4‘145151
).
._ r

 

‘ T _ _Ti11§ is to certify that til?
43,314 Lg. 8.5-3. L.

. msmwmm-dissertation entitled

‘

a“; WT?" WM '57 “Me” ”ti”?
Premdby -

Min/é 'SHW 72W

has been accepted towards fulfillment
of the requirements for

Wan: in CM

Major professor

 

Date (93. L9 1, mg

 

AfSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

PL. .cE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or betore date due.

DATE DUE DATE DUE DATE DUE

t

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Indltution

r-‘K

N r“
2‘ hi 0 O\

[XX 2‘81
0 .

   

 

CLAY CATALYSTS PILLARED 3v
‘ METAL HYDROXY POLYMERS

By

Ming-Shin Tzou

A DISSERTATION

Submitted to
Michigan State University
in partiai fulfiiiment of the requirements

for the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry

1983

 

ABSTRACT

CLAY CATALYSTS PILLARED BY HYDROXY METAL POLYMERS

by

Ming-Shin Tzou

By the intercalation oi polynuclear hydroxy cations
of chromium(III) and iron(III) in montmorillonite, new
pillared clay catalysts with pore size larger than 10 A
have been synthesized and characterized. Alumina pillared
clays derived from 50 wt % chlorhydroxy aluminum solution
(Al(0H)2.SCl0.5), marked under the tradename chlorhydrol,
have been studied also.

The 27Al-NMR measurements of freshly diluted 50 wt Z
chlorhydrol solution indicate the presence of a great
amount of rod-shape polymers; most of the polymers disasso-
ciate to "Al13" keggin ions upon aging at room temperature.
The pillared clay products, prepared from the diluted
chlorhydrol solution which has been aged at room temperature
for 2-6 weeks, have higher surface area and thermal

stability than those prepared from unaged solution. The

 

 

model of the pillar species made from fresh and aged
solutions are proposed.

Long aging times and aging temperature are essential
to Synthesize large polynuclear hydroxy chromium cations
in aqueous solution at OH/Cr = l.5-2.0 and to obtain
pillared clays with high basal spacing and large pore size.
At least one of the dimensions of hydroxy-chromium cations
determined by XRD is about l7 A. Upon calcination at 300°-
400°C in air, the hydroxy-chromium-montmorillonite undergoes
collapse. However, high spacing are retained if the clay
is heated in vacuo or in an inert atmosphere. The clays
pillared by chromia are active for the dehydrogenation of
cyclohexane with high selectivity.

The synthesis of iron pillared clays with basal
spacings larger than 25 A is achieved by interlayering the
polynuclear hydroxy iron cations which is prepared by the
hydrolysis of ferric ions (OH/Fe = 2.00) at room tempera-
ture. At the initial stage of hydrolysis (1.5 h), the
polycations are believed to be spherical with a diameter
of about l5 A. These polycations afford pillared products
with BET sdrface area near 350 mz/g. Upon aging at room
temperature for 7 days, the spherical polycations appear to
coalesce_and to form nods and/or rafts. The rods or rafts
in the interlayer cover more silicate surface area and

give pillared products with much lower BET surface area.

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to
Professor Thomas J. Pinnavaia for his very able guidance,
counseling, encouragement and support throughout the course
of this investigation.

I deeply appreciate Dr. Carl H. Brubaker, Jr. for
serving as my second reader. I am also thankful to Dr.

Max M. Mortland for providing special research facilities.

In addition, I wish to thank my colleagues Han-Min
Chang, Steve Christiano, Emmanuel Giannelis, Ivy Johnson,
Abbas Kadkhodayan, Edward Keller, Steven Landau, Kevin
Martin and Rasik Raythatha for their cooperation, dis-
cussions and friendship.

Many thanks go to Mr. Kermit Johnson for his assistance

of 27

Al NMR spectra, Ms. Beth Martin at AHDL and Agr.
Experimental Station of Michigan State University for
elemental analysis, Mr. Jason Lee for his patient proof
reading of this manuscript and Ms. Sharon Corner for her
expert preparation of the typescript.

I would like to thank the friends in Chinese Christian

Fellowship of Michigan State University for their concerns

and deep friendship in past years.

ii

Above all, I am also grateful to my family for their
considerable support and continuous encouragement pursuant

to this dissertation.

To My Mother

iv

TABLE OF CONTENTS

CHAPTER Page
LIST OF TABLES ......................................... v
LIST OF FIGURES ....................................... xi
CHAPTER I - INTRODUCTION .............................. l
I-A. Introduction ................................ l
I-B. Objectives ...................... , ............ 23
References .................................. 25
CHAPTER II - EXPERIMENTAL ............................. 29
II-A. Purification of Smectite Clays ............. 29
11-3. General Procedures for Pillaring
Reactions .................................. 30
II-C. X-ray Powder Diffraction Measurements ...... 3l
II-D. BET Surface Area Measurements .............. 3l
II-E. Thermal Analysis ........................... 32
II-F. Adsorption of Molecular Probes for
Pore Size Determination .................... 32
11-6. Chemical Analysis ........................... 32
II-H. Magnetic Resonance Spectroscopy ............ 33
11-1. UV-Visible and Infrared Spectra ............ 34

 

CHAPTER

CHAPTER III - ALUMINUM PILLARED CLAY CATALYSTS ........

III-A.
III-B.

Objectives ................................
Results and Discussions. ..................

l.

\1 05 01 ~15 (a)
O O

10.

ll.

Pillaring Reagents and Synthetic

Methods. ...............................

. Chemical Composition of Aluminum.
Pillared Clay Catalysts I & II --------

XRD and Surface Area ..................
DSC Studies .......... . ................
. Hydrogen Adsorption Isotherm ..........
. Hydrolysis Chemistry of Aluminum ......

. Al-27 NMR Spectra of Fresh and
Aged Chlorhydrol Solutions ............

7-a.
7-b.

. Al-27 NMR Spectra After Pillaring
Reaction ..............................

8-a.
8-b.

Proposed Model for Pillars in

Interlayer ............................

Attempts at Synthesizing Pillared

Clays with "A113" as pillar only ......

lO-a. Synthesis of "Al13" Keggin

Ions from Al + .................

lO-b. Selective Intercalation by

Ca2+-Montmorillonite ............

lO-c. Collection of "Al13" Keggin

Ions by Dialysis Method --------
The Effect of cec on Pillared Products -

vi

Fresh Chlorhydrol Solution ......
Aged Chlorhydrol Solution .......

Fresh Chlorhydrol Solution ......
Aged Chlorhydrol Solution .......

PAGE

35
35
38

38

39
41
47
47

52
55
55
6O
64~

64
64

67

69

69

69

7T
7T

CHAPTER PAGE

III-C. Conclusions ............................... 73
References ................................. 75

Chapter IV CHROMIUM PILLARED CLAY CATALYSTS ........... 78
IV-A. Objectives ................................... 78
IV-B. Results and Discussions..... ................ 79

l. Preparation of Hydroxychromium
Polymers ................................ 79

2. Pillaring Reactions of Na+-
Montmorillonite with Chromium

Solutions ............................... 81
3- Chemical Analysis ....................... 86
4. Thermal Treatment of Hydroxy-
Chromium-Montmorillonite.. .............. 88
5. Surface Area Results. ................... 96
6. Infrared Spectra and Pyridine ‘
Adsorption .............................. '96
7. Stability in O.l N HCl Solution ......... lOO
8. Sorption of Hydrocarbons . ............... 100
9. Catalytic Dehydrogenation
Reactions ............................... 102
l0. Dehydrogenation of Cyclohexane .......... 103
IV-C. Conclusions ................................. 108
References .... ................ . .............. 110
CHAPTER V IRON PILLARED CLAY CATALYSTS ................ 112
V-A. Objectives ................... . ............... 112
V-B. Hydrolysis Chemistry of Iron(III) ............. 113

CHAPTER

V-C.

V-D.

PAGE
Results and Discussions ................... 118
l. Synthesis of Hydroxy Iron(III)
PolymerS.............. ................. 118
2. Pillaring Reactions of NaT-Mont-
morillonite with Hydroxy Iron(III)
Polymers ............ . ......... . ........ 119
3. Aging Effects............. ............. l23
V-C-3-a. Aging for Different Time
Periods ..................... l26
V-C-3-b. Hydrolyzing at Low Tempera-
ture ......................... 130
V-C-3-c. Hydrolyzing at High Tempera-
ture ......................... 131
4. Chemical Composition, Thermal Proper-
ties and Hydrocarbon Adsorption of
Iron Pillared Clays .................... 135
5. Anion Effects .......................... l4l
Conclusions ................................ 146
References ................................. l49

viii

TABLE OF TABLES

Table Page
I-l Structural Properties of Some Important '
Zeolite Catalysts .......... .... ..... . ......... 3
1-2 Classification Scheme for Phyllosilicates
Related to Clay Minerals ........ ...... ........ 8
1-3 Idealized Structural Formulas for Some
Dioctahedral and Trioctahedral 2:1 Phyllo-
silicates ...... . ................ . ............ 10

III-l Chemical Composition of 50% MIN Chlorhydrol
Solution. ........ . ............ . .............. . 38

III-2 Chemical Composition and Chemical Formulas
of Wyoming Na+- Montmorillonite and Its
Pillared Products I and II Made From Fresh

and Aged Diluted Chlorhydrol Solutions ........ 40
III-3 XRD Basal Spacings and Surface Area of Alumina

Pillared Clays I and II .......... ..... ..... ... 45
III-4 Sorption of Probe Molecules by Alumina

Pillared Clays I and II..... .................. 49
III-5 Surface Area of Pillared Clays Made From

Smectite Clays with Different CEC ..... . ...... . 73
IV-l pH and Visible Adsorption Maximum of

Chromium Nitrate Solution Under Various

Conditions .................................... 83

IV-Z pH Values as a Function of Aging Condition.... 84

IV-3 Basal Spacings (A) of Air-Dried Chromium
Pillared Clays ...... ............. ............. 87

ix

Table Page

IV-4 Chemical Composition of Wyoming Na+-Mont-
morillonite and Its Pillared Product Made
From a Chromium Solution with 0H/Cr = 2.00
and Aged at 95°C for 36 Hours ..... . ........ .... 89

IV-5 Basal Spacings of Thermally Treated Chromium
Pillared Clays Made from Solution with 0H/Cr =
2.00 Aged at 95°C for 36 Hours.. ....... ........

IV-6 Assignment of Vibrational Modes of Pyridine
(Py) Adsorbed on Bronsted (H+) or Lewis (L)
Sites. ...... O ...... O ....... O 000000000000 O ...... 99

IV-7 Sorption of Probe Molecules of Chromia
Pillared Clays (Freeze-Dry) (OH/Cr = 2.00,

aged at 95°C for 36 Hours lOl

Amax (nm) for Hydrolyzed Iron(III) Solutions...ll9
V-2 Basal S acin s of Iron I I Pillared Cla 5
Made frgm Thgee Iron(IiI) olution ...... ¥ ...... ‘22

V-3 Basal Spacings of Iron(III) Pillared Clays
Made from 0.2'fl FeCl3 Aged for Various Time
Periods and at Different Temperature ........... l34

V-4 Chemical Composition and Unit Cell Formulas
of NaT-Montmorillonite and Iron(III) Pillared
Clay ........................................... 136

V-5 Sorption of Probe Molecules on Iron(III)
Pillared Clay. ................................. l37

V-6 Properties of Iron(III) Pillared Clays Pre-
pared by the Hydrolysis of Different Iron(III)
Salts ....... . ....... .. ......................... 144

III-l.

III-2

III-3

III-4

III-5

III-6

III-7

III-8

III-9

TABLE OF FIGURES

Framework Structure of Faujastic Zeolite .......
The Framework Structure of Smectite Clays ......

The Scheme of Metal Ion Hydrolysis
Mechanisms .................... . ................

X-Ray Diffraction Patterns of Alumina
Pillared Clay I. a) air dry; b) 350° C;
c) 700° C .............. . . ............. ... .....

X-Ray Diffraction Patterns of Alumina Pillared
Clay II. a) air dry; b) 350°C; c) 700°C .......

X-ray Diffraction Patterns of Alumina Pillared
Clay II. a) air dry; b) exposing the air day
under 100% humidity for 48 hours ............ ...

Differential Scanning Calorimetry Curves of
a) Na+- Montmorillonite; b) Alumina Pillared
Clay I. ........... . ............... ....... ......

Representation of [A11304(0H)24(H20)12]7

Keggin Type Ion ..... . ...... . .. .............

The Chemical Shift of a) Octahedral Aluminum

Al(H20)53+ and b) Tetrahedral Aluminum Al(0H)4‘.

The Proposed Hydrolysis Mechanism of Alumium

Ion From Hsu. .................... . ..............

The 27Al NMR Spectra of Freshly Diluted

Chlorhydrol Solution .......... .. . ............

The 27Al NMR Spectra of Diluted Chlorhydrol
Solutions a) Freshly Prepared; b) Aged at
Room Temperature for 1 week; c) 2 weeks;

d) 6 weeks; e) 3 months (colloidal particles

was removed from solution).. ...... ......... .....

xi

Page

4
ll

Zl

42

43

45

48

51

53

56

III-ll

III-12

III-l3

'111-14

IV-l

IV-2

IV-3

IV-4

IV-5

The 27AT NMR Spectra of Freshly Diluted

Chlohydrol Solution a) before the pillar-

in Reaction (diluted to same volume as in

(bi); b) after pillaring reaction., ............. 65

The 27Al NMR Spectra of Aged Chlorhydrol

Solution at 25° C for 2 weeks a) before the
pillaring Reaction (solution was diluted to

the same volume as in (b)); b) after pillar-

ing reaction ...................... . ......... ... 66

Proposed Models of Hydroxy Aluminum Polymers
in the Interlayer Region of Pillared Clays
I and II ........... ...... ...................... 68

The 27Al NMR Spectra of Aluminum0 Chloride
Solutions with OH/Al Ratio a) 0.0; b) l. 00;
C) T. 50; d) 2.00; e) 2. 25; and f) 2. 42. ........ 70

The 27Al NMR spectra of solution which was

aged at 25°C for 2 weeks and then dialyzed.

The NMR spectra is the solution outside

of the membrane tube. ............... . ..... ..... 72

UV- Visible Spectra of 0. l M Chromium Nitrate
Solutions (a) n = OH/Cr = U. 0, freshly pre

pared, b) n = l. 5, freshly prepared; c) n =

l. 5, aged at 95° C for 36 h; d) n = 2.0 freshly
prepared; e)n = 2.0, aged at 95°C for 36 h.... 82

DSC Curves of Chromia Pillared Clays a) in the
presence of air; b) under flowing of nitrogen
gas ............................................ 9l

Basal Spacings at Elevated Temperatures of

Chromia Pillared Clays Which Were Prepared

from 0. l M Chromium Solution with OH/Cr =

2.00 and Aged at 95° C for l h (X- X- X); 6 h

(A- --A o); and 33 h (D- D-Cl) ...................... 93

Basal Spacings at Elevated Temperatures of

Chromia Pillared Clays which were Prepared from

0. l M Chromium Solution Aged at 95° C for 36 h

with OH/Cr ratio 0.0 (X- X- X); 0.5 (0- 0- 0

l. 0 (0-0 -0); l. 5 (A- A- A); and 2.0 (D- -D- -D .... 94

XRD of Chromia Pillared Clay, Which was Pre-

pared from O.l M Chromium Nitrate Solution

With 0H/Cr = L 00 Aged at 95° C for 36 h,

Heated Under the Flowing of Argon at a) 25° C;

b) 200° C; c) 350°C; and d) 500° C ............... 95

xii

Ir

itali“,

 

Figure
IV-6

IV-7

IV-8
IV-9
V-l

V-2

V-3

v-4

V-5

Infrared Spectra of Montmorillonite Clays ......

a) Natural sediment montmorillonite;

b) The sample in (a) treat with pH 5 sodium
acetate;

c) Intercalated montmorillonite with hydroxy-
chromium polymers (air dry);

d) the sample in (c) was heated at vacuum
at 350°C for 3 h;

e) After adsorption of pyridine into (d)
sample for 6 h and then evacuated for
l h at 25°C;

f) After heating (e) at ll0°C for 3 h then
300°C for 3 h in vacuum..... ...... . .........

Hydrogenation-dehydrogenation-iSomerization
Reaction Scheme of Cyclohexane on Reforming
Catalysts Containing Acidic Sites ..............

Conversion yield of Cyclohexane Over Chromia
Pillared Clay Catalysts ........................

Adsorption Isotherm of Cyclohexane in the
Used Chromia Pillared Clay Catalyst. ...........

Proposed Structure of Hydroxy Iron(III)
Polymer, Saltman-Spiro Ball Model ..............

Visible Adsorption Spectra of 0.2 M FeCl3
Solutions Aged at 25° C for 24 h and with 0H/Fe
Ratio a) 0.0; b) 1.00 and; c) 2.00 ............

XRD of Iron Pillared Clays Made from 0.2 M
FeCl3 A ed at 25° C for 24 h and with OH/Fe
Ratio a 0.0; b) 0.5; c) l. 0; d) l. 5; e) 2.0;
and f) 2.5 .....................................

The Variation of pH Values of Iron(III) Solu-
tions and the dog] Spacing and A20 of 002
Diffraction Lines of Pillared Products with
Sgagg Time Periods of 0.2 M FeCl3 with 0H/Fe

XRD of Air Dry IronflII) Pillared Clays Which
are Prepared from 0.2 M FeCl3 Solution with
OH/Fe = 2.00 and Aged at 25° C for a) l. 5 h;

b) 10 h; c) 30 h; d) 50 h and; e) 75 h .........

xiii

Page

105

107

116

.120

128

Figure Page

V-6 XRD of Air Dry Iron(III) Pillared Clays Which
are Prepared from 0-2 M FeCl3 Solution with
OH/Fe = 2.00. The Hydrolysis Temperatures are
a) 0°C; b) 4°C; and c) l0°C. The XRD of (d)
is of the sample of (a) heated at ll0°C for 132
24 h .......................... . ................

V-7 XRD of Iron(III) Pillared Clays Which is Made
from 0.2 M FeCl Solution with OH/Fe = 2.00
and Aged at 25° for 24 h, at a) 25°C; '

b) 350°C and; c) 550°C ......................... l38
V-8 DSC Curve of Iron(III) Pillared Clay Which is

Made from 0.2 M FeCl Solution with OH/Fe =

2.00 and Aged At 25° for 24 h ................. 140
V-9“ XRD of Iron(III) Pillared Clay Whish is Made

From 0.l M Fe(N03)3 Solution with 0H/Fe = 2.00

And Aged At 25°C for 7 days, at Elevated
Temperature a) 25°C; b) ll0°C; c) 350°C;

d) 550°C and; e) after expose the (d) in air

for 7 months ................................... l47

CHAPTER I
INTRODUCTION

I.A. Introduction

Acid-modified clays were used extensively as petro-
leum cracking catalysts prior to World War II. By the
mid-l960's, zeolite containing catalysts became the main
cracking catalysts due in part to their unusually high
activity, selectivity and thermal stability. Smectite
clays are still used today as commercial catalysts but
only in minor quantities. The impact of zeolites on the
petroleum industry is comparable to that of the Haber
ammonia synthesis and Ziegler-Natta polymerization catalysts
that resulted in the award of Nobel prizes. It has been
estimated that in l970 more than l00,000 tons per year
of zeolites was consumed in petroleum cracking, which

‘0 liters

resulted in a conservation of more than 3.2 x 10
per year of crude oil in U.S.A.

The framework structure of zeolites are composed of
three-dimensional networks of Si04 and A104 tetrahedra
linked together through common oxygen atoms. The tetra-

hedra are arranged so that the zeolites have an open

framework structure, which defines a pore structure with
a high surface area. The pore structure varies greatly
from one-zeolite to another. In all zeolites, pore diame-
ters are determined by the free aperture resulting from
4-, 6-, 8-, 10-, or lZ-membered rings of oxygen atoms
which have maximum pore size values calculated to be 2.6,
3.6, 4.2, 6.3, and 7.4 A, respectivelyl’z.
The isomorphic substitution of Si by Al gives rise to
a net negative charge on the oxygen framework which is

compensated by cations. Zeolites generally can be repre-

sented by the formula
Mn/20.A1203.XS102.YH20

where M is the compensating cation with valency n. There
are 34 natural zeolites and nearly 100 synthetic zeolitesz.
However, based on their catalytic applications, the zeo-
lites of industrial importance are, in decreasing order:
synthetic faujasite (synthetic analogs types X and Y),
ZSM-5, mordenite, synthetic zeolite type A and erionite.
The unit composition, porosity and critical dimension of
these zeolites are listed in Table I-l. The framework

structure of faujasite is described by the line drawing in

Figure I-l.

 

LNANQLmVAANo_<Vmaz

 

m ~.m x m o.m mecca m ooLee_cm
m N.a domed m ~_A~o_mV~_ANo.<V~_az <
m 0.x x m K.o mecca NP m.mmA~oFmvh.mA~o_<ve.maz ooLeooeoz
m m.m mecca o_ omAwoLmva~o_<vmaz m-sz
m m.~ momaoeoesm ~_ om_A~o_mvmmA~o_<vmmaz x
m m.~ momaoeoasm N_ eo_5~o_mvomA~o_<vomaz x

newcono zou:_z Auwmogoa Lo maxh mcwz cmmxxo coppwmoasou ppmo u_== max»

 

 

mumz—mumu mu_~ooN u:mugoa5~ msom mo mo_ugmqogm Fagzuozgum ._-H o_nah

HEXAGONAL
PRISM

 

 

 

 

 

 

 

The main advantages of zeolites as catalysts in the
'petroleum industry are related to two important zeolitic
properties: the shape-selectivity and the extremely high
activity of these solids. The shape-selective properties
of zeolite catalysts, which were enunciated by pioneers
Weisz and Frilette3, result from their small, uniform
intracrystalline cavities and pores. The internal geometry
of zeolites permits facile sorption of reactant molecules
with dimensions less than those of the pore diameters.
Futhermore, only products whose dimensions are sufficiently
small to allow diffusion out of the intracrystal pores can
be formed.

The_greater catalytic activity of zeolite catalysts
relative to amorphous alumina-silica or acid-treated clays
arises primarily from the greater concentration of acid
sites in thepore cavity and the greater effective concentra-
tion of hydrocarbon in the vicinity of the active sites.
The latter factor results from the strong adsorption of
organic molecules in the micropore structure of the
zeolitesz. The electric field in the zeolite pores which
enhances the formation and reaction of carbocations through
polarization of C-H bonds also is an important factor.

For example, CaA zeolite with 5 A pore opening ex-
hibits selective cracking of linear paraffins but not

4

branched paraffins , dehydration of primary but not branched

alcohol4

, and hydrogenation of only n-olefins in a mixture

 

with iso-olefinss. Also, the disproprotionation and alky-
lation of toluene over ZSM-5 gives 90% p-xylene even though
the para isomer is thermodynamically disfavored relative to
the ortho- and meta-6.

In the cracking of crude oil faujasite-type zeolites
yield about 25% more gasoline than amorphous alumina-

silica catalysts7. Based on the molecular size selectivity

3, it is expected that large

principle described by Weisz
molecules might be difficult to diffuse into the pore
cavity. Indeed, in the cracking of hydro treated heavy
vacuum gas oil which contains C18-C25 (or higher) mono-,
di-, and trinuclear aromatics, the residual of polynuclear
aromatics was found to have limited accessibility to
zeolite poress. Usually, about 30% of the gas oil remains

after the cracking reaction.

Clay Structure and Properties

Clay minerals belong to the family of minerals known
as phyllosilicates. The principal building elements of the
clay minerals are two-dimensional layers of silicon-oxygen
tetrahedra and two-dimensional layers of aluminum- or
magnesium-oxygen—hydroxyl octahedra. If one octahedral
layer is condensed with one tetrahedral layer, it gives
rise to the 1:1 layer clay structure of kandites, e.g.
kaolinite and halloysite, with composition A12$i205(0H)49.
If one octahedral layer reacts with two tetrahedral layers,

a 2:1 clay layer mineral is the result. The latter clays

have the general composition A12S140]0(0H)2 and may be
found with various isomorphous substitution in talc,

pyrophyllite, micas, vermiculites and smectites1o.

In a unit cell of a 2:1 phyllosilicate formed from
20 oxygen atoms and four hydroxyl groups there are eight
tetrahedral holes and six octahedral holes. When two-
thirds of the octahedral holes are occupied by cations, the
smectite minerals are classified as octaphyllites or
dioctahedra (ngg, montmorillonite and beidellite). While
those smectite minerals in which all octahedra holes are
filled are called heptaphyllites or trioctahedra (eggg,
talc and hectorite).

Isomorphous substitution is observed among phyllosili-

4+ 13+

cate structures. The Si in tetrahedral holes and/or A

2+ in octahedral holes are substituted by cations with

or Mg
similar size but lower valency. As a result of isomorphous
substitution, many silicate sheets are negatively charged.
This positive charge deficiency is compensated by cations
located in interlamellar regions which may or may not be
exchangable. The amount of isomorphous replacement
determines the surface charge density on the silicate sheets
and influences the surface and swelling properties of the
layered silicates. The phyllosilicate classes which are
based on the composition, type of structure, and cation
exchange capacity of the clay minerals are given in

Table 1-21]. Each group of minerals is divided into two

 

 

 

mupcou:_pu meow: o_uumgm —mgcosmuuopsh
mavgomcmz meow: opuupgm pmgumgauoopc mo_: mpup_gm o.~
mapqomopga mmu_z Fmsumsmuuopgh .
ouw>oumzz mou_z —aguogmuoo_o new: o.p
ouw~:u_sgm>
pocvmgmuuoPLh mmpw_=u_sgm> ngumgmuuopsh
mu_p:opeLo>
Fmsumgmuuo_a mmu__:op5cm> _cgum;ouo0wa muwp=o_sgm> m.oro.o
mawcoosmm
mavgouom: mmp_coamm
.mumcoaam Lo mou_uoosm ~ogumsmuoowgh
ouwcogucoz ouwcoqmm
.ouFP—muwmm mu_=opp_gosu:oz -mu_:opp_gosp=oz
ma_:o___gosucoz so mmu_uoosm _mgoogmpuowo so muwuuosm m.c-m~.o
o—oh mo—ah pmguogmpuopgh
ope»
oo_PF»;aocsa mocwp_»;aocsa .aeoegaooo_a -oo_PF»;aocsa o .um
opwuomxggu mm:_u:magmm —mgcm=muuo_gh
oupcwpomx mou_:v_oax ~mgumzmuoopo mucvpomx o pup
a_=: mpzsgom
mmpomam azogmaam aaogc Log omgmsu camp
ppmpmgmcwz ampu ca nonpoma mmuau_FFmop—xza com osmgum :owumowm_mmmpu .~-_ m—am»

 

 

subgroups: di- and tri-octahedral.

Among these clay minerals, smectite is the most
attractive species in catalysis research because of its
swelling and exchangable cation properties. The struc-
tural formulas of smectite, as well as the formula of
talc for comparison, are listed in Table I-3.

The schematic framework structure of smectite is
illustrated in Fig. 172. As shown in this figure, the thick-
ness of the rigid two-dimensional silicate sheets is
approximately 9.6 A. The interlayer surface area of
smectite, which is calculated from the unit-cell weight,
770 g/mole and unit-cell surface area 5.25 A x 9.18 A, is
about 750 mzlg.

ESR studies of Cuz+

and other metal ion exchanged
smectites show that at low degrees of interlayer salvation
(EMgg, 1-3 water layers controlled by humidity) these
solvated metal ions are in a dynamic state and undergo
anisotropic rotation about specific molecular axes. Unco-
ordinated water molecules between the solvated cations are
capable of translational diffusion within and between the
"cages" defined by the solvated cations and the silicate

12. These results reveal that the interlamellar

layers
region is.effectively a two-dimensional solutions. As the
smectite interlayers are swelled beyond the dimensions of
the coordination sphere of the aquated metal ion, the com-

plex becomes solvent -separated from the surface and begins

10

 

efizcvomo
Ao.N_<c.o_mvno.em=H~¥

eazovo.moao~_<o.ewmvmo.e_<uwx

 

mupaomo—xsm oa_>oum:: mmu_z
«AzovomoAx_<x-o.m_mv
Ho.sadgo I». e\x=
ou_=ogu:oz
«AzovomoAx_<x-o.mLmv eazovomoAxp<x-o.m_mv
: x . .
no. emzuom I». \ :2 Ho aP<HoN=».eH“z
oo_=eaam aoP._oe_om
eAL.=ovo~vo.mLmv aAzovowvo.m_mv
m Lox- o. omzHON z». e\xz fix-o.e_<uo~ I». e\x=
muvgouum: mpvcop—_sosu:oz mmp_uomam
o_a» oopppseaecsa o_eh - oow__»=aocsa
.mgcosmuoovgh Pogumgmuoo_o aaogw Fmgm:_z

 

ngcmgmuoopgh wen

.mouao_._me__aea FUN
pmgvmzmuoowo meow sow mmP:ELom pmgzuusgum umNPPmouH

.muu m—nmh

 

11

Figure I-2. The Framework Structure of Smectite Clays

12

BEIDELLITE

 

 

 

NONTRONITE

040
A13. FOR 31

 

13

to tumble rapidly13. Indeed, in the water, the interlayer

spacing of Na+-smectite can be swollen up to infinite such
that each layer is totally solvated by water molecules in

14. However, the

the form as “house-of-cards" structure
extent of interlayer expansion of smectite varies from one
solvent to another solvent. For instance, the lattice
expansion in the 001 direction of Na+-montmorillonite when
swolled with methylene chloride, methanol and benzene is
10.0 A, 7.7 A and 5.7 A respectively. In the catalytic
hydrogenation reaction of alkynes and alkenes in methanol
over intercalated Rh(PPh3)2+, it is found that the spatial
region available for reaction is primarily restricted by
the thickness of the adsorbed methanol layer. Thus the
interlayer spacing relative to the substrate size influ—
ences the catalytic selectivity of the Rh(PPh3)2+ catalyst.
As mentioned above, even if the solvent is water, the
interlayer spacings of smectite can be separated by one,
two or three molecular layers of water by controlling its
environment humidity., Under these restricted interlayer
conditions, hydrated multivalent metal ions such as Al3+,

Cr3+ or Fe3+

are active for proton assisted reactions of
organic molecules, (e.g., ester formation from carboxylic

acid and alkene and elimination of hydrogen sulfide from
16)

thiols Apparently, substrate molecules can penetrate
into the restricted interlayer to accomplish these kinds of

chemical conversion.

14

When the guest molecules such as water or organic
solvent molecules are removed, the silicate layers come
together (van der Waals contact) and the interlayer space
no longer is available for adsorption. Thus interlayer
catalysis cannot occur. However, the interlayer region
can be permanently propped open by insertion of robust
cations acting as molecular props or pillars to allow the
expanded clay minerals to sorb guest molecules and to
function zeolites or molecular sieves.

d17 showed that if the

In 1955, Barrer and McLeo
.4.
(CH3)4N

O
tites, silicate sheets were permanently separated by 4.1 A

and (C2H5)4N f were exchanged into the smec-

and 4.5 A, respectively. The space between the interlayer
alkyl ammonium cations provided intracrystalline volume
readily accessible to various non-polar as well as polar
molecules.

In 1976, Mortland g£_gl. and Shabtai g£_gl. indepen-
dently introduced a divalent cage-like ammonium ion, the
diportonated complex of 1,4-diazabicyclo [2,2,2] octane,
also called triethylene diamine (TED), into montmoril-

18,19 18

lonite and vermiculite ‘as molecular props to expand

the interlayer. The d(001) spacing was 14.8 A, which
I 0
corresponds to a lattice expansion of 5.2 A. The surface

+ + . .
2 2 -verm1cu11te were

area of TED -montmorillonite and TED
280 and 144 mz/g, respectively, and the free distance

0
between two adjacent cations in montmorillonite was 3 6 A.

15

Both clays were shown to be active in the catalytic con-
version of acetonitrile to acetamide and esterification
of carboxylic acid with alcohol.

Another exchange cation which was used as a pillar
to expand smectite and vermiculite is M(phen)32+, where
M is Fe, Co, Ni or Cu and phen is 1,10-phenanthroline.
The basal spacings of the intercalated and intersalated

AZO’ZI, respectively.

montmorillonite were 17.3 A and 29.0

Because the pillars in the systems contain C-C
bonds, they are not stable at high temperature and their
basal spacings and surface area decline at 250° 5 350°.
However, these pillared systems can be utilized as mole-
cular sieve for selective adsorption and as catalysts
below their decomposition temperature.

A more attractive and heat-stable expanded-clay based
on cationic oxypolymer have been investigated. ThoSe are:
A113o4TOH)§{_x22‘27, zr4(ou)f128, [Cr(0H)3_x]:X+29,
[Si(0H)4]x30’31 and Bi6(0H)$;32. These hydroxy metal poly-
mer-smectite materials posses several novel features.
First, after calcination, the intercalated metal hydroxy
polymer is dehydrated to become a metal oxide, which main-
tain the lattice expansion even at high temperature.
Apparently, these expanded clays are more thermally stable
than those expanded by organic pillars. Second, the syn-

thetic conditions can be varied to make metal-hydroxy poly-

mer of different size and thereby used to change the

16

apparent pore size of the pillared clay. Third, the metal
oxide pillars can potentially be activated by oxidation/
reduction or other treatments to become an active site for
catalytic reactions.
Generally, there are three methods for synthesizing

hydroxy-metal polymer smectites.

1. Metal Complex Hydrolysis

In this method an intercalated metal complex
precursor is hydrolyzed to form a hydroxy-metal polymer in
the interlamellar region. The method is demonstrated by

the Si(acac)3+-smectite system30’3].

 

 

+

a ———> Si(acac)3+ + N + (l)

. +
Si(acac)3 + N a

 

 

 

 

Si(acac)3+ + H20 ——-.» s~;(ou)4 + H” + 3(c113c0)2cr12 (2)

 

 

where acac is acetylacetonate. When the intercalated
Si(acac)3+ is hydrolyzed to Si(0H)4, another Si(acac)3+
will move to the interlayer to replace the bound proton and
subsequently be hydrolyzed. Aggregation of the Si(0H)4
occurs to form an interlayer silicic acid oligomer.

2. In Situ Hydrolysis orTitration Method

In this method a hydroxy-metal polymer is formed
in the interlayer by the addition of alkali solution to a
mixture of metal salt and a clay suspension. The reaction

can be illustrated by the following scheme.

17

 

 

n+

Mn+ OH' M(0H)(n-l)+ + H+ lgnM

 

 

 

«(oh)‘"“)+ + l/nM"+ (t) (3)

 

 

a .... ——> Th<om,.1£"'“”‘* + H" <4)

 

Materials made by this method contain monolayer metal-
hydroxy polymers in the interlayer with basal spacing,
d(001) near 14.8 A. Eventually, the two-dimensional
metal-hydroxy polymer will fill the interlayer space to
form a brucite-type layer or chlorite-type structure33’34.
When the base to metal ion ratio (meq/mmole) is close to
the oxidation state of of the metal ion, metal-hydroxide
precipitates can deposit on the external surface of clay
sheets. This latter result is generally undesirable be-
cause the pillared clay becomes contaminated by the co-
precepitated metal hydroxide.

3. Direct Exchange Method

Cationic hydroxy-metal polymer can be synthesized
by the hydrolysis method and directly exchanged into the
interlayer of smectites.

-——:

n+ ~ '——_'—_33 +
Mx(0H)y + N ———e» MX(OH) + n N (5)

a Z a

 

18

In principle, the lattice expansion of the pillared
clay can be increased in proportion to the size of the
hydroxy-metal polymers. To obtain a regular expanded-clay,
the oxymetal polymer complexes should be in high concen-
tration and its molecular weight or size should be ‘
narrowly distributed. If the shape of the polymers are
two-dimensional, they will occupy most of the interlayer
surface. Subsequently, they leave small internal surface
area and pore volume for adsorption and catalysis. 0n the
other hand, if the polymers are grown in one-dimension,
i.e. chain-like polymers, they will occupy most of the
interlayer surface and give an irregular pore size dis-
tribution.

In the present work, the direct exchange method is
employed to synthesize pillared clays. The hydrolysis of
the metal ions to form polymeric species becomes the most
important step in the strategy leading to new pillared
clays.

When the salts of metallic elements, especially
those of 4+, 3+, and small 2+ ions, are dissolved in water,
the positive charge on the metal ions will facilitate the
loss of a water proton. The process can be visualized by
following reaction equations:

+' - + -
Mayxby-CHZO e——--‘ a [M(H20)d] y(aq) + b X 2(en) (6)

19

[M(H20)d]y+ + H20 ;::= [M(H20)d_1(OH)]+(y']) + H30+(7)

It is possible for reaction (2) to continue until all
of the water molecules Coordinated to the metal ion have

lost a proton.

[M(H20)d_](0H)]+(y’1) + (d-l) H20 :

[M(0H)d]*(y‘°) + (d-l) H30+ ’ 5(3)

Such a reaction will not lead to cationic aggregation.
This is brought about by the interaction of two or more
partially hydrolyzed cationic species, as depicted in
equation (4).

0H +(Y'1) 0H +(Y'1)
(H o)~ M + [(H 0) M ] ;===.
[ 2 d-z \\0H2 2 d-2 \\0H2
0H 2(Y'])+
,/ \\
[(H20)d_2M\OH/ "(1120)” ] + 21120 (9)

The elimination of two molecules of water between two
partially hydrolyzed metal ions leads to the formation of
a dinuclear hydroxo-bridged entity "olation". An extension

of this process could lead to infinite polynuclear species.

20

’0”\ /0H (2y-3)+ +(y-l)

[(H20)d_2M\OH/M(0H2)d_4 ] + [M(H20)d_1(0H)] :2:
\
0H2
0H 0H +(3y‘4)
z’ \\ z’ ‘\
H o M M H o M H o __,_

U 2 )d'2 \on/ ( 2 )d-4\0H/ ( 2 )d-2 .— ---- (10)

Through the influence of high temperature, prolonged aging
and/or high pH, the hydroxo-bridges of polynuclear species
may undergo "oxolation" reactions as in equation (6) and

(7)

x+ x+

OH OH
__ / \ / \ __ __s ___ - _ _ _ __- +
[M\0H/M\OH/M ] .— [ M o M o M ] .H20 (11)

x+ (x-4)+
[---M”0H\‘M”0H\\M---] ;:: ---M’/O\\M’/O\‘M---
\OH/ \OH/ ,' \o/ \o/
+ 411“ (12)

When the degree of aggregation is large enough to form
high molecular weight polymers, colloidal particles will
form. Eventually the metal-hydroxide will precipitate upon
further hydrolysis. The overall scheme for the metal-ion
hydrolysis processes can be depicted as in Figure I-3.

A number of methods both chemical and physical have
been used to establish the presence of polynuclear species.

The methods employed most are spectrophotometric

21

Figure I-3. The Scheme of Metal Ion Hydrolysis
Mechanisms

22

 

 

8.3.3: .222 safes:

 

 

la>0 IQDVA—I-OVO 2
3.83500
Loo-oacron.

032:4

....‘I

 

 

 

 

O O O O —+> > O 0
ll 7592i soill ......

 

_+.. a II x ..
1.1.7339 2 £0.

0 OOOT+ANII>nV

N

+3.»...

n n
£9 2i 17539 2

L

N

N

N N
i«.539 2.1: f9 .2

+ :LN.
333800 622523. 2:230

 

.8388 32355.2

uouobetbbv

‘
V

 

 

 

£2.05»:

 

23

methods35-37 38-40

41-44

, ultracentrifu-
45-47

, light-scattering methods

, polynuclear magnetic resonance and

d48‘50 .

gation
coagulation metho The light-scattering method and
ultracentrifugation methods are the typical techniques to the
study of high polymers. Both provide molecular weight infor-
mation. From the coagulation experiment, one may estimate
the change on the predominant complex ion present in
.solution.

Sillen and his co-workers have developed an elegant
method, based on e.m.f. measurements, for demonstrating

51’52. The method

the existence of polynuclear species
is the so called "core plus links" hypothesis and assumes
that the composition of all polynuclear complexes formed
in a given solution can be explained in terms of a °core"
plus a certain number of 91inks" and that no other complexes
are formed. Extensive studies revealed that most systems

fit or approximately fit this simple mode153'56.

I-B. Objectives of Thesis Research

The crystalline aluminosilicates known as molecular
sieve zeolites, have been widely used as adsorbents and
catalysts partly due to their characteristic pore openings
(31,, Section I-A). The largest pore opening found among
the known zeolites is a ring composed of 12 oxygen atoms
and an aperature of 7.5 A. Because the pore openings are
limited, the utility of zeolites is also limited for the

catalytic conversion of small molecules.

24

Recently, the successful synthesis of pillared clay
has received wide attention due to the large pore open-
ing and Bronsted acidity that make them potential
catalysts for proton assisted reaction of large molecules,
as in the cracking of gas oil. Alumina pillared clay has
been studied in greatest detail. However, the nature and
composition of aluminum-hydroxy polymers in the interlayer
region and their effects on the properties of pillared
products are not understood. One of the objectives of this
research is to investigate the factors that influence the
pore size ranges andlayer ordering in alumina pillared clays.

The framework of zeolites are mainly composed of A1
and Si, though in some zeolites aluminum may be substituted

58-60. Conversely, pillared

by iron, born and germanium
clays can have a wide variety of substituents, since many
different metal oxides can act as pillars. As alluded to
in Section I-A, the crucial step in the synthesis of
pillared clays is synthesizing large cationic hydroxy-
metal polymers, of the type MX(OH);+. Many of these
polymers are known. Thus, the known pillared clays, e.g.
Al, Zr, Bi and Cr could be the initial members of a large
family of materials. The other objective of this thesis
research is to synthesize new pillared clays to add new

members to this class of catalysts and to examine their

properties.

11.

12.

13.
14.

25

REFERENCES

Barrer, R.M. Chem. Ind. (London), 1968, 1203.

Breck, D.W. "Zeolite Molecular Sieves", Wiley, New
York, 1974.

Weisz, P.B.; Frilette, V.J. J. Phys. Chem. g4, 382
(1960)

Weisz, P.B.; Frilette, V.J.; Maatman, R.W.; Morver,
E.B. J. Catal. l, 307 (1962).

Weisz, P.B. Chem. Tech. 498 (1973).

Chen, N.Y.; Kaeding, W.W.; and Dwyer, F.G. J. Amer.
Chem. Soc. 101, 6783 (1979).

Nace, D.M. Ind. Eng. Chem. Prod. Res. Dev. 8(1),
24 (1969).

Owen, H.; Snyder, P.W.; Venuto, P.B. Proc. Sixth
Int. Congr. Catal. 2, 1071 (1977).

Pauling L. Proc. Nat. Acad. Sci. USA lg, 578 (1930).

Grim, R.E. "Clay Mineralogy" 2nd Ed., McGraw-Hill,
New York, 1968.

Theng, B.K.G. "The Chemistry of Clay-Organic Reactions"
Tohn Wiley, New York, 1974.

Hall, P.L. in "Advanced Techniques for Clay Mineral
Analysis", Fripiat, J.J., Ed., Elsevier, New York,
1981. pp. 51-75.

Woessner, D.E. J. Magn. Res. g2, 297 (1980).

Van 01phin,rL "An Introduction to Clay Colloid Chemis-
try“, 2nd Ed., John Wiley, NY 1977.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

26

Pinnavaia, T.J.; Raythatha, R.; Lee, J.G.S.; Halloran,
L.J.; Hoffman, J.F. J. Am. Chem; Soc. 101, 6891 (1979).

Thomas, J.M. in “Intercalation Chemistry", Ed. by
Whittingham, M.S. and Jacobson, A.J. Academic Press,
New York, 1982, Chapt. 3.

Barrer, R.M.; MacLeod, D.M. Trans. Faraday Soc. 21,
1290 (1955).

Mortland, M.M.; Berkheiser, V.E. Clays and Clay Miner-
als 23, 60 (1976).

Shabtai, J.; Frydman, N.; Lazar, R. Proc. 6th Intern.
Cong. on Catalysis, London, July 1976, B-5.

Berkheiser, V.E.; Mortland, M.M. Clays and Clay
Minerals 2;, 105 (1977).

Loeppert, R.M.; Mortland, M.M.; Pinnavaia, T.J. Clays
and Clay Minerals 21, 201 (1979).

Vaughan, D.E.W.; Lussier, R.J. 5th Int. Conf. Zeolites,
Naples, Italy, June 2-6 1980.

Lussier, R.J.; Magee, J.S.; Vaughan, D.E.W. 7th Cana-
dian Symposium on Catalysis, Edmonton, Alberta, Oct.
19-22, 1980.

Lahav, N.; Shani, U.; Shabtai, J. Clays and Clay
Minerals 22, 107 (1978).

Shabtai, J.;.Lazar, R.; Oblad, A.G. Proc. 7th Intern.
Cong. Catalysis, Tokyo, Japan, July 1-4, 1980,
paper B8.

Brindley, G.W.; Sempels, R.E. Clay Minerals 12, 229
(1977).

Occelli, M.L.; Tindwa, R.M. Clays and Clay Minerals
21, 22 (1983).

Yamanaka, 5.; Brindley, G.W. Clays and Clay Minerals
‘21, 119 (1979).

Brindley, G.W.; Yamanaka, S. American Mineralogist
25, 830 (1979).

Endo, T.; Mortland, M.M.; Pinnavaia, T.J. Clays and
Clay Minerals 22, 105 (1980).

Endo, T.; Mortland, M.M.; Pinnavaia, T.J. Clays and
Clay Minerals 22, 153 (1981).

32.

33.

34.

35.

36.
37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

27

Yamanaka, S.; Yamashita, G.; Hattori, M. Clays and
Clay Minerals 22, 281 (1980).

Yamanaka, 5.; Brindley, G.W. Clays and Clay Minerals
‘22, 21 (1978).

Brindley, G.W.; Kao, C.C. Clays and Clay Minerals 22,
435 (1980).

Milburn, R.M.; Vosburgh, W.C. J. Amer. Chem. Soc. 11,
1352 (1955).

Milburn, R.M..J. Amer. Chem. Soc. 12, 537 (1957).

Sutcliffe, L.H.; Weber, J.R. J. Inorg. Nuclear Chem.
12, 281 (1959).

Tobias, R.S.; Tyree, S.Y., Jr. J. Amer. Chem. Soc.
21, 6385 (1959).

Tobias, R.S.; Tyree, S.Y., Jr. J. Amer. Chem. Soc.
22, 3244 (1960).

Gimblett, F.G.R. J. Inorg. Nucl. Chem. 22, 2887
(1971).

Holmberg, R.W.; Kraus, K.A.; Johnson, J.S. J. Amer.
Chem. Soc. 12, 5506 (1978).

Johnson, J.S.; Krans, K.A.; Scatchard, G. J. Phys.
Chem. 22, 1034 (1954).

Johnson, J.S.; Scatchard, G.; Kraus, K.A. J. Phys.
Chem. 22. 787 (1959).

Aveston, J.E.; Anacker, E.W.; Johnson, J.S. Inorg.
Chem. 2, 735 (1964).

Momii, R.K.; Nachtrieb, N.H. Inorg. Chem. 2, 1189
(1967).

Akitt, J.W.; Farthing, A. J. Magn. Reson. 22, 345
(1978).

Bottero, J.Y.; Cases, J.M.; Fiessinger, F.; Poirier,
J.E. J. Phys. Chem. 21, 2933 (1980).

Matijevic, E.; Janauer, G.E. J. Coll. Interf. Sci.
‘21, 197 (1966).

Matijevic, E.; Stryker, L.J. J. Coll. Interf. Sci.
22, 68 (1966). '

50.

51.
52.
53.

54.

55.

56.

57.

58.

59.

60.

28

Matijevic, E.; Mathai, K.G.; Kevker, M. J. Phys.
Chem. 22, 1799 (1966).

Sillen, L.G. Acta Chem. Scand. _3_, 299 (1954).
Sillen, L.G. Acta Chem. Scand. B, 318 (1954).

Hietanen, S.; Sillen, L.G. Acta Chem. Scand 2, 1607
(1954).

Ingri, N.; Lagerstrom, G.; Frydman, M.; Sillen, L.G.
Acta Chem. Scand. ll 1034 (1957).

Sillen, L.G. Acta Chem. Scand. 12, 186 (1956); 11,
1034 (1957).

Hietanen, S.; Sillen, L.G. Acta Chem. Scand. 22, 265

Vaughan, D.E.W. Abstracts, 18lst ACS Annual Meeting,
Atlanta, March 1981.

Marosi, L.; Stabenow, J.; Schwarzmann, M. Ger. Offen.
2,831,611. . .

Barrer, R.M.; Freund, E.F. J. Chem. SOC. Dalton,
1049 (1974).

Barrer, R.M.; Freund, E.F. J. Chem. Soc. Dalton,
2123 (1974).

CHAPTER II
EXPERIMENTAL

II-A. Purification of Smectite Clays

Ca2+

-montmorillonite from Apache County, Arizona,
contains carbonates and free ferric oxide which can be
eliminated by treatment with sodium acetate buffer solu-
tion and sodium dithionite/sodium citrate, respectively.
The mineral is converted to the sodium ion exchanged form
by stirring it with 1 M NaCl solution for one day. The
excess salt is removed by centrifugation and then dialysis.
To each 5 gram portion of clay in 1% suspension is added
150 m1~l N sodium acetate which is buffered to pH 5 by
acetic acid. The solution is digested for 3 hours at

75° - 80°C with sitrring and then centrifuged to remove
excess salt. Usually, the addition of some sodium chloride
makes the centrifugation process easier. The solution is
then redispersed in enough water to make a 1% suspension
and 40 m1 oflLBN Na-citrate and 5 m1 of l N sodium bicar-
bonate is added. This suspension is warmed carefully to
75° - 80°C and 1 gram of sodium dithionite, Na25204, is

added with stirring. The suspension immediately turns from

29:

30

reddish-brown to milk -white upon the addition of sodium
dithionite. This solution is then kept at a temperature
below 80°C for 30 minutes. If the temperature is higher
than 80°C, it will cause the black precipitation of
ferrous sulfite. The mineral is washed with deionized
water and then dialyzed to remove all possible free ions.
Finally, it is allowed to sediment to remove impurities
and the fraction of particles less than Zu are collected.
The montmbrillonite from Wyoming and hectorite from
California occur as the sodium ion exchanged form. Both
minerals contain carbonates and trace amounts of free ferric
oxide. They can be purified by treatment with sodium
acetate buffer solution, using the same procedure described
above for Arizona montmorillonite. The fraction of parti-
cles less than Zu was collected by sedimentation prior to

the purification treatment.

II-B. General Procedures for Pillaring Reactions

A 1% suspension of the purified smectite clay is added
slowly (100 mL/0.5 h into the desired base-hydrolyzed
metal ion solution containing hydroxy-metal polymers.
Usually the ratio of metal (mmol) to clay (meq) is be-
tween 10 and 50. After complete addition of the smectite
suspension, the mixture is stirred for another two hours
to allow complete exchange. It is then repeatedly washed
with deionized water and centrifuged until all excess

salt is removed. The washing procedure may have to be

31

repeated ten or more times. The resulting intercalation

compounds are then air-dired or freeze-dried, as desired.

II-C. x-ray Powder Diffraction Measurements

A Philips or Siemens X-ray diffractometer with Ni-
filtered CuKa radiation (A = 1.5405 A) was used to measure
basal spacings. One ml of freshly prepared expanded-clay
solution (0.5-1.0 wt%) was allowed to air-dry directly on
a 1 inch x 1 inch glass slide. The diffractograms were
usually recorded from 20 = 2° to 35°. Basal spacings were
determined for samples dried both at room temperature and at
'elevated temperature for 3 hours. The peak positions in
the angle 26 were converted to d spacings with a standard

chart (Bragg equation).

II-D. BET Surface Area Measurements

Surface areas were determined with a Perkin-Elmer
Shall Model 212-8 Sorptometer at liquid nitrogen tempera-
ture using N2 as the adsorbate and He as the carrier gas.
All samples were degassed at 350°C for 3 hrs under flowing
argon unless otherwise noted. Approximately 75 mg of
expanded clay was needed for each measurement. The adsorp-
tion data were plotted according to the BET equation.
Usually, 3 data points were collected and the slope and
intercept for the BET equation was determined by a least

squares method.

32

II-E. Thermal Analysis

DSC analysis of pillared clays was obtained using a
DuPont 990 thermal analyzer which was operated in a dif-
ferential scanning calorimetry mode. The samples were
prepared by using non-hermitic, high thermal conductive
aluminum pans and covers. DuPont thermal analyzer grade
aluminum oxide was used as a reference. Most of the scans
were carried out at a constant heating rate of 5°/min in
the range of 25° to 500°C. Usually a 10 mg sample and

reference were needed for each analysis.

II-F._ Adsorption of Molecular Prdbes for Pore Size
Determination

The apparent pore dimensions of pillared clays.were
determined by the adsorption of hydrocarbon molecules
having various kinetic diameters. The probe molecules
employed were benzene, neopentane, 1,3,5-triethyl-benzene
and perflorotributylamine whose kinetic diameters are
5.8 A, 6.2 A, 9.2 A and 10.2 A, respectively. A conven-
tional McBain balance equipped with quartz glaSs Springs
was used to measure the amounts of probe molecules adsorbed
at room temperature. The samples were activated at 350°C

in a vacuum for 3 hrs prior to the adsorption measurements.

II-G. Chemical Analysis
Elemental analyses of clay samples were determined
using a Jarrell-Ash 955 Atom-Comp instrument marketed by

Fisher-Scientific Company. Standards for the analysis of

33

Si, Al, Fe, Mg, Na, Ca and Cr were J.T. Baker instra-
analyzed grade reagents. All clay samples were completely
fused with lithium metaborate (LiBO2 from Aldrich) at
1000°. General procedures were as follows. A clay sample
(50 mg) was mixed with 0.3 g LiBOz. This mixture was
transferred to a preignited graphite fusion crucible and
placed in a muffle furnace at 1000°C for 12 min. The
molten material was poured into a 400 mL beaker which con-
tained 40 mL 3% NH03. Approximately 10 min was allowed for
dissolution of the fused samples. The samples were then
diluted to 100 mL with deionized water. To avoid possible
contamination, all the containers such as graduated cylin-
ders, beakers, volumatric flasks and bottles were made of
teflon or polypropylene. Because of the difficulties
encountered in dissolving fused chrominum and iron pillared
clays, a mixed solution of 30 mL 3% HNO3 and 5 mL 15% HCl

was used in place of 3% HN03.

II-H. Magnetic Resonance Spectroscopy

The 27A1 magnetic resonance spectra of aluminum solu-
tions containing hydroxy-a1uminum-polymers were obtained
with a Bruker WH-180 spectrometer. The spectra were cali-
brated in parts per millon relative to l M A1(N03)3 with
a chemical shift of 0.0 ppm. A1(0H)4' (4.63 x 10'3M) was
placed in 10 mm o.d. tube and used as an external
reference. The resonance of Al(0H)4', is located at 80 ppm

downfield from the signal of A1(H20)63+. Deuterium oxide

34

in a small tube (5 mm o.d.) was used for the field

frequency lock.

II-I. UV-Visible and Infrared Spectra

Electronic spectra for chrominum and ferric solutions
were obtained on a Varian Associates Model Cary-17 Spec-
trometer. Infrared spectra of pillared clays were obtained
with a Perkin-Elmer Model 457 gratting Spectrophotometer.
A self-supported clay film, which was obtained by drying
0.5% clay solution in air, was placed in a T-shape vacuum
cell containing two NaCl windows. The clay sample was
activated at one arm of the IR cell under vacuum and then

transferred to the other arm to record the IR spectra.

35

CHAPTER III
ALUMINUM-PILLARED CLAY CATALYSTS

III-A. Objectives

Since zeolites were introduced in late 1954 as adsorb-
ents for industrial separations and purifications, they
have been extensively applied in catalytic cracking,
hydroisomerization, selective forming, hydrocracking and
transformations of aromatics hydrocarbons in petroleum
refining and petrochemical industries. During the past 30
years, about 120 zeolites were synthesized]. Commercial
catalysts are only of the faujasite type (X and Y),
mordenite, ZSM-S, A and erionitez. Faujasite type zeolites
have the largest pore sizes in the zeolite family (m 8 A).
When dispersed in an amorphous silica-alumina matrix they
constitute the main class of catalysts for hydrocarbon
cracking, which is the largest volume catalytic process
practiced in industry3. The limited pore size of zeolites,
however, limits the utility of zeolites as catalysts for
the conversion of higher molecular weight hydrocarbons

4

found in heavy oils . Therefore, for practical purposes,

it would be of considerable interest to synthesize ordered,

36

porous materials with pore size larger than 8 A.

A new class of molecular sieve-like materials with
pore size range (6-40 A) larger than that of faujasite
type zeolites has been synthesized by interlayering
expandable clay minerals with large cations. Several ex-
panded smectites with cationic organic species or complexes

2+_6,7

have been prepared, e.g., NR4+-5, HZDABCO and

ML§+-8’9 where R = CH3 or C2H5’ DABCO = 1,4-diazabicyclo-

[2,2,2]octane and L = bipyridyl or 1,10-phenanthroline.
These materials are thermally stable to temperature below
350°C. Highly stable, high surface area expanded clays

can best be prepared by interlayering the clays with
cationic hydroxy-metal polymers derived from the hydrolysis

of polyvalent metal ions, such as A13""]0"14 Zr4+ 15,

3+ 17

Cr3+ 16 and Bi It is generally believed that the
pillar species of hydroxy-aluminum polymer is

7+ . .
A11304(0H)24(H20)12 because the expanded basal spac1ng 15
closed to the size of its sulfate salt determined by X-ray

‘8 though Brindley et al.‘1

crystallography methdd and

12

Lahav et a1. suggested that the pillars might be composed

. 6+
of two units of [Al6(0H)12(H20)]2] .

19

Vaughan and Lussier reported that dehydrated A113-

0
montmorillonites possesses pore opening m 8.0 A, average

pore diameter around 22 A and surface area between 300-500
m2/g. It has been disclosed that the catalyst is capable

of cracking more gas oil than conventional cracking

37

14 20

catalysts , enhancing light cycle oil yields when they
are utilized as petroleum cracking catalysts. Also, mont-
morillonite pillared with alumina is reported to adsrob

molecules as large as mesitylene, but not isodurene, which
have kinetic diameters of 7.6 A and 8.0 A, respectively1o.

The highly selective molecular sieving properties of

clays pillared by hydroxy metal cations requires a regular
distribution of pillars and pores in the interlayer region.
However, it is known that the layer charge distribution in

21-24

smectite clays is highly irregular , with the layer

charge varying by as much as a factor of two from interlayer
to interlayer23. Thus it is expected that pillared clays
exhibit a range of pore sizes if ion exchange is the sole
driving force for intercalation. Nevertheless, the effect
of charge density of smectite clays on the pore size and
other physical properties are not known. In order to
understand these points, we have investigated the pillar-
ing reaction of hydroxy aluminum polymer and smectite

clays. In this study, we have found that the pore size
distribution is greatly dependent on the size, shape and

distribution of hydroxy aluminum polymers, which are not

just "A113" keggin ion only.

38

III-B. Results and Discussions

l. Pillaring Reagents and Synthetic Methods

The active ingredient of antiperspirants, 50% w/w
chlorhydrol solution, obtained directly from Reheis
Chemical Company is used as the pillaring reagent. Its
chemical formula is reported as [Al2(0H)5Cl] and its

chemical composition is listed in Table 111-1.

 

 

Table III-l.
Chemical Composition of 50% w/w Chlorhydrol Solution

 

A1203 23.7 %

Cl 8.2l%

Al : Cl (mole/mole) 2.01 : l
504 <0.025 %
Pb <l0 ppm
Fe 42 ppm
pH 4.20
Specific gravity l.337

 

 

The preferred method of industrial preparation of
aluminum chlorhydrol is by dissolving aluminum metal in

acid as described in the following over simplified equation:

2Al + HCl + 5H 0 —-> "Al2(0H)5+" + 2H2+ + or

2

39

The original concentration of 50% w/w chlorhydrol
solution was 6.ll M Al which was too high to be suitable
for the pillaring reaction. The chlorhydrol was usually
diluted to 0.l2 M Al prior to reaction by mixing 2.6 g
50% chlorhydrol solution with 50 ml deionized water for
each gram of Wyoming Na-montmorillonite utilized in the
reaction. The relative quantity of Al to clay was 7.9
mmole/meq. The general procedures used for the pillaring
reaction are the same as described in Chapter II of the
Experimental Section. The freshly diluted chlorhydrol
solution had pH 4.20. However, if the solution was aged
at room temperature for two weeks, its pH value was in-
creased to 4.75. Two different pillared clays I and II
were synthesized from the FRESH and AGED pillaring solu-
tions respectively. The properties of the resulting inter-

calates are described in the following section.

2. Chemical Composition of Aluminum Pillared Clay
Catalysts I and II.

The chemical composition of the clay minerals before
and after the pillaring reaction are shown in Table III-2.
For the Na+-montmorillonite, it was assumed that each unit
cell has 22 oxides and a total metal cation valence of
+44. However, this assumption is not valid for determining
the chemical formula of the pillared clays, because the
interlayered hydroxy aluminum polymers contain various

unknown amounts of Al, 0 and 0H. Thus, the formulas for

4O

 

Table III-2. Chemical Composition and Chemical Formulas 0f
Wyoming Na+-Montmorillonite and Its Pillared
Products I & II Made from fresh & aged
Diluted Chlorhydrol Solutions.

 

1. Chemical Composition

 

Na+-montmorillonite Al-Pillared Al-Pillared
Clay I Clay II
5102 53.94% 39.42% 45.14%
A1203 18.78 25.82 24.72
F6203 3.83 2.83 2.82
M90 2.23 l.60 1.63
NaZO 2.05 0.0 0.0

2. Chemical Formulas

 

Na+-montmorillonite

(”30.58)[A13.11Feo.42”90.48](517.88A‘0.i2)°2o(°”)4

Al-Pillared Clay I

+0.53
[A‘(°")2.80]2.87 [A]3.llFeo.42M90.48](Si7.88A10.l2)020(0H)4

Al-Pillared Clay II

+0.58 .
[A‘(°“)2.7532.31 [A‘3.11Feo.42"90.483(5‘7.88A‘0.12)°20(°")4

 

41

the pillared clays were obtained by assuming that the num-
ber of Si and Al ions in tetrahedral holes and Al, Mg, Fe
ions in octahedral holes remains unchanged after reaction.
Thus, the composition of interlayer pillars was obtained
by substracting the aluminum content in the layer lattice
from the total content of aluminum.

As shown in Table 2, the sodium ion in the inter-
layer corresponds to the cation exchange capacity of
77 meq/l00 g clay. These interlayer Na+ ions are replaced
by cationic hydroxy-aluminum polymers. The difference
between pillared clay I and II is that pillared clay II
has a lower aluminum content.

‘3. XRD and Surface Area

X-ray diffraction patterns of pillared clay I treated
at different temperatures are shown in Fig. III-l. There
are only a few first order peaks for the air dried sample.
However, when the sample was heated at ll0°C, up to six
higher order peaks appeared, due to the removal of inter-
layer water molecules and the formation of a more ordered
interlayered structure. When heated at higher tempera-
tures (g;g;, 550°C) this material remains a high 00l spac-
ing but the higher order peaks disappear. Usually, the
first order peak is very narrow and strong while the higher
order peaks are broad and weak. The XRD patterns of
pillared clay II are more or less the same as I, and are
shown in Figure III-2. All the results for both pillared

clays, summarized in Table III-3 give the average values

42

l9.84 A

@538

 

938

 

3108 1681
wow mom
L l
30 25

 

Figure III-l. X-Ray Diffraction Patterns of Alumina
Pillared Clay I. a) air dry; b) 350°C;
c) 700°C.

43

2Ioaii

 

le$a

 

 

 

l l l 1 L l

30 25 20 :5 IO 5 26

Figure III-2. X-Ray Diffraction Patterns of Alumina
Pillared Clay II. a) air dry; b) 350°C;
c) 700°C.

44

of c1009v as determined from a plot of l/d li- 2. Comparing
their XRD results, we.see that pillared clay II is more
thermally stable because its basal spacing is larger than
I after thermal treatment at the same temperature. The
air dried sample of pillared clay II can be swelled to
22.7 X after being exposed to 100% humidity for 48 h as
shown in Figure III-3 where the 2nd order peak is shifted
from 20 = 3.4° (10.53 X) to 7.4°_(11.95 K). On the other
hand, its spacing is not changed under the same condition
if it has been previously heated at 200°C.

Table III-3 also includes the surface area of pillared
clays that were heated at 350°C and 700°C. The native
Na+-montmorillonite readily collapses to a spacing near
9.6 A when heated at 200°C and exhibits an external.
surface area of approximately 30 mz/gzs. The eXpected
internal and external surface area is about 750 m2/g by
calcUlation. Therefore, the surface area of the pillared
clays indicated in Table III-3 is due to the micropore
structure of the interlayered space. The decreased sur-
face area following calcination at 200°C may have been due

to partial collapse of the clays. The surface area of

pillared clay II is 50% more than that of 1, This is con-

sistent with the chemical analysis results that pillared

clay II contains less of interlayered aluminum.

45

21.04 A
"’(00 I)

 

 

 

 

 

Figure III-3. X-Ray Diffraction Patterns of Alumina
Pillared Clay II. a) air dry; b) expos-
ing the air dry under ]00% humidity for

48 hours.

46

 

oem m\~s omm Au\~sv mmsq momegzm

as.“ me.“ e¢.m _e.m m ee.,_ Amy we
oo.2_ mm.~_ em.m_ Fo.mp m eo._~ Mev e

HM sepo eeee_P_e
cop. m\~E cow Am\wsv mmg< mumugsm
“3.2 m~.e mm.m o~.m m e~.cp any ea
no.e_ mm.e_ mm.m_ cm.m_ m em.m. Amv e

H sepu eeeep_we

 

oooon ooomm ooomm coo—p comm

 

HH a H aepu eeeepPee e=e_e=_< to eee< eeeceem eee mme_eeem _emem max .m-HHH epeeh

 

47

4. DSC Studies

The differential scanning calorimetry for the Na+-
montmorillonite and aluminum pillared montmorillonite I
are shown in Fig. III-4 (a) and (b), respectively. Both
samples were pre-equilibrated at 50% humidity. The endo-
thermic peak centered near l15°C for Na+-montmorillonite
corresponds to the loss of sorbed water in interlayer. No
other peak appears between l50°C and 550°C. Early gt_gl.26
studied the thermal properties of montmorillonite clay
(caly spur, Wyoming) and found that the loss of hydroxy
groups was beyond 600°C. For the pillared montmorillonite,
the endothermic peak at 125°C corresponds to the loss of
water in interlayer. There is another small endothermic
peak at 400°C, starting from about 350°C. It was reported
that gibbsite, a-Al(0H)3 and bayerite, y-Al(0H)3 are endo-
thermically changed to X-Alz)3 and n-Al203, respectively,

at about 350°C with a weight loss of 32%27'29.'

Thus, the
small endothermic peak in (b) of Figure_III-4 might be
due to the dehydroxylation of the pillars, i.e. the
hydroxy-aluminum polymers.

5. Hydrocarbon Adsorption Isotherm

The results for the adsorption of probe molecules by
pillared clays I and II are summarized in Table III-4.
Adsorption can occur only when the pore openings are larger

than the kinetic diameter of the probe molecule. As seen

in Table III-4, pillared clay I adsorbs not only the small

48

.H sepu eeee_Ppe ee_e=_< An
“mu_=o_~wgosu=ozi+mz Am mo mm>cau xgpmewgopmu m=_:=aum —mwu:mgm$$wo

Uo 00m 00? 00m CON 00.

u u _ - 4 q - 1 -

 

3v

 

3

.e-HHH ee=m_e

49

 

 

o.o o.o. e~.o . ¢.o_ -.o ee_5ep»u=eweeeee_ceee
o.o No.0 m~.o N.m m..o eee~eee.»eee_eo-m.m.p
m~.~ e~._ oe.p ~.e mm“ eeeeeeeeez
m.m em._ we._ m.m o“ eeeNeem
>ez _ H_ H A<v ee_meee_a Aeeeuv a

11Aa\m_ossv vmnsomn< u:=oE<.Iu . e pmvuvgu mmpaumpoz mace;

flH a H mxepu umLmPFPQIE=:_E:—< An mwpaumpoz mace; mo :owungom .e-_HH epoch

 

50

molecules but also the large molecules. The basal spacing
of the dehydrated clay at 350°C is 8.93 K, which is
smaller than the molecular dimension of perflorotributy-
amine. Apparently the pore window or the distance between
two adjacent pillars is larger than 10.4 A.

Pillared clay II adsorbs small molecules, e.g. benzene
and neopentane and only a trace amount of l,3,5-triethyl-
benzene, but totally excludes the large molecule perfloro-
tributylamine, which indicates that the interlayer micro-
pore openings are less than 10.4 8-

Both pillared clays dehydrated at ll0°C have inter-
layer thickness of about 9.2 A which is close to the
dimension of the cation in [Al1304(0H)24(H20)12]-3.5 SO4
whose structure is shown in Figure III-5. Also, the
interlayer thickness is consistent with the results of

10 and Occelli et al.14. However, the

Vaughan gt_gl.
thermal stability, surface area and pore opening of these
two pillared clays are extremely different even though
they were made from the same diluted 50% w/w chlorhydrol
solution.

The reasons for these differences might be related to
the size of the hydroxy aluminum polymers in the interlayer.
Reaction might happen during the aging period which leads
to changes in the size or shape of the polymers in solution.

In order to understant these reactions during aging, we

employed Al-27 NMR to examine the solution.

51

 

Figure III-5.

Representation of [Al1 0 (0H) 4(H 0) ]7+
Keggin Type Ion. The 8e tereg Teéralgdral
Coordinated Al in Surrounded by l2 Octahedral
Coordinated Al in Equal Distance (Ref. 18).

52

6. Hydrolysis Chemistry of Aluminum
. The 27Al-NMR can only distinguish the aluminum exist-
ing in octahedral and tetrahedral coordination geometries.
Octahedral species such as Al(H20)g+ and [AlMo6OZTJB' can
be observed at a chemical shift near 0 ppm, whereas tetra-
hedral complexes like Al(0H)4' and [AlW1204015'
observed further downfield near 80 ppm and 72 ppm, respec-

30 27

tively The

AleNMR spectra of Al(H20)g+ and
Al(0H)4' are shown in Figure III-6. As seen In Figure 2
of Chapter I, the hydrolysis process is very complicated
and there are many hydrolyzed species present in solution.
Although 27Al-NMR might not be able to distinguish all of
the possible octahedral and tetrahedral species, the
chemical shift of octahedral aluminum should be close to
0 ppm while the chemical shift of tetrahedral coordinated
aluminum should appear at a downfield region near 80 ppm.

The empirical formula of commerical chlorhydrol com-
pound, [Al2(0H)5Cl]-2H20, is known31, but its structure
has not been characterized. The chemistry of partially
neutraliZed aluminum chloride solution has been studied
extensively and various aluminum species have been pro-
posed.

In addition to monomer, Al(0H)(H20)§+ and dimer,

5
Alz(0H)2 (H 20):+ , other proposed hydroxy-aluminum polymers

include the following Al 5(0H)3+ 32, Al 6(0H)5+ 5
4+ 32- 34 8+ 35
A18(0H)m , Al1o(0H)22(H2 0)16 , Al 13 04(0H)24-

(+120)7+ 3° 37, and Al 224(0H)60(H 0)12+ 37. Follwing a

53

MW 20)3+
Al (on);

 

) (b)

80 ppm Huh...

(0) )

Figure III-6. The Chemical Shift of a) Octahedral
Aluminum Al(H20)53+ and b) Tetrahedral
Aluminum Al(OH)4'.

 

 

Opprn

54

series of dialysis experiments and chemical methods of

identification, Hsu and Bates35

proposed a continuous
series of polymers which were aggregated from the basic
ring unit, [Al6(0H)12(H20)]2]6+ to form two-dimensional
crystalline structures, upon the addition of base as shown
in Figure III-7.

38-41 using nuclear magnetic

Recently, Akitt g£_31,
resonance, concluded that rapid hydrolysis or hydrolysis
at lower temperatures appears to proceed via a series of
equilibria which involve only three Species, as equation

(l) indicates:

3 ___. 4 ....
[A‘<“2°>51 " ,._. [A12(0H)2(H20)81 " ....

7
[A113°4(°”)24(”2°)12] + (1)

The final product, the "Al13" keggin ion, whose structure
is shown in Fig. III-5, is not stable in acidic medium
(pH m 4.0) and can undergo slow transformation, probably
by a variety of pathways to form other species.
Fitzgerald's studies of various aluminum solutions
with 27Al-NMR also reveal that at least three octahedral
and two tetrahedral aluminum sites as well as monomeric

. . 42
and dimeric are present 1n solut1on .

55

7. Al-27 NMR Spectra of Fresh and Aged Chlorhydrol
Solutions
7-a. Fresh Chlorhydrol Solution

The Al-27 NMR spectra of the freshly diluted chlor-
hydrol solution used to prepare pillared clay I consisted
of three distinct signals as shown in Fig. III-8. Exter-
nal standard, Al(0H)4-, whose chemical shift is at 80.0 ppm
was used to quantitatively calibrate the spectra. The
broad peak with a half width of m 320 Hz at m 0 ppm is
corresponding to the peak of aluminum ions in octahedral
coordination3o. This signal corresponds to aluminum ions
in monomeric Species of the type Al(H20)g+, Al(0H)(H20)g+
and Al(0H)2(H20): and dimeric species Al2(OH)2(H20)g+. The
broad resonance may be due to chemical exchange between

38. 27Al is a quadrapolar

these species Also, since
nuclear with a spin of 5/2, Quadrapolar broadening is ex-
pected for species which do not have a center of symmetry.
The small peak at 62.8 ppm is attributed to the tetra-
hedrally coordinated aluminum ion, Al04, located at the
center of the polymer A11V04A112(0H)24(H20)(;. The 12
other aluminums are presumed not to be in'a symmetrical
environment. The electric field gradient at their level
is therefore relatively high, and the octahedron may be
distorted. All this causes the peaks to broaden38’43.
Another much broader peak at l0.6 ppm with half-width

of m 1850 Hz is probably due to the aluminum in a high

56

Figure III-7. The Proposed Hydrolysis Mechanism of
Aluminum Ion From Hsu.

57

mu_xogua:
:5: E: E 4|.
w:__—oumxgu

M.Mio.m u PK
:omz

 

 

 

om em
+m. _< +N_A=ov _<

+5-

+m

om m—
+mA=ov _<

 

 

Aeom— .zmzv mus—xomo>: zaz~z24< mo hzm2¢oam>mo oumoaoma

 

mm c_
+mA=ov _<
+m
_.N-m.o u mwmm

58

 

 

ARCH);
Ociohedrol
AI Monomers
Ociohedrol
AI Polymers
N 7...
l A104 A1|2(OH)24
80 62.8 I08 0 ppm

Figure III-8. The 27Al NMR Spectra of Freshly Diluted
Chlorhydrol Solution.

59

molecular weight polymer where the aluminum is coordianted
with oxygen in an octahedral site of not very distorted

symmetry. Akitt et al.38'41

obtained the same pattern

of 27Al-NMR spectra from the hydrolysis of Al(III) and
aluminum metal at high concentration, [Al3+] 3 l.0 M.
After detailed studies which included gel-permeation chro-
matography, they found that aging leads to the production
of larger polymers. The structure of these polymers was
suggested to consist of partial "Al13" units with octa-
hedral units disposed as flexible chains and/or forming
cross links.

Recently, Bottero et_gl.43'44 has studied partially
neutralized aluminum chloride solutions by NMR and small-
angle x-ray scattering. They reported that when
0H/Al = 2.00, this solution was almost entirely made up of
"A113" keggin ion with an experimental radius of gyration
9.8 A. In aluminum solution, 0H/Al = 2.50, colloidal
particles with chemical composition similar to that of the
trihydroxide and rod-shaped polymers of various size were
found along with "Al13" polymer. The particle morphology
of the colloidal species changes as a function of time.
After aging l.5 h, the particles are cylindrical with
a hydrated radius of about l5 A and a length of 3l0 A.
After 24 h aging, the cylinders have agglomerated into

more homogeneous platelets of diameters 500 A and thickness

60 A.

60

The scanning electron photomicrographs of aluminum
chlorhydrate showed platey structure45.

From the above evidence, it seems that the 50%
chlorhydrol solution from Reheis Chem. Co. contains mainly
high molecular weight polyoxocation of aluminum and minor
amounts of Al13 polymer. Indeed, Teagarden gt_al.45 had
ever studied the aluminum chlorhydrate 50% w/w solution
and found the distribution of aluminum species as 4%
monomeric, 8% small polycations and 88% large polycations.

7-b. Aged Chlohydrol Solution

When the diluted chlorhydrol solution is aged for
various time periods at room temperature, the nmr spectra
dramatically changes, as shown in Figure III-9. When the
solution is allowed to age for one Week, the amount of
"Al13" keggin ion increases, because the integral ratio of
the AlO4 moiety (62.8 ppm) of Al13 keggin ion to the external
reference (80 ppm) increases from 0.6 to 3.0. The monomeric
(-0 ppm) and highly polymeric species (l0.8 ppm) are still
present after one week aging; but their amount declines
dramatically.

If the aging periods is lengthened to 6 weeks, the
integral ratio of "Al13" to Al(OH)4‘ increases gradually
to 3.4; the hump peak at l0.8 ppm does not change too much.
The pH values remain in the range of 4.65 - 4.75. The pH
of freshly diluted chlorhydrol solution was 4.20. Combined

with the observation of increased pH value, the increase of

Figure III-9.

"61

The 27Al NMR Spectra of Diluted Chlorhydrol
Solutions a) Freshly Prepared; b) Aged at
Room Temperature for l week; c) 2 weeks;

d) 6 weeks; e) 3 months (colloidal particles

was removed from solution).

62

(o)

  

(pH=420)

 

 

 

 

63

"A113" kegging ions might come from the disassociation of
high polymer by release of OH' as the following specula-

tive equations suggest:

[(A113):+C1;] 1;: (A113):+ + a 01' (2)
H
(A113)n_1 0 (A113)a+ + 2H20 32

0(a-x+1)+ + (1120-11113)X+ + 0H'(3)

(A1 H

13)n-1 2

3+ A 4+ A ll H +
Al(H20)6 ... A12(0H)2(H20)8 ... A113 X (4)

The charges on the high polymers as well as their structure
and molecular weight in above equations are not known.

The aggregation of monomer and dimers to form Al13 keggin
ion as in equation (4) will release proton. However, the
0H" released in equation (3) could be more than the depro-
tonation in equation (4) leading to the observed increase
of pH.

When the solution is aged longer than 3 months, a
colloidal gel appears and the solution becomes turbid. The
white gel was identified as biggsite by XRD. The pH of
the solution decreased to 4.10. The 27Al-NMR spectra of
this solution after the movement of gel particle is shown
in Figure III-9 (e). The amount of Al13 keggin ions in
this solution has decreased greatly. This decrease would

be due to the aggregation of "A113" keggin ions to form high

64

polymer (gibbsite) through the deprotonation reaction of the
general hydrolysis process. Teagarden g£_gl. also observed
that the pH of diluted chlorhydrol solution was first in-
creased and then decreased46.

8. Al-27 NMR Spectra after Pillaring Reaction

8-a. Fresh Chlorhydrol Solution

The nmr spectrum of the supernatant solution after the
pillaring reaction with freshly diluted chlorhydrol solu-
tion is shown in Figure III-10b. (For the convenience of
comparing the spectra change before and after pillaring
reactions, the NMR spectra of fresh chlorhydrol solution
diluted with deionized water to the same volume is shown
in Figure III-10a. Both spectra were taken within four
hours after the 50% w/w chlorhydrol solutions were diluted.
Examining these two Spectra, we find the ratio of integra-
tion of "A113" peak at 62.8 ppm to external reference,
Al(OH);, at 80 ppm and the area of broad hump peak at
10.8 ppm of spectra (b) is smaller than those of spectrum
(a) and suggests that both "A113" keggin ions and high-
polymers are intercalated into the clay interlayer.

8-b. Aged Chlorhydrol Solution

The Al-27 NMR Spectrum of the supernatant solution is
shown in Figure III-11b. Again the aged solution was
diluted to the same volume with deionized water. The
Spectrum is shown in Figure III-11a. Comparing the Spec-

tra before and after pillaring reaction, we find that both

 

 

 

65

 

(b)

 

80 628 I080 ppm

Figure III-10.

The 27Al NMR Spectra of Freshly Diluted
Chlohydrol Solution a) before the pillar-
ing Reaction (diluted to same volume as
in (b)); b) after pillaring reaction.

66

(o)

 

 

 

 

 

 

80 62.8 10.86 ppm

Figure III-ll. The 27A1 NMR Spectra of Aged Chlorhydrol
Solution at 25°C for 2 weeks a) before the
Pillaring Reaction (solution was diluted to
the same volume as in (b)); b) after pillar-
ing reaction.

67

the "A113" keggin ion and the higher polymer are absorbed

into the clay interlayer.

9. Proposed Models for Pillars in Interlayer

It is clear from the nmr studies that the two pillar-
ing solutions, fresh and aged chlorhydrol solutions, con-
tain monomeric, dimeric, "A113" and high polymers in
different relative quantities. All of these species are
adsorbed into the interlayer region as pillars.

AS mentioned above, the high polymer of oxoaluminum
cations are probably in the shape of cylindrical or rod-
like structures with a diameter of 9 A as measured by
x-ray diffractometer (Table III-3). They are probably
randomly distributed in the interlayer of montmorillonite.
The possible arrangement of pillars in I is proposed as
having a large pore Size distribution and large pore
openings to absorb big probe molecules.

0n the other hand, the rod-shaped polymers disassociate
to form "A113" spherical polymer units in the aged
solution. A possible pillar arrangement in the pillared
clay II is porposed in Figure III-12(b) where rod-shaped
polymers are also included but in lower amount than in
pillared clay I, Figure III-12(a). Even though a small
amount of high polymer was present, "A113" spherical keggin

innsstill can pack in such way as to result in small pore

size structure becoming zeolitic pillared-clays.

 

 

/— rod- shaped Polymer

//0§ / @1222: ( hydrated diameter

 

(o)

 

i/iafi O 0‘ _hydroted
o AIl3 Keggin Ions

(a) o o

 

 

 

 

 

 

 

Figure III-12. Proposed Models of Hydroxy Aluminum Polymers
in the Interlayer Region of Pillared Clays
I and II.

 

 

69

10. Attempts at Synthesizing Pillared Clays with
"A113" as Pillar Only
Several attempts were made to synthesize aluminum
pillared clays containing exclusively "A113" keggin ions
as the pillaring Species.
lO-a. Synthesis of "A113" Keggin Ions From Al3+
In order to exclude the high polymer in pillaring
solution, "A113" was made from AlCl3 and NaOH solutions.
Various amounts of 0.1 N NaOH were added to 0.2 M AlCl3
to give solutions with 0H/Al ratio (n) from 0.0 to 2.42.
The Al-27-nmr spectra in Figure III-13 illustrate that
the amount of "A113“ keggin ion is increased with n values
and the amount of monomer and dimer decreased. When the
n value is 2.42, there is only one peak at 62.8 ppm which
corresponds to "A113" ions. It seems likely that “A113"
was the major species in solution. However,.after pillar-
ing reaction at n = 2.42, the surface area was only
202 m2/g at 350°C. Thus, there might be some nmr unobser-
vable high-polymer of aluminum hydroxy in the solution.

41

In fact, it has been suggested by Akitt et al. that some

NMR un-observable aluminum polymers do exist in solution.
lO-b. Selective Intercalation by Ca2+-

Montmorillonite

Ca2+ ion exchanged montmorillonite has a basal spacing
d002 = 19.2 A or interlayer spacing Adoox = 9.6 A in water47.

Under this condition, the tendency for Ca2+-montmorillonite

 

70

1 /

(’—
(m%F=oo r”——— /i::K~J //jKNmflflp

{ fl) ‘/

 

 

 

 

 

 

 

 

 

 

 

(b) IOO
(c) 1.50
(”2.42 k
b M
Sb 62.8 (3 ppm 810 62.8 0 ppm

Figure 111-13. The 27A1 NMR Spectra of Aluminum Chloride
Solutions with 0h/Al Ratio a) 0.0; b) 1.00;
c) 1.50; d) 2.00; e) 2.25; and f) 2.42.

 

71

to bind the small Spherical "A113" keggin ions with diameter
0
m 9 A might be greater than for the high polymers. The

2+montmorillonite

surface area of pillared clays mady by Ca
and aged chlorhydrol solution was 394 mZ/g.
10-c. Collection of Al13 Keggin Ions by Dialysis
Method

An aged chlorhydrol solution (150 ml) (2 weeks) in
a cellous dialysis membrane tube was immersed in 300 m1 of
deionized water for 5 days at room temperature. The mem-
brane was shaken occasSionally to mix the solution. The
pH of the dialygate was 4.70 after the 5 days of aging.
The 27Al-nmr Spectra of the solution in bottle is shown
in Fig. III-14. There were two signals besides the exter-
nal standard (80 ppm): 62.8 ppm for "A113" and 0-11 ppm
hump for high polymers. Apparently, not only the small
spherical ions "A113" but also the high polymers can
penetrate through the membrane tube. The reaction of
these solutions with Na+-montmorillonite gave a pillared
clay with a surface area of 411 mzlg.

11. The E-fect of CEC on Pillared Products

In order to understand the influence of the silicate
surface charge density on the intercalation of hydroxy-
aluminum polymers, the pore size distribution and other
physical properties, two other smectite clays with higher

cationic exchange capacity (CEC) were investigated and the

results were compared with those obtained for Wyoming

72

a)

80 62.8 ppm

k

Figure 111-14. The 27A1 NMR Spectra of Solution which was
Aged at 25°C for 2 Weeks and Then Dialyzed.
The NMR Spectra is the Solution Outside

of the Membrane Tube.

 

73

Na+-montmorillonite. The surface area including the

pillared Wyoming montmorillonite are listed in Table III-5.

 

Table IV-S.
Surface Area of Pillared Clays Made From

Smectites with Different CEC

 

Pillared Pillared
Clay I Clay II

Wyoming-Montmorillonite 77 meq/100 g 260 m2/g 390 mZ/g
Arizona-Montmorillonite 102 240 340
Fluoro Hectorite 187 140 115

'—

 

 

As shown in Table III-5, increasing the CEC decreases the
surface area of the pillared clays. It seems that the

amount of intercalated pillar polymers is proportional to

the CEC value. Thus, if the smectite had a higher cationic
exchange capacity, it could adsorb more pillar polymers to
occupy more interlayer surface area and subsequently to

give lower BET surface area. In this sence, one could expect
the pillaring ions to pack closer together and form

materials with smaller pore sizes.

III-C. Conclusions '

The production of pillared clays with high thermal
stability and internal pore volume for adsorption is
achieved by interlayering smectite clays with cationic

hydroxy aluminum polymers. Nitrogen adsorption shows high

 

74

surface area which is due to the adSorption by a intra-
crystal micropore structure. 4

Their pore sizes and surface areas depend intimately
on the degree of polymerization of hydroxy cations as
well as the CEC of smectite clays. The ideal pillars
would be the large spherical cationic hydroxy metal poly-
mers, rather than one-dimensional chain polymers or two-
dimensional platey polymers. Such one and two-dimensional
platey polymers will occupy most interlayer surfaces and
give rise to low surface area and large pore size distri-

bution pillared products.

10.

11.

12.

13.

 

 

75

REFERENCES

Whan, D.A. Chem. in Britain, 532 (1981).

Jacobs, P.A. "Carboniogenic Activity of Zeolites",
Elsevier Sci. Publ. Co., New York, 1977.

Heinemann, H. in "Catalysis-Science and Technology",
ed. Anderson, J.R. and Boudart, M. Vol. 1, Springer-
Verlag, NY 1981, Chapter 1.

Satterfield, C.N. “Heterogeneous Catalysis in Prac-
tice", Chapter 9, McGraw Hill Book Co., NY, 1980.

Barrer, R.M.; Millington, A.D. J. Colloid Interf. Sci.
2g, 359 (1967).

Mortland, M.M.; Berkheiser, V. Clays and Clay Minerals
21, 60 (1976). -

Shabtai, J.; Frydman, N.; Lazar, R. Proc. 6th Int.
Congr. Catal. 85, 1-7.

Berkheiser, V.E.; Mortland, M.M. Clays and Clay
Minerals 2;, 105 (1977).

Loeppert, R.H.; Mortland, M.M.; Pinnavaia, T.J. Clays
and Clay Minerals, 22, 201 (1979).

Vaughan, D.E.W.; Lussier, R.-J., preprints, 5th Int.
Conf. Zeolites, Naples, Italy, June 2-6, 1980.

Brindley, G.W.; Sempels, R.E. Clay Minerals 12, 229
(1977).

' Hahav, N.; Shani, U.; Shabtai, J. Clays and Clay

Minerals 29, 107 (1978).

Shabtai, J.; Lazar, R.; 0b1ad, A.G. Proc. 7th Int.
Congr. Catal. 88, Tokyo, Japan, July 1-4, 1980.

14,

15.

16.

17.

18.
19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

 

 

76

Occelli, M.L.; Tindwa, R.M. Clays and Clay Minerals
21, 22 (1983).

Yamanaka, S.; Brindley, G.W. Clays and Clay Minerals
22, 119 (1979).

Brindley, G.W.; Yamanaka, S. Amer. Mineral 21, 830
(1979).

Yamanaka, S.; Yamashita, G.; Hattori, M. Clays and
Clay Mineral 22, 281 (1980).

Johansson, G. Acta Chem. Scand. 11, 771 (1960).

Vaughan, D.E.W.; Lussier, R.J.; Magee, J.S. U.S.
Patent 4,176,090.

Lussier, R.J.; Magee, J.S.; Vaughan, D.E.W., pre-
prints, 7th Canadian Symposium on Catalysis,
Edmonton, Alberta, October 19-22, 1980.

Stul, M.S.; Wortier, W.J. Clays and Clay Miner.
22, 391 (1974).

Peigneur, P.; Maes, A.; Cremers, A. Clays and Clay
Miner. 22, 71 (1975).

Lagaly, G.; Weiss, A. Proc. Int. Clay Conf., 1975,
Mexico City, Bailey, S.W. Ed., Applied Publishing
Ltd., Wilmette, IL, 1976, 157-172. ‘

Talibudeen, 0.; Goulding, K.W.T. Clays and Clay
Miner. 21,‘37 (1983).

Grim, R.E. "Clay Mineralogy", McGraw-Hill, New York,
1968. PP. 265.

Early, J.W.; Osthans, 8.8.; Milne, I.H. Amer. Miner.
22, 707 (1953).

Mackenzie, R.C. in "The Differential Thermal Investi-
gation of Clays" Ed. by Mackenzie, R.C., Mineralogical
Society, London, 1957, page 311.

Wefers K.; Bell, G.M. "Oxides and Hydroxides of
Aluminum", Technical Paper No. 19, Aluminum Company
of America, Pittsburgh, PA 1972.

MacZur, G.; Goodboy, K.P.; Koenig, 0.0. in Kirk-0thmer
“Encyclopedia of Chemical Technology", Vol. 2, 3rd. Ed.
John Wiley & Sons, Inc. NY 1978, p. 218.

 

80.
31.
32.
33.
34.

35,
36.
37.
38.
39.

40.

41.
42.

43.
44.
45.
46.

47.

77

Akitt, J.W.; Farthing, A. J. Chem. Soc., Dalton
Trans. 1981, 1615.

a. Reheis Chemical Co. Certificate information
b. Merk Index, 9th Ed., Merck 8 Co., Inc. N.J., 342.

Brosset, C.; Bidermann, G.; Sillen, L.G. Acta. Chem.
Scand. 2, 1917 (1954).

Matijevic, E.; Mathai, K.G.; Ottwill, R.H.; Kerker,
M. J. Phys. Chem. 2;, 826 (1961).

Matijevic, E.; Janauer, G.E.; Kerker, M. J. Colloid
Sci. 12, 333 (1964).

Hsu. P.H; Bates, T.F. Min. Mag. 22, 749 (1964).
JohanSson, G. Acta Chem. Scand. 11, 771 (1960).
Rausch, N.; Bale, H.D. J. Chem. Phys. 19, 3891 (1964).
Akitt, J.W.; Farthing, A. J. Mag. Res. 22, 345 (1978).

Akitt, J.W.; Farthing, A. J. Chem. Soc., Dalton Fara-
day 1606 (1981).

Akitt, J.W.; Farthing, A.; Howarth, G.W. ibid. 1609
(1981).

Akitt, J.W.; Farthing, A.; ibid. 1617 (1981).

Fitzgerald, J.J. Division of Inorg. Chem. Abstract 264,
185th ACS Natl. Meeting, Seattle, WA, March 20-25, 1983.

Bottero, J.Y.; Cases, J.M.; Fiessinger, F.; Poirier,
J.E. J. Phys. Chem.'§5, 2933 (1980).

Bottero, J.Y.; Tchoubar, 0.; Cases, J.M.; Fiessinger,
F. J. Phys. Chem. 22, 3667 (1982).

Teagarden, D.L.; Radavich, J.F.; White, J.L.; Hem,
S.L. J. Pharm. Sci. 1g, 762 (1981).

Teagarden, D.L.; White, J.L.; Hem, S.L. ibid. 29,
808 (1981).

Berkheiser, V.; Mortland, M.M. Clays and Clay Miner.
22, 404 (1975).

CHAPTER IV
CHROMIUM-PILLARED CLAY CATALYSTS

IV-A. Objectives

The investigation of zeolite-like pillared clays pre-l
pared by interlayering smectite clays with large hydroxy
metal cations, e.g. A11304(0H)24(H20);; and Zr4(0H)%;, has
recently received considerable attention due to their pro-
perties as cracking catalysts. In chapter III, the influ-
ence of the size of pillaring Species on the properties of
pillared clays is described. It would be interesting to
extend the pillared clay systems by using other metal oxides
as pillars such that the highly isolated metal oxide clus-
ters would have specific activity in the catalysis reactions.

1 had synthesized chrominum-

Brindley and Yamanaga
hydroxy montmorillonite previously. However, maximum basal
spacing of the intercalate was about 16.5 A at 200°C and
m 10 A (totally collapsed) at 300°C. A prerequisite for
preparing chromium pillared clays is the presence of

cationic chromium hydroxy polymers in high concentration

with narrow molecular weight distribution.

78

 

79

Studies of the hydrolysis chemistry of chromium are
very limited. In 1908, Bjerrum2 postulated the existence
of 6r6(08)$; and or]2(08)§3 in addition to the presence of
dimeric species, Cr2(0H)g+. .In 1964, Thompson3 isolated
the chromium dimer Cr2(0H)g+ and trimer Cr3(0H)2+ from
hydrolytic chromium perchlorate solution by ion-exchange
methods. The structures of these two hydroxyl-bridged
Species were porposed based on magnetic susceptibility
measurements. Even though the cationic chromium hydroxy
polymers are not known, Brindley's results strongly
suggest the presence of polymeric species in hydro-
lytic chromium solution.

Chromium based catalysts are used in many significant
catalytic reactions, including hydrogenation-dehydrogena-

4a 4b

tion ,naphthareforming4c, ethylene

, dehydrocyclization
polymerizationsa’b and N0 reduction with 005C. The objec-
tive of this work is to synthesize chromium pillared clay
with large basal spacing and to study its dehydrogenation

activity of cyclohexane.

IV-B. Results and Discussions
B-l. Preparation of Chromium-hydroxy-polymers
Appropriate amounts of sodium carbonate were added in-
to 0.1 M chromium nitrate solution to obtain several basic
chromium solutions with various 6 = OH/Cr ratios. The range
3+

of i was 0.0 to 2.5 mole of 0H" per mole of Cr These

solutions were degassed with N2 gas (when the hydrolysis

 

80

was carried out at 25°C) or, alternatively heated at low
temperature (4 40°C) for solution to be aged at elevated
temperature to allow complete reaction of carbonate and
to remove all C02 gas. Also, they were aged at 60°C or
95°C for different time periods. Solutions were occasional-
ly shaken during aging. When the sodium carbonate was
added, some precipitate was formed in the solution at H =
2.00; but it disappeared after being aged for 6 hours and
35 hoursat 95°C. There was precipitate in the solution
prepared at H = 2.50, but the amount of precipitate was
dramatically decreased after the solution was allowed to
age at 95°C for 36 hours. The precipitate was removed by
centrifugation if any was present prior to the pillaring
reaction. There are advantages of using sodium carbonate
as base instead of sodium hydroXide solution.. The solid
Na2C03 prevents a high local concentration of base which
would produce a chromium hydroxide precipitates, thus mak-
ing it easier to keep the concentration of Cr(III) constant.
When sodium carbonate was added to Cr(N03)3-9H20 at
25°C, the pH values of the solutions increased and Amax
shifted further to longer wavelength. After aging at.
elevated temperature, the pH values dropped further and
Amax shifted further to longer wavelength. Also, the absor-
bance decreased. These features might indicate the occurance
of hydrolytic polymerization reaction upon the addition of

base and heat treatment. Laswick and Plane report similar ob-

servations for chromium perchlorate solution heated to reflux

81

temperature6. The pH values and).max of freshly prepared
and aged chromium solutions are listed in Tables IV-l and
IV-Z. The UV-visible absorbance of selected chromium solu-

tions are shown in Figure IV-l.

B-2. Pillaring Reaction of Na+-montmorillonite with
Chromium Solutions

The pillaring reaction was carried out at room tempera-
ture by the same procedures as described in chapter II.
The ratio of Cr(III) to clay (millimole of Cr(III) to
milliequivalent of clay) was maintained at 50 to insure a
large excess of pillaring reagent. Usually, five to ten
washings with deionized water was necessary to cause the
clay particle to flocculate. If the clay is not washed com-
pletely free of electrolyte, the suspension retains the
turbidity typical of a completely dispersed clay. The
pillared products made from high 6 solutions required a
larger number of washings. Generally, only two orders of
002 reflection were observed in the XRD pattern for pillared
clays with d001 3 17 A. However, higher order reflections
were observed for those intercalated clays with d001 < 17 A.
The basal spacings determined from the first order reflec-
tion for all air dried pillared products are Shown in Table
IV-3. A typical XRD pattern is shown in Figure IV-S.

The variation of chromium (mmole) to clay (meq) ratio
affects the expanded Spacings of pillared clays. When the
amount of chromium-was reduced in the series, Cr/clay =

50, 25, 15 and 10 mmole/meq, the basal spacings of the

 

82

 

 

Figure IV-l.

UV-Visible Spectra of 0.1 M Chromium Nitrate

Solutions (a) fi'= 0H/Cr = 5.0, freshly pre-
pared; b) n = 1.5, freshly prepared; c) n =
1.5, aged at 95°C for 36 h; d) n = 2.0 freshly
prepared; e) n = 2.0, aged at 95°C for 36 h.

 

 

83

.mmea we mugaom msu we now: we: moummz

 

owe
mpe
m—v
ape
Noe

Aacv

mmm
mmm
omm
mxm
mum

xms
A

e m~.¢

o_.N
oo.~
mm.p
mm.p
om.~

In

 

86mm ca 62361 on eoma

m—e
owe
owe
npc
v—e
No¢

Ascv

mum
omm
me
sum
mmm
mmm

me

«

a
.m=o_uzpom mmmgu cw um>gmmno we: mmucg_g_uamca :mmcm <e
a mo.m a ~¢.e m.~
c mo.~ ope mkm m mo.m o.~
mm.~ mpe mum m~.m m._
opm mpe mum mp.m o._
om.— mow mum o_.m m.o
mm.. Moe mum mp.~ o.o
:8 Asev xeea ze e c0\:o

 

comm oe atop: m com<

 

comm um vmgnmmga xagmmgu

 

eoeospom 68826.2

Ezpeocsu z ~.o mo Eaevxmz cowaagomc< m_n.m.> use In

meoFaPucou maoPLc> cane:
.—1>H m—nmh

 

 

 

84

 

Table IV-Z. The Change of pH of Chromium
Solution (OH/Cr = 2.00) as a

Function of Aging Condition .

 

 

Aging Condition pH

freshly prepared at 25° 3.63
25°C, 7 hours 3.50
25°C, 5 months 2.80
95°C, 6 hours 2.75
95°C, 36 hours 2.10

95°C, 240 hours . ’_ 2.00

 

 

85

air-dried pillared clays decreased to 27.6 A, 26.0 A,

o o i
25.5 A, and 22.0 A, respectively. In these latter experi-
ments, the hydroxy-chromium solutions were prepared at 0H/Cr =

2.00 and these solutions were aged for 36 hours at 95°C.
If a chromium solution aged for 36 hours at 95°C-and
fi = 2.00 was allowed to cool and aged further at room tem-
perature for 24 hours, the pH increased from 2.10 to 2.40.
The lattice expansion of the pillared product was about
1 A smaller for products prepared from the latter solution.
This result implies that the hydroxy-chromium polymers are
somewhat unstable and disassociate to smaller polymers at
lower temperatures. But, the extent of disassociation at
25°C is not great. In order to interlayer the large poly-
mers, the pillaring reaction was undertaken when the aged
chromium solution was still hot or within 1 hour after aging.
Smectite clays have been used as a molecular calipers
to measure the size of intercalated molecules7’8. Thus,
the expanded spacings of smectite clays should be equal or
close to the Size of interlayered molecules at least in one
dimension. As shown in Table IV-3, the tendency for the
001 spacings to increase with 0H/Cr ratio, aging tempera-
ture and aging time period strongly suggest that the degree
of hydrolytic polymerization of Cr(III) is directly related
to these three factors. With small additions of Na2C03,
the clays were only expanded by m 7 A. Evidently, the

interlayered chromium hydroxy species are not large com—

plexes. With larger additions of Na2C03, the clays were

 

86

expanded to higher spacings, a result which reflects the
larger size of interlayered hydroxy chromium polymers. The
results in Table IV-3 also indicate that heating the Cr3+
solutions with higher a values promotes considerably the
agglomerization of hydroxy chromium to form larger cationic
complexes. Only small hydroxy chromium oligomers, e.g.,

+ + +
Cr2(on)§ , Cr3(0H)g 3,Cr4(OH)g 9, or6(0H)$‘2‘ and

Cr]2(0H)gg 2 have been isolated or porposed previously,

Obviously, the interlayered hydroxy chromium polymers
must be much larger than these species.

Matijevic g£_gl, studied the hydrolytic reaction of
Cr(III), and found that pH, heating temperature and heating
time can promote the hydrolysis reaction of Cr(III) to form
identified chromium hydroxy polymer523. Laswick and Plane6
also found the formation of a colloidal precipitate in re-~
.fluxed chromium perchlorate solution, containing a small
amount of added base. I

Unlike the alumina pillared clays, the basal spacing
of hydroxy chromium-montmorillonite is not altered when it
is exposed to l00% humidity for 48 hours. However, when
the clay is exposed to ethylene glycol vapor, the basal

O O
spacing expands from an initial value of 27.6 A to 29.75 A.

B-3. Chemical Analysis
The elemental analysis of chromium pillared products
was done by the lithium metaborate fusion method as described

in Chapter II. The solvent used to dissolve the fused

 

87

 

Table IV-3.

Basal Spacings (A) of Air-Dried Chromium
Pillared Clays

 

OHZCr
0.

0.
1.

010010010

Aging Conditions for Hydroxy Chromium Solutions

25°C

3 weeks

14.48
15.5
17.3
l7.7

 

 

 

 

50°c 95°C 95°C 95°C
__2_hr_. 1 .h_r _6_h_:_ _3_5_£L_
14.5 A 13.38 A 15.0 A 15.3 A
14.5 14.0 15.3 17.0
15.0 14.7 - 17.3 18.5
15.7 18.5 | 23.9 27.5
17.7 21.0 8“ 25.0 27.5
20.0 23.2 24.5 27.5

 

 

 

88

sample was a mixture of 1.2% HNO3 and l.5% HCl instead of
1.2% HNO3 used for the aluminum pillared clay samples.
The procedure used to obtain the unit cell formula of
chromium pillared clays from elemental analysis is the
same as described in Chapter III. The chemical composition
and unit cell formula of chromium pillared clay are listed
in Table IV-4 along with the unit cell formula of
Brindley and Yamanaka's chromium-hydroxy montmorillonite].
Comparing the unit cell formulas of the two chromium
pillared clays in Table IV-4, we see that the

[Cr(OH)20é;7]+g'gg-montmorillonite synthesized in this

work contains more hydroxy chromium pillar species than that
of Brindley's material. Also, the interlayer lattice expan-
sion is higher, i.e. 27.6 A is l7.7 A for air-dried

samples. Therefore, the products represented by the data
contained in the box in Table IV-3 represent new composi-

tions of matter which should exhibit catalytic properties

distinctly different from those of Brindley's material.

B-4. Thermal Treatment of Hydroxy-Chromium Pillared
Montmorillonite
B-4a. Heated in the Presence of Air
Table 5 shows the basal spacings of calcined pillared
clays which were made from a Cr(III) solution at 0H/Cr =
2.00 and aged at 95° for 36 hr. After being heated at
110°C in air, the spacing decreased from 27.6 A to 25.2 A

which probably is due to the loss of solvating water

89

 

Table IV-4 Chemical Composition of Wyoming Na+-Montm0rillonite
and its Pillared Product Made from a Chromium
Solution with 0H/Crr=2.00 and Aged at 95°C for

 

 

 

36 Hours
Na-Montmorillonite Crx(0Hly-Montmorillonite
3102 53.94 Nt% 33.26 Wt%
A1203 18.86 10.71
Fe203 3.85 2.29
M90 2.18 1.09
Na20 2.13 0.02
Cr203 0 18.44

Chemical Formula
Na-Montmorillonite

Fe

Na0.5[A‘3.23 0.42"90.47](517.87A‘0.13)°20(°“)4

Crx(0H)y-M0ntmorillonite

+0.17
2.83 13.53EA‘3.23

020(0”)4

[Cr(OH) Fe A

0.42“90.47](517.87 10.13)'

Brindley and Yamanaka's chromium-hydroxy Montmorillonite

.34 .
[C”‘°”)2.55]1.88EA‘3.14Feo.40M90.483(5‘7.84A‘0.14)°20(°”)4

 

90

molecules. The pillared clays still exhibited a d002 at
22 A at 300°C (Table IV-5), but the clay was almost totally
collapsed at 400°C (d00£ = 10.3 A). The differential
scanning calorimetry analysis of this pillared clay, as
shown in Figure IV-2 (a), had a strong endothermic peak at
m 140°C which was probably due to the dehydration of the
pillared clay. Another broad endothermic peak at 390°C,
starting from 4 315°C, might be due to the oxidation of
chromium(III) oxide, perhaps the +6 oxidation state,
causing the migration of chromium from the interlayer to
the external surface. A broad endothermic peak at m 460°C
could be due to a further oXidation reaction or a phase
transition. It has been reported that oxidation of Cr203
occurs at about 350°C with the formation of mobile charged

species‘o.

 

Table IV-S. Basal Spacings of Thermally Treated Chromium
Pillared Clays Made from Solution. 0H/Cr =
2.00, Aged at 95°C for 36 Hours.

Temperature 25°C 100°C 200°C 300°C 400°C 500°C

 

 

In Air
5001 (A) 27.5 X 25.2 24.0 22.0 10.27

In Vacuum or Inert gas

000] (A) 27.5 X. .25.0 24.2 23.5 22.5 21.0

S

 

 

91

.m8@ :mmogu_c mo m:_3o_$ gmuca An
”own we mocmmmca as» aw Am mam—u umgmpp_m m_sogso mo mm>czu own

0.. 000 00¢ 00m CON 00.
q 1

3 d . #

 

an:

A3

.N->_ 823828

mm

 

92

B-4b. Heated In Vacuum or Under Inert Gases

In contrast to the above thermal results in the
presence of air, the chromia-pillared clays are consider-
ably more stable in the absence of oxygen. The basal spac-
ing results of a series of pillared clays made from
0h/Cr = 2.00 solution aged at 95°C for 1 hr., 6 hr. and
36 hr. and heated at different temperatures in the absence
of air are shown in Figure IV-3. Figure IV-4.summarizes
the spacing results of heated pi11ared clays made from the
solutions at 0H/Cr = 0.0 to 2.0 and aged at 95°C for 36 hrs.
Based on the estimated ionic radius of oxide ion, 1.2 A,
the interlayered chromium oxide clusters are composed
approximately five close packed oxide layers, at least when
the chromia pillared clays are made from solutions at
0H/Cr = 1.50 and 2.00 and aged at 95°C for 35 hours.

The XRD patterns of chromium pillared clays, which
were heated at various temperatures, and were made from
0H/Cr = 2.00 solution aged at 95°C for 36 hours are shown in
Figure IV-S. As shown in Figures IV-3, 4 and 5, these
chromium pillared clays still retain high spacing after
high temperature treatment.

The differential scanning calorimetry analysis of
Cr-PILC (OH/Cr = 2.00, aged at 95°C for 36 hr) under
flowing nitrogen gas, as shown in Figure IV-2(b), has only

an endothermic peak which is indicative of dehydration.

There is no other relevant endothermic or exothermic peak

93

 

[-5
1

 

 

““kk “*0 95°c 36hr
“‘~~A 95°c 6hr

1hr

 

Figure IV-3.

L 1 l l 1
ICK) ZIXD' 3(X) ‘ 14(X3 SIX) -
Temp (°C)

Basal Spacings at Elevated Temperatures of
Chromia Pillared Clays Which Were Prepared
from 0.1 M Chromium Solution with OH/Cr =
2.00 and Aged at 95°C for 1 h (X-X-X); 6 h
(A-A-A); and 36 h (1:1-0-0).

 

94

 

 

 

 

Figure IV-4.

 

Basal Spacings at Elevated Temperatures of
Chromia Pillared Clays Which were Prepared
from 0.1 fl Chromium Solution Aged at 95°C
for 36 h with OH/Cr ratio 0.0 (X-X-X); 0.5
(0-0-0); 1.0 (0-0-0); 1.5 (A-A-A); and 2.0
(o-u-o).

 

95

 

 

 

 

 

27.63
24
140$
(002) '
(o)
H.811 23d4
«xxa' oi
(b) 1
- (C)
W /
10.8A
(002)
h“ ..- _ # #(d)
35 2'5 2‘5 .5 15 ‘
26

Figure IV-5. XRD of Chromia Pillared Clay, Which was

Prepared from 0.1 fl Chromium Nitrate Solu-
tion With 0H/Cr = 2.00 Aged at 95°C for 36

h, Heated Under the Flowing of Argon at

a) 25°C; b) 200°C; c) 350°C; and d) 500°C.

 

96

at higher temperature. However, the thermal diagram has an
endothermic curvature extending to high temperature which
might reflect the continuous dehydroxylation of OH group

on the pillars.

B-5. Surface Area Results

The surface areas of chromium pillared clays, as
measured by the BET method, are 378 mz/g (OH/Cr = 1.50) and
350 m2/g (OH/Cr = 2.00). Both samples are made from the
solutions aged at 95°C for 36 hours. These results are
in agreement with published data on clays pillared with
inorganic oxides. Brindley and Yamanaga reported their
hydroxy chromium pillared montmorillonite to have BET
surface area of 280 m2/g when dehydrated at 200°C]. For

comparison, the alumina pillared clays described in Chapter

III have surface area of 250 to 400 m2/g.

B-6. Infrared Spectra and Pyridine Adsorption
The infrared adsorption spectra of Na+-montmorillonite

purified by sedimentation is shown in Figure IV-6(a). The

1

absorption peak at 3615 cm' is due to the stretching of

the lattice hydroxy group. The broad band at m 3400 cm"1

represents the 0H stretching of H20 in the interlayer and

1

the 1630 cm' peak represents the bending frequency of HOH.

Both of these disappear after calcination. A strong peak at
1420 cm"1 is indicative of the distinct adsorption of carbon-
ate which was removed after treatment at pHESsodium acetate
solution (Fig. IV-6(b)). The frequencies between 1150 to 250

cm"1 are assigned as the lattice vibrations arising from the

 

    

 

 

40001 ‘3000

1111111111L£/1\11
\W

2000 1500' 1200 1800 400
WAVE NUMBER (cm")

Figure IV-6. Infrared Spectra of Montmorillonite Clays

61)
b)

C)
d)

e)

f)

Natural sediment montmorillonite;

The sample in (a) treated with pH

5 sodium acetate;

Intercalated montmorillonite with
hydroxy-chromium polymers (air dry);
The sample in (c) was heated at vacuum
at 350°C for 3 h;

After adsorption of pyridine into (d)
sample for 6 h and then evacuated for
l h at 25°C;
After heatin

g (e) at 110°C for 3 h then
300°C for 3 h i

n vacuum.

 

 

98

vibrations of Si-0, Si-O-Al, Si-O—Fe, Si-O-Mg and H0-A1“.
The spectra of a self-supported film sample of hydroxy
chromium montmorillonite are shown in Figure IV-6(c) and
(d). The vibrations of the hydroxy group on chromium and
Cr-O-Cr may overlap with the vibrations of silicate lattice
and be difficult to distinguish. The broad band between

3400-3000 cm“ 1

and the strong peak at 1630 cm' arising
from the stretching of 0H and bending disappear after

heating in vacuum at 350°C (Figure 5-(d)). There is an

additional peak at 1440 cm“. 0n the basis of Coblentz
Society Spectra Cr203 and Cr203-2H2012, the 1440 cm'1 most
likely is not the vibrational peak of the pillaring chromium

(III) oxide species. Instead, the band may arise from the

‘ presence of a small amount of carbonate which come from
the binding of carbonate with cationic chromium-hydroxy
polymers during synthesis. The presence of carbonate,
however, is not regarded as essential for the syn-

thesis of the new composition of matters. Almost any non-
coordinating base can be used in place of carbonate pro-
vided high local concentration are avoided to prevent
precipitation of the hydroxide.

The adsorption of pyridine onto the chromia pillared
clays was employed to study its acidic properties because
this technique has been well developed and has become
almost a routine means of characterizing the acidity of

solid surfacesl3. Pyridine is a weak Bronsted base and

 

99

thus should interact only with the stronger, more cataly-
tically interesting proton sites. Furthermore, the bands
of the adsorbed species are sharp, allowing easy distinc-
tion between pyridinium ion and pyridine to Lewis acid
sties. The assignment of the vibrational modes of
pyridine adsorbed on Bronsted and Lewis acid sites are

listed in Table IV-614.

 

Table IV-6. Assignment of Vibrational Modes of Pyridine
(Py) Adsorbed on Bronsted or Lewis (L) Sites
Intensity: m = medium, 5 = strong,

vs = very strong

 

...

 

Vibrational Mode PyH Py:L
8 a Vcc (N) (A1) .. 1655 s ~1595 vs
8b Vcc (N) (B1) 1627 s 1575 m
19a vcc (N) (A1) 1490 vs 1490 5
19b ”cc (N) (B1) 1550 m 1455-1442 vs

 

Spectrum (e) in Figure IV-6 was taken after exposure
to pyridine vapor at room temperature for 6 hours and

evacuation at ambient temperature for 1 hours. The bands

1

at 1440, 1485, and 1590 cm' are the characteristics of

Lewis-bound pyridine. The distinctive band for pyridinium

1 is absent 'and indicates that surface

ion at 1550 cm'
acidity of crhomia-pillared clays was mostly of the Lewis

type.

100

By evacuating the pyridine-loaded sample at 300°C
(Figure IV-6(f)), some adsorbed pyridine was removed
because the band intensity decreased and the vibrational
‘bands shifted in frequency to 1445, 1480, and 1510 cm".
These frequency shifts suggest that the interaction becomes

stronger as the surface coverage is lowered.

B-7. Stability in 0.1 N HCl Solution

The acid stability of Cr-PILC was tested by exposing
the intercalate to 0.1 N HCl solution. The freeze-dried
Cr-PILC (60 mg), which was preheated at 300°C in vacuum,
was mixed with 5 ml 0.1 N HCl for 1 hours. Its XRD
pattern and spacing are not changed upon treatment with the
HCl solution. In contrast, the framework structure of NaY
was totally decomposed under the same condition. Jacobs14
described that some silica-rich zeolites, e.g. mordenite
and offretite, are stable in 0.1 N HCl solution.

The air-dried Al-PILC is not stable under analogous
acid conditions because the spacing decreases from N 20.0 A
to 15.78 A. This is due to the dissolution of the pillaring
hydroxy aluminum polymers in HCl solution. 0n the other

hand, if the Al-PILC is calcined at 350°C, its spacing

0
does not change and remains around 18.5 A.

B-8. Sorption of Hydrocarbons
The adsorption of probe molecules which have different

kinetic diameters was also employed to investigate the pore

opening of chromium pillared clays. The results are

101

summarized in Table IV-7.

Table IV-7. Sorption of Probe Molecules of Chromia
Pillared Clays (freeze-dry) (OH/Cr = 2.00,
aged at 95°C for 36 hours)

 

 

Kinetic Amount
Probe Molecules ' P ~ Diameter Adsorbed
O
Benzene 70 torr . 5.8 A 2.58 mmole/g
Neopentane 728 6.2 1.56
1,3,5-triethylbenzene 0.18 , 9.2 0.91
Perflorotributylamine 0.22 10.2 0.74

 

As shown in Table IV-7, the Cr-PILC can adsorb the
largest molecule, perflorotributylamine, with a kinetic
diameter of 10.2 A. This adsorption implies that the pore
size is large enough for perflorotributylamine molecule to
diffuse into the intracrystalline pore cavity. The molecule
might be adsorbed in such an orientation that the lone
pair of electron on nitrogen atom is perpendicular to the
C-axis of the clay framework structure because at 350°C

Ad = 13.5 K.

001
The shape of 1,3,5-triethylbenzene molecule has a pal-

mate or pallet type structure. Its kinetic diameter is

9.2 A and it can diffuse into Cr-PILC in an analogous way

to a benzene ring that is parallel to c-axis of clay. 0n

the other hand, the neopentane molecule has a spherical type

0
structure with kinetic diameter 6.2 A. Thus, conservatively,

102

O

the interpillar distance is larger than 6.2 A and the pore

opening is larger than 6.2 A x 13-5 A.

B-9. Catalytic Dehydrogenation Reaction

The dehydrogenation of alkanes and alkenes are pro-
cesses of industrial importance. The dehydrogenation of
butane to butene and butadiene and of isopentane to olefins
and ultimately to isoprene are essential processes for the
manufacture of synthetic rubber. Straight-chain C12 to
C18 alkanes are dehydrogenated to alkenes which are employed
in the production of alkylbenzenes used in the manufacture
of biodegradable detergents. Ethylbenzene is dehydrogenated
catalytically to styrene which is used in the manufacture
of rubber and plasticsIs’IG.

Two types of catalysts are Used for the dehydrogenation
reactions: reduced metals and metal oxides. The reduced
metal catalysts are those of Group VIII in the Periodic
Table, preferably nickel or platinum deposited on supports
such as alumina, silica or activated carbon. The catalytic
metal oxides are mainly chromia and molybdena.

Usually, supported metal catalysts are used for aroma-
tization reactions and oxide catalysts, especially
chromium oxide on alumina are preferred catalysts in the
dehydrogenation of butane and isopentane. Supported metal

catalysts, e.g. Pt/Alzo3 are less suitable for the dehydro-

genation of lower alkanes since at the temperatures required

for dehydrogenation they produce a secondary hydrogenolysis

 

 

103

reaction. The oxides as such undergo rapid deactivation
during aromatization; however, the oxides deposited on
alumina by impregnation are more resistant to rapid
deactivation17.

Chromium pillared clay has a porous-structure in
which the pillars, formally Cr203 clusters, are more or
less regularly distributed on the silicate surface. As
mentioned above, supported chromia is a well known
catalyst for dehydrogenation of saturated hydrocarbons.
The small chromium oxide clusters in Cr-PILC‘might be
active in the dehydrogenation reaction. Preliminary studies
carried out by Steven Landau in this laboratory indicate

that chromia pillared clays are indeed active dehydrogena-

tion catalysts.

B-10. Dehydrogenation of Cyclohexane
For comparison, Brindley's chromium-hydroxy-montmoril-
lonite and chromia alone were prepared according to the

1 18 respectively. In

method of Brindley and Burwell
Landau's studies all these three catalysts were dehydrated

at 550°C in a flow of helium gas for 1 hours. At 550°C,
hydrogen was passed over the pillared clay for 1 hours.
Cyclohexane was introduced to the reactor at 550°C by using a
syringe pump and swept through the reactor with helium gas.
The weight-hourly-space-velocity (WHSV) was adjusted to

about 1.0.

104

The pretreatment of catalysts under Hz at high tempera-

19,20

ture is essential. It has been found that the hydro-

gen pretreatment generates divalent chromium ion (Cr2+) as

the active site upon catalyst reduction:

3+ _ .2+

n-l 2

Upon adsorption, the saturated hydrocarbon, is activated

via the cleavage of C-H bond. The alkyl group is bound

by a Cr2+ ion on the catalyst surface and an H atom is
transferred to an oxygen neighbors20
2 02‘ OH-
- .2+ 2- 2+ -
0 ' 0 '

The dehydrogenation pathway of cyclohexane on bifunc-
tional catalysts is schematically shown in Figure 721’22.
As shown in the reaction scheme, cyclohexane is converted

to cyclohexene on a dehydrogenation site where further
dehydrogenation to cyclohexadiene and finally benzene takes
place. The cyclohexene migrates to a neighboring acid

center; there it is protonated to give a carbonium ion,

which can then isomerize to form methylcyclopentene. The
latter compound may migrate to a hydrogenation-dehydrogena-
tion site to form methylcyclopentane or methylcyclopentadiene.
Further parallel reaction of methylcyclopentane to alkenes

and alkanes can occur on acid centers and hydrogenation-

dehydrogenation centers.

 

105
A\
NV /Y\
. 11 11
2% g ‘<:>, (£7 ;;2: /A\/~\4> :2: 47\r~\
835» 11 11
as; 0 = <5
5% 11 11
$3; <3 6:, o
i: 11
Q

 

 

VI

Transition on ocidic centers

Figure IV-7. Dehydrogenation-isomerization Reaction
Scheme of Cyclohexane on Reforming Catalysts
Containing Acidic Sites.

 

106

The reaction product of dehydrogenation of cyclohexane
is mainly benzene. This observation is consistent with
the results of pyridine adsorption on dehydrated Cr-PILC
which reveals that Bronsted acid sites on Cr-PILC are very
limited. Thus, cyclohexane is directly converted to ben-
zene rather than being isomerized to methylcyclopentene;
the benzene selectivity here is essentially 100%.

The results for cyclohexane dehydrogenation over the
Cr3.5-montmorillonite clay is shown in Figure IV-B. The
chromia pillared clay of this work (Cr3°5-mont) retains high
conversion activity after two hours reaction. In contrast
the chromia gel exhibits an induction period prior to
reaching maximum activity24. However, it is deactivated
very rapidly. The conversion yeild over Brindley's
chromium-hydroxy-montmorillonite (Cr].9-mont) is almost
undetectable. .

The differences between the two chromia pillared clays
(Cr3.5-mont and Cr].9-mont) may be due to their different
pore structures. After being heated at 550°C in vacuum,
the Cr-PILC have a basal spacing 13.7 3 (or Ad00] = 4.1 A
for CrIOQ-mont) and 19.0 A (or Ad001 = 9.4 A for
Cr3.5-mont). Obviously, the pore opening of latter
pillared clay is larger than the kinetic diameter of
cyclohexane, 6.0 A, and allows the cyclohexane molecule

to diffuse into the intracrystalline pore cavity and react

with the intercalated Cr203. The ease of intrapore

107

C6 le Dehydrogenation of 550°C

'°° ' LO 9 06H .2 /g Crz Oa/hr.

90-
80-
70"
60 '-
50-

 

% CONVERSION

3O
20 '-
IO

r

 

l l l l I I 1 1

I5 30 45 60 75 90 |05 IZO l3

 

TIME (MIN)

Figure IV-8. Conversion Yield of Cyclohexane Over Chromia
Pillared Clay Catalysts.

108

diffusion is supported by the adsorption isotherms of

cyclohexane on these two Cr-PILC as shown in Figure IV-9.

IV-C. Conclusions

The present results have clearly shown that the forma-
tion of thermally stable porous pillared smectite clays
can be achieved by interlayering the clays with hydroxy
chromium polymers. It has been demonstrated that the
expanded spacings are dependent on the sizes of cationic
hydroxychromium polymers which are produced by the
hydrolysis of chromium ion under various conditions of
aging temperature and aging time period.

The results also provide for the first time evidence
that chromium forms hydroxy-polymers in solution with a
size of m l7 3 in one dimension. It is not possible to
describe the shape and composition of these polymers from
the present results. However, it is evident that the
aggregation of hydroxy chromium polymer is three dimen-
sional.

The dehydrated chromium pillared clays have a pore
cavity accessible to the adsorption of hydrocarbon and are
active catalysts for the dehydrogenation of cyclohexane in
high selectivity. The implications are that these new
intercalation compounds may be useful catalysts for the

reforming of hydrocarbons to high octane gasoline.

109

2 0L C6 Hl2 Uptake Curves

    

\ Crag-Mont offer rxn. at 550°C
(com: 913)
I5

MMOL PER GRAM
6

0.5

 

l L l l L l l n l g l

6 8 IO 12

4
fimm’z

Figure IV-9. Adsorption Isotherm of Cyclohexane in
the Used Chromia Pillared Clay Catalyst.

11.

110

REFERENCES

Brindley, G.W.; Yamanaka, S. Amer. Mineral. 64, 830-
835 (1979). -

Bjerrum, N. Ph.D. Dissertation, Capenhagen, 1908.

a) Sister Gertrude Thompson, C.S.J., Ph.D. Thesis,
Lawrence Radiation Laboratory Report UCRL-ll410,
University of California, Berkeley, June 1964;

b) Thompson, M.; Connick, R.E. Inorg.hem. 20, 2279-85
(1981); c) Finholt, J.E.; Thompson, M.E.; Connick, R.E.
Inorg. Chem. 29, 4151-4155 (1981).

a) Kearby, K.K. "Catalytic Dehydrogenation" in
"Catalysis" Emmett, P.H., Ed., Vol. III, Reinhold, NY
1955, pp. 453-491; b) Pines, H.; Csicsery, S.M. J.
Catal. l, 313 (1962); c) Ciapetta, F.G.; Dobres, R.M.;
Baker, R.M. in "Catalysis" Emmett, P.H., Ed., Vol. VI
Reinho1d: NY 1958. PP. 495.

a) Clark, A. Catal. Rev. 3, 145 (1968); b) McDaniel,
M.P. J. Catal. 1g, 37 (1902); c) She1ef, M.; Otto, K.;
Gandhi, H. J. Catal. 1g, 361 (1968).

Laswick, J.A.; Plane, R.A. J. Amer. Chem. Soc. g1,
3564 (1959).

Green-Kelley, R. Trans. Faraday Soc. 51, 412 (1955).

Brindley, G.W.; Hoffmann, R.M. Clays Clay Miner. g,
546 (1960).

Stunzi, H.; Marty, N. Inorg. Chem. 22, 2145 (1983).

Hope, G.A.; Ritchie, I.M. J. Chem. Soc. Faraday Trans.
I, 11, 2621 (1981).

Grim, R.E. "Clay Mineralogy" 2nd Ed. McGraw-Hill, NY
1968. PP. 459.

12.

13.

14.

15.

16.
17.

18.

19.

20.

21.

22.

23.

24.

111

Coblentz Society Spectra, No. 4699 and 4700, Sadtler
Research Lab. Inc. 1975.

Benesi, H.A.; Hinquist, B.H.C. Adv. Catal. 21, 97 (1978).
Jacobs, P.A. "Carboniogenic Activity of Zeolites",
Elsevier Scientific Publishing Company, Amsterdam,
1977.

Hiseman, P. "An Introduction to Industrial Organic
Chemistry", John Wiley & Sons, NY, 1972.

Kaeding, H.H. Catalysis Review B, 307 (1973).

Pines, H. "The Chemistry of Catalytic Hydrocarbon
Conversions" Academic Press, NY, 1981.

Burwell, R.L., Jr.; Taylor, H.S. J. Am. Chem. Soc.
§§, 697 (1936).

Burwell, R.L, Jr.; Haller, G.L.;Taylor, K.C.; Read,
J.F. Adv. in Catalysis 20, 2 (1969).

VanReijen, L. L. ; Sachtler, H. M. H. ; Cossee, P.;
BrouHer, D. M. 3rd Inter. Cong. Catalysis, 829 (1965).

Mills, G.A.; Heinemann, H.; Milliken, T.H.; 0b1ad,
A.G. Ind. Eng. Chem. 45, 134 (193

Satterfield, C.N. "Heterogeneous Catalysis in
Practice" McGraw-Hill Book Co. New York, 1980.

Kratohvil, S.; Matijevic, E. J. Coll and Interf. Sci.

2_4, 47 (1967).

Landau, S.; Pinnavaia, T.J. unpublished results.

CHAPTER V
CLAY CATALYSTS PILLARED BY HYDROXY IRON(III) POLYMER

V-A. Objectives

Pillared clay catalysts represent a new class of
microporous inorganic solids with the potential to be as
useful as the aluminosilicate zeolite. It has been demon-
strated that smectite clays pillared by alumina and zirconia
are capable of cracking more gas oil1 and have higher reac-
tion rates and conversion in the dealkylation of cumene and

2’3 than Y-type zeolites due to the

l-isopropylnapthalene
sterically unhindered intrasorption of such bulky sub-
strates in the pillared clays.

In Chapter III, it has been shown that the certain
properties of aluminum pillared clays such as surface area,
thermal stability and pore size are highly dependent on
the shape of the pillaring hydroxy aluminum polymers (1.3;,
spherical keggin ions or rod-shape polymers). These
I known metal-hydroxy-polymer smectites, including those of

A1, Zr, and Bi, might represent the initial members of a

broader family pillared clay catalysts. In an effort to

112

113

extend the pillared clay catalysts family, we have success-
fully synthesized a new pillared clay, Cr-PILC, as illus-
trated in Chapter IV.. These results have shown that the
expanded basal spacings and catalytic dehydrogenation
properties of the intercalate are dependent on the size of
the three-dimensional hydroxy chromium polymers which formed
during hydrolysis. In this chapter, we report the synthe-
sis and characterization of another new pillared clay
catalyst which is based on the intercalation of hydroxy

iron(III) polymers as the pillaring agent.

V-B. Hydrolysis Chemistry of Iron(III)

Ferric oxides and their hydrates occur widely in soil.
Their properties in aqueous solutions is a valuable adjunct
to the study of the whole soil. A large volume of work is
available on quantitative studies of the composition of
aqueous ferric salt solutions.

It has been generally recognized that when an iron(III)
salt is dissolved in water, the initial reaction products
are true solutions species containing low molecular weight
hydrolysis products, such as Fe(0H)2+, Fe(0H);, Fe2(0H)g+

1 2 reviewed this area.

and Fe3(0H)2+. Bjerrum'g£_al. and Sylva
Gelatinous ferric hydroxide precipitates can be observed as

the final reaction product if the solution is aged either at
room temperature for a long time period (at least two months)

or at 75°C for a few hours.

114

These hydrolysis reactions of aqueous ferric(III) salts
have been studied by a variety of means, such as electro-

metric measurement3, conductometric titrations4’5,

spectrophotometric measurements including UV-visible6'9

10 11

and Raman spectroscopy, magnetic measurements and

EXAFS‘Z.

The hydrolyzed species formed in basic iron(III)-
hydroxy solutions are very complex. Considerable amounts
of base can be added to iron(III) solutions of moderate
concentrations, up to a OH/Fe molar ratio of at least 2.5,
without causing immediate precipitation‘4. As the 0H/Fe
ratio increases, the color of the solution changes from
yellow to orange to deep red-brown. It is doubtless that
these color changes are the result of continuing polymeriza-
tion reactions. As the degree of polymerization of the
polynuclear species increases, further deprotonation of
coordinated water molecules (olation) and hydroxy-groups
(oxalation) can be expected to occur.. These are common
reactions of many metal ions but such systems are extremely
'difficult to investigate quantitatively becauselyftheir
complexity and the absence of equilibrium conditions.

Diffusion measurements with partially hydrolyzed
solutions of iron(III) perchlorate led Jander and Hinkel13
to suggest that polynuclear iron-hydroxy complexes were

present. These complexes were thought to contain 1 to 50

metal atoms.

115

‘4 published a detailed investiga-

In 1966, Spiro et_al,
tion of hydrolyzed iron(III) solutions. Varying amounts of
potassium carbonate were added to an 0.3 M iron(III) nitrate
solution up to a 0H/Fe ratio of 2.5. Analytical ultracen-
trifuge studies demonstrated the presence of a reasonably
discrete polymer fraction. The extent of formation of the
polymer in fresh solutions increased linearly with increas-
ing values of 0H/Fe. Measured sedimentation coefficients
indicated that the molecular weight of the polymer
remained relatively constant for values of 0H/Fe in the
range of 0.5 to 2.0, but the molecular weight increased
considerably if the ratio is increasedto 2.5.

By using the gel-filtration method, they were able to
isolate hydroxy iron polymers from 0.3 M ferric nitrate
solution at a OH/Fe value of 2.00. The composition of the
isolated polymer was found to be [Fe00.75(0H)(N03)0.5]~
0.36H20 with average molecular weight of 1.4 x 105.

A structural model, called the "Saltman-Spiro Ball"
model was proposed for the hydroxy iron(III) polymer as
shown in Figure V-lls. Each internal Fe atom in the poly-
mer is tetrahedrally coordinated to two 0 atoms and two
OH groups. The external Fe atoms are bound to one 0
atom, two OH groups and one water molecule. The nitrate
ions appear to be hydrogen-bonded to the water molecules.

After a careful reexamination of the electronic adsorp-

tion spectrum of the Saltman-Sprio Ball polymer, Gray16

116

@ @

Coo@u®e@e@@

@

@@o- g -@e©e@@

0 ® @

”@mmwe ©e s®®®©

o
@@M@_@e

m m @
a- -@M©®®©

117

concluded that the hydroxy iron(III) polymer contains dis-

torted polymeric networks of [Fe(III)06] units primarily

oct
linked by bent oxobridges.

Dousma et al.18"20

observed the basic hydrolysis of
titrated iron(III) nitrate solutions by several techniques,
including pH relaxation, optical density measurements, light
scattering, ultracentrifugation and electromicroscopy.
They concluded that the hydrolysis process occurs in two
steps18'20:
a) Hydrolysis - polymerization leading to oxolation,
nucleation and growth into small particles with
maximum size about 40 A;
b) Aggregation to large particles (200-500 A) and
anisotropic flocculation to linear chains followed
by slow sedimentation.

21'24 on the characterization

In a series of publications
of partially neutralized ferric salt solutions, Murphy 91:31,
demonstrated by electron microscopy and density gradient
ultracentrifugation techniques, that discrete spherical
polycations with 15-30 A diameter were formed at the initial
stage of hydrolysis. These spheres were linked during aging
to form short rods (that were composed of 2-6 spheres. As

aging continued, the spheres coalsced and the rods increased

0
in both length and thickness to give rafts m 200 A x 30 A.

118

V-C. Results and Discussions
C-l. Synthesis of Hydroxy Iron Polycations at Room
Temperature

The synthetic procedures used to pillar montmorillonite
with hydroxy iron polycations are related to those used in
'the chromium system. Sodium carbonate is used as the base
to hydrolyze Fe(III) solution to form hydroxy iron(III)
polymers. Various amounts of carbonate are added into
0.2 M FeCl3 or Fe(N03)3 solutions to give different 0H/Fe
ratios (meq of base per mmole of Fe3+) in the range of
0.0 to 2.5 without causing any precipitation. These basic
ferric salt solutions are flushed with nitrogen to remove
all C02 gas. As the 0H/Fe ratio is increased, the color
of the solutions changes from yellow'(0H7Fe = 0.0) to
orange (OH/Fe = 0.5) to deep-brown (OH/Fe 1 1.0). The
color change is believed to be the result of a continuing
polymerization reaction14.

When ferric sulfate solutions are used, a precipitate
is formed at 0H/Fe = 1.5 and more precipitate is formed
at higher ratios of 0H/Fe. Upon the removal of ferric
hydroxide precipitates, the colors of all Sulfate solutions
were a similar yellow-brown.

The electronic adsorption spectra for the nitrate,
chloride, and sulfate salt solutidns at 0H/Fe = 0.0 to 2.5
after aging at room temperature for 24 hours exhibited a
A in the range of 870 ;: 30 nm. The).max of these

max
same solutions increases with 0H/Fe ratio as shown in

119

Table V-l and Figure V-2. The).max at high OH/Fe values
(1.5 to 2.5) are near 900 nm, which is close to the Amax
of the Saltman-Spiro Ball‘s. It seems that these solutions
might have polycation species similar to the isolated poly-

mers of Spiro et al.14.

 

Table V-l. A (nm) for Hydrolyzed Ferric Solutions

 

 

max
OH/Fe FeCl3 Fe(M03)3 Fe2(SO4)3
0.0 810 nm 810 nm 820 nm
0.5 840 860 830
1.0 875 -860 840
1.5 880 870 840
2.0 885 890 840
2.5 895 840

 

[Fe3+] = 0.2 M; cell path = 1 cm; base = Na2C03;
Solutions are aged at room temperature for 24 h
before measurements.

C-2. Exchange Reaction of Hydroxy Iron Polycations
and Na-Montmorillonite
The pillaring reactions were conducted in the same way
as described in Chapter II. The ratio of iron to clay was
70 mmole/meq clay. The excess hydroxy iron polymers were
washed with deionized water by centrifugation. It takes
about ten washings to obtain a flocculated,reddish-brown

pillared product.

120

ABSORBANCE

 

 

 

l_ l l_

400 500 600 700 800 900 mm IIOO I200
x (nm)

Figure V-2. Absorption Spectra of 0. 2 M FeCl3 Solutions
A ed at 25° C for 24 h and with OH/Fe Ratio
610 .00 and; c) 2. 00.

.121

Table V-2 provides the basal spacings of the pillared
products obtained from the three different ferric salt
solutions, as well as the pH values of pillaring solutions.
Generally, the pH of solutions containing the same
anion increases with the addition of carbonate, but the
rise in pH is only a few-tenths of a unit for each consecu-
tive 0.5 unit increase in OH/Fe ratio. The products in
Table V-2 with d001«3 18 A have lattice expansions which
are far greater than previously described Fe(III) inter?

17, and, therefore, they represent

calates of smectite clay
a new composition of matter.

At 0H7Fe = 0.00, the solutions give a basal spacing
of 12.3 A suggesting that there may only be relatively
small oligomers such as, Fe(0H)2+, Fe2(0H)g+ and Fe3(0H)2+
present in solution. In the FeCl3 system the d001 spacings
increase rapidly in the low DH/Fe ratio range (5 = 0.0 to
1.0), and then increase more slowly between a = 1.0 and 2.0.
This suggests the fprmation of a discrete fraction of
hydroxy iron(III) polycation near 5 = 1.0. The linear
increase of basal spacing with a in the range of 1.0 and
2.0 reflects the linear growth of the polycation with the
amount of added carbonate. This is consistent with the
results of analytical ultracentrifuge studies for
similar solution514. The expanded spacing of Ad001 =
15.6 A for the pillared clay made from FeCl3 solution with

h = 2.00 is in the range of sizes for hydroxy-iron(III)

122

 

.mcowaapom wmmzu cw cmsccm pan I

 

 

 

 

 

 

..N 2 ae mm game cease m

.cowpummc mcowmn meson em cow assumcmasmu soc; um ummm men mcowuapom "moummz "mmmm m

.cme\oposs cu n amFU\mu

meeapspem mAeomVNea Lee 2 _.o eea «Amozvec eea m_oac sec 2 N.o m_ +mec to .eeee m
- e.N_ mom.~ ma.e_ ..NNK mo.N mo.- m..~ m.~
- e.N_ mmN.~ mm.m_ m.m~ Nm._ e.N_ N.mN N~._ o.~
- m.N_ mm~.~ N.m_ N.m~ m~._ o.~_ m.e~ oe.F m.—
- m.- m_.~ o.m_ ~.m~ we.P mm.m_ m.m~ mm._ o._
- m.~_ mm._ - e.mp ~m.. m.NF m.m_ oe._ m.o
- mm.- mm.p - m~.~_ mN._ - m~.~_ mm._ 0.9

«so Pace za. Nose. .ooe. =a Amv Noon Amy .ooa 2a m ec\=o

omvmec mfimozvec n.9ec

Incowaspom :owc< owcgwm mags» Ease mum: mxnpu voLmPFPQ coca we macromgm pummm .~-> mpamh

a

 

123

polymers characterized by electronmicroscopyzz.

The XRD (Figure V-4(b)) of the pillared product made
from FeCl3 solution with 0H/Fe = 0.50 has a weak diffrac-
tion peak at 19.8 A and an intensive one at 12.5 A. This
indicates that this material is a mixture with compo-
sitions with d001 = 19.8 X and d001 = 12.5 K. This
might happen in the pillared clays made from FeCl3 solutions
with 0H/Feratio between 0.0 and 1.0.

Interstratified products were obtained when mont-
morillonite clay was exchanged with a fi = 2.50 FeCl3 solution
(3:. Table V-2 and Figure V-3). X-ray diffraction has a
broad peak between 29 = 4° and 7°. The interstratification
could be the result of the formation of a wide size distri-
bution of hydroxy iron(III) polymers at this 0H/Fe ratio.
The XRD of air-dry pillared clays made from FeCl3 solutions-
with a = 0.0 to 2.5 are shown in Figure V-3-

The pillared clays made from ferric nitrate solution
have XRD patterns similar to those made from ferric
chloride solutions. Surprisingly the ferric sulfate solu-
tions give products with low basal spacing. This result

will be discussed in greater detail in Section C-5.

C-3. Aging Effect

Aging time and aging temperature are very important
factors in hydrolysis chemistry. These parameters influ-
ence the formation of hydroxy iron(III) polymers and the

spacings of pillared clay products.

124

Figure V-3.. XRD of Iron Pillared Clays Made from 0.2 M
FeCl3 A ed at 25°C for 24 h and with 0H/Fe
Ratio a 0.0; b) 0.5; c) 1.0; d) 1.5;
e) 2.0 and; f) 2.5.

125

I2.28&'\

(001)
I2.5A
(002)
19.8A
(a) (con
12.6A
(b) (00»
RCA ,(9) 4.5%. I
(002) 001
W 25213“
(001)
I253
(002)

(e)

l l 1 l I 1 1 1 I l

 

 

l
242220|8l6l412|0864
26

126

C-3a. Aging for Different Time Periods

When 0.2 M ferric chloride (5 = 2.00) was aged at room
temperature, the pH value decreased and the basal spacings
of pillared clays increased with the aging time period
as illustrated in Figure V-4. The solution becomes more.
acidic with aging time, probably because the increasing
extent of the olation and oxolation steps in the hydrolysis
processes (gfi., Chapter I).

The decrease in pH is greatest in the first 30 hours of
aging. Also, the basal spacings (d001) of the pillared
products increase from 23.8 A after 1.5 hours of aging
to 25.2 A after 5-30 hours of aging and then increases fur-
ther to a value of 26.8 A. The XRD patterns of these
pillared clay products are illustrated in Figure V-5. Still
larger inCreases in basal spacing may occur at longer aging
times. These results suggest that the hydroxy iron(III)
polymers increase in size to 15 A in 1.5 hours. When the
solution is allowed to age 1.5 to 30 hours, the hydrolysis
reaction occurs continuously, causing a smooth decrease in
pH values. Thus, it seems that during this aging time
period (1.5 hours to 30 hours), the extent of hydrolysis
increases, but the size of polycations does not increase.
Also, the size distributions probably become more uniform,
because the half-width of the 002 diffraction lines become
smaller as shown in Figure V-4.

These results are consistent with the conclusions

drawn from previously reported hydrolysis studies of ferric

127

 

 

 

 

 

E 2.1 ‘27 25 i
2.0 r’ . ‘26, ~20:
v ‘ g
S 1.9 I» ~25 -15 3
s 1 ‘ a
,0, 1.8 -24€ ~10 g '
an 423 g (15°
1.7 "" 0.—
LE 1, o
o.- .C
o 1.6 ‘22 '0 .32"
3

£45 -21 g
l l 1 l l l 1 a

10 20 30 40 so so 70 N

Time (hour) 4

Figure V-4.

The Variation of pH Values of Iron(III)
Solutions and the dog] Spacing and A26 of
002 Diffraction Lines of Pillared Products
With Aging Time Periods of 0.2 M FeCl3 with
0H/Fe = 2.00.

128

Figure V-5. XRD of Air Dry Iron(IV) Pillared Clays Which
are Prepared from 0.2 M FeCl Solution with
OH/Fe = 2.00 and Aged at 25° for a) 1.5 h;
b) 10 h; c) 30 h; d) 50 h and; e) 75 h.

(o)

(b)

(c)

 

 

(d) .
I 1
(e)
l 1 l 1 I l
12 lo 8 6 4

 

130

ions, namely, that in extensively hydrolyzed solutions,
the polymer particles grow rapidly until a limiting value
of molecular weight is reached14.
When the solution is aged beyond 30 hours, the pH
values decrease and the (1001 Spacings increased, and
might be due to further polymerization to form larger hydroxy

iron(III) polymers.

C-3b. Hydrolysis at Low Temperature

In general, the hydrolysis of metal ions is promoted
at high temperatures. This observation can be useful for
obtaining pillared products with higher d spacings. For
instance, montmorillonite can be expanded from 17 A to
27.6 A if a chromium nitrate solution at 0H/Cr = 2.00
is heated at 95°C for 36 hours. On the other hand, if
the ferric ion is hydrolyzed at low temperature, it should
be possible to synthesize small size polycations by carry-
ing out the hydrolysis at lower temperatures. Thus, a low
temperature hydrolysis experiment was carried out with a
ferric solution to examine the possibility of obtaining
more regular expanded spacings. Systems with regular
spacings are apt to be more highly shape-selective as
catalysts.

Ferric chloride solutions (0.2 M) were hydrolyzed
with sodium carbonate (OH/Fe = 2.00) at three different
temperatures, 0°C, 4°C and 10°C, and were allowed to age

at these temperatures for 24 hours prior to being used in

131

the pillaring reaction. The montmorillonite suspensions
were cooled to the same temperature before they were added
into the hydrolyzed ferric solutions. The pillaring
reaction and washing procedures were also carried out at
these three low temperatures.

The XRD of the air-dry pillared clays prepared at 0°C,
4°C and 10°C are shown in Figure V-6. The d001 spacings
obtained are 21.0 A, 22.1 A, and 24.871, respectively. Obviously
polycations with a size of about 11 A are formed even at
0°C. The polycations are only slightly smaller than those
formed at 25°C. If the pillared clay prepared at 0°C is
heated at 110°C for 24 hours, the spacing decreased to

d = 17.7 A. Its XRD patterns are shown in (d) of

001
Figure V-7. Clearly, these products are no better
ordered than those obtained by hydrolysis of ferric ion
solutions at 25°C.
C-3c. Hydrolysis at Elevated Temperature

Hydrolysis of metal ions at elevated temperature
should promote the formation of higher molecular weight
species. A 0.2 M FeCl3 solution (OH/Fe = 2.00) which has
been aged at 25°C for 10 hours only a pillared product with
d001 = 12.3 A. When the solution is heated at 75°C for
several hours, the hydrolysis reaction is greatly extented,
as indicated by the greater decrease in pH and the formation
of a precipitate (See Table V-3). These solutions also

give clays with d001 = 12.3 A, which suggests that

mainly small cation oligomers are present in such

Figure V-6.

132

XRD of Air Dry Iron(III) Pillared Clays Which
are Prepared from 0.2 fl FeCl3 Solution with
0H/Fe = 2,00. The Hydrolysis Tempratures are
a) 0°C; b) 4°C and; c) 10°C. The XRD of (d)
is of the sample of (a) heated at l00°C for
24 h.

ich

are
d)

133

Ian/S ‘
(002) i885

 

 

 

(o)
IZBK ,
. (002) 2481‘ p
(b) l l . (Ob!)
, y j ‘ '
(9532) w'
mil
(C) M (000 ‘
. \ {I
it i
(d) '
“I.M....

I2 IO 8 6 4

134

 

 

 

 

m=o_aou ..~_ o~._ o.o_
gone ..N_ o~._ c.m
agom a N.o~ om._ m.P 00mm
ago: N.m~ m~., e.g.
ago: ~.mm om._ o.m
«so: < «.mm om.~ m._ comm oo.~
gaze NN.~_ oo.~ o.o_
meow NN.N_ op._ o.m
m_uu__ < N~.~. o~._ m._ 00mm
ago: ¢¢.~p ~¢.. e.g.
ago: m.~_ o¢._ o.m
acac < mN.N_ w¢._ m._ UOmN oo.o
use a .pooo In a: .mE_H mewm< .aamp me_m< aa\=o
mmgaamgmqam» acmgomw_o u< muowgma weph maowgm> so» umm< mpuwm
z ~.o sate mum: manpu omtm___a Afimhvcoeu cc mm=.u~am .ammm .m-> m_nah

 

135

extensively hydrolyzed solutions.

Heating a ferric chloride solution (OH/Fe = 2.00),

at 75°C for l.5 hours promotes the hydorlysis reaction and

increases the basal spacing from 23.8 A to 26.7 A. How-
ever, some precipitates are formed. If the solution is
heated longer, more precipitate is formed and the exten-
sively hydrolyzed solution gives only product with a
lZ.l A spacing.

C-4. Chemical Composition, Thermal Properties and
Hydrocarbon Adsorption

As discussed in the previous section, the basal spac-
ings of pillared montmorillonite vary with the 0H/Fe
ratio (V-C-2). Even at a fixed 0H/Fe ratio, the basal
spacings depend on the hydrolysis time period (V-C-3a).
The pillared montmorillonite product synthesized from
0.2 M FeCl3 at 0H/Fe = 2.00 and aged at room temperature
for 24 hours was chosen to study thermal properties,
chemical composition and hydrocarbon adsorption.

The chemical composition and unit cell formula is
given in Table V-4. Each unit cell contains about 10
iron atoms, which is much higher than in Al-system
(2.87 Al per unit cell) and Cr-system (3.53 Cr per unit
cell).

After being calcined at high temperature, the pillared
clay reatins a high spacing. The X%ray diffractograms of

sample heated at different temperature are shown in

136

 

Table V-4. Chemical Composition and Unit Cell Formula of
Wyoming Na+-montmorillonite and Iron(III) hydroxy
Polymer-montmorillonite Made From a 0.1 M FeCl3
Aged at Room Temperature for 24 hours at OH/Fe =
2.00

 

l. Chemical Composition

+ . 0+
Na -montmorillon1te Fex(OH)y-montmorillonite
8102 53.94 wt. % 26.28 wt. %
A1203 18.78 ' ' 8.89
Fe203 3.83 43.37
M90 2.23 0.97
Na20 2.05 0.017

2. Unit Cell Formula

Na+-montmorillonite

("°o.58)[A‘3.1iF°0.42”90.48]‘517.83A‘o.121°20‘0”)4

Fe-Pillared Clay

0.06 +0.58 -
[Fe(o“)2.9 9.75 [A‘3.1iF°0.42”90.4a](5‘7.88A1o.12)°20‘

(on)4

 

137

0

Figure V-7. At 350°C, its basal spacing d001 is 22.6 A

0
(Ad = l3.2 A), which corresponds to five to six oxide

001
layers between the two adjacent silicate sheets.

The DSC analysis of this Fe-PILC (Figure V-9) has an
endothermic peak at ll0°C, which corresponds to the dehy-
dration of the interlayer. Comparing the DSC diagrams
for Al-, Cr-, and Fe-PILC, we find that the dehydration
temperature is lower for Fe than for the other two (Al-PILC
at 135°C and Cr-PILC at l55°C).

The pore opening of Fe-PILC was investigated by the
adsorption of probe molecules having different kinetic

diameters. The adsorption results are summarized in

Table V-S.

 

Table V-5. Sorpiton of Probe Molecules on Fe-PILC at 25°01

 

 

Probe Molecules Pressure Sigfigigr Agggaged
Benzene 70 mmHg 5.8 A 2.64 mmole/g
Neopentane 730 6.2 1.68
l,3,5-triethylbenzene 0.18 9.2 l.23
perflorotributylamine 0.22 10.2 0.96

 

1 The sample was freeze-dried, then dehydrated at 350°C in
vacuum for 3 hours.

 

As shown in Table V-S, the pore opening of Fe-PILC is
big enough for the adsorption of the largest probe mole-

cule, perflorotributylamine, whose kinetic diameter is

138

Figure V-7. XRD of Iron(III) Pillared Clays Which is
Made from 0.2 M FeCl3 Solution with 0H/Fe =
2.00 and Aged at 25°C for 24 h, at a) 25°C;
b) 350°C and; c) 550°C.

139

fi.

 

 

 

 

140

 

 

 

 

2_________-‘-‘~‘
F
i.
1 4‘1 ' 1 l l I l l l 1 1
I00 200 300 400 500

°C

Figure V-8. DSC Curve of Iron(III) Pillared Clay Which
is Made from 0.2 fl FeCl Solution with 0H/
Fe = 2.00 and Aged At 2 °C for 24 h.

 

l4l

l0.2 A. The adsorption capacity of this Fe-PILC is higher
than that of the other two pillared clay systems (Al and
Cr).

C-5. Anion Effect

In the previous sections (V-C-l, 2 and 3), the effect
of hydrolytic conditions on the spacings of pillared clays
has been discussed. The polycations used in the majority
of the studies reported here were produced from ferric

14 and Murphy et al.21"24

chloride solutions. Spiro 33431.
have concluded that the sedimentation coefficient and the
rate of formation of ferric hydroxide precipitations are
dependent on the anion of the ferric salts. These points
reveal the influence of the anion on the molecular weight of
the polycations. In order to investigate the anion effect

on the hydroxy-iron polymers and the subsequent effect on

the properties of the pillared products, we have used three
additional ferric salts (nitrate, perchlorate and sulfate)

to prepare hydroxy iron pillared clays.

The pr0perties of pillared clays prepared from these
four Fe(III) solutions are summarized in Table V-6. As we can
see from Tables V-6 and V-2, the basal spacings of pillared
clays made from the chloride, nitrate and perchlorate
salt solutions are in the same range. However, the pro-
ducts made from the sulfate salt behaves differently. The

low spacings suggest that the pillaring species formed

by hydrolysis ferric sulfate solution are smaller than

 

142

those formed from the other salts. 1
In hydrolysis chemistry of metal ions, the competition

between hydroxide ions and other anions for positions in

the first coordination sphere of the metal cation has been

observedzs. This phenomenon and the realted replacement

of coordinated groups such as water or anions by another

25

anion has been referred to as "anion penetration" Rela-

tive to the other three anions, the sulfate anion is strongly

 

bound to the hydrolyzed species (hydroxy iron oligomers

or polymers) causing the decrease of surface charge. The

 

greater penetration leads to greater association of the ions
and gives rise to high molecular weight hydroxy species
which end up as neutral colloids and precipitates. There-
fore, the concentration of polycations in the sulfate solu-
tion is very small and probably only the small oligomers
remain to give low spacing products.

The anion penetration effect of sulfate also occurs
in both aluminum and chromium sulfate solutions. The
aluminum solution with OH/Al = 2.00 and chromium solution
with 0H/Cr = 2.00 and aged at 95°C for l4 hours gave
exchanged montmorillonite with basal spacing only 15.3 A
(air-dry). Apparently, there are only small hydroxy
oligomers in these two hydrolyzed solutions.

As we have seen in Figure V-S, an increase in aging
time of FeCl3 solution (OH/Fe = 2.00) allows for further
hydrolysis reaction and larger polymers. This process

also occurs in nitrate and perchlorate solutions

143

(OH/Fe = 2.00 for both solutions) as shown in Table V-6. . A
When the aging time of a iron nitrate solution is increased

from 3 hours to 7 days,_the pH is decreased by 0.2 unit,

the size of the polycations is increased, and the basal

spacing of the pillared clay product is increased from

27.2 A to 29.45 A. At the same time, the iron content of

the pillared clay is increased from 6.83 iron atoms to

8.83 iron atoms per unit cell and the surface area is

._ P.» £6.11” —

decreased dramatically.

mi '14'

These observations can be explained in terms of the

 

FAD-5-.

polymerization processes during aging. The electron micro-
scopy studies of partially neutralized ferric nitrate
solutions indicated that the initial polycations are dis-
crete spheres 1.5 nm - 3.0 nm2]. 0n aging, the spherical

polycations coalesce to form short rods, many of which

subsequently form rafts. >To minimize the repulsion, the.
polycations undergo side-by-side association. Therefore,
rafts are built up. This aggregation will decrease the
charge on the polycations and result in a decrease in the
pH of the aged solutions (deprotonation)26’27.

The decrease in the charge on the polycations (or the
charge density per iron) upon aging, leads to an increase
in the amount of iron needed to balance the negative charge
on silicate surface. This deduction is consistent with the
increase in iron content of the pillared clays with increas-

ing aging time of the iron solution. InCidently, the

 

 

144

 

» F? itisfl
m.~. - - mN.~ meson cm ImAeomVNwa
Am._,v Am.__v Am.~PV Ae.m_v u
m.m~ a.m~ m.¢~ ~.m~ - oe.owm@mmfizovoag ¢~._ msau N
Am.F_v Ao.m_v Am.m_v
- 4m.¢~ we.w m.m~ mam Np.mmmwnm zovmau mm._ maze; m mmaeo_uvma
Ao.__v Am._PV Ac.mcv A¢.mpv
o.o~ 9.5N m.m~ m¢.m~ m.o- om.mflmwnmu=ovmau mm._ axon a
Ame._.v Ao.m_v Ae.mpv .
- em.¢~ ¢~.mm ~.~N _mm mm.cnmmnm zovmau Na._ maze; m unAmozvaa
A_.N_V .A~.mpv
C.NN m.- - ~.m~ «em Wmum+mwmumAzovmau -._ mesa; em
. u - e.mm cum mmmmmmmmnmAzovmuH om._ acac; m.P \ m.m_omu
comm comm oOPP 0mm mAm\Nsv. |__au “we: .2:
m Amoocv —oou mcwumam .mmmm mmc< muwccsm mom smp__m In ms_h m:_m< mapmm u_cgmm

 

.mp_am A__HV=OEH
ucmcmwcpo we mwmxpogv»: as“ an cmgmamga mxmpo umcoppwa Ammmvcogm yo mmpucmgoca

.on> mpnmh

 

145

—IF...:..

.cosxw—ose om

.ams\o—oes on

»m_0\ma
»a_u\mu

n.

.m

.m

.Nz co mcpzo_c Laue: ; m Lot ooomm an catasvpoa

.oo.~ wu\:o m: p.o mw +mmu mo .ucou
mm\:o m: —.o my +mmu mo .ucou
mm\:o m2 ~.o m_ +m

czosm my gmxmpcmucw on» =_ ngPFq on» Apco

oozmz mp mmmn moo.~

oumwz m_ amen moo.~ am we .ucoo

1“ ul 1” ml

MI

 

um==_u=ou .c-> a_nah

 

146

elemental analysis of these two iron pillared clays shows . L
that all sodium ions in the interlayer had been replaced.
Based on related considerations, we may expect that
raft-shape polycations will cover to more interlayer sur-
face area, and therby decrease the BET surface area.
The differences in the basal spacings of the iron
pillared clays made from ferric nitrate solutions in
Tables V-2 and V-6 might be the result of differences in

the ferric concentrations and the bases used in the hydro-

5. .-qfixtnma u ‘-

lysis reaction. Upon calcination at 550°C, the iron 1

 

pillared clay made from 0.l M Fe(N03)3 solution with

OH/Fe = 2.00 and aged at room temperature for 7 days re-
tains a basal spacing d001 = 26.0 A. The XRD of this

sample, as well as those of the same sample heated at various
temperatures, are shOwn in Figure V-9. After being calcined
at 550°C, the iron oxide clusters in the interlayer are not
very stable. The interlayer structure of the product

heated to 550°C may rearrange at 25°C because the XRD
patterns is changed (Figure V-l0(e)) after 7 months of

aging at 25°C.

V-D. Conclusions

These results demonstrate that the smectite clays can
be expanded by hydroxy iron polycations to form new
pillared clays with basal spacings as high as 30 A. The
pillared clays do not collapse when heated to a temperature

of 550°C. The nitrate, chloride and perchlorate salts of

147

 

 

29.458
(00:)

I343

002)
1"
1;.
l 1'
E1

 

 

 

L A .
so 26 22 ' I8 l4 IO 6 2
26

Figure V-9. XRD of Iron(III) Pillared Clay Which is Made
From 0.1 fl Fe(N03)3 Solution with OH/Fe = 2.00
and Aged at 25°C for 7 days, at Elevated
Temperature a) 25°C; b) 110°C; c) 350°C;

d) 550°C and; e) after expose the (d) in air
for 7 months.

148

Fe3+

upon hydrolysis form suitable polycations for the K
pillaring reaction but the sulfate Salt, wherein sulfate
binds strongly to the metal cations is not suitable.
Preliminary studies carried out by Edward Keller in
this laboratory indicate that the iron pillared clays are
active catalysts in Fischer-Tropsch synthesis. These
pillared clays may contain spherical hydroxy iron(III)
polycations, rods or rafts and may possess different pore

structure, and different shape selectivity.

The hydrolysis conditions needed to produce polymers

 

of one metal ion may not be suitable for another metal ion.
Aging at high temperature (m 100°C) tends to favor forma-
tion of larger hydroxy chromium polycations and the forma-
tion of chromia pillared clay with high spacings. However,
heating is not preferred in the iron(III) system.

The findings in this thesis suggest that new expanded
clay catalysts pillared by other metal oxides can be
formed, and the compounds reported here may only be part

of a larger family of pillared clays.

10.

11.

12.

13.

14.

149

REFERENCES

Bjerrum, J.; Schwarzenbach, G.; Sillen, L.G.
Stability Constants of Metal-ion Complexes Spec. Publ.
Soc. Lond. No. 17 (1964).

Sylva, R.M. Rev. Pure Appl. Chem. 22, 115-32 (1972).

'IBWI

 

 

Hedstrom, B.0.A. Arkiv. Kemi. 6, 1 (1953).

Biedermann, unpublished reSults cited by Schlyter, K.
Trans. Roy. Inst. Tech., Stockhom. No. 196 (1962).

Zvyagintsev, D.L.; Lyakhmanov, S.B., Russian J. Inorg.
Chem. 1;, 643 (1968).

Sutton, J. Nature 169, 71 (1952).
Jorgensen, C.K. Acta. Chem. Scand. g, 1502 (1954).

Olson, A.R.; SimOnson, T.R. J. Chem. Phys. 11, 1322
(1949).

Milburn, R.M.; Vosburgh, W.C. J. Amer. Chem. Soc. 11,
1352 (1955).

Jones, M.M.; Jones, E.A.; Harman, D.E.; Semmes, R.T.
J. Amer. Chem. Soc. 83, 2038 (1961).

Mulay, L.N.; Selwood, P.W. J. Amer. Chem. Soc. 21,
2693 (1955).

Morrison, T.I.; Shenoy, C.K.; Nielsen, L. Inorg. Chem.
20, 3565 (1981).

Jander, G.; Winkel, A. Z. Anorg. Allgem. Chem.'193,
1 (1930).

Spiro, T.G.; Allerton, S.E.; Renner, J.; Terzis, A.;
8ils,)R.; Saltman, P. J. Amer. Chem. Soc. 8Q, 2721
1966’.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.
27.

ii» 527

150

Brady, G.W.; Kurkjian, C.R.; Lyden, E.F.X.; Robin,
M.B.; Saltman, P.; Spiro, T.; Terzis, A. Biochem. l,
2185 (1968).

Gray, H.B. ACS, Adv. in Chem. Ser. 100, 365 (1971).

Barnhisel, R. I. 'Chlorides and Hydroxy Inter1ayered
Vermiculite and Smectite' in "Minerals in Soil
Environments" Dixon, J. B. ; Weed, S. B. Ed. , Soil Science
Society of America, Madison, WI 1977, Chapter 10,

pp. 345.

; de Bruyn, P.L. J. Colloid. Interface Sci.
1976).

Dousma,

J.

(
Dousma, J. ; de Bruyn, P.L° J. Coiloid. Interface Sci.
66,154(1978).

Dousma, J. ; de Bruyn, P.L. J. Colloid. Interface Sci.

1;, 314 (1979).

Murphy, P.J.; Posner, A.M. irk, J.P. J. Colloid.
Interface Sci. 66, 270 (197 .

Murphy, P.J.; Posner, A. M. uirk, J.P. J. Colloid.

;Q
6
30
Interface Sci. 66, 284 (197 6).
Q
6)
Q
)

Murphy, P.J.; Posner, A.M.; uirk, J.P. J. Colloid.

Interface Sci. 66, 298 (197

Murphy, P.J.; Posner, A.M.; uirk, J.P. J. Colloid.
Interface Sci. 66, 312 (1976 .

Graham, S.P.; Thomas, A.W. J. Amer. Chem. Soc. 62,
816 (1947).

Hsu, P.H.; Ragone, S.E. J. Soil Sci. 66, 17 (1972).

Murphy, P.J.;Posner. A.M.; Quirk, J.F. Aust. J. 5011
Res. 1;. 139 (1975).

r4
Iq.;§1_.'1:vvg.,1;v,..-