolo Venturello, and E. Angeletti, J. Amer. Chem. Soc. 104, 6551-6555 (1982). P.C.H. Posher, Angew. Chem. Int. Ed. Engl. 31, 487 (1978). (a) S.L. Regen, S. Quici, S. Liaw, J. Org. Chem. 44, 2029 (1979); (b) G. Bram, T. Fillebeen-Khan, NT Geraghyt, Synth. Comm. 10, 279-289 (1980); (c) E. Angeletti, P. TundoT—P. Venturello, J. Chem. Soc., Chem. Comm. 1127 (1980). and M.M. Egorov et al., Zh. Fiz. Khim. 33, 1882 (1962). R.K. Iler, J. Colloid Interface Sci. 33, 25 (1976). T.J. Pinnavaia, R. Raythatha, J.G.S. Lee, L.J. Holloran, and J.F. Hoffman, J. Amer. Chem. Soc. 101, 6891 (1979). R. Raythatha and T.J. Pinnavaia, J. Catal. 31, 80 (1983). J.M. Thomas in "Intercalation Chemistry", M.S. Whittingham and A.J. Jacobson, Eds., Academic Press, New York (1982) Chapter 3. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. ~167- P. Monsef-Mirzai and W.R. McWhinnie, Inorg. Chem. Acta 33, 211 (1981). F.T. Bingham, J.R. Sims and A.L. Page, Soil Sci. Soc. Am. Proc. 33, 670 (1965). C. McAuliffe, M.S. Hall, L.A. Dean, and S.B. Hendricks, Soil Sci. Am. Proc. 33, 119 (1947). J.L. Ahlrichs, Bonding of Organic Polyanions to Clay Minerals, Dissertation Abstracts 22, 2121 (1962). _— S.K. De and R.K. Jaim, J. Indian Chem. Soc. 33, 379 (1964). M.B. McBride, T.J. Pinnavaia and M.M. Mortland, Am. Mineral 33, 66 (1975). G. Jacobs, F. Speeke, Acta Cryst. 3, 67 (1955). (a) R.A. Inskeep, J. Inorg. Nucl. Chem. 33, 763 (1962); (b) Inorganic Syntheses, vol. p. 228. The iron reagents, third edition, published by G.F. Smith Chemical Company, 1980. G.F. Smith, Anal. Chem. 33, 925 (1951). Lewis J. Clark, Anal. Chem. 33, 348 (1962). Man-Sheung Chan, J. Barry Deroos and Arthur C. Wahl, J. Phys. Chem. 31, 2163 (1973). R.S. Vagg, R.N. Warrener and E.C. Walton, Aust. J. Chem. 33, 1841 (1967). P.J. Taylor and A.A. Schilt, Inorganica Chimica Acta 3, 691 (1971). . (a) T.D. George and W.W. Wendlandt, J. Inorg. Nucl. Chem. 33, 395 (1963; (b) A. Braver, Ed., "Handbuch der Praparativen Anorganischen Chemie" Ferdinand Enke Verlag, Stuttgart, Germany (1954). E.H. Burstall and R.S. Nyholm, J. Chem. Soc. 3570-3579 (1952). Inorganic Synthesis 3, 37 (1939). 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. -168- R.A. Schoonheydt, J. Pelgrims, Y. Heroes, and J.B. Uytterhoeven, Clay Minerals 33, 435 (1978). A. Wada, N. Sakake, and J. Tanaka, Acta Cryst. B32, 1121 (1976). J. Baker, L.M. Englehardt, B.N. Figgis, and A.H. White, JCS Dalton, 530 (1975). P. Krumholz, Inorg. Chem. 3, 612 (1965). R.A. Palmer and T.S. Piper, Inorg. Chem. 3, 864 (1966). J. Hidaka and B. Douglas, Inorg. Chem. 3, 1180 (1964). P. Day and N. Sanders, J. Chem. Soc. 1530 and 1536 (1967). (A) . T. Ito, N. Tanaka, I. Hanazaki, and S. Nagakura, Bull. Chem. Soc. Japan 33, 365 (1968). S.F Mason, Inorgamica Chemica Acta Reviews 1-3, 89 (1968). D.J. Greenland and J.P. Quirk, Clays and Clay Minerals 3, 484 (1962). M. Mesrogli and G. Lagaly, Proceedings of the International Clay Conference, Italy, p. 201 (1982). D.M. Clementz, T.J. Pinnavaia, and M.M. Mortland, J. Phys. Chem. 11, 196 (1973). M.B. McBride, T.J. Pinnavaia, and M.M. Mortland, Am. Mineral. 33, 66 (1975). W.H. Quayle and T.J. Pinnavaia, Inorganic Chemistry 33, 10 (1979). M.B. McBride, J. Phys. Chem. 33, 196 (1976). R.G. Kooser, W.V. Volland, and J.H. Freed, J. Chem. Phys. 33, 5243 (1969). W.H. Schreurs and G.K. Fraenkel, J. Chem. Phys. 33, 756 (1961). Z. Luz, B.L. Silver, and C. Eden, J. Chem. Phys. 33, 4421 (1966). 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. -169- L.J. Berliner, ”Spin Labeling, Theory and Application, Academic Press, New York, 1976. T.J. Stone, T. Buckman, P.L. Nordio, and H.M. McConnell, Proc. Nat'l. Acad. Sci. USA 33, 1010 (1965). S.A. Goldman, G.V. Bruno, C.F. Polnaszek, and J.H. Freed, J. Chem. Phys. 33, 716 (1972). G.H. Schmid and D.G. Garratt, Can. J. Chem. 33, 1807 (1974). G.B. Schmid, V.M. Csizmadia, V.J. Nowlan, and D.G. Garratt, Can. J. Chem. 33, 2457 (1972). S.L. Regen, J. Heh, and J. McLick, J. Org. Chem. 33, 1961 (1979). F.T. Bingham, J.R. Sims, and A.L. Page, Soil Sci. Soc. Amer. Proc., 33, 670 (1965). C.L. Liolta, H.P. Harris, M. McDermott, T. Gonzalez, and K. Smith, Tetrahedron Lett., 2417 (1974). K.F. Clare, Nature 160, 828 (1947). F.X. Grossey, and L.J. Woolsey, Ind. Eng. Chem. 31, 2253 (1955). P. Harold Klug and Leroy E. Alexander,"X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials", John Wiley and Sons, Inc., New York, 1974. ' 93 03142 412 (n(((I((((((((((hflfmim(((((((((