MSU

LIBRARIES
an.

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES wiII
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

 

 

 

 

GALLERY AND SURFACE PROPERTIES OF
LAYERED DOUBLE HYDROXIDES

BY

Kevin Julius Martin

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1986

ABSTRACT

GALLERY AND SURFACE PROPERTIES OF
LAYERED DOUBLE HYDROXIDES

BY

Kevin Julius Martin

Layered double hydroxides (LDHs) are a class of
compounds of general formula [M(II)1_XM(IIIIXKH02] [x/zY'
°nH20], where x ranges from 0.16-0.33. The gallery anions
are exchangeable, making LDHs negative analogs to clay
minerals. The structural properties of LDHs are well docu-
mented. Very little has been done to characterize the
properties of the gallery anions or the surface properties.
Most LDHs that are synthesized are dehydroxylated to form
oxide catalysts. No reactions have been reported for LDHs
in their pristine form. This dissertation will describe the
investigations into the properties of the gallery anions of
LDHS.

LDHs containing a paramagnetic anion asaispin probe
were synthesized. Because the ~SO3 groups of the spin
probe K2(SO3)2NO bind to surface hydroxyl groups,ru>spin
probe was intercalated even when it*was added to the syn-
thetic mixture. Therefore, no information about the mobil-
ity of gallery anions was obtained from EPR studies.

The availability of gallery anions of [ZnZCr(OH)6]
[Y‘nHZO], where Y’=I, C1, F, for nucleophilic displacement
reactions was investigated. The reaction rate of iodide
substitution of alkyl bromides under solution-solid

conditions was independent of the alkyl chain length. The

Kevin Jul ius Martin

reaction rate depended on the surface properties of the
LDH. The rates also depended on the nature of the solvent
in which the reaction occurred. From adsorption studies it
was deduced that the more tightly a solvent was bound to
the solid, the slower the reaction.

Chloride substitution reactions did not occur under
solution-solid conditions, but did occur under gas-solid
conditions. Fluoride substitution did not occur under
either conditions. A detailed mechanism is proposed in
which the bromo group of the reactant molecule must insert
into the gallery of the reactant LDH. The mechanism is
consistent with all observations regarding the kinetics

behavior and adsorptive properties of the Zn-Cr LDH.

 

TO JOHN CLIFFORD MARTIN

1944 - 1974

H

ACKNOWLEDGMENTS

I would like to thank Dr.1LJ. Pinnavaia for patience
and guidance down the long road of graduate school. I look
forward to future interactions.

I also wish to thank Dr. C.H. Brubaker for careful
editing of this manuscript. I hope we got all the typos
and grammar mistakes.

I would like to express my gratitude to all past and
present group members with whom I have come in contact.
Your support and questions have helped a lot.

I would like to thank my mother, siblings and nephews
for emotional and financial support. You helped me keep my
sense of humor through it all.

Finally, thanks to my mother-in-law Norma Donovan for
use of the word processor, and to Susan and Elizabeth,
words alone cannot express what I owe to you in the com-
pletion of this work. Maybe now we can live like normal

people.

111

TABLE OF CONTENTS
a e
LIST OF TABLES 0.0.0.0000...O...OOOOOOOOOOOOOOOOOOOOOVI
LIST OF FIGURES OIOOOOOOOOOOOOO00.0.0000...OOOOOOOOOOViii
CHAPTER 1 0 INTRODUCTION 0 I O O O O O O O O O O O O O O O O O O O O I C O I O O O 1

A. Synthesis and properties of layered double
hYdIOXides 00.0.0.0...0.00.00.00.000000000001

B. Supported reagents .........................23

C. Objectives of current research .............34
CHAPTER 2. EXPERIMENTAL SECTION .....................35
A. Synthesis of materials .....................35

1. [ZnZCr(OH)6][C1'nHZO] ....................35

2. [Ca2A1(OH)6][C1°nHZO] ....................36

3. [LiA12(0H)6][0.5504-nH201 ................36

3a. Spin-labelled [LiA12(OH)6][0.5804'nH20] .37

4. Ion exchange reactions ...................38

B. Displacement reactions over [ZnZCr(OH)6]
[Y.nHZO] 0..0....0...0...0.0.00.000000000000038

1. Reactions under triphase catalytic
conditions ...............................38

2. Reactions under stoichiometric
conditions OOOOOOOOOOOOOOOOOOOOOO0.0.0....39

3. Gas-solid reactions ......................40
C. Physical methods ............................41
1. X-ray diffraction ........................4l
2. EPR studies ..............................42

3. IR StUdies OOOOOIOOOOOOOOOOOOOOOO0.0.0....42
1v

CHAPTER 3.

 

page

Gas chromatography .......................42
surface areas 0 O O O O O O O O O O O 0 O O O O O O O O O O O O O O O 43
Adsorption studies 0 O O O O O O O O O O O O O O O O O O O O O .43

RESULTS AND DISCUSSION 0.0.0.000000000000046

A. Properties Of LDHS 0.00.00.00.00.0.00.00.00.046

1.

2.

Exchange and swelling properties .........46

EPR StUdies 0......COCOOOOOOOOOOOOOOOOO00.55

B. Nucleophilic displacement reactions .........66

1.

Reactions under triphase catalytic
conditions OOOOOOOOOOOOOOOOOCOOO0.0.00.0..66

Stoichiometric biphase reactions in
t01uene 0.00....0.00.0.0....0.0.0.0000000070

Adsorption properties of Zn-Cr LDHs ......86

Reaction and sdsorption studies in other
SCI-vents .0.0.00.0.0...0.0.0.0000000000000102

Thermodynamics of reaction ...............199

Proposed detailed mechanism ..............112

C. conCIUSions .0.0...0...0...0.0.0.0000000000000124

LIST OF REFERENCES OOOOOOOOOOOOO00.00.00.000000000000127

 

LIST OF TABLES

Table Page
1 Some Natural LDH Minerals...............3
2 Phases Formed Upon Aging Ni-Al LDHs.....12
3 Variation of Basal Spacing In

[Ca2A1(OH)6 ][Y' 2820] and [an Cr(OH)5 ]
[Y nHZO] For Various Y Species.u.6u.u.l3

4 Dependence of Lattice Parameters on
Octahedral Layer Composition for Some
LDH SpeCieSOOOOOOOOOOOOOOOOIOOCOOOOOOOOOl4

S Dehydration Temperaturesa for Some
Mg-Al and Ni-Al LDHSoooooooooooooooooooo18

6 Dehydroxylation Temperaturesa'b For
some LDH carbonates.C...................19

7 Swelling of [Zn Cr(OH)6][C12H250803]
in Alcohols an Amines..................2l
8 Reactions Utilizing Polymer-Bound
ucleophiles of the Form
r-CHNMeX- .00...0.0.00.0000000000026
2 3
9 EPR Correlation Times for PADS-doped

[Ii-A1 LDH MateriaISOOOOOOOOOOOOOOOOOOOOO61

10 Conversion of Alkyl Halides by the
Reaction RX + Y'-9RY'+ X' Under
Triphase Consitions.....................67

11 Results of Blank Reactions for Halide
EXChangeoooooooooooooooooo0000000000000069

12 Conversions of Nucleophilic Displacement
Reactions of C4H98r Under Modified
conditiOUSOO0.0.00.0...0.0.0.0000000000071

l3 Calculated Pseudo First Order Rate
Constants, for Iodide Substi-
tution of AlEyI Bromides over A. D.
[ZnZCr(OH)6][I nHZO]....................72

v1

 

Table Page
14 Pseudo First Order Rate Constants, kobs'

for Iodide Substitution of Alkyl

Bromides Over O.D. [ZnZCr(OH)6]

[I' OHZO]...............................74

 

15 Activation Parameters for Some Alk 1
Halide Iodide Substitutions at 90 C....76

16 Conversions of Gas-solid Iodide Substi-
tutions of l-Bromobutane Over .
[znzcr(OH)6][I]oooooooooooooooooooooooo.89

l7 Conversions of Gas-Solid Chloride
Substitutions of l-Bromobutane Over
[zn2CI(OH)6] [C1]........................8g

18 Maximum Adsorption Capacities and
Desorption Experiment Results for
Various Adsorbates on O.D.
[ZnZCr(OH)6][I].........................106

l9 Pseudo First Order Rate Constants for
' Iodide Substitution of l-Bromobutane
with 0.0. Zn/Cr/I LDH at 90°C in
Various Solvents........................107

20 Activation Parameters for Iodide
Substitution of l-Bromobutane at 90 °C
by O.D. [ZnZCr(OH)6][I] in various
Solvents................................108

21 Values of Der - DC and 24.2 dEGI for
eac

Various Alky Hal1 e Exchange tions
RX + LDH(Y)#RY + LDH‘X)000000000000000111

v11

 

LIST OF FIGURES

Figure Page
1 Structure of layered double hydroxides
(water molecules omitted for clarity)u.2

2 Diagram of the coordination geometry
around the calcium centers in
[Ca Al(OH)6]HN12H 0] showing the
dis ortion6 caused y coordination to
gallery water. ..........................6

3 The structure of montmorillonite, a
clay mineral of the smectite group......7

4 a.) The effect of aging on the product
distribution of coprecipitated Ni-Al LDH
for a solution value of x=0.25. b.)The
effect of the initial product distri-
bution on the shape of x-ray powder
diffraction peaks. Adapted from
reference S8............................l7

5 Structure of an alkylammonium-exchanged
smectite swollen with a linear alcohol.
From reference 2S.......................22

6 A conceptual diagram of triphase
cataIYSiSOOOOOOOOOOO0.0.0.0000...0.0.00028

7 A diagram of a smectite clay with
gallery ions as a result of a.)
homoionic intercalation of cations,
and b.) intersalation of two equiva-
lents of cation and one of anion.
Structure b acts as an anion exchangeru32

8 Powder x-ray diffractograms of (a).
nCr(OH)6 ][C1' 2H O];~(b). [LiAl (OH)6
SO4°nH2601; and2 (c). [Ca2A1(OI-l%6]
[0. SCO43'ZHZO]...........................47

9 Powder x-ray diffractograms of

[Ca Al(OH) ][G.SCO3'2H O] (a). pressed
pow er; b) oriente film.u.u.n.u.n49

V111

Figure Page
10 Powder x-ray diffractograms of the

dodecylsulfate-exchanged forms of (a).
Zn-Cr LDH; (b).Ca-Al LDH; and (c).
Li'Al LDHOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.S]-

ll Powder x-ray diffractograms of
n-octanol-swollen DDS-exchanged forms
of (a). Ca-Al LDH; and (b). Li-Al LDH...52

12 Infrared spectra from 4000 cm'1 to
woo curl of [ZnZCr(OH)6][Cl'nH20]
(a). pristine material; and (b). exposed
to 0.1 M NaDDS solution.................S4

l3 EPR spectra of K PADS (a). 10"5 M
solution in DMS ; and (b). frozen in

DZOOOOOOOOCOOOOOOOOOOOO0.00.00.00.00000056

14 EPR spectrum of [Zn Cr(OH)6][C1°nH201.
Frequency = 9.14 Hz....................58
15 EPR spectra of PADS-doped [LiAlg(OH)6]
)

[Cl'nH O] (a). damp material; (

after 2 hr in vacuum dessicator; and

(c). after 12 hr in vacuum dessicator.
Frequency = 9.14 GHz....................S9

16 EPR spectra of an Li-Al LDH synthesized
in the presence of NaDDS and K PADS
(a). damp material; (b). after 4 hr in
vacuum dessicator; and (c). after 12 hr
in vacuum dessicator. Frequency =
9.14 GHz.62

1? Powder x—ray diffractograms of unaged
Li-Al LDH product (a). filter-washed;
(b). dialyzed until free of
electrolyte.............................64

l8 Powder x-ray diffractogram of product
formed by reaction of LiCl solution
and gibbSiteOOOOOOOOOOOOOOOOOOOOOOOOO...65

19 First order kinetic plots for iodide
substitution of alkyl bromides over
[ZnZCr(OH)6] [I’2.3H20]: ll-bromo-
butane; Al-bromohexane; and Ol-bromo-
3-methy1butane..........................73

20 IR spectra from 4000 cm’1 to 1000 cm"1

of (a). [ZnZCr(OH)6][I°2.3HZO]; (b).
[ZDZCI(OH)6] [1.6H20]00000000000000.000075

1x

Figure Page
21 First order plots for iodide substi-

tution of l-bromobutane over OJL
Zn/Cr/I LDI-l: (a). 122 °C; (b). 1001 °C;
(c). 90 °C; ((1). 80 °C; and (e). 5e °c..77

22 Arrhenius activation energy plot,
1n kobs vs. l/T, for iodide substi-
tution of l-bromopentane over OJL
Zn/Cr/I LDH.............................78

23 First order kinetic plot for the
extended iodide substitution of 1-bromo-
butane over (LD. Zn/Cr/I LDHuu.n.u.u82

24 First order kinetic plot for iodide
substitution of 1-bromobutane by OJL
Zn/Cr/I LDH at RBr/LDH ratios of I4:1;
03:1; 02:1; A1:1, plotted as
functions of (a). -1n fI-; (b) -1n fRBr'84

25 Isotherms for adsorption of toluene
onto: (a). A.D. Zn/Cr/I LDH at .0 °C;
022 °C; (b). 0.0. Zn/Cr/I LDH at
A0 °C; 322 °C88

26 Isotherms for adsorption of 1-iodo-
butane'onto: (a). A.D. Zn/Cr/I LDH at
on °C; 022 °C; (b). o.o. Zn/Cr/I LDH
at no °C; I 22 °C......................89

27 Isotherms for adsorption of 1-bromo-
butane onto: (a). A.D. Zn/Cr/Br LDH at
.0 °C; 022 0C; (b). O.D. Zn/Cr/BrrLDH
at A0 °C; I 22 °c......................90

28 An adsorption isotherm showing the
location of Point B, the attainment of
monolayer coverage......................92

29 Heat curves, qS vs. V/V , for adsorp-
tion of toluene onto [ZnZCr(OH)6]
[1.nH20] ‘AOD0.0.D...OOOOOOOOOOOOO00.093

30 Heat curves, q t vs. V/Vm for adsorp-
tion of l-iodogutane onto [ZnZCr(OH)6]
[I.nHZO] ‘AoDo .O.D.-00000000000000.0094

31 Heat curves, qst vs. V/Vm, for adsorp-
tion of l-bromobutane onto [ZnZCr(OH)6]
[1.nH201 ‘AOD0.00DOO0.0.00.0000000000095

 

Figure Page
32 Plot of the enthalpy of liquid adsorp-

tion,AH ads' vs. V/V for toluene
adsorbed onto [ZnZCrTOH)6][I'nH20] A.A.D.

.O.D....O0.00000000000000000000000000.96

33 Plots of AHlads vs. V/Vm for adsorption
of l-iodobutane onto [ZnZCr(OH)6]
[I'nHZO] QA.D.OO.D.; and 1-bromo-
butane onto [ZnZCr(OH)6][I'nH20] I A.D.

‘O.D....OOOOOOOOOOOOOOOOIOOOO00.0.000097

34 Adsorption of toluene onto O.D.
[ZnZCr(OH) ][I‘nH O] as a function of
pressure a? A0 0 I22 °C..............99

35 Plots of AHla 3 vs. V/Vm for O octane
and Acarbon getrachloride adsorption
onto O.D. [znzcr (OH)6][I'DH2010000000000163

36 Plots of AHlads vs. V/Vm for adsorption
of.tchlorobenzene and . toluene onto
O.D. [ancr(OH)6] [I'DHZO]ooooooooooooooolg4

37 Possible bonding interactions of
hydroxyl group orbitals with an
aromatic system. (a).e2 +6‘+ -
9 OH
no net overlap; (b). e2 + H2p -
p-orbital too high in eaergy;
(c). 0.5 e2 +5+OH - probable
interaCtio 0.000000000IOOOOOIOO0.0.0.000114

x1

Chapter 1. Introduction

A. Synthesis and properties of layered double hydroxides.
There are a large number of compounds that can be
classified as layered double hydroxides1 (LDH). This class
of materials, also known as hydrotalcite-like compounds
(HT)2 and Feitknecht compounds3, has both synthetic and
naturally occurring members. Most of these species, of
general formula [M(II)1_XM(III)X(OH)2][x/zYz"nH20]1, con—
sist of brucite-like octahedral sheets with trivalent metal
cations periodically substituted for the divalent metal
cations. This substitution imparts a net positive charge
on the sheets which is compensated by anions intercalated
between the stacked hydroxide layers together with n water
molecules (Figure 1)l. These compounds will be the subject
of this dissertation. Related compounds exist with only

3 anala-

divalent cations in hydroxide-carbonate structures
gous to those of malachite, Cu2(OH)2CO3; hydrozincates,
Zn5(OH)6(CO3)2; and azuriteé, Cu3(OH)2(CO3)2. These mixed-
metal hydroxides with no trivalent cations will not be
considered in this work. Also, natural minerals with more
than one octahedral layer per anionic layer, such as coa-
lingite7 and carrboyditea, will not be considered.

Only a small percentage of the known LDH species occur

in nature, the most common of which is hydrotalcite,

m. w. m.
H\I.H H \I..H
owom. wMo

M A M

 

Structure of layered double hydroxides (water

molecules omitted for clarity).

Figure l.

3

 

 

Table 1. Some Natural LDH Minerals.a
Name Idealized Formula
hydrotalcite, [Mg6A12(OH)16][CO3'4HZO]
mannasseite
pyroaurite, [M96F92(0H)16][C03'4H201
sjogrenite
stichtite, [Mg6Cr2(OH)16] [C03'4H20]
barbertonite
reevesite [Ni6Fe2 (OH) 6][CO3' 4H20]
eardleyite [N16 Al (OH) [CO3 4H
meixnerite [M96 AI «n16][(OH)2%H20]
takovite [Ni6 Al OH)1[(OH)2'4H20]
hydrocalumite [Ca62Af(OH)6]6 [OH 2H 25]

 

aFrom reference 9.

 

[Mg6A12(OH)16] [CO3°4H2019. The other formulae in Table l

with two entries in the name column exist in two

polymorphic formsg. The most extensively studied natural
LDHs from a structural standpoint are the pyroaurite/sjog-
renite polymorphs, [Mg6Fe2(OH)16] [CO3'4H20]. Early powder
diffraction work10 showed that sjogrenite is hexagonal with
a=3.13 A and c=15.65 A. Pyroaurite is rhombohedral with
a=3.l3 A and c=23.47 A. Single crystal studies of sjogren-
ite and’pyroaurite confirmed these data11‘13. The hydroxyl
groups of neighboring octahedral layers stack on top of
each other. The rhombohedral pyroaurite has three octahe-
dral layers per unit cell stacked in the sequence BC-CA-
A812'13. The AB-BA-AB stacking sequence of sjogrenite
leads to a hexagonal unit cell with two octahedral layers
per unit ce11.11'12 In the samples examined, there was no
the Mg(II) and

order to the cation site distribution, i.e.,

Fe(III) centers were statistically distributed throughout

 

the octahedral siteslz.

Much of the early attention focused upon synthetic
LDHs resulted because a calcium-aluminum compound formed as
an intermediate in the setting of Portland cement was in
fact a Ca-Al LDH14. A report from 192915 showed that two
metastable compounds of "calcium sulphoaluminate" were
formed by interaction of calcium aluminate and calcium
sulfate. The formula of one of them, given as 3CaO°A12 3'
CaSO4'12H20, can be rewritten as [Ca2A1(OH)6][1/2304'3HZO].
No structural data were given,lnn:the formula falls into
the category of a LDH. Later studies16 of the Ca-Al system
confirmed that it was an LDH.

In the following years, many reports appeared on the
synthesis of these basic double salts, as they were called
before the structure was known17. From x-ray powder dif-
fraction data came the knowledge that the 001 parameter of
the materials varied as the size of the anion used in the
synthesis or exchanged into the structurela.’ A double
layer structure was proposed19 in which the divalent metal
formed a hydroxide layer, like brucite, with the trivalent
metal and anions intercalated between these sheets. One
group20 has used infrared spectroscopy results to support
this structural argument. Specifically, the C032' vibra-
tions were consistent with Al—OCOz-Al bridges in the gal-
leries between brucite-like sheets. This structure is
today not generally accepted today because of the single

crystal work reported more recently. Synthetic analogues

of the natural mineral hydrocalumite, [Ca2A1(OH)6]
HNP2H20]9, have been investigated by single crystal x-ray
diffraction21‘24. The results of these investigations are
in essential agreement with the reported structures of
pyroaurite and sjogrenite in that the metal cations are
contained in octahedral hydroxide layers with anions inter-

21422. Because of the large size of

calated between them
Ca(II) compared to Al(III), the metal cations were ster-
ically limited to an‘ordered arrangement23. This result
was different from the random distribution of Mg(II) and
Fe(III)iJIpyroaurite and sjogrenite».12 The<difference in
size and charge allows the more polar Al(III) centers to
distort the hydroxyl groups away from the Ca(II) centers.
Because of its large size and the free space created by the
distortion, each calcium becomes 7-coordinate by bonding to

23'24 (Figure 2).

a water molecule in the interlayer region
The results from the single-crystal studies have been used
to infer structural details of the many synthetic LDHs of
which crystals large enough for single crystal studies
cannot be grown. .

The LDHs can be compared to the clay mineralszs. Both
have a layered structure consisting of metal cations in an
oxide (hydroxide) framework, as can be seen in Figures 1
and 3, with a net charge distributed through the framework
due to substitutions of metal cations within the layers.

Both clay minerals and LDHs have ions intercalated between

these layers to balance the charge on the framework. Clay

Figure 2.

 

C: A O

 

1.00 A 3 on

. 0.57 A Ca

 

-1.00 i 3 OH

r

Diagram of the-coordination geometry around the
calcium centers in [Ca2A1(OH)6][OH'2H20] showing
the distortion caused by coordination to gallery
water. Adapted from reference 1.

  

2' roll Al”

Figure 3. The layered structure of montmorillonite, a clay
mineral in the smectite group.

 

8

minerals have a net negative charge on the oxide-hydroxide
layers while LDH have a positive charge. The interlayer
cations of many clay minerals are exchangeable,25 as are
the anions of most LDHs.l Because of the possibility of
substitutions in the octahedral and tetrahedral sheets of
the clays, there is a greater variation in the properties
of the clay minerals than is possible through substitutions
in LDH compounds. Also, the swelling properties of smec-
tite clays in various solvents cannot be duplicated in the
LDH due to differences in solvation of anions vs.
cationszs. The swelling of certain LDHs will be discussed
in a later section.

Three general methods of synthesis have been devised.
They are: 1) solid-solid reactions; 2) solid-solution reac-
tions; and 3) solution-solution reactions or
coprecipitation. There is only one example of the first
type of synthesisza. In this synthesis MgO and A1203 were
suspended in water and heated together at 80 °C for one
week to give LDH products. This technique was especially
suited for preparing carbonate-free materialsZG.
The second method was used in the preparation of Ca-Al LDHs
for crystallographic studie521'22. In these syntheses
solid Al(OH)3 and Ca(OH)2 were suspended in an aqueous
solution of the calcium salt of the desired gallery anion.
After 200 hours at 150 °C, or 3 months at room temperature,
crystals of [Ca2A1(OH)6][Y°nHZO] large enough for single

crystal studies were produced24. As another example of

 

 

9

this synthetic approach, in studies of aluminum-containing
soils and clay minerals,27 alkaline earth cations combined
with various forms of hydrated alumina solids. In the case
of Mg(II), these materials could be aged to give Mg-Al
LDHs. The speed and extent of the reactions were dependent
upon the crystallinity of the aluminum-containing material,
with the "fresher", more amorphous materials reacting
faster27'28. Gels of hydrated iron oxides also adsorb
alkaline earth cations from solution29. Reactions of this
sort have been postulated as possible routes in nature to
LDH minerals and chloritic interlayers in certain clay
minera1327'29.

This concept of reaction of a hydroxide gel with
another cation has been extended to a general synthetic
technique called induced hydrolysis30. Aldivalent metal
salt solution is added to a freshly-prepared M(III)(OH)3
gel at a controlled pH above that at which the M(II)(OH)2
precipitates. The fresh gel seems to "induce" the divalent
cation to condense into the gel structure to give the LDH
product.30

Another example of the solid-solution approach is the
synthesis of [ZnZCr(OH)6][Y'nH20131'32. This compound was
first noted in a study of the effects of chromium salt
solutions upon ZnO oxidation catalysts33. Later workers

identified the lavender products as LDHs.31

In the syn-
thesis of the Zn—Cr LDHs, ZnO was exposed to a l M solution

of a Cr(III) salt until all traces of ZnO disappear from

10

the x-ray diffractogram of the product. It was reported
that this same reaction also worked for Cr(III) and CuO.31
The great majority of LDHs have been prepared by
coprecipitation reactions. Coprecipitation is the tech-
nique of simultaneously precipitating at least two species
within a solution.34 In the original "calcium sulpho-
aluminate" synthesis,15 the strongly alkaline calcium
aluminate solution provided the required hydroxide ion
concentration. For most other LDHs, an alkaline solution
is usually added to a solution of the desired metal salts
to precipitate the double hydroxide salt. This technique
was exploited by Feitknecht in extensive studies of the Ca-
A135’37 and Mg-A116'17'38'40 LDH systems. Coprecipitation
has been extended to hundreds of LDHs reported in a series

41-44 45‘47 patents. The compounds

of German and Japanese
contained these M(II) species (among others): Ba, Be, Ca,
Cd, Cu, Fe, Mg, Mn, Ni, Pb, Pt, Sn and Zn, with more than
one M(II) species in a single compound in many cases.
These M(III) species were utilized (among others): Al, Am,
Au, Bi, Ce, Co, Cr, Fe, Ga, In, La, Mn, Nd, Ni, Rh, Ti, V
and Y. The LDHs were synthesized with or had exchanged
into them a wide variety of simple mono-, di- and trivalent
gallery anions. The range of x values reported for
[M(II)1_xM(III)x(0H)2][x/zYz"nH20] was 0.1195030,
preferably with x between 0.24 and 0.45.41‘47 The reported

values of n were between 0.25 and 1.

These patents are unsurpassed in sheer numbers of

11

compounds presented. The properties of synthetic LDHs have
been reported in detail for the Mg-Al,48‘54 Mg-Fe55 and Ni-
A150,56-58 systems. One of the reaction parameters studied
was the effect of the variation of the value of x in the
reaction systems on the identity and properties of the
products. Ewn:the Mg-Al compounds,.48'50 values of x<0.l7
led to the formation of brucite along with an Mg-Al LDH.
Values of x>0.33 led to the formation of Al(OH)3 phases
along with the LDH product. Similar results have been
reported for aged Ni-Al products.57 For the Mg-Fe

system,55

pure product was formed in the range of
0.27ng0.15. For the Mg-Al and Ni-Al compOunds the values
of x indicate an upper limit on the M(II)/M(III) ratio of 5
and a lower limit of 2. The limits for the Mg-Fe system
appear to be 5>M(II)/M(III)>3. All these data are for
materials aged either for 4-5 days at room temperatures“,
or 5 days at 60 0C.48"55'57

When freshly-precipitated Ni-Al materials were
examined results different from the aged LDHs were
reported.57 For freshly-prepared materials, single phases
were found for values 0.5Zx20.15, as shown in Table 2.
When the compounds were aged hydrothermally for 2 days at
150 0C, the samples with x>0.33 or x<0.25 contained LDHs
along with an additional phase, as indicated in the table.
This was interpreted as evidence that the compounds formed

initially when x>0.33 or x<0.25 were metastable, and that

they changed into the more stable LDHs and either bayerite

12

Table 2. Phases Formed Upon Aging Ni=Al LDHs.‘a

 

 

Solution xb Before Aging After Aging
0.85 LDH LDH, Ni (0102
0.80 LDH LDH, Ni (0H)2
0.75 LDH LDH
0.66 LDH LDH
0.60 LDH LDH, boehmite
0.50 LDH LDH, boehmite
0.40 LDH, —-

bayerite

 

:From reference 57.
X=[M(III)]/([M(II)]+[M(III)])o

 

or Ni(OH)2 upon hydrothermal treatment.57 This conclusion
was:h1accord with general observations concerning fresh
precipitates of low crystallinity,34 that they are often
metastable under their conditions of preparation and that
their composition is usually flexible. This points to the
importance of aging coprecipitated LDHs, a step mentioned
in every synthesis of this type. The importance of impro-
ving the crystallinity of the product can be seen in the
Ca-Al example.15'21"24 The material prepared by coprecipi-
tation and unaged was metastable in its reaction
mixture.15 The same compound formed under hydrothermal
conditions,21"24 with aging built into its synthesis, was
very stable. So improving crystallinity is not only impor-
tant for structural studies, but also for increased stabi-
lity of the precipitates.

An extensive literature exists on pOwder x-ray dif-
fraction of synthetic LDHs. Almost all the patterns

recorded can be indexed on the basis of either a hexagonal

13
or rhombohedral unit cell. As discussed earlier, this in-
dexing corresponds to two layers and three layers per unit
cell, respectively.12 In the following discussion, the
rhombohedral species will be discussed on the basis of

hexagonal unit cell parameters _a_ and _c_ For simplicity,

0.
9f, the actual layer repeat distance, will be used instead

of c since 2' is independent of symmetry group. In

—o'
Table 3 the values of g' for two LDH series of constant
octahedral layer composition are given. The 3' parameter
varies as a direct function of the radius of the gallery
anion.l If the octahedral layer thickness (4.8 A= two
layers of hydroxidesg) is subtracted from gf, the result is
close to the diameter of the anion. Based on this

assumption, the values of g‘ for the C032" and N03 forms
of the Zn-Cr LDH should be approximately the same. For the
C0327 form, 5y-4.9 A = 2.7 A, which is the approximate

diameter of an oxide group.59 This data agrees well with

Table 3. Variation of Basal Spacing In [Ca Al(OH)6 ]
[Y ZHZO] and [Zn2Cr(OH)6][Y nHZO] or Various
Y Species.

 

 

Metals Y' 2' (A) References
Ca, Al OH 7.4 23

Ca, A1 C1 7.9 22

Ca, Al 0. 5CO3 7.6 8

Ca, Al 0. 5504 8.9 21

Zn, Cr F 7.51 31,32
Zn, Cr Cl 7.73 ,3l,32
Zn, Cr Br 7.85 31,32
Zn, Cr 0. 5CO3 7.60 31,32
Zn, Cr No3 8.88 31,32

14

the single crystal results12 that show carbonates in an
orientation parallel to the hydroxide sheets. Apparently
since there are twice as many nitrates per unit cell as
carbonates, there is not room for them to orient in the
same manner as the carbonates.32 Therefore, they must
stand on end to all fit into the interlayer. This effect
has been observed for other IJHL54'57

Table 4 shows the dependence of g-and E} on the
constitution of the octahedral layers for several carbonate
intercalated LDHs. There are several observations which can
be made about the data presented in the table. For a given
pair of metal cations, as the value of x increases, the
value ofgy decreases. This decrease was said postulated
as the result of increased electrostatic attraction between
the octahedral layers and anionic interlayers as the den-
sity of charge centers increases.5g This effect was not
only observed in the Mg--Al"’9'50 and Ni-A150'57, but also in
the Mg-Fe55 LDH series.

Table 4. Dependence of Lattice Parameters on Octahedral
Layer Composition for Some LDH Species.

 

 

M(II) M(III) xa 3(A) g'(A) Reference
Mg A1 0.16 3.07 7.9 49

Mg A1 0.25 3.06 7.90 54

Mg A1 0.30 3.05 7.73 50

Mg Al 0.34 3.04 7.57 54

Ni Al 0.17 3.05 7.86 50

Ni A1 0.22 3.04 7.76 50

Ni A1 0.28 3.02 7.58 50

 

ax=lM<III)]/([M(II)1+[M(III)J)

15

Observations about the 3 parameter can also be made.
In the Mg-Al series, the Value»of 2 decreases from 3.07 A
to 3.04 A as x increases from 0.16 to 0.34.56“54 This
trend also holds in the Ni-Al series.57 The shrinkage of
the unit cell was due to the increasing substitution of
smaller Al(III) ions (r=0.675 A)” into Mg(OH)2 (Mg(II)
r=0.86 A)59 and Ni(OH)2 (Ni(II) r=0.83 A)59 lattices. This
is evidence, independent of the single crystal studies,
that the M(III) species was incorporated into the octa-
hedral layer, not in the gallery as originally suggested.19

When precipitates are hydrothermally aged, there is,
in general, a narrowing of the x-ray diffraction lines.34
This effect is usually attributed to an increase in the
crystallinity of the sample. Another factor which can
account for line narrowing is related to the method of
synthesis.34 In coprecipitation reactions, when an alka-
line solution is added to the mixture of salt solutions,
the pH begins to rise toward the final desired value. As
it does, the precipitate formed first will be rich in that
metal ion component of the LDH wh'ose hydroxide forms at a
lower pH. Likewise, at higher pH, the solid formed will be
rich in the second component since the ratio of the two
metal cations in solution has shifted in that direction
during the earlier precipitation. As shown above, the x—
ray parameters of LDHs vary with the composition of the

octahedral layers. X-ray powder diffraction of the fresh

materials reveals broad peaks. These could be made up of

16
reflections from two or more different materials, over-
lapped to give a broad peak, as shown in Figure 4 for a Ni-
Al example.58 Upon aging the material becomes more homo-
geneous in composition and more crystalline, both of which
serve to sharpen the x-ray peaks.

The results of thermal analysis of different LDHs are
very similar. In almost all of them the first change upon
heating is loss of molecular water, both from the surface
of the particles and from interlayer regions,32 followed by
dehydroxylation of the octahedral layers and thermal degra-
dation of the interlayer anions. The last two steps often
occur-simultaneously.6g

Differential thermal analysis (DTA) and thermogravi-
metric analysis (TGA) of pyroaurite and sjogrenite66 showed
that molecular water was lost below 200 0C, but that the
process was reversible with no change in structure. By 450
°C, dehydroxylation and loss of C02 were complete. X-ray
diffraction showed the formation of MgO and the spinel
MgFe204. These products became more crystalline as the
sample was heated to 750 oC.6G

Early stuidies of synthetic Ca-Al LDHs indicated that
up to 150°C the structure of the LDHs remained intact. The
gf distance was shortened due to loss of interlayer water.
At temperatures above 150 0C, the x-ray diffraction pattern
grew into that of a mixed calcium-aluminum oxide.

Among the systems studied in detail were the Zn-Cr32,

Mg-Alsg'54 and Ni—Al57 series. In the Zn-Cr system, DTA

Intensity

Figure 4.

17

 

 

 

 

 

 

(I
x value in
l‘;r"”’parent solution
I
5 l
o Precipitate after
‘3 "/,/ ’ aging (more
3 I homogenous)
“ l
5 l
g Precipitate before
0 I .////’aging (x value
0 l varies)
w
m l
(U
f I
O.
l
l
l
1 4
1.00 0.25 0.50 0.00
x value
001 reflections
—--—--—- Nickel-rich phases
-——- - - - Aluminum-rich phases

-— -----—- Resultant profile of
inhomogenous
precipitate

/”\

 

 

 

 

7.8 7.5 3.9 3.75 d-spacing (A)

a.) The effect of aging on the product disa
tribution of coprecipitated NihAl LDH for a
solution value of x=0.25. b.)The effect of the
initial product distribution on the shape of
xhray powder diffraction peaks. Adapted from

reference 58.

18

results showed a dependence of the temperature of tran-
sition on the gallery anion. Of the halide forms, only the
fluoride species showed a well-defined endothermic peak
resulting from the loss of gallery and surface water from
the solid. The next step, partial dehydroxylation, occur-
red at a lower temperature for the fluoride and iodide than
for the bromide and chloride.32 The chloride and iodide
showed another endotherm. These were interpreted as loss
of halide as either HX or X2. The final forms of the LDHs
were assumed to be a mixture of ZnO and chromium oxides or
a spinel, as can be formed by other LDHs.

The Ni-Al LDHs57 also showed a dependence of
dehydration peakcnigallery anion. luaparticular, carbo-
nate forms showed a higher first dehydration peak in DTA
results (Table 5). The carbonate forms of Mg-Al LDHs show
the same effect.51'54 This result was rationalized on the
basis of single-crystal data reported for carbonate

minerals.12 The carbonate was seen to be a "good fit" in

Table 5. Dehydration Temperaturesa for Some MgbAl and
Ni-Al LDHs.

 

 

Dehydration
Metals xb Anion Temperature(°C) Reference
Mg, A1 0.25 NO3'2'_ 235 51
Mg, Al 0.25 co3 ‘ 245 54
Ni, A1 0.25 N03; 100 57
Ni, A1 0.25 co3 ' 245 57

 

:Center point of dehydration peak of TGA.
X’[M(III)]/([M(II)]+[M(III)])o

19

the gallery. The interlayer spacing, dependent on the
oxide group diameter, provided a good fit for the oxygen of
water.9 The oxide groups of the carbonate galley anions
also provided strong hydrogen-bonding to the water, further
hindering dehydration of the galleries.1

A comparison of dehydroxylation temperatures among the
various LDHs in Table 6 shows trends associated with the
metal centers in the octahedral layer. The species con-
taining'Zn(II)[dehydroxylated at lower temperatures than
the other compounds in accord with the reported dehydroxy-
lation temperatures of the pure divalent metal
hydroxides.62

One of the interesting properties of smectite clays is
their ability to undergo unidimensional swelling.25 They '
can imbibe a number of different solvents, expanding in the
3 direction. In the extreme case in water and other sol-

vents, the sheets may be completely exfoliated, expanding

the g parameter to essentially infinity.

Table 6. Dehydroxylation Temperaturesa'b For Some LDH

 

 

Carbonates.
Dehydroxylation
Metals xc Temperature(°C) Reference
Ni, Al 0.25 345 ' 57
Ni, A1 0.33 357 57
Mg, A1 0.25 440 54
Mg, A1 0.33 420 54
Zn, Cr 0.33 300 32

 

aTemperatures from center of dehydroxylation peak from TGA.
Dehydroxylation temperatures(°C) for comparison: Zn(OH) :
125; Ni(OH)2: 230; Mg(OH) : 350. Values from reference 2.

°x=[M(III)l/([M(II)]+[M(III)]).

20

For LDHs, however, there have been only two reported
examples of swelling behavior. The first of these was
reported for a Ca-Al LDH63. Two forms of the hydroxide-
exchanged LDH were reported. [Ca2A1(OH)6] [OH‘2HZO] had a _<_:_'
parameter of 7.4 A, reflecting the presence of one layer of
water in the galleries with the hydroxide. [Ca2A1(OH)6]
[OH'6HZO] exhibited an interlayer spacing of 10.7 A, enough
additional space for a second layer of molecular water. No
other LDH has shown this ability to imbibe extra water
layers. It has been suggested that the behavior of the
Ca-Al LDH is related to its unique structure outlined
earlier. According to the latest hypotheses,25 smectites
intercalated with monovalent cations swell due1x>inter-
actions of the clay surface with the cation and its.
strongly-held solvation sphere. Gallery anions of LDHs
lack this type of solvation sphere. Therefore, LDHs would
not be expected to swell.in the same manner as smectites.
In Ca-Al LDHs, however, because of its large radius, the
Ca(II) is exposed to the interlayer region. In the crystal
structure of [Ca2A1(OH)6] [OH'ZHéO], a water molecule is
coordinated strongly enough to the Ca(II) center to distort
its geometry-from octahedral.]‘2"13 It was suggested that
this cation was exposed enough to hold an extra layer of
water in the same way the cations in smectites attract
extra layers. That this phenomenon was observed only when
OH‘ was the gallery anion suggests hydrogen bonding inter-

actions as important factors in the swelling.

21

The other example of swelling in LDHs was reported for
Zn-Cr LDHs.31 When an alkyl sulfate exchanged form of the
material was exposed to either a linear primary alcohol or
amine, the interlayer spacing was found to increase» The
magnitude of the swelling was dependent upon the hydro-
carbon chain length of the swelling agent, as shown in
Table 7. This swelling phenomenon was said to be the
result of solvation of the alkyl chains of the anion to
give an approximate double layer structure.31 This be-
havior was compared to that of alkyl ammonium substituted
layer silicates.25 A model of this interaction is shown in
Figure 5.

Among the patented uses of LDHs are applications as ion
41 64 in-

exchangers, carriers for anionic pharmaceuticals,

hibitors of thermal and ultraviolet degradation of thermo-

65

plastic resins and scavengers of Cl’ from Ziegler-Natta

Table 7. Swelling of [ZnZCr(OH)6][C12H250803] in Alcohols
and Amines.a

 

 

Swelling Agent Interlayer Spacing (A)
noneb 26.2
C4H90H 29.2
C6H130H 30.8
C10H210H 38.2
C12HZSOH 41.1
C12H25NH2 41.7

 

:From reference 31.
Exchanged in water, air-dried.

22

\

.\\
NH3 on on m3'( NH3

1\\\\\\\\\\\\\\

 

 

 

 

 

Figure 5. Structure of an alkylammonium-exchanged smectite
swollen with a linear alcohol. From reference
25.

23

66

catalyst residue in plastics. By far the greatest utili-

zation of LDHs has come in the field of catalyst

preparation.34.56-58,67-71

LDHs formed by coprecipitation
have been calcined to form oxides to catalyze aldol
condensations.7g'71 Ni-Al LDHs have been calcined, then
\reduced, to give supported nickel catalysts for steam

67 Cu-Zn—Al LDHs have been treated in a similar

reforming.
manner to produce catalysts for methanol synthesis.69 Part
of the extensive literature on the preparation of these
catalysts is concerned with the characterization of the
catalyst precursors (LDHs), and much of the data quoted in
this work comes out these reports.55'58'67'68 To date,
however, no reactions involving uncalcined LDHs have been
reported.
2. Supported reagents.

One of the major advances in organic chemistry in the
recent past has been the use of supported reagents in

synthesis72. In this technique, reactants for organic

reactions are bonded to a polymeric organic backbone73, or
somehow attched to an inorganic support72'74. This can be
accomplished by covalent bonding, adsorption or
intercalation. These supported reagents offer several

72-75. (1) The effective surface area for

advantages
reaction is increased. This accelerates the reaction rate
by increasing the number of reactant pairs brought
together. (2) Solids always have microscopic pores on

their surfaces. These pores may serve to constrain the

24

substrate and reactant, and therefore lower the entropy of
activation of the reaction. (3) Product workup after reac-
tion is usually simplified. This is especially true where,
without a solid support, a mixture of solvents must be used
to bring the reactants together; This situation can lead
to complex distillation schemes or azeotrope problems.
With the supported reagents the solvent for the solution-
phase reactant can be chosen so that separation problems
are kept to a minimum, or in some cases no solvent is
needed at all. (4) In many cases the supported reagent can
be easily regenerated.

Among the solid supports which have been utilized in
the formulation of supported reagents are:72'73 Celite,‘
silica, alumina, graphite, activated carbon, montmoril-
lonite, molecular seives, polymers and resins. Of the
inorganic supports, Celite-supported silver carbonate was
among the first to be utilized72. Acid and base forms of
ion exchange resins were among the first organic solid

76. Since these early

reagents to be used in synthesis
uses, many other reagents supported on solids have been
devised.

The range of reactions studied encompasses all classes
of organic synthesis. The focus here will be on nucleo-
philic substitution reactions by anionic species. In
general, these reactions are carried out by preparing an

anion exchange resin in the form of the desired nucleo-

phile, then mixing the substrate with the resin at the

25

desired temperature73. One of the first reactions to be

reported using this technique was the cyanation of benzyl

halides to their corresponding cyano compounds77

. Among the
problems associated with this reaction in the solution
phase was the formation of side products by competing
elimination reactions. The polar solvent used in these
reactions, usually acetonitrile, promoted these side
reactions. With the supported reagent, only the substrate
benzyl halide need be dissolved so that a nonpolar solvent
could be utilized. The problem of elimination reactions
was solved in most of the systems studied.76 A partial
listing of other nucleophilic displacement reactions which
have been carried out using polymer-supported reagents is
given in Table 8. As can be seen, a number of reactions
have been reported. Many of the reactions shown were
difficult to carry out in the solutions phase. The polymer
support seemed to increase the nucleophilicity of such

79 86

reactants as F“ and sulfonates, allowing reactions

such as this to be performed simply.

One application of supported reagents is the technique

of triphase catalysi387'88,

89,90.

which is an extension of phase
transfer catalysis Phase transfer catalysis provides
a way to react two mutually—insoluble reagents without the
use of a cosolvent. All phase transfer reactions involve

9“ (1) transfer of one reagent from its

at least two steps:
"normal" phase into the second phase; and (2) reaction of

the transferred reagent with the nontransferred reagent.

26

Table 8. R actions Utilizin Pol mereBound Nucleo hiles of
t e Form ~CH g+Me §a p
2 3

 

X“ Reaction System References

 

F conversion of sulfonyl chlorides and ' 78,79
alkyl halides to fluorides

Cl,Br,I halogen exchange with alkyl halides 79,80

CN alkyl halides to nitriles 77,81

SCN alkyl halides to thiocyanates or isoe 81,82
thiocyanates

OCN alkyl halides to ureas and urethanes 82

N02 . alkyl halides to nitroalkanes and 81,83
nitrite esters

OCOR alkyl halides to esters 84

OH condensation reactions 85

OAr alkyl halides to ethers 83

OZSPh synthesis of alkyl sulfones 86

 

27

Some of the species which have been found to function as

90'91 quaternary ammonium and

phase transfer catalysts are
phosphonium salts, crown ethers, cryptands and linear poly-
ethers. Details of the mechanism of reaction vary with
these different catalyst systems, but the simplest case
involved SNZ-type nucleophilic displacement reactions as
shown in Scheme 1.89 The catalyst provides a way of fer-
rying the nucleophile to the organic reagent, then regene-
rating in thelaqueous phase. More than 1000 publications
and patents on phase transfer catalysis attest to its util-

ity in these and many other reactions.88'90

or + RX .\.-=-_-‘: RY + ox (organic)

QY + MX : MY + QX (aqueous)

Scheme 1

More recently, a synthetic approach has been developed
that simplifies catalyst recovery and avoids the problem of
emulsion formation sometimes encountered in phase transfer
catalysi387'88. This approach, termed triphase catalysis,
is an extention of phase transfer catalysis. A conceptual
diagram of triphase catalysis is given in Figure 6. Most
triphase catalyst systems have involved the attachment of
one of the active phase transfer species to an insoluble

support. The supports utilized include polystyrene-

divinylbenzene polymers87 and silica gel92'93. Smectite

28

 

AQUEOUS ORGANIC

 

 

 

 

Figure 6. A conceptual diagram of triphase catalysis.

29

clays modified so that they became anion exchangers94 have
also been utilized In order for these materials to function
as triphase catalysts, they must provide(): OJ anion ex-
change capacity; (2) an environment in which these aqueous
anion exchanges can take place (idh, a degree of hydro-
philicity); and (3) an environment in which the organic
reagent has access to the nucleophile (idh, a degree of
organophilicity). The way in which these criteria were met
in the aforementioned systems will now be described.

By'far the greatest number of reports have concerned

themselves with polymer-based catalysts.88

In these sys-
tems the active sites consist of quaternary ammonium or
phosphonium groups bound to the polymer matrix through a
hydrocarbon chain. Crown ethers and cryptands have also

been effectively utilized as active species.95

The polymer
matrix employed in these studies is usually polystyrene
crosslinked with 1-4% divinylbenzene. The organic matrix
and hydrocarbon groups surrounding the anion exchange site
make the solid organophilic enough for the organic reagent
to interact with the nucleophile. At the same time the
anion exchange site is exposed enough to be accessible to
the aqueous solution of nucleophile that regenerates the
active sites. One of the problems associated with the use
of these supports is matching solvent with polymer.90
These polymers swell to differing degrees in different

solvents as a function of their percentage of crosslinking.

In the cyanation of l-bromooctane, the activity of the

30

catalyst depends on the swelling properties of the resin

9“ Specifically, the resins that swell more in the

support.
organic phase than in water showed greater activity. This
effect can be explained on the basis of diffusion. If the
resin were swelled with organic phase, there would be a
large amount of reactant dispersed in the resin. Under
these conditions diffusion of the organic reactant to the
active site would be more favorable than if the aqueous
phase predominated in the resin. Since the reaction rates
were higher when the organic phase was present, the inter-
pretation was that the rate is controlled by diffusion of
the substrate to the active site on the catalyst.

Another way in which the ammonium and phosphonium
salts have been supported is by attachment to silica
gel.92’93 Attachment was accomplished by way of a
functionalized organosilane coupling agent. These catalysts
were effective in many of the same reactions as the more
prevalent resin-bound materials. These solids provided
access to both organic and aqueous phases in much the same
way that the polystyrene-supported materials did. There was

96'97 as there

no problem with swelling in these catalysts,
was with the resins. Therefore, in order to simplify
product workup, the reactions could be run without an
organic solvent. There was a problem, however, with water
adsorbed on the surface of the silica that created a

hydrophilic environment around the active sites of the

catalyst under certain conditions. These conditions arose

3]

if the hydrocarbon chain on the silane group was too short,
or long enough to allow the active site to be bent to the
silica surface.92 Also, with very long chains, phenyl-
substituted substrates were solvated by the long chains.
These solvated molecules were unavailable for reaction and

93

consequently the rate of reaction was slower. The use

of smectite clays as triphase catalyst supports has been

(L94'98 Montmorillonite intersalated with

recently reporte
benzyltri-n-butylammonium bromide catalyzed the reaction of
phenol with l-bromobutane in the presense of base.98 The
environment of the solid seems to have enhanced the nucleo-
philicity of the phenoxy group, as the conversion with the
supported system was greater than that using an analagous
homogeneous phase transfer catalyst system.

In other work that utilized clays, hectorite was
intersalated with Ni(phen)3SO4 (phen=l,l0-phenanthroline)

and with cetylpyridinium bromide.94'99

A comparison of
intercalation and intersalation is given in Figure 7. The
net effect of intersalation was to induce anion exchange

94 These materials were utilized

capacity in the smectites.
as triphase catalysts in halide exchange reactions on alkyl
halides. ApparentLy, the use of organometallic complexes
and organic cations in the galleries of the clays provided
the organophilicity that, along with the aforementioned
anion exchange capacity, are the two requirements for pro-

ducing a solid catalyst for triphase reactions. In the

halide exchange reactions, the [Ni(phen)3]2+ intersalates

32

 

H

 

 

 

 

 

 

 

 

 

«I» 2+
0 Mu): M(L)3
Homoionic, l8A°
[ ____ .____ ____ ____.
NHL)? NHL)? NHL)? NHL)?
.
NHL)? NHL)? NHL)? NHL)?

 

_

1._J

Biloyer lntersoldtion,~30A°

Figure 7. A diagram of a smectite clay with gallery ions
as a result of a.) homoionic intercalation of
cations, and b.) intersalation of two equiva-
lents of cation and one of anion. Structure b
acts as an anion exchanger.

33

were more effective than tflua cetylpyridinium

99

intersalates. Also, in the caserof the halide exchange

reactions using the nickel-based material, the catalyst was

94 The rate of con-

somewhat size and shape selective.
version of l-chloropentane to its corresponding bromide was
reported to be much greater than any of the larger, more
bulky alkyl chlorides. The reactions proceeded at almost
identical rates under homogeneous phase transfer conditions
that used tricaprylmethylammonium chloride as the
catalyst.94 Inherent differences in reactivity of these
primary alkyl chlorides were apparently not responsible for
the differences in conversion using the solid catalyst.
There were two possible explanations put forth for the
differences in reactivity.99 The first one stated that the
differences arose from the layered structure'of the clay.
There was only a limited amount of space in the interlayer
region of the clay, allowing the smaller, unbranched mole-
cules to diffuse more rapidly into the interlayer region
where the nucleophile was immobilized. This explanation
was made with the assumption that the reaction took place
within the clay layers. The other explanation started out
with the hypothesis that the reactions took place in an
interface between the two liquid phases on the surface of

99 Since the nucleophiles were contained within

the clay.
the aqueous phase, the smaller, less hydrophobic alkyl
would react more quickly by virtue of their greater ability

to diffuse into this aqueous phase. This still did not

34

explain why the branched hydrocarbons reacted so much more
slowly than their straight-chain analogs. Perhaps the true

mechanism was some combination of these two.

Objectives of Current Research.

As outlined above, the physical properties of layered
double hydroxides have been relatively well explored. The
chemical properties of gallery anions and surface proper-
ties of the materials have not been adequately addressed in
the past. In this work, the goal of the research was an
understanding of these properties. The properties of the
gallery anions and surface were to be explored by use of
LDHs as sources of nucleophile under triphase catalytic
conditions and under stoichiometric conditions. The inter-
actions of LDHs with aqueous and organic phases under these
conditions would provide data about the availability of
gallery anions for reaction and the relationship of surface

properties to reactivity.

Chapter 2. Experimental Section.

A. Synthesis of Materials.

1. [ZnZCr(OH)6][Cl'nH20].

This compound was prepared by a modified version32 of
the original reported procedure. A slurry of 45 g (0.55
mol) ZnO was prepared by adding water to the powder with
stirring. To this slurry was added 150 mL of a 1.0 M CrCl3
solution (0.15 mol Cr(III)) in 50 mL water. This addition
brought the pH of the solution to 3.4. The mixture was
stirred at 60 0C until the solution became colorless,
indicating that all of the Cr(III) had been consumed. The
clear solution was decanted from the solid, and a further
150 mL of l M CrCl3 and 50 mL water added to the solid,
bringing the total Zn/Cr ratio of the mixture to slightly
less than 2, the small excess of Cr(III) assuring complete
reaction of all ZnO. The mixture was stirred at 60°C for 5
h. The solid product was washed with distilled water by
centrifugation until the wash water was free of C1“ (AgN03
test). The lavender-gray product was resuspended in water
and poured onto a glass sheet to dry as a 5-10% suspension
and allowed to air dry. The product was identified by x-
ray powder diffraction (5001:7'735‘)‘ Completeness of
reaction was verified by the absense of any peaks

attributable to 260.100

35

36

2. [Ca2A1(OH)6][Cl'nH20].

In accord with one of the published methods for prepa-
ration22 5.97 g (54.0 mmol) anhydrous CaC12 was dissolved
in approx. 75 mL distilled water. To this solution was
added 6.96 g (89.0 mmol) Al(OH)3 and 10.0 g (135 mmol)
Ca(OH)2, resulting in a Ca/Al ratio of 2.10, a slight
excess of Ca(II). The suspension was placed in a Parr
stainless steel pressure reactor equipped with a Pyrex
liner and a stirrer. The mixture was heated to 150 oC.
The temperature was monitored and controlled through use of
a temperature controller and a built-in thermocouple. The
reaction was allowed to proceed for 100IL. The resulting
white product was washed with distilled water by cen-
trifugation until the wash water was free of Cl' (AgNO3
test). The material was resuspended and poured onto a
glass plate to dry. Identity of the product was confirmed
by x-ray powder diffraction through comparison with pub—
lished data(d661= 7.9 A)22 and by the absence of any peaks
associated with the starting materials.106
3. [LiA12(OH)6][SO4‘nHZO].

This compound was prepared by following the procedure

al.101 A saturated solution of LiOH was

of Serna et.
prepared with freshly boiled deionized water to remove
traces of carbonate. This solution was placed in a three-
neck flask that had been purged with NZ to remove traces of

C02 that could be absorbed by the strongly basic (pH=l3.0)

solution. These precautions were necessary because of the

37

reported preference of LDHs for carbonate ion.l As N2 was
bubbled through the Solution, a 0.5 M solution of A12(SO4)3
was added dropwise. The pH of the solution was monitored
during the addition. When a value of pH=l0 was reached,
the addition was stopped. The resulting powder product and
its mother liquor were placed in a Parr pressure reactor
and hydrothermally treated for 3 days at 130 °C. The
resulting product was washed on a Buchner funnel under N2
until the wash water was free of Li+ (flame test). The
identity of the product was confirmed by x-ray powder
diffraction by comparison of the basal spacings with liter-
ature values (d061=l0.45 A)102and the absence of peaks from
any other Al or Li containing phases.
3a. Synthesis of spin-labelled [LiA12(OH)6][Cl'nH20].
Nitroxide-doped samples of [LiA12(OH)6][Cl‘nHZO] were
prepared for EPR analysis with K2N0(SO3)2, peroxylamine
disulfonate (PADS), present in the reaction mixture. Since
the PADSZ' species was stable in basic solution and
unstable in acid, 0.064 g (0.24 mmol) KZPADS was placed in
5 mL water with 1.0 g (24 mmol) LiOH'HZO. A 0.5 M solution
of A1C13 was added as before. The solid was allowed to age
at room temperature for 4 h and then washed under N2. The
hydrothermal treatment used on the sulfate sample earlier
would have caused the decomposition of the spin probe. The
solid was stored under Ar in a dessicator to prevent the

.decomposition of PADSZ‘ by oxygen.

38

4. Ion exchange reactions.

In a typical ion exchange reaction, 1 g of solid was
immersed in enough 1 M solution of the desired exchange
anion to give a 30-fold molar excess of anion. The mixture
was stirred at 60 °C for 2-3 h. The solution was decanted
and the process repeated. Usually, one repetition was
enough to make theloriginal LDH undetectable in the x-ray
diffraction pattern. lecases where solubility problems
precluded use of a 1 M solution, as in the case of sodium
dodecylsulfate exchange, or thermal instability prevented
reaction at 60 °C, as in attempted exchanges with PADS,
Llonger exchange times and repeated applications of fresh

solution were used to effect complete exchange.
B. Displacement Reactions Over [ZnZCr(OH)6][Y'nHZO].

1. Reactions Under Triphase Catalytic Conditions.

The conditions utilized were similar to those used by
earlier workers in triphase catalysis in order to facili-
tate direct comparison of results. Iodide substitution
reactions of alkyl halides were investigated. In these
reactions the aqueous phase consisted of 0.899 g (6 mmol)
NaI in 3.0 mL water. The organic phase was composed of 1
mmol alkyl halide in 2.0 mL toluene. In all cases a 20:1
molar ratio of alkyl halide:LDH was utilized. So for the
reactions, 0.023 g (0.05 mmol) [ZnZCr(OH)6][I'2.3 H20],

prepared by anion exchange with the original chloride

39

material, was added to catalyze the reaction. A dodecyl-
sulfate-exchanged form (DDS, C12H250803') was also used in
some reactions to test whether the presence of an organic
species in the gallery would affect reaction yields. In
these cases, 0.029 g of the DDS form of the solid was used
in the reaction mixtures. All of these reactions were
carried out in Pyrex culture tubes sealed with a Teflon-
lined cap. The tubes were placed in a constant temperature
(Omega Engineering model 4000 temperature controller) oil
bath maintained at 90 °C. A magnetic stir plate and spin
bars in the individual tubes provided agitation of the
reaction mixtures. The reactions were allowed to proceed
for 24 h. The products were analyzed by GC. Blank
reactions were also run which included all the ingredients
listed except the solid LDH. These were carried out and
analyzed in the same manner as the triphase reactions.

2. Reactions Under Stoichiometric Conditions.

The same reactions carried out under triphase catalytic
conditions were also attempted in the absence of an aqueous
phase. The iodide-exchanged form of the Zn-Cr LDH was
again utilized. The air-dried (AJL) form with n=2.332 and
a form dried at 150 0C for 2 h under Ar, the oven-dried
(OJL) form with n=0, were tested in the reactions. In
both cases 1 mmol equivalent of intercalated I" “L453 9
A.D. LDH or 0.412 g O.D. LDH) was added to 3.0 mL organic
solvent in a pyrex culture tube with Teflon-lined cap. To

this mixture were added 1 mmol alkyl halide and a magnetic

40

spin bar. The culture tubes were placed in the temperature
controlled oil bath at the desired reaction temperature,
typically 90 °C. A magnetic stir plate provided agitation
of the reaction mixtures. The tubes were withdrawn from
the oil bath every 10 min, cooled in an ice bath to quench
the reaction and ca. 10 pl of liquid reaction mixture
withdrawn. The culture tubes were quickly resealed and
returned to the oil bath after each extraction of material.
The samples were analyzed by GC to calculate percentage
conversions.

In some cases the RBr:I' ratio was varied to study the
effect on the reactions. Ratios of 2, 3, and 4 were used
in addition to the typical 1:1 ratio.

3. Gas-Solid Reactions.

A vertically-mounted, fixed bed continuous flow micro-

reactor previously used to test for catalytic activity of

clay minerals162

was utilized to study the reactivity of
LDHs toward vapor phase alkyl halides in nucleophilic dis-
placement reactions. The reactor consisted of 7 mm I.D.
quartz tubing encased in a tube furnace heated and control-
led by a three stage temperature controller (Eurotherm
model 919A). The alkyl halide was introduced with a
syringe pump (Sage Instruments, model 341A) into a pre-
heater zone at the top of the reactor. The vector gas, He,
was purified with BASF R3-ll catalyst to remove oxygen and

4A molecular seives to remove water. The He flow rate was

controlled through a flowmeter.

41

In a typical reaction, 1 mmol Zn-Cr LDH was loaded into
the reactor tube. Several 2 mm glass beads were placed on
top of the solid to ensure complete vaporization of the
alkyl halide. The solid was pretreated at 150 °C for 2 h
under a flow of He to remove water from the structure. The
syringe pump delivered 1 mmol alkyl halide in 4.8 min into
a He flow of 3.01nL/min. This combination of parameters
gave a contact time on the solid of 0.20 seconds. The
product mixture was collected at liquid N2 temperatures in
a quartz trap at the bottom of the reactor for 30 min to
ensure collection of all reaction products. The products

were diluted with toluene and analyzed by CC.

C. Physical Methods.
1. X-ray Diffraction.

Powder x-ray diffractograms were recorded using either
a Siemens Crystalloflex-4 or Philips x-ray diffractometer.
Both instruments utilized Ni-filtered Cu K“ radiation.
Some samples were prepared by placing approx. 1 mL of a l-
5% suspension of the solid to be examined on a microscope
slide and allowing it to dry. The slide was then placed in
the goniometer for examination. For samples such as DDS-
exchanged species that could not be resuspended or would
not form a film which adhered to the slide, a different
technique was used. Double sided tape was placed on the
microscope slide. The powder was applied to the tape and

then pressed on‘with another slide. 'The second slide was

42

removed, leaving a pressed powder sample. The first tech-
nique, which gave films with the LDH platelets preferen-
tially oriented parallel to the slide, was the preferred
method of preparation since the °001 reflections were
accentuated in these diffractograms. Typically, the sam-
ples were scanned from a 2 value of 2° to 30°. The inten-
sity of the signal was recorded on a strip chart recorder.
The 2 values were converted to d-spacings using a standard
chart based on Bragg's equation for Cu K3 radiation (A
=1.5405 A).

2. EPR Studies.

EPR spectra were recorded using a Varian E-4 X-band
spectrometer. The synthesized materials containing the -
PADS spin probe were placed in a quartz tube while damp and
forced to the bottom of the tube by a Teflon rod. EPR
spectra were recorded while the sample was damp and after
drying for various periods of time. Standard pitch served
as a reference material (g=2.0023).

3. Infrared Studies.

Infrared spectra were recorded by use of a Perkin Elmer
model 457 or model 533 spectrophotometers. Both were
double beam instruments with diffraction gratings for wave-
length selection. The spectra were recorded from 4000 cm"1
to 250 cm'1 on samples prepared as KBr pellets.

4. Gas Chromatography.
Reaction product mixtures were analyzed by using gas

liquid chromatography (GC). Some samples were analyzed by

43

use of a Varian Associates model 920 CC with thermal con-
ductivity detection. A 10' x 0.25" 5% SE-30 on Chromosorb-
W column separated the components of the sample. The
chromatograms were recorded by use of a Linear Instruments
model 252A recorder-integrator which was calibrated with
standand mixtures to insure accuracy of integration. All
reaction mixtures involved in kinetic studies were analyzed
by use of a Hewlett-Packard model 5890A capillary GC using
a 60'm x 0.25 mm dimethyl polysiloxane (Supelco SPB-l)
column and flame ionization detection. The chromatograms
were recorded and analyzed by a Hewlett-Packard model 3392A
programmable plotter/integrator which was capable of custo-
mized data handling. All peaks were identified by
comparison of retention times with authentic samples of
reaction mixture components.

5. Surface Areas.

Surface areas were determined using a Perkin-Elmer-
Shell model 212B sorptometer with N2 as the adsorbing gas.
These surface areas are referred to as B.E.T. surface
areas, in honor of the developers of the technique,
Brunauer, Emmett, and Teller. The samples were either
outgassed for l h at room temperature under dry Ar or at
150 °c for 2 h under Ar.

6. Adsorption Studies.
The adsortion capacities of Zn-Cr LDHs were measured by

103

using a McBain balance arrangemen of quartz springs in a

vacuum system. The vacuum line was constructed of quartz.

44

Teflon stopcocks and valves were used to keep the system
grease-free. Vacuum was achieved with rotary pumps and a
CVC model PVMS-3l diffusion pump. Pressure was measured
with a Datametrics model 1400 electronic manometer.

In a typical experiment, 75 mg of an LDH was placed in
a quartz bucket and suspended from a pre-calibrated quartz
spring. The position of the bucket was measured using a
telescope on a height-calibrated mount. The portion of the
vacuum system containing the solid was then evacuated.
Samples were either evacuated to le0’3 torr at room tem-
perature before adsorption, or heated to 100 °C for 90 min
under vacuum to remove water from the structure. Nhile the
samples were being treated, the solvent to be used as
adsorbate was degassed in a vacuum isolated from the solid
samples. Heat treated samples were allowed to equilibrate
for l h before adsorption studies began. The position was
noted so that the final weight of the solid material could
be calculated and used in subsequent data treatment.

The adsorptions were carried out at 2 temperatures,
22°C and 0‘WL An ice bath was employed to cool the sam-
ples to 0‘NL These samples were allowed to equilibrate
for l h before beginning adsorption. The adsorbate was
also maintained at 0 °C to minimize condensation on th
solid at high values of p/po. After all temperatures were
equilibrated the system was isolated from the vacuum pumps.
A small quantity of adsorbate was leaked into the system

from a smaller manifold. The- system was allowed to

45

equilibrate approx. 5 min, by which time the sample had
stopped adsorbing. The position of the bucket was recorded
to a precision of 0.05 mm by using the telescope. The
pressure was also recorded. The procedure was continued up

to a value of p=po for the adsorbate at 0 °C.

Chapter 3. Results and Discussion.

A. Properties of LDHs.
1. Exchange and Swelling Properties.

The compounds [ZnZCr(OH)6][Cl'2H20], [Ca2A1(OH)6]
[l/2CO3'2H20] and [LiAlZ(OH)6][0.5SO4'2HZO] were
synthesized. The carbonate form of the Ca-Al LDH was
formed in the initial synthesis due to improper precautions
against C02 absorption. Proper precautions were taken in
subsequent syntheses. The powder x-ray diffractograms of
these materials are shown in Figure 8. As can be seen, the
patterns are nearly identical, as the structural components
of these species were very similar. When anion exchange
reactions were attempted with these species, the Zn-Cr LDH
exchanged more readily than the other two species. This
difference may have been due to some adventitious carbonate
in the structure of the Ca-Al and Li-Al LDHs, even when
precautions against C02 adsorption were taken. It has been
reported that carbonate forms of LDHs are difficult, if not
impossible to exchange. This difficulty is due to the good
fit of carbonate in the gallery and the -2 charge. The
acidic conditions under which the Zn-Cr LDH was formed
preclude the uptake of C02 by the reaction mixture.

As can be seen from the diffractograms in Figure 8 the

46

47

 

 

 

 

 

 

 

 

 

 

 

 

1
i
a 1
Leax 1
HL4SA
'Lssi ' flpflWNP
5.271
352A
1_ 1 j (
. ‘ V I
C
0
378A
./ \ k'\de\Aw-v*Ffi-
fl v
25 20 15 10 . 5

degrees 29

Figure 8. Powder x-ray diffractograms of (a). [ZnZCr(0H)6]
[C1'2H20]; (b). [LIA12(OH)6][0.5304'01‘120]; and
(C). [C32A1(0H)6][0.5C03'2H20].

48

peaks from Zn-Cr LDH were broader than those of the other
species and was partially due to the method of sample
preparation for x-ray examination. The Ca-Al and Li-Al
species formed oriented films on microscope slides. This
preparation tends to minimize layer disorder that can
broaden diffraction peaks. An x-ray diffractogram of a Ca-
Al sample prepared by the pressed powder method is shown in
Figure 9 along with that of the oriented film sample. Some
broadening of the °00l peak is observed in the pressed
powder sample. The Zn-Cr LDH did not form oriented films
and therefore was examined in the diffractometer as a pres?
sed pdwder sample. 'This random orientation of platelets
probably helped broaden its peaks.

Another factor which determines diffraction peak width
is particle size. The synthesis of the Ca-Al and Li-Al
LDHs featured a hydrothermal aging step. Aging increased
particle size and in turn lead to sharper diffraction
lines. The synthesis of the Zn-Cr LDH included no such
aging step. Indeed, other workers31'32 have found that the
particles of this compound cannot be made larger through
aging. The small particle size, along with the absence of
carbonate, probably contributes to making Zn-Cr LDHs the
most facile anion exchangers.

In a previous section the swelling behavior of alkyl
sulfate intercalates of Zn-Cr LDH was discussed. The swel-
ling behavior is a general property of the LDHs tested.

Anion exchange reactions were carried out by using a 0.5 M

49

 

 

 

7.56 A
(I
3.78 A
i
l
i
M
i
'3 Ii
’1
!
1
-¢~4rw
25 20 15 10 5
degrees 20
Figure 9.

Powder x-ray diffractograms of [Ca2A1(OH)6]

[0.5CO3°2HZO] (a). pressed powder; (b) oriented
film.

50

solution of NaDDS with the three LDHs. X-ray diffrac-
togranusof the three intercalates are shown in Figure 10.
Again the diffractograms of different LDHs with the same
gallery anion are practically identical. The small dif-
ferences (dual=22.6—24.5 A; gallery heights=l7.6-19.5 A)
were probably due to instrumental uncertainties at low
angles and to so-called kink block structures.1°4 These
result from differences in the carbon chain arrangement
when long chain hydrocarbon species are intercalated in
layer structures. When the three DDS-LDHs were immersed in
n—octanol, all were swollen as had been previously reported
for the Zn-Cr LDH. The gallery heights ranged from 30.4 to
33.9 A. The x-ray diffractograms for the swollen species
are given in Figure 11.

There was a dramatic change in surface properties of
the three LDHs upon exposure to NaDDS solutions. The DDS’
seemed to have a strong affinity for the surfaces of the
LDHs. The affinity was first indicated in washing the
materials after exchange of DOS“ into the galleries.
Removal of the excess electrolyte by centrifugation nor-
mally required 5-10 washings of the material after
exchange. After exposure to DDS', however, 15-20 washings
were required before foaming of the wash solutions ceased,
indicating that excess NaDDS had been removed from the
system. Upon drying, the DDS-exchanged materials were
found to be hydrophobic to the extent that they floated on

water. This behavior was completely opposite that of

51

22.6 A

 

 

 

 

 

23.9 A
b 8.2 A 12.1 A
C
15 10 5 2
degrees 25
Figure 10. Powder x-ray diffractograms of the dodecyl-

sulfate-exchanged forms of (a). Zn-Cr LDH; (b).

Ca-Al LDH; and (c). Li-Al LDH.

52

3L9 A

 

O
30.4 A

10.53

13.3A

10

degrees 29

Powder x-ray diffractograms of n-octanol—swollen
DDS—exchanged forms of (a). Ca-Al LDH; and (b).

Li-Al LDH.

Figure 11.

53

inorganic-anion exchanged LDHs. The pristine materials
were all found to disperse in water, and also to be wetted
by toluene and alcohols. When placed in the presence of
both toluene and water, the water was preferred. The DDS
materials dispersed in organic solvents but not in water.
This behavior persisted even upon repeated washing in etha-
nol and acetone, sonification and repeated washing to
remove the DDS from the surface.

The tenacity with which the DDS anion was attached to
the surface of the LDHs is not unprecedented. It has been
reportedla4 that DDS“ was strongly bound to -alumina
through interaction of the -SO3' tail of the anion with the
hydroxide surface. The "fit" of the -SO3' group to the
surface groups was found to be very good, encouraging a
strong electrostatic interaction. LDHs have a similar ar-
rangement of hydroxide groups on the surface, plus a net
positive charge in the layers not completely compensated by
the gallery anions. Therefore, an even stronger attraction
for DDS' probably exists here than in the alumina.
Additionally, [ZnZCr(OH)6][Cl'2H20] was exposed to a 0.1 M
solution of NaDDS for 30 seconds, then removed and washed
as if an exchange had taken place. X-ray diffraction
indicated that no exchange had occurred, but the material
exhibited the same hydrophobic properties as the exchanged
material had, indicating that the source of the change in
properties was due to surface interactions. The IR spectra

of pristine and DDS-exposed materials shown in Figure 12.

54

4000 3500 3000 2500 2000

P

2000 1800 1600 1400 1200 1000

wavenumber (cm'l)

Figure 12. Infrared spectra from 4000 cm‘1 to 1000 cm'1 of
[ZnZCr(OH)6] [Cl'nHZO] (a). pristine material;
and (b). exposed to 0.1 M NaDDS solution.

55a

The presence of DDS is indicated by the C-H stretching
bands near 3000 cm-1 and bands at 1100 cm'1 and 1175 om'l
due to S-O stretches of bound DDS. These IR data compare
with those given for DDS on alumina, with S-O bandsat 1200
cm'1 and 1240 cm’l.104 The shift to lower energy of the S-
O stretch in the present study is probably an indication of
stronger binding of the -S03 group to LDHs. The authors
also stated that after the first layer of DDS' was attached
to the surface, additional layers were adsorbed by cosol-
vation of the alkyl groups of bound and unbound anions.
This phenomenon could account for the large number of
washings required to remove excess NaDDS from an exchange
mixture. The implications of this type of interaction will
play a role in the interpretation of EPR data to be
presented.
2. EPR Studies.

EPR spectroscopy has provided a probe into the mobility
.of gallery cations and anions of layered compounds. For
smectite clays, paramagnetic metal cations such as Cu(II)

and MMU) have been used as probes,l°5'106

as well as the
protonated form of the radical 4-amino-2,2,6,64tetramethyl-
piperidine-N-oxide (HITEMPO).1°7 The PADS dianion has been
used as a probe for intersalated smectites. The mobilities
of these species in the galleries has led to work in which
gallery cations in clays were used as catalysts,108 and the

triphase catalysis using intersalated clays outlined

earlier.

55b

In the present work, attempts were made to use the PADS
dianion as a spin probe in LDHs. If LDHs were to be effec-
tive as solid reagents or solid phase transfer catalysts
for nucleophilic displacement, freedom of movement of the
gallery anions would be necessary. The nitroxide spin
probes in general have been observed to tumble rapidly
enough at 25 °C to completely average the anisotropies in g
tensors and hyperfine values. As the tumbling of the anion
in solution is restricted, the spectrum becomes increasing-
ly anisotropic. The two limits of anion mobility give the
two spectra shown in Figure 13.

The spectra obtained in this study were analyzed fol-
lowing an approach used by biophysicists studying mobility
of spin probes in muscle tissue under conditions of varying

109

water content. The relationships for the spin corre-

lation times are shown in Equations 1 and 2. In these

’13.: 0.65W0(R+-2) (2)
R+ = [(hg/h+1)1/2] + [(hg/h-1)l/2] (3)

equations, and are tumbling times in nanoseconds, W9 is
the linewidth of the central peak in G and H0 is thelmag-
netic field in G of the central peak. The symbols 110' h+1
and h_1 represent the heights of the middle, low and high
field peaks, respectively; The correlation time» c' was
the average of the two values from Equations 1 and 2. The
equations used to obtain the correlation times are strictly

valid only for isotropic rotation, but they are useful for

 

 

 

(I
o—N( :3
b
20 GAUSS
F—H H
—'-
Figure 13.

EPR spectra of KZPADS (a). 10"5 M solution in
DMSO; and (b). frozen in D20.

57

comparative purposes for studying anisotropic rotation at
surfaces.

The first attempt to prepare samples for EPR analysis
was made by trying an anion exchange reaction of a 0.1 M
solution of KZPADS into [ZnZCr(OH)6][C1'2H20]. The EPR
spectrum of the resulting material is shown inFigure 14.
As can be seen, if any PADSZ‘ were present, its signal was
swamped out by the huge peak due to Cr(III). This spectrum
is identical to one of the pristine Zn-Cr LDH. It was
obvious from this experiment that an LDH with only diamag-
netic constituents in the octahedral layers was need for
EPR experiments.

The [LiA12(OH)6][Cl'nH20] synthesis with the PADS-doped
LiOH solution was undertaken to provide just such a
material. The EPR spectra of the material obtained are
shown on Figure 15. Analysis of the top spectrum taken
while the material was still damp gave a correlation time
of 1.8 ns. This value was essentially the same as that
observed in the doped hectorite intersalates maintained a
98% humidity. Drying the LDH material in a vacuum dessi-
cator for 411and overnight produced more anisotropic EPR
signals and correspondingly longer correlation times of
12.0 and 40.1 ns respectively. That the figures for damp
material compared favorably with those of intersalated
hectorite seemed to be a good sign that reaction of the
gallery anions could take place in the LDHs as they had in

the intersalates.

58

0' 2.00

 

Figure 14. EPR spectrum of [ZnZCr(OH)5] [Cl'nHZOI-
Frequency = 9.14 GHz.

 

 

 

 

 

 

 

 

 

J fl
4

—_—’

EPR spectra of PADS-doped [LiA12(OH)6] [Cl'nHZO]
(a). damp material; (b). after 4 hr in vacuum
dessicator; and (c). after 12 hr in vacuum

dessicator. Frequency = 9.14 GHz.

Figure 15.

60

The numbers were not reasonable, however, when the
basal spacings of the species involved were considered.
The °00l spacing of the intersalated hectorite was 31.5 A,
which corresponded to a gallery height of 6 A when the
thicknesses of the main clay layer and two layers of com-
plex cation were subtracted. That a material with a basal
spacing of 7.7 A (gallery height= 2.9 A) could exhibit the
same degree of tumbling freedom seemed unreasonable. There
was no doubt that the PADS species was present in the Li-Al
LDH system since the blue color persisted through numerous
washings and the material gave an EPR spectrum. The answer
to the problem lies in the structure of PADS, shown in
Figure 13. It contains -SO3' groups of the same orien-
tation as those found in DDS“. The PADS dianion would bind
more strongly than the DDS due to greater electostatic
attraction. As the material was dried, surface water was
removed, restricting the freedom of movement of the PADS on
the surface. Even if some of the spin probe made it into
the galleries, its signal was overpowered by that of the
surface species.

To try to alleviate this adsorption problem, another
synthesis of the Li-Al material was attempted. In this
case, the LiOH solution was not only doped with PADS but
also was 0.1 M in NaDDS. It was hoped that the greater
concentration of DDS would block adsorption of PADS on the
surface by blocking edge sites. In this way any PADS

present in the product would be in the galleries. The EPR

61

spectra of the resulting materials are given in Figure 16,
and a comparison of the correlation times obtained for this
sample and the previously—prepared sample is given in Table
9. As can be seen the correlation times for the two simi-
larly treated products were practically identical. Also,
x-ray diffraction of the samples showed that Cl’ from the
AlCl3 solution had been intercalated in both cases. The
DDS concentration was not high enough for the larger DDS
anion to be favored over the smaller C1". The dianionic
PADS should have been preferred over Cl", since dianions
have been found to possess a higher affinity for LDHs than

1 Even with the surface of the LDH covered with

monoanions.
DDS the surface concentration of PADS was larger than that
of the gallery, if any were exchanged at all.

Later experiments cast further doubt on the validity of

the interpretation of this work with Li-Al LDHs. When

Table 9. EPR Correlation Times for PADS-doped LibAl LDH
Materials.

 

 

Materiala Drying Timeb c(nanoseconds)
LieAthI damp 1.8
Li-Al-Cl 4 hr 12.0
LihAl-Cl 12 hrc 40.1
Li-AleDDS damp 1.8
LiFAlflDDS 4 hr ' 11.8
Li-Al-DDS 12 hrc 43.0

 

aPrepared by coprecipitation. Li-Ala-DDS material had LiOH
gynthesis solution 0.1 M in NaDDS.

In vacuum dessicator.
cOvernight.

62

o- 2.00

 

 

 

 

 

 

 

 

 

 

 

a
i
\ b
c
zoo
I!

Figure 16. EPR spectra of an Li-Al LDH synthesized in the
presence of NaDDS and KZPADS (a). damp material;
(b). after 4 hr in vacuum dessicator; and (c).

after 12 hr in vacuum dessicator. Frequency =
9.14 GHz. '

 

 

63

freshly—prepared Li-Al LDH was placed in dialysis tubing to
be washed free of excess electrolyte, only Al(OH)3
remained, as shown by the x-ray diffraction patterns in
Figure 17. The diffractogram of fresh material shows the
expected LDH pattern. When this material was dialysed, the
second diffractogram results. Elemental analysis confirms
that >98% of the Li+ had been removed. Material which had
been hydrothermally treated after synthesis did not lose
Li+ upon dialysis. Gibbsite intercalated with lithium
salts have been prepared110 by heating LiX solutions with
gibbsite. A powder x-ray diffractogram of a material
prepared in this manner from gibbsite and LiCl is shown in
Figure 18. It looks very much like the patterns obtained
from LDHs. The Li+ in such species is removeable by
washing with fresh water. From this data it was concluded
that upon coprecipitation of Al(III) and Li(I) salts, an
intercalated aluminum hydroxide is formed. When this
material is treated hydrothermally, the Li+ ion is driven
into the octahedral sheet. Reactions of this sort111 also
occur in montmorillonite, a smectite clay mineral with an
empty site in its octahedral layer (Figure 3). These later
results would have raised doubts as to the validity of any
correlation times obtained had the PADS been successfully
intercalated.

In summary, the reactions with PADS did not provide a
mechanism to examine anion mobility in LDHs. A majority of

the spin probe was apparently adsorbed on the surface of

7.63 A

“—

 

'11

1

9w“ Mfiwk . WNW”
MflummwflNAWNTfifibflu AWMNHWNNWMAA

4.5.4.8 A

"VA”
1
\NM ., W)“ m ”WWW

25 20 15 10 5

b 1’
«flwhm°v°1’°'¥‘o‘pfil~mvw\/

degrees 26

Powder xaray diffractograms of unaged Li-Al LDH

Figure 17.
filter—washed; (b). dialyzed until

product (a).
free of electrolyte.

65

7.63 A

 

3.80 A

 

 

 

 

 

4.3=4.5 A

 

25 20 15 10 5

degrees 20

Figure 18. Powder x-ray diffractogram of product formed by
reaction of LiCl solution and gibbsite.

66

the LDH in a manner analogous. As a result of the attempts
to synthesize samples of an Li-Al LDH with PADS present,
the variable structure of unaged Li-Al LDHs was discovered.
EPR analysis could be helpful in studying the gallery if a
different probe could be used, one which did no interact
appreciably with the surface, such as an anionic complex of

a paramagnetic transition metal.

B. Nucleophilic Displacement Reactions.
1. Reactions Under Triphase Catalytic Conditions.

One of the main interests in developing the chemistry
of anion exchangers was working toward an effective system
for triphase catalysis. As outlined earlier, the require-
ments for a successful system include anion exchange capa-
city and sturdiness under reaction conditions. The LDH
materials seemed to fit these two requirements. Questions
remained as to the availability of the anions for nucleo-
philic attack under reaction conditions and about the
ability of the solid to be regenerated under these con-
ditions. Also, since the nucleophile for the reaction
would be in a restricted environment, the possibility of
size and/or shape selectivity existed. It was postulated
that access to the anions might be more restricted for
larger, bulkier alkyl halides than for smaller ones. With
these thoughts in mind the triphase reactions were carried

out.

67

The choice of reactions to study was very large, as
outlined earlier. Halide substitution reactions were cho-
sen for study because of their simplicity and because of
expertise developed in the group in previous work of this

type.99

Iodide substitution reactions of alkyl halides
were chosen as the most likely to succeed because aL.the
nucleophilicity of the 1' species is greater than that of
Cl’ and Br“; and b). I“ substituted Zn-Cr LDH provides a
larger basal spacing than Br‘ or Cl' forms and conceivably
more space for a reaction to take place.

The results of the reactions are shown in Table
10. These reactions were carried out for 24 h at 90°C and

were vigorously stirred. The data seemed to indicate that

some of the predicted behaviors were occurring. The larger

Table 10. Conversion of Alkyl Halides by the Reaction Rx +
Y’—9»RY + X‘ Under Triphase Consitions.a

 

 

RX Y“ Solidb % Conversion

n-C4898r I 1 85.4
n~C 5H1 13: I 1 54.9
nI-CS I 1 40.7
H38”}3H3)CH2CH23I I 1 30.4
n-C4flgBr Cl 1 9.2
naCGRl Br c1 1 1.6
D“C4H9 I I 2 88.6
“*CSHIIBI I 2 56.3

 

3Reaction conditions: reaction time: 24 hr; temperature:
90 oC; Y‘/RX mole ratio: 6.0; RX/solid mole ration: 20.0;
[Y ] 2. 0 M; 2. 0 mL toluene; 3. 0 mL H O.

b30113 1: [2n2 Cr(OH) III 2. 3H 0] when Y = I; [Zn Cr(OH)6 1
[C1'2HZO] when Y' 81. Soli 2: [ZnZCr(OH)5][DD§ nHZO].

68

alkyl halides were reacting more slowly than the smaller 1-
bromobutane. Also, the branched alkyl halide, l-bromo-B-
methylbutane reacted more slowly than the straight-chain
counterparts. The conversion of RBr to RCl was slower than
conversion of RBr to RI. The low conversion for the chlo-
ride substitution reactions could be attributed to the
energetic requirements for the replacement of intercalated
C1' by Br“, a process requiring lattice expansion of the
LDH. All of these observations were interpreted as evidence
that anions for reaction were coming from the galleries of
the LDH.

However, the above data were open to another interpre-
tation, one that was shown to be correct when blank experi-
ments were carried out. The same reactions were performed
with no solid in the mixture, but otherwise duplicating the
conditions of the triphase reactions. Table~ll gives the
results of the blanks. A comparison of the numbers in
Tables 10 and ll shows that the conversions are virtually
identical. The conclusion drawn from these data was that
the reactions were taking place much more rapidly at the
organic-aqueous interface than at the solid. The selec-
tivity toward smaller alkyl groups can_be explained on the
basis of decreased miscibility of longer alkyl groups with
the aqueous phase. The observed selectivity toward iodide
substitution reactions over chloride substitutions appar-
ently arises from the greater miscibility with organic

solvents of iodide over chloride.

69

Table 11. Results of Blank Reactions for Halide Exchange.

 

 

Rx I“ % Conversion
nhC4H98r I 86.2
naC H Br I 57.9
CH H}&H3)CH2CHZBr I 31.2
0" 489Br C]. 8.9

 

aReaction conditions: reaction time: 24 hr; temperature:
90 oC; Y'/RX molar ratio: 6.0; [Y-laq: 2.0 g; 2.0 mL
toluene; 3.0 mL H20.

 

The interpretation of these results was that the:sur-
face of the LDH was too hydrophilic to allow interaction of
the organic reagent with the gallery anion under the
reaction conditions. Since it was known that exposure to
NaDDS solution made the surface more hydrophobic, some of
the Cl' form of Zn-Cr LDH that had been so treated was used
in a series of triphase reactions. It was thought that the
I“ would quickly displace the Cl' under reaction conditions
and that iodide substitution would proceed more readily on
the hydrophobic surface. The results of the reactions are
given in Table 10. The conversions with the DDS-treated
materials were slightly higher than those of the plain 1"
form. However, x-ray diffraction showed that even after 24
h at 90 °C the Cl' had not been displaced from the gallery.
Therefore none of the reaction was taking place with the
gallery anions; indeed, none of the iodide even made it
into the galleries. The increase in conversion when the
DDS-treated LDH was used may have been due to increased

interfacial surface area. The surface area of the DDS-

70

treated solid was found to be 15.0 m2/g versus 7.7 mz/g for
the pristine material under air-dried conditions. The
result of this experiment was to transform a material that
was too hydrophilic for access by organic reagent into one
that was too hydrophobic for interaction with the aqueous
phase in the reaction mixture.

Under triphase reaction conditions with.the untreated
LDH the anions were unavailable for reaction with alkyl
halides. The surface properties of the LDH seemed to be
responsible. In order to test this hypothesis, another
series of reactions was undertaken. In these reactions,
the components of the triphase mixture were removed one or
two at a time to study the effect on conversion. The
conditions of reaction and conversions are listed in Table
12. These reactions were carried out under stoiciometric
conditions. A dramatic improvement in conversion was noted
when the water was removed from the system (11.5%
vs.76.6%). This seemed to confirm that water on the sur-
face was blocking access of the organic substrate to the
anions. A series of experiments characterizing the avail-
ability and reactivity of gallery anions in relation to the”

surface properties of the LDHs proceeded from this point.

2. Stoichiometric Biphase Reactions in Toluene.
The probe reactions used in this portion of the
research were the same as those used in the triphase

reactions. Toluene was chosen as the organic solvent to

71

Table 12. Conversions of Nucleophilic Displacement
Reactions of C4H9Br Under Modified Conditions.a

 

Run H20 toluene 1 mmol NaI 1 mmol LDHb % conversion

 

1 yes yes yes no 51.6
2 no yes yes no 0

3 yes yes no yes 11.5
4 no yes no yes 76.6

 

‘31.0 mmol C H Br present initially in all runs. 3.0 mL H20,
3.0 mL toluene present where indicated. Reactions run 24 hr
gt 90 °C.

[ZnZCr(OH)6][I'2.3H20].

 

keep some continuity with previous work in the triphase
area. The experimental procedure was designed to allow the
conversion to product to be followed over the course of the
reactions. Analysis of the conversion data would provide
information about the kinetics of reaction and possibly a
greater understanding of the mechanism of the reactions
[taking place.

Data treatment was simplified by the programmable
plotter/integrator from which the percent conversion could
be obtained directly. A flame ionization detector was used
with this particular GC and meant that all conversions
reported by the integrator were weight percent values. The
weight percentages were converted to molar percentages by a
routine on a programmable calculator before any subsequent
data treatment.

Air-dried (AJLJ [ZnZCr(OH)6][I’2.3 H20] was used in
the initial studies of displacement reactions. The data
collected from these reactions fit a first order kinetic

expression with a high degree of correlation. From plots

72

of '1n(fRBr) vs. time, 1where fRBr=the fraction alkyl
bromide remaining unconverted, pseudo-first order rate
constants were determined. Several kinetics plots are

shown in Figure 19 with the values of Rob and their 90%

s
confidence limits given in Table 13. The data in this

table indicate that the the values of kob did not vary to

s
a significant extent with the length of the alkyl chain of
RBr. The branched l-bromo-3-methy1butane was the slowest
of the reactions, suggesting that there was some steric
hindrance in the reaction mechanism.

In the triphase systems, reactions on the solid were
seemingly inhibited by the presence of water in the system.
Since A.D. Zn-Cr LDHs had water on their surfaces,32 it
seemed reasonable that removing this surface water would
accelerate the bdphase reactions. Previous studies32 had
indicated that all the water could be removed from both the

surface and galleries of the I' Zn-Cr LDH by heating the
material to 150 0C under argon. Treatment at this
Table 13. Calculated Pseudo First Order Rate Constants,

kobs' for Iodide Substitution of Alkyl
Bromides Over A.D. [ZnZCr(OH)6][I'nHH20].

 

 

RBr 105 kobs(s“1)b
labromobutane ’ 3.79 + 0.28
l-bromopentane 3.66 + 0.59
lhbromo-3-methylbutane 2.08 + 0.14
lbbromohexane 2.89 + 0.34
lbbromooctane 3.03 + 0.21

 

aReaction conditions: temperature: 90 0C; 0.453 g (1.00
mmol) [ZnZCr(OH)6][I'2.3HZO]; 1.00 mmol RBr; 3.0 mL toluene;
gampled every 10 min for 1 hr. .

+ 90% confidence limit based on linearity of data fit.

73

 

 

0.3:;
0.2:
I
0.1 ‘
O
A
O
00!] 1— ; : ; g t
0 600 1200 1800 2400 3000 3600
time (s)

Figure 19. First order kinetic plots for iodide substi=
tution of alkyl bromides over [ZnZCr(OH)6]
[11.31120]: labromobutane;

lhbromohexane;
and l-brom0h3kmethylbutane.

74

temperature apparently did not cause any dehydroxylation of
the octahedral layers. With these thoughts in mind, some
of the I" Zn-Cr LDH was treated at 150 °C for 2 h under Ar.
Part of the IR spectrum of this material, along with that
of an A.D. sample, is given in Figure 20. A comparison of
the region from 4000 cm"1 to 1000 cm’1 shows the virtual
disappearance of the bending modes of water around 1600 cm‘
1 after heat treating. The OH-stretching region of the
spectra, due mainly to the hydroxide groups of the octahe-
dral layers, remains almost identical upon heat treatment.
When this OJL.material was used as the iodide source
in the conversion of alky bromides, the kinetics results
were similar to those obtained with the A.D. material. The
data could be fit with a high degree of linearity to a

first order rate expression. The values of for Rob ob-

s
tained for CLD. materials are given in Table 14 along with

their 90% confidence limits. The values obtained using

OJL LDH are approximately 3-4 times those of their AJL

Table 14. Pseudo First Order Rate Constants, kob , for
Iodide Substitution of Alkyl Bromides Sver O.D.
[ZnZCr(OH)6][I' 01120].a

 

 

RBr 10S kobs (s'l)b
labromobutane 13.2 t 0.92
lebromopentane 12.7 i 0.43
lhbromoa3amethylbutane 9.40 1.0.26
lhbromohexane 12.1 i 0.23
1rbromooctane 9.39 i 0.11

 

aReaction conditions: 90 0C; 1.00 mmol RBr; [Zn Cr(OH)6]
[I]; 3.0mL toluene; sampled every 10 ndJi for l r.
b+ 90% confidence limits based on linearity of data fit.

75

L
r
1

4000 3500 3000 2500 2000

{

 

L J I 1 ' L
F t r ' '

2000 1800 1600 1400 1200 1000

A.

wavenumber (cm'l)

Figure 20. IR spectra from 4000 cm'1 to 1000 cm‘1 of (a).
[thCI(OH)6][I°2.3HZO]; (b). [ZnZCr(OH)6]
[I'NGH20].

76

counterparts. The increase in surface area upon drying the
material was only approximately 25% (7.7 mz/g A.D. vs. 9.9
m2/g O.D.), so increased surface area could not adequately
account for the 3-4 fold increase in rates. This seemed to
confirm the proposition that removing water from the sur-
face did improve access to the gallery anion.

One of the experiments carried out to characterize the
reactions measured the rate constants at different tempera-
tures in order to calculate an Arrhenius activation energy

112

and thermodynamic parameters of activation. The plots

of kinetics data used to calculate activation parameters
for l-bromobutane are shown in Figure 21. The iodide
substitution of l-brompentane and 1-bromo-3-methylbutane
were also carried out at different temperatures. The plot
of In kobs v3.1JW'for 1-bromopentane is shown in Figure
22. The values of the activation parameters for the three

alkyl halides are given in Table 15, along with some values

 

 

Table 15. Activation Parameters for Some Alkyl Halide
Iodide Substitutions at 90 °C.a
RX 1“ source AH‘(kcal/mol) AS’(cal/mol°K) ref
n~C4HgBr LDHb 13.1 1 1.0 ~40.6 1 3.0 this work
nacsallar LDHb 14.5 1 1.3 -36.8 1 3.7 this work
isoscsHllBr LDHb 15.1 1 0.6 -35.7 1 2.7 this work
CH3Br solution 15.8 1 1.0 a7.6 1 3 113
CH CHZBr solution 16.3 i 2.0 516.4 i 6 113
n‘33H7C1 solution 20.3 515.4 114

 

3+ 90% confidence limits for values from this work;

of limits not stated in referenced material.

O.D. [ZnZCr(OH)6][I].

source

 

Figure 21.

77

L l
f r ' ' V

600 1200 1800 2400 3000 3600

time(s)

First order plots for iodide substitution of
lubromobutane over CLD. Zn/Cr/I LDH: (a).
129 °C; (b). 100 °C; (c). 99 °c; (d). 80 °C;
and (e). 50 0C.

78

144»

 

 

‘14 e 1
V V '_1

0.0 1.0 2.0 3.0

 

103(1/T)(K‘1)

Figure 22. Arrhenius activation energy plot, 1n kobs vs.
l/T, for iodide substitution of labromopentane
over O.D. Zn/Cr/I LDH.

79

calculated for similar reactions in solution.]'l3'114

The values.of AH’ foundxfor the reactions over solids
and in solution were similar. This similarity was inter-
preted as evidence that the reactions over the solid were
following a similar mechanism to that in solution, 8N2
displacement. The negative values of AS‘ indicate a coming
together of particles in the transition state, supporting
the picture of a nucleophilic attack. The magnitude of
entropy change was significantly larger for the reactions
over solids as compared with their homogeneous cogeners.
This difference was interpreted as evidence for the
reaction taking place on the surface of the solid. The
arranging of an adsorbed molecule into the special orien-
tation required for reaction on the solid would conceivably
be a more ordered state than the coming together of two
particles in solution.

The activation data argued against another mechanism
that couldbe postulated. That was one in which an equili-
brium concentration of halide ion existed in the toluene
solution and was responsible for the nucleophilic displace-
ment reactions occurring. If this mechanism were
operating, it seems that the ASt data would be approxi-
mately the same as that for the homogeneous reactions.
Instead, the values are much more negative.

The data from gas-solid reactions also argue against a
soluble nucleophile mechanism. As can be seen from the

conversions for these reactions listed in Table 16 the

80

Table 16. Conversions of Gasssolid Iodide Substitutions of
1~Bromobutane Over [ZnZCr(OH)6][I].a

 

 

Temperature(°C) % Conversion
90 38
130 64
140 74
150 80

 

aReaction conditions: He folw rate: 3.0 mL/min; contact
time: 0.20 3; total C4H9Br: 1.0 mmol; 1.0 mmol LDH treated 2
hr at 150 0C in He flow before reaction.

 

reaction proceeds in the gas phase from 90 °C to 150 °C.
That the reactions proceeded in the gas phase at all where
there was no solvent present suggests that the soluble
nucleophile mechanism was certainly not necessary'to ex-
plain the conversions in the solution phase.

Gas-solhfl chloride substitution reactions of lsbromo-
butane were also performed. The conditions employed were
identical with the iodide substitutions except that
[ZnZCr(OH)6][Cl'2HZO] was the solid phase in the reactor.
The conversions are given in Table 17. The values are very
close to the iodide substitution results. Chloride substi-

tution of l-bromobutane under solution-solid conditions

Table 17. Conversions of GaseSolid Chlorinations of
lvBromobutane Over [ZnZCr(OH)6][Cl].a

 

 

Temperature(°C) : % Conversion
90 33
130 60
140 71
150 80

 

aReaction conditions: He flow rate; 3.0 mL/min; contact
time; 0.20 8% total C H9Br: 1.0 mmol; 1.0 mmol LDH treated
2 hr at 150 C under e before reaction.

8]

gave a conversion of approximately 1% after 1 h, compared
with 30-35% iodide substitution conversion. When fluoride
substitution of 1-bromobutane with [ZnZCr(OH)6][F] was
attempted under solution-solid and gas-solid reaction con-
ditions, no conversion was noted. The significance of
these data to the proposed mechanism of reaction mechanism
will be discussed in a later section.

From this evidence, then, it appeared that the
mechanism was one of adsorption of RBr onto the solid,
followed by the rate determining step of nucleophilic dis-
placement, then desorption of product. This chain of

events is illustrated by Scheme 2.

Kads kobs Kdes
____> __—_\. ——.L
RB“:soln‘ RBrads< RIads ‘ RIsoln
Scheme 2

The questions that remained at this point concerned the
relative availability of all the anions in the solid and
the importance of the adsorption properties of the solid to
the course of the reaction. These problems will be
addressed.

In the reactions at 90 °C of alkyl bromides with
toluene as the solvent, pseudo-first order kinetics
behavior was observed for the first hour of reaction. When
reactions were carried out for longer periods of time,
deviations from first order behavior occurred. The first
order plot for extended iodide substitution of l-bromo-

butane is shown in Figure 23. The graph clearly shows the

82

 

 

. o
o
o
0161' .
o
O
A 0.4+
1.:
CD
a:
“-4 o
c 4,
H
.I
0.2«r
4»
0.0 _$ i ‘: +f : f + j'
0 1000 2000 3000 4000 5000 6000 7000 8000
time(s)

Figure 23. First order kinetic plot for the extended
iodide substitution of l-bromobutane over OJL
Zn/Cr/I LDH.

83

deviation from first order behavior. The data from
reactions carried out a higher temperatures also sheds some
light on the situation. An examination of the data given
in Figure 21 for reactions a 100 oC and 120 OC reveals
deviation from first order behavior at values of '1n(fRBr)
of 0.35-0.40, corresponding to conversions of 30-35%. This
was also the region of the plots at 90 °C at which first
order behavior was no longer observed.

Two explanations could be advanced to explain this
behavior. The first has to do with the availability of
gallery anions. It seemed reasonable to assume that as
more of the I’ was consumed and replaced by Br' in the
galleries of the LDH, the remaining 1' would be relatively
inaccessable compared with that near the outside edges of
the particles. Perhaps 35% conversion, the point at which
deviation from first-order behavior became pronounced, was
the value at which access to I' became the rate controlling
factor of the reaction.

This hypothesis was easily tested. Reactions were run
at 90 °C with excess alkyl bromide, but were otherwise
identical to previous reactions. The data from the
reactions run a 2-, 3-, and 4-fold excess of l-bromobutane
plotted as first order data are shown in Figure 24. The
data was not very revealing until it was plotted in a
different manner. Since the stoichiometry of the reaction
was known, as was the initial ratio of RBr to I', it was

possible to calculate the fraction of the initial amount of

‘1n(fRBr)4

Figure 24.

2.4”

1.

 

84

64>

 

 

 

 

I
r I, '
O
I
L O .
I O
O 0
.2A a
1' . ‘ ‘
A
o
A
4» '
O
o
1
0 s 4" fig ? ¢ 1
0 600 1200 1800 2400 3000 3600

time (5)

First order kinetic plot for iodide substitution
of 1-bromobutane by O.D. Zn/Cr/I LDH at RBr/LDH
ratios of I4:l; 03:1; 02:1; A1:1, plotted as
functions of (a). Eln £13; (b) -1n fRBr'

85

1' remaining (fI-). When the data were plotted as -ln(fI-)
vs. time, the results shown in Figure 24a were obtained.
These plots were linear out1x>a consumption of 85-90% of
gallery iodide when a 4-fold excess of RBr was used. These
data suggested that the rate of iodide substitution was not
related to what was going on in the galleries of the LDH.

The linearity of the plots calculated from fI- indi-
cated that the proper statement of reaction kinetics was
that the reaction was pseudo-first order in fI- for the
first hour of reaction. Under the original reaction con-
ditions, where fI- =fRBr' this distinction was not made.
It now seemed that the reaction wduld follow pseudo-first
order kinetics as long as the surface concentration of RBr
was kept approximately constant.

The data supported the other possible explanation for
the deviation from first order behavior after 30-35%
conversion. This explanation had to do with the adsorptive
properties of the LDH. In the first stages of the
reaction, the reactant molecules were competing for ad-
sorption sites on the solid with the solvent. Later in the
reaction, as iodide substitution product was accumulated,
it would begin to compete for adsorption sites also. If
the affinity of the product for the solid were comparable
to that of the reactant, which seemed likely given the
similarities in structure, fewer reactive sites would be
available for the reactant, slowing the reaction.

Knowledge of the adsorptive properties of the solids

86

involved would be useful in understanding the reactions
taking place. Therefore, adsorption studies of the Zn-Cr

LDHs were undertaken.

3. Adsorption Properties of Zn-Cr LDHs.

In general the affinities of solids for vapors can be
determined by measuring adsorption isotherms for adsorption
at two different temperatures. The Clausius-Clapeyron
equation is then applied to calculate an isosteric heat of
adsortion, qst' as shown below:

qst a -R 5(10 P)
(6 (UT) 9 (4)
In this equation, ais the surface coverage. From ad-
sorption isotherms, pressure and temperature data are
entered into the equation for points from the two isotherms
of the same surface coverage. For the isotherms described
here surface coverage was expressed in mmoles adsorbed per
mg of solid. Physical adsorption is a spontaneous process
and as such the free energy change is negative. Since the
change in entropy is always positive when a gas is changed
tmaa more ordered adsorbed state, the enthalpy of adsorp-
tion must always be negative. The values calculated using
by Equation 4 are positive for physical adsorption. The
sign convention was adopted so that the numbers to be dealt
with would be positive. To convert to an isosteric enthal—
py of adsorption one must change the sign on qst'
The adsorption of a vapor can be viewed as a two-step

process, as shown ix: Scheme 3.

87

The value of AHgad is the inthalpy change that is

calculated from the two adsorption isotherms by using the

Clausius—Clapeyron equation. The value which has bearing

 

A(9) ’ A(1) 'AHvap
l
A(1) —7 A(ads) AH ads
A(9) 1 A(016$) AHgads (3 ‘qst)
Scheme 3

on the reactions under study that are carried out under
solution-solid conditions is Afllads' the enthalpy of ad-
sorption of a liquid phase species onto the surface of a
solid. The equations in Scheme 3 may be rearranged to

calculate AHlad as shown in Scheme 4.

 

5

Am ——* Am AHvap

A(9) -__) A(ads) AI"gads

A(1) '—-—9 A(ads) AHlads=AHgads+AHvap
Scheme 4

Adsorption studies of toluene and 1-iodobutane by
[Zn2Cr(OH)6] [£‘nH20] were undertaken. These isotherms for
both the AJL and O.D.Inaterials are shown in Figures 25
and 26. The isotherms for l-bromobutane shown in Figure 27
were measured for adsorption onto the Br' form since
adsorption onto AJL 1' material resulted in reaction to
form l-iodobutane. The solvent source was maintained at 18
°C for those isotherms measured at room temperature (22 °C)
and at 0 °C for isotherms recorded at 0 °C. The po in the
p/po expressions refers to the saturation pressure of the

adsorbate at the temperature of the adsorbent.

88

 

 

 

 

o
6.01
o
O
o O
.0
00
. o
4.0» ° °
0 ‘ °
".4 . O
H o
o I
m .58.
m I
5 2.0::
H
(E) D
Q.
G
H
'0 $00 A 4
Q)
.0
H
O
m
E 6.0“
0)
E
3
H
o A
>
4.00 . I
. I
I 3 . ‘ ‘
I ' ‘ .
2.01 . . ‘
I . ‘
I .1
I...
"A
0.0 ,1 * 1 ‘ i
0.0 0.2 0.4 0.6 0.8 1.0
p/po

Figure 25. Isotherms for adsorption of toluene onto: (a).
A.D. Zn/Cr/I LDH at .0 °C; 022 °C; (b). O.D.
Zn/Cr/I LDH at 'A0 °C; I22 oC.

89

 

 

 

 

600"
4.0"
8 o
2:: O
8 , ° ° -
0‘ . ... o. . . 0
E 2.0 . .. 5 °
r-l .09
o 38
E '°
Q‘ 0
a O
.-4
0.0 4 ‘7
'O
(D
.Q
:4
O
8
m 6.01
(D
E
:3
H A
O
>
4.01
2.00 . .
‘ I I
.1 . I .
:5. . . I
.I
I
a.“ f A i
0.0 0.2 0.4 0.6 0.8 1.0
p/po
Figure 26.

Isotherms for adsorption of l-iodobutane onto:
(a). A.D. Zn/Cr/I LDH at .0 0C; 022 °C; (b).
O.D. Zn/Cr/I LDH at I0 °C; I22 °C.

90

 

 

 

 

6.0
O
4.0
A °.
'U o I
"-0 O
H . ° 9
a - .-
‘ O
m 00 '
E 2.0" .0.
\ 0".
H
‘53 i‘
Q‘
&
H
V 00“ L i1
'5
CD
.0
H
8
,0 6.0»
(U
0)
E
5
0-1
0
>
4.0
200" ‘
I ‘ '
I I . a . ‘ . .- I '
..I..‘A ‘ ‘ ‘
0.“. t 4
0.0 0.2 0.4 0.6 0.8 1.0
p/po

Figure 27. Isotherms for adsorption of l-bromobutane onto:
(a). A.D. Zn/Cr/Br LDH at 90 °C; .22 °C; (b).
O.D. Zn/Cr/Br LDH at 10 °C; I22 °C.

91

Isosteric heats of adsorption were calculated for these
systems. For a given specific adsorption volume at 0 0C, a
value of pressure was interpolated from the room temper-
ature isotherm. This procedure produced a (p,V) point from
each curve so that a value of qst could be calculated at
that point. The procedure was repeated for each point on
both curves to generate pairs of (p,V) data. These values
were plotted as so-called heat curves as function of V/Vm,
the fraction of monolayer coverage. The values of Vm' the
volume of adsorbate required for monolayer coverage of the
adsorbent, were calculated for each adsorbate by using the
isotherms. .Adsorption isotherms can generally be broken
down into three regions (Figure 28): (1) a non-linear
region at low values of p/poj (2)1ailinear region at mid—
range values of p/po; and (3) another non-linear region at
high partial pressures. The transition point between the
first two regions, the so-called Point B, has been
postulated117 to be the point at which monolayer coverage
is completed since the uptake of adsorbate is changing in
character most rapidly at this point. Point B analysis has
proved successful in calculating monolayer coverages,
yielding good agreement in surface area measurements with

N2 B.E.T. measurements.115' 117

The agreement was good for
a variety of adsorbates on different adsorbents. The
values for the present study were estimated from the iso-
therms at the two temperatures studied and averaged. The

heat curves are given in Figures 29-31. Plots of Afilads

volume adsorbed

 

Figure 28.

92

Point B

An adsorption isotherm showing the location of
Point B, the attainment of monolayer coverage.

93

 

 

 

 

101
A...
(.111 7‘ A
A
A
50> ‘
A
001’
1 o
o
3 I
e
D 91
8 ‘ o
as r 0
tn) . . I
m . o
O
o
-51} I
o
o
O
.01
~10 ; : ’ t 1 : 1 fl
0.4 0.8 1.2 1.6
V/Vru
Figure 29. Heat curves, qst vs. V/Vm, for adsorption of
‘AOD. .OID.

toluene onto [ZnZCr(OH)6][I°nHZO]

94

 

 

. .
I . .
O
220 . I
I
I O
‘P
18‘?
. I
I
'3' ' :
d
E 14 P
\
H
‘0
O
.2
V I
u 0 “.‘. ‘
m A
A A
‘ A
100
O
A
1* A‘A
A
A
A
64-
; ‘1— 1 *1 + 4
0.6 1.0 1.4 1.8
v/vm

Figure 30. Heat curves, qst vs. V/Vm for adsorption of
1-iodobutane onto [ZnZCr(OH)6][I'nHZO] AA.D.
OO.D.

qst(kcal/mol)

96

 

 

‘A
10-r A. A
A
A.
‘ A A
1‘ ‘
A
I

5.. A

A

I
I
A
1 I
I
I
I
01> .
I
I I .
I
I

I

'5 ‘fi VP ‘1 1 .L 1* 1
0.6 1.0 1.4 1.8
V/Vm
Figure 31. Heat curves, qst vs. V/Vm, for adsorption of
AA.D.

1-bromobutane onto [ZnZCr(OH)6][I°nHZO]
O O.D. ‘

 

 

1)-
l60
.0
0- O.
I
I
I
I
8*. . I
L ‘ °
I
I
I
E ‘ ‘ ‘ .0
E . . I
b I
(U A .
3 “f I‘_
To ‘ .
. A‘A A
'3 I ‘ “ A‘IA
P4
3:
G
-31
0.4 0.8 1.2 1.6
V/Vm

Figure 32. Plot of the enthalpy of liquid adsorption,
Afllads' vs. V/Vm for toluene adsorbed onto

[ZnZCr(OH)6][I'nH20] IA.D. IO.D.

97

 

 

121+
‘ A
I
I
8.»
I I A
I
I
I
I
41 I ‘ I
I
I
A . I
H n
2, '0'.
1b
:5 I '0 U
I
1’ ' - ' g '
. I.
.34 ‘IIIA~ ’. . .
(U I .I
H
:1:
Q . I
-841 I
I I
~12‘L
I
I
0 I
I
I I
-16+r .
0.4 0.8 1.2 1.6
V/vm

Figure 33. Plots of Afllads vs. V/Vm for adsorption of
l-iodobutane onto [ZnZCr(OH)5][I'nH20] OA.D.
O O.D.; and l-bromobutane onto [ZnZCr(OH)6]
[I'm-120] IA.D. AO.D.

vs. V/Vm are given in Figures 32 and 33.
From the plots it can be seen that the values of qst

are sometimes negative, giving a positive AHgad Even

where‘Aflgad is sometimes

is negative, the value of AHlad

S 5

-positive once AHva is added in. For simple physical

P
adsorption, these values are not possible. When positive
values for these parameters are calculated from adsorption
isotherm data, it means that an activated chemisorption is
taking place rather than or along with physical adsorption.

Most often, chemisorption is thought of as formation of
directed chemical bonds between a solid and an adsorbed
species. The most common example of chemisorption are
found in catalytic systems where simple gases such as H2,
CO and C2H4 are involved. These gases can be chemisorbed
by a variety of metals and oxide surfaces, either forming
covalent bonds through interactions involving pi-electron
systems or empty orbitals (CO, C2H4), or through dissoci-

ation and bonding (H2),113:119

Chemisorption may also
involve the formation of ionic bonds, or simply dipolar
interactions stronger than the van der Waals interaction
responsible for simple physical adsorption.118

Simple physical adsorption is an exoergonic process
with essentially no activation energy. The process is
thermodynamically controlled, so a cooler sample should
always adsorb more vapor than a warmer one. A plot of the

volume of toluene adsorbed onto O.D. Zn/Cr/I LDH as a

function of pressure is given in Figure 34. The source of

 

99

 

A 6IG‘b
'o
"-1
H
0
U)
A

61
E
b
E 4.0" ‘
V‘ I
a A
H A
v . . I
'00, g A. . I
.0 I '
3 2.0» . ' .
In I I
'0 A
(U I..A
o l‘
5 1
H
o
>

0.0 A f Ar 7L 4' v:

0.0 2.0 4.0 6.0
p(torr)
Figure 34. Adsorption of toluene onto O.D. [ZnZCr(OH)6]

[I'nHZO] as a function of pressure at I0 °C
I22 °C.

100

the positive values ofAHgadS calculated for the adsorption
of toluene onto OJL.Zn/Cr/I LDH can be seen in this plot.
For the range of pressures illustrated here, the warmer LDH
sample adsorbed more. Therefore, for a given value of V/Vm
the pressure of toluene over the solid is greater for the
cooler sample than for the warmer one. These values, when
plugged in the Clausius-Clapeyron equation produce the

positive values of AHgad The greater adsorption at

3'
higher temperature is indicative of a chemical reaction
taking place upon adsorption. In a given amount of time, a
reaction taking place at a higher temperature will produce
more product than one at a lower temperature. The product
in this case is a chemisorbed molecule. A physical inter-
pretation of the activation of the solid for adsorption
will be suggested in a later section.

Values of qst for physical adsorption are usually close

to the value of AHva for a particular adsorbate. The

P
values of qst for l-iodobutane on O.D. Zn/Cr/I LDH are in
the vicinity of 20 kcal/mol at V/Vm values near 1.
Calculated values of qst of this magnitude are usually,
though not always indicative of chemisorption.119 The
difference between this type of chemisorption and that
discussed earlier is the activation barrier. So-called
activated chemisorption has au1 activation barrier to
adsorption which is surmounted by significantly more

molecules at 22 0C than at 0 OC. The activation barrier

for the second type of chemisorption observed, like that of

101

l-bromobutane on LDH, is so low that there is no signifi-
cant difference between the number of interactions which
occur at 0 °C and 22 oC.

and AH9

The values of AHI calculated by using

ads ads

isotherm data are not valid in an absolute sense when
activated chemisorption occurs. A true value of the heat
of adsorption and chemisorption could be obtained calori-
metrically were such a device available.

Some of the data from the adsorption experiments is
useful as given. The curves for the adsorption of l-bromo-
butane and l-iodobutane on AAD. material, shown in Figures
26 and 27, represent physical adsorption for much of the

curve. The calculated values of AHlad for the two species

5
are very close and help confirm the speculation about the
deviation of the iodide substitution reactions away from
first order behavior. Since the affinity of the solid for
the two compounds is approximately the same, when the
concentration of RI reaches a significant fraction of the
RBr concentration, RI is able compete for adsorption sites
on the solid. The accumulation of product on the surface
would inhibit reaction and give the observed deviation from
first order behavior.

The data for the alkyl halides on O.D. material is
harder to interpret. 'Thelnagnitude ofAHladS for l-iodo-
butane on the O.D. material puts it in the realm of chemi-
sorption and indicates a great affinity of the solid for

the alkyl iodide. The positive value of AHlad for l-

S

102

bromobutane indicates that chemisorption is also occurring
here. Some idea of the relative affinity of the solid for
these molecules can be inferred from the adsorption iso-
therms. At room temperature, the O.D. Zn-Cr LDH adsorbed
2.05x10‘4 mmol/mg of l-bromobutane at p/po = 0.90 while
2.27X10'4mmol/mg of l-iodobutane was adsorbed at a partial
pressure of 0.92. These data indicate that the affinity of
the solid for the two compunds was about the same.

The differences in reactivity between the AmD. and CLD.
materials are harder to reconcile by using the adsorption
data. The AND. material adsorbed more of each of the
adsorbates than did the O.D. material (Figures 25-27).
This observation was surprising given the larger surface
area of the O.D. LDH vs. the A.D. material. The so-called
AWD. material which was used for adsorption studies was
somewhat different from that used in the reactions. The
reaction material was air-dried solid scraped off a glass
sheet and used in the reaction mixture without further
treatment. The material referred to as "AJL" that was
used in adsorption studies was outgassed at room temper-
ature under le0'3 torr before adsorption commenced. This
treatment removed loosely-held surface water that would
otherwise have volatilized during adsorption measurements,
resulting in inaccuracies in pressure measurements. Also,
water would have condensed in the reservoir of adsorbate,
causing contamination of the liquid being used. The out-

gassing procedure resulted in a 4.4%_weight loss, corre-

103

sponding to loss of 1.1 water molecules per mole of LDH out
of the original ZJLJZ Of the original 2.3 waters, 0.3
were considered surface-bound, with 2.0 gallery waters.
Therefore, not only surface water was removed, but also
some gallery waters from near the edge of the particles.
The surface of the AmD. LDH used in reactions was different
from that of the "AJLN solid used for adsorption studies.
The difference in surface composition may have lead to
results that are hard to reconcile with observed reactivity
trends. The importance of these observations will be dis-
cussed later in relation to a proposed detailed reaction

mechanism.

4. Reaction and Adsorption Studies in Other Solvents.
Studies of other solvent were undertaken to elucidate
the detailed mechanism of reaction and to ascertain the
importance of surface properties. The CLD. solid material
was used in these studies because the solid used in ad-
sorption studies could be treated to give a material very
similar to that used in the reaction mixtures. Zn/Cr/I LDH
treated at 100 0C for 90 min under le0'3-1Xl0'4 torr
experienced an average weight loss of 7.5%, corresponding

32 This loss

to 1.9 water molecules of the original 2.3.
compares to a loss of 8.0-8.5% upon treatment at 150 0C
under argon. The slight difference in composition was

accepted because more severe heating in a vacuum caused

partial dehydroxylation of the solid as indicated by color

104

change (lavender-gray to green) and weight loss.
Adsorption studies of chlorobenzene, n-octane and
carbon tetrachloride were undertaken in the same manner as
in the experiments just described. As before, fresh
samples were used for each isotherm. The isotherms were

obtained and from these plots of AHlad vs. V/Vm were

s
generated. These plots are shown in Figures 35 and 36.
The heat curve of chlorobenzene is very similar to that of
toluene, showing maxima and minima at the same values of
partial monolayer coverage. Octane also showed activated
chemisorption behavior, though judging from the magnitude

of the positive value of AH1 , not to as great an extent

ads
as toluene and chlorobenzene. Carbon tetrachloride showed
no signs of chemisorption, based on the heat curve that was
generated.

An additional experiment was performed with all the
samples used for adsorption experiments. After the final
point on the adsorption isotherm was recorded, the solvent
source was closed off. The system containing the solid was
evacuated to 1X10“3 torr for 5 min. The amount of adsor-
bate remaining on the solid after this treatment was calcu-
lated from the position of the quartz spring. The amount
of adsorbate remaining on the solid after the treatment for
O.D. materials at 22 0C is given in Table 18 along with
their maximum measured capacities at the partial pressures

given. This crude desorption experiment provides another

measure of the strength of interaction between the solid

105

12+

AHlads(kcal/mol)

 

Figure 35. Plots of AHlads vs. V/Vm for I octane and
Icarbon tetrachloride adsorption onto O.D.

[ZnZCr(OH)6][I°nH20].

«J.-

106'

 

 

.10
121 O.
I
I
I
‘ I
I .II
? z . A ‘ . I
A ‘g‘
:. . .
g 4* I‘I
\. I ‘l.
3 ’ I I
U
.3
3 ad.
10
0-1
I
<3
{1.4-0
.581»
0.4 0.8 1.2 1.6
V/Vm

Figure 36. Plots of ‘Hlads vs. V/Vm for adsorption of
Ichlorobenzene and Itoluene onto O.D.

[Zn2Cr (0H) 6] [I'flHzO] .

107

Table 18. Maximum Adsorption Capacities and Desorption
Experiment Results for Various Adsorbates on OJL

[ancr (0H) 6] [I] .

 

 

Ma imum Capacitya Af er Desorptionb

Adsorbate (1a mmol/mg solid) (10 mmol/mg solid)
toluene 4.03 1.21
lliodobutane 2.27 6.58
lubromobutane 2.05 ' 6.63
chlorobenzene 4.93 ’ 1.33
nnoctane 2.11 9.13
carbon tetrachloride 2.41 0.G5

 

aAt 22°C. Value corresponds to the last point on the
gdsorption isotherm (p/p = 9.88b0.92).

Evacuated for 5 min at 2X10‘3 torr after last adsorption
point.

 

and adsorbate. The amount of adsorbate remaining after 5'
min is.dependent on the rate of desorption, the more
tightly held adsorbates being removed more slowly than
other more loosely bound molecules. These data will be
utilized in the interpretation of the reaction results to
be discussed next.

Iodide substitution reactions were carried out in these
solvents under the same conditions used for the reactions
in toluene. l-Bromobutane was the substrate in these
reactions in which LEG mmol O.D. solid and 1.00 mmol 1-
bromobutane were allowed to react in 3.0 ml solvent at 90
°C. The values of kobs for these reactions are listed in
Table 19.

The rate constants for iodide substitutions in toluene
and chlorobenzene were nearly identical. This similarity

was not surprising given the similarity in adsorption iso-

therms and heat curves of the two adsorbates. The results

108

Table 19. Pseudo First Order Rate Constants for Iodide
Substitution of lsBromobutane with OJL Zn/Cr/I
LDH at 90°C in various Solvents.a

 

 

Solvent 104kobs(8“1)b
toluene 1.32 i 0.09
chlorobenzene 1.34 i 0.10
neoctane 10.4 i 1.0

carbon tetrachloride 2.89 i 0.17

 

a1.00 mmol lsbromobutane; 1.00 mmol [ZnZCr(OI-I)6][I] (O.D.);
.0 mL solvent. ‘
i 90% confidence limits.

 

of adsorption measurements of octane could also be corre-
lated to reactivity. Evacuation after adsorption showed
that a much smaller amount of octane remained on the sur-
face than toluene (1.31X10'5mmol octane/mg solid vs.
1.21Xl0‘4mmol toluene/mg solid), indicating that adsorbed
octane was more easily removed. Since the octane was more
easily displaced, the rate of reaction should be faster
since more 1-bromobutane could be adsorbed on the surface.
The reaction which did not fit the adsorption data was that
carried out in carbon tetrachloride. There was less carbon
tetrachloride left on the surface after evacuation than any
other adsorbate»(5.06Xl0’6mmol CC14/mg solid). This obser-
vation was consistent with the observation that no acti-
vated chemisorption was noted in the adsorption of carbon
tetrachloride, indicating that there was no extra energy
barrier to desorption either. The equilibrium concen-
tration of carbon tetrachloride on the surface should have
therefore been lower than that of any of the-other

solvents, producing a faster reaction. Ixafact, the rate

109a

constant for reaction in carbon tetrachloride was lower
than octane and not much above toluene and chlorobenzene
(Table 19). An activation energy study of these reactions
was undertaken to help resolve the discrepency inadsorp-
tion and reaction results.

Iodide substitution of l-bromobutane was carried out at
several temperatures in each of the solvents. Rate con-
stants were obtained for the initial linear portion of the
reactions and Arrhenius plots prepared. The calculated
values of the activation parameters for reactions in the
different solvents are given in Table 20. The activation
parameters for reactions in toluene and chlorobenzene were
nearly identical, as expected. The values of 48* for
reactions in toluene, chlorobenzene and octane were essen-
tially the same, perhaps indicating a similar structure in
the transition state of the reaction. The activation para-
meter values obtained for reactions in carbon tetrachloride
were different from the others. The enthalpy of activation

is lower than that of the other solvents, but the entropy

Table 20. Activation Parameters for Iodide Substitution of
l-Bromobutane at 90 °c by 0.0. [ZnZCr (om61 [I]
in Various Solvents.a

 

 

solvent 4H*(kca1/mol) AS’(cal/mol°K)
chlorobenzene 13.1 1 1.1 -38.7 i 3.5
n-octane 11.7 1 0.9 -38.4 i 2.9
carbon tetrachloride 8.6 i 0.8 -49.6 i 3.2

 

aFrom Arrhenius plots. Reaction temperatures 50-120 °C.
1.00 mmol 1-bromobutane; 1.00 mmol LDH; 3.0 mL solvent.

109b

of activation is more negative, indicating perhaps that
more order was required in the transition state. A
detailed mechanism in which these points are addressed will
be suggested after a section dealing with the thermo-

dynamics of halide exchange reactions over LDHs.

5. Thermodynamics of Reaction.

The thermodynamics of the halide exchange reactions can
be examined by looking at the individual steps involved and
analyzing the known data. The individual steps for
solution-solid reactions, along with expressions for the
enthalpy change associated with them, are given first,
followed by those for gas-solid reactions.

(a) Rx(solv)—’ RX(ads) (5)

AHa = Afllads(Rx) ' AHsolv(RX) ’
AHl

adS[(solvent)LDH(Y)] (6)

where AHl (RX) is the enthalpy of adsorption of liquid RX

ads
onto the LDH, AHsolv(RX) is the enthalpy of solvation of RX

in the solvent utilized, and AH1 [(solvent)LDH(Y)] is the

ads
enthalpy of adsorption of liquid solvent onto Y-substituted
LDH. This model assumes that the surface of the LDH is
initially'saturated with solvent molecules. This satur-
ation a likely occurrence given the excess solvent in the
system and because the solvent was always added to the
solid before the bromobutane.

AHb a DCX " DCY + (Aflladswy) ' AHlads(RX) (8)

110

where DCX and DCY are the bond dissociation energies of
carbon-X and carbon-Y bonds, respectively. The AHlads
terms must be added to account for any differences in the
enthalpies of adsorption of the product and reactant. The
conversion of the substrate is accompanied by the conver-
sion of the LDH.
(c) LDH(Y) —-) LDH (X) (9)
ABC = 24.2 kcal mol-1 A’1[Ad901{LDH(X-Y)] (10)
where Adaal{LDH(X—Y)} = [d091{LDH(X)} — dggl{LDHun}] and
24.2 kcal mol'1 A'1 is a calculated value for the change in
enthalpy for changing the gallery height of LDl-Is.1]'8 The
difference in energy for different size gallery anions
results from changing the distance between the centers of
positive charge in the octahedral layer and the anions.

The final step in the reaction sequence is the desorption

of product and solvation in the reaction mixture.

65d = -AH1ads(RY) + AHsolv(RY) +
Afllads[(solvent)LDH(X)] . (12)

Totalling the enthalpy terms for reaction in the solution
phase gives.
Aern = -AHsolv(RX) + DCX - DCY + 24.2[AdMl{LDH(x-Y)}]
+ AHsoivaY) + {AHlads[(solvent)LDH(X)] -
AHlads[(solvent)LDH(Y)]} (13)
For gas-solid reactions, steps (b) and (c) are the same
as those for solution phase» The initial steps and final

steps are slightly different.

111

Ana. = Augadsmm = -AHvap(RX) + Anladsmm (15)

and. = -AHgads(RY) = AH (RY) - AHI

vap um (17)

ads

The sum of the enthalpies for vapor phase reaction is

Ann“. = -AHvap(RX) + DCX - DCY + 24.2[Ad001{.LDH(X-Y)}]

+ AHvap(RY) (18)

If the enthalpies of solvation of RX and RY are assumed to
be equal, and the enthalpies of solvent adsorption onto
LDH(X) and LDH(Y) are close to each other, the expression
for the solution-solid reaction reduces to the middle terms
concerning the bond strengths and energy change of the LDH.
For the gas-solid reactions a similar value of the enthal-
pies of vaporization can be assumed.62 Eliminating those
two terms gives the same expression as for the solution-
solid reaction, namely

Aern = DCX - BC! + 24.2[Adggl{LDH(X-Y)}] (19)
The values of (DCx - DCY) and 24.2[Ad601{LDH(X-Y)}] for
some of the reactions studied are listed in Table 21. The

importance of the thermodynamics in relation to the

Table 21. Values of D - DCY and 24.2 600 for Various
Alkyl Halide Exchange Reactionslkx + LDH(Y)—e»RY

 

 

+ LDH(X).a
RX Y‘ DCX - DCYb 24.2Adaol(X—Y) Total
RBr I 17.0 -11.0 6.0

 

aAll values in kcal/mol.
bBased on bond dissociation energies from reference .

112

proposed detailed mechanism will be discussed next.
5. Proposed Detailed Mechanism.

A.detailed mechanisnlof the reactions under study can
be proposed which is consistent with the results obtained
from the present study. The simple mechanism proposed in
Scheme 2 will be presented in light of all the data col-
lected and discussed in some detail.

The first step of the reaction is adsorption of the
reactant alkyl halide. This adsorption occurs after dis-
placement of a solvent molecule from the surface in
solution-solid reactions, but occurs unhindered in gas-
solid reactions. The next step of the reaction is nucleo-
philic attack by a gallery ion. After this step, the
product alkyl halide is either retained on the solid or
replaced by solvent or by reactant alkyl halide.

One of the questions surrounding the first step of the
reaction concerns the nature of the adsorption of solvent
and alkyl halide. The interaction of various adsorbates
with hydroxyl surfaces has been studied in det:ail.12]"122
IR data obtained from the study of substances adsorbed on
porous Vycor glass indicate that even nonpolar molecules
such as alkanes interact with hydroxyl groups, shifting the
O-H stretching frequency to lower energy; The magnitude of
the'O-H band shift is a direct indication of the strength
of interaction between the adsorbate and the hydroxyl
group. The results indicated that, for a series of similar

compounds, such as a aromatic compounds or alkanes, the

113

adsorbate with the lowest ionization potential caused the
greatest shift in the hydroxyl band, indicating the
strongest interaction. Detailed data of this type cannot
be obtained for LDHs due to the presense of broad, overlap-
ping bands in the'hydroxyl stretch region. The data from
the other studies can be applied to LDHs.

The data in the published reports indicate that n-octane
should bind more strongly than carbon tetrachloride,121 as
is borne out in the present study since more octane than
carbon tetrachloride remained on the surface of the solid
after desorption at 1X10"3 torr. Toluene is bound more
tightly than either alkanes or carbon tetrachloride as
would be expected from its lower ionization potential [8.82
eV vs. 11447(CC14); 9.90 eV’(octane)].62 One of the pd
orbitals is also of the proper symmetry to overlap the
sigma bonding orbital on the hydroxyl H-aton1(Figure 37).
Chlorobenzene should bind in the same manner, but the
interaction will be slightly weaker due to its slightly

larger ionization energy (9.07 eV).62

The adsorption of
the two aromatic species was similar according to the
adsorption data from the present study. There are few data
available on the adsorption of alkyl halides on hydroxyl-
groups, but the polar halide group should provide a site
for binding to hydroxyls. The ionization potentials for
the two alkyl halides of interest indicate that l-iodo-

butane should bind more strongly than 1-bromobutane (ioni-

zation potentials: 10.13 eV C4H98r; 9.21 eV C4H91).

114

 

Figure 37. Possible bonding interactions of hydroxyl
group orbitals with an aromatic system. (a).
e29 + CFO“ - no net overlap; (b). e2g + H29 -
p-orbital too high in energy; (c). 0.5 e2g +

C7+OH - probable interaction.

115

The entire surface of LDHs is covered with hydroxyl
groups, including the exterior basal planes, the edges of
the particles and the surface of the hydroxide layers
within the galleries. In the LDH samples that were adsor-
bents in the present study, surface water and some gallery
water, probably from near the edges of the particles, had
been removed prior to the adsorption studies. The removal
of water allows the hydroxyl groups to interact with adsor-
bates. It seems likely that adsorption takes place not
only on the exterior surface of the LDHs but also within
the galleries. Not only are the gallery hydroxyl groups
available for interaction, but the net positive charge of
the octahedral layers is concentrated in the galleries by
the presence of 2 charged layers. Polar groups, such as
the halide groups of the reactant and product molecules,
would be attracted to the positive charge and would bind
through ionic-dipole type interactions. Adsorption of the
halide group of a reactant alkyl halide is probably neces-
sary for reaction to occur. The results of the experiments
will be discussed with that assumption in mind.

(aL. Under solution-solid conditions, iodide substi-
tution of alkyl bromides occurs at a much greater rate than
chlorinations even though chloride substitution is thermo-
dynamically favored over iodide substitution (Table 21).
Under gas-solid reaction conditions in the temperature
range of 90 °C to 150 °C, chloride substitution and iodide

substitution of l-bromobutane by LDH gallery anions give

116

similar conversions (Tables 16 and 17). Fluoride
substitution of vapor phase l-bromobutane does not occur at
150 °C, however. The pattern of reactivity is related to
the gallery heightq the free space between the octahedral
layers, of the LDHs involved. This pattern is consistent
with the assertion that the reactant bromide group must
insert into the gallery in order for reaction to take
place. The O.D. iodide-exchanged LDH has a gallery height
of 3.4 A (6001 - 4.8 A, the approximate thickness of a
brucite sheet). The chloride exchanged form has a gallery
height of‘2.7 A; the fluoride LDH, 2.5 A. The diameter of
the bromo group of l-bromobutane lies somewhere between the
non-polar diameter of 2.28 A and the ionic diameter of 3.90
3.59 Assuming an intermediate value of 3.0 A for the bromo
group diameter, the gallery heights of the chloride and
fluoride LDHs would have to expand to allow the bromo group
to enter. The iodide gallery would admit the bromo group
without further expansion. If the estimate of 3.0 A is a
reasonable one for the bromo group diameter, the energy
required to expand a chloride gallery to accomodate bromo
group is approximately 7 kcal/mod (24.2 kcal mol"1 A‘1 X
0.3 A). In order to expand the fluoride lattice the
required 0.5 A, approximately 12 kcal/mol is required. The

value of AHgad for l-bromobutane is 6-7 kcal/mol more

S

exothermic at reaction temperatures than AHlad The dif—

S.

ference is AHva of l-bromobutane which is liberated when

P
the vapor condenses on the solid prior to adsorption. The

117

extra energy given off upon vapor phase adsorption is
enough to activate the chloride-substituted LDH in the gas-
solid reactions. Recall that chloride substitution did not
proceed at an appreciable rate under solution-solid
conditions. The extra energy of adsorption is not enough
to activate the fluoride substitution reaction since it is
not enough to expand the fluoride gallery enough to(allow
insertion of the bromide group.

There is a parallel between these observations and the
behavior of LDHs in gallery anion replacement reactions in
aqueous solution. When the reactions are carried out, the
platelets of LDHs do not separate and disperse as do those

25 The anion ex-

of smectite clays under going exchange.
change reactions apparently involve the adsorption of the
exchange anion to the edge of the platelet, incorporation
into the gallery and expulsion of the starting anion from
the gallery. The new anion is then mixed into the interior
of the gallery, away from the edge of the platelet. When
the exchange anion is larger that the anion present in the
galleries, exchange is accomplished only by use of a large

31'32 These con-

excess of anion and high temperatures.
ditions are apparently necessary to overcome the activation
energy of exchange, the expansion of the gallery to accomo-
date the larger anions. Exchange of smaller anions and
multiply-charged anions into LDHs is more easily

accomplished, the reactions proceeding rapidly at room

temperature with“, moderate concentrations of anion in

118

solution.

(b). The values of the entropy of activation for
reactions over the solid material are more negative than
those for reactions in homogenous solution. The ordering
of the system for reactions is greater than that in
solution. Not only are two particles required to come
together to make the reaction proceed, but the arrangement
of molecules needed to for the reaction on the solid is
more specific than that required in solution. The arrange-
ment must allow the halide group of the alkyl halide to be
partially extracted while the gallery anion moves to
replace it.

(c). The values of‘AH' calculated for reactions over
the solid are of the same order of magnitude as those of
homogenous reactions, but are consistently lower. This
observation reflects the assisted bond-breaking of the C-X
bond in the gallery region. Part of the AH' of a nucleo-
philic substitution reaction goes into breaking the bond
between the attacked carbon and the leaving group. When
the bond is weakened by interaction with another species
the barrier to breakage is lowered.

When AH‘ and AS’ are combined to give AG4 values for
solution-solid iodide substitutions and the homogenous
reactions listed in Table 15, the values are lower for
homogenous reactions than for solution-solid reactions.
The advantages of a lower enthalpy of activation on an LDH

are overshadowed by the entropic changes of the solid

119

system.

.There is another element in the activation energy. The
gallery nucleophile is located in a potential energy well
from which it must move»to react. Reactions that used an
excess of RBr showed that approximately 90% of the inter-
calated iodide was available for reaction. That the
reaction rate depended only on the fraction of iodide
remaining and not its location in the solid (iéh, the
distance from the end of the gallery) suggested that the
barrier to gallery diffusion was surpassed at 90 oC.
Aqueous anion exchange reactions that occur without
particle delamination support the supposition that the
barrier to movement is relatively low. This discussion is
also related to the next point to be considered.

(d). When l-bromobutane was adsorbed onto O.D.
[ZnZCr(OH)6][I] at 22 °C, normal adsorption behavior was
observed. When the same experiment was tried with A.D.
material, a reaction occurred in which the weight of the
material decreased. Analysis of,the adsorbate revealed
that l-iodobutane had been formed in the vacuum line and
had condensed in the reservoir of l—bromobutane that was
open to the vacuum line. That reaction occurred on the
A.D. solid and not on the O.D. solid is at first surprising
since the O.D. material was more reactive in solution
studies. The treatment of the samples must be kept in
mind, however. Before adsorption, the so-called AJL

material was equilibrated at a pressure of le0"3 torr.

120

This treatment caused a weight loss of approximately 4.4%,
enough to remove the surface water and a small amount of
gallery water. In the adsorption studies, the difference
between the samples examined was mainly a difference in
gallery water and not surface water. The solids used in
reaction studies were different in surface and gallery
composition.

The differences in behavior of materials used in ad-
sorption studies was related to the gallery properties.
The d00l spacing of an.AJL.iodide-exchanged Zn-Cr LDH is
8.34 A vs. 8.21 A for an O.D. sample. Assuming an octa-
hedral°layer thickness of 4.8 A, the gallery heights of the
O.D. and A.D. samples are 3.4 and 3.5 A, respectively. The
diameter of the iodide ion is 4.32 A, larger than the gal-
lery height. The octahedral layers somehow distort to fit
more tightly around the gallery anions. The electron cloud
of the anions could also be distorted to decrease the dis-
tance between charge centers in their respective layers.
If the anions are located over a metal center in the octa-
hedral layer, the three hydroxyl groups could conceivably
distort to allow closer approach of anion. The gallery
water of AND. samples apparently props the layers apart to
some extent since the gallery height of an A.D. sample is
greater than that of an O.D. sample. The propping action
of the water molecules is not the same as a column holding
up a roof of a building. The water functions more as a

dielectric. The oxygen of water compensates for some of

121

the positive charge of the octahedral layers, decreasing
the amount of distortion that the iodide undergoes.
Because of the distortion of the anions, they are in
physical "wells" in addition to the potential energy wells
caused by electrostatic attraction. In order for reaction
between a gallery anion and an alkyl halide to occur, the
anion must come out of its physical well, i.e., either the
anion must further distort or the gallery height must
expand. Either process requires energy. 131the reaction
that occurred in the vacuum line, the A.D. sample had a
gallery height that was 0.13 A larger than the O.D. sample.
The difference, which corresponds to a difference of 3.1
kcal/mol, *was enough to inhibit reaction on the (LD.
material while it was permitted on the~AnD. material.

A question that remains concerns the nature of the
activation step in the species that exhibit activatied
chemisorption. The activation barrier for chemisorption
involves some rearrangement of the adsorbent to enable the

interaction to occur. Examination of the AHlad curve for

s
toluene indicates that activated chemisorption occurred in
the early stages of adsorption on A.D. LDH (Figure 32).
The extent of chemisorption was much greater for CLD. LDH.
Since the adsorption interaction presumably takes place
through hydroxyl groups, the activation step at least par-
tially involves preparing the hydroxyl groups for binding
by adsorbate. The preparation may involve breaking hydro-

gen bonds that exist between particles.32 Since hydrogen

122

bonds are weak interactions, a larger number of them could
conceivably be broken at 22 0C than at 0 oC. The greater
extent of activated chemisorption on O.D. samples indicates
that, compared to the AJL material, a greater number of
the hydroxyl groups are unavailable at 0 °C than at 22 °C.
There could be more hydrogen bonding between particles in
OJL.LDHs, or adsorption into the galleries could require
an activation in O.D. solids not required for A.D.
material. This difference in activation required for
chemisorption, when examined in terms of gallery heights,
is additional evidence for adsorption into galleries by
adsorbates.

The absence of chemisorptive behavior of carbon tetra—
chloride also suggests that adsorption into the galleries
takes place. Carbon tetrachloride is unique among the
adsorbates studied because it is the only one too large to
fit into the gallery an iodide-exchanged LDH. The strength
of interaction between carbon tetrachloride and hydroxides
is predicted to be the weakest of the adsorbates examined.
Since the interaction of carbon tetrachloride is only
slightly weaker than that of octane, according to reports,
some chemisorption behavior might have been predicted for
carbon tetrachloride. These data can be interpreted as
evidence for chemisorption into the galleries since carbon
tetrachloride alone doesn't chemisorb onto CLD. LDH.

In the discussion of differences in reaction rates and

activation parameters in different solvents, theigreater

123

rate in octane has been attributed to greater access of
reactant alkyl bromide to the galleries. However, if a
greater number of collisions alone were responsible for the
increased reaction rate, the value of AS' for reactions in
octane should have been less negative than that for
reactions in toluene and chlorobenzene. The value of AS‘
is derived from the A parameter of the Arrhenius expression
and is a reflection of the frequency'of reactions leading
to reaction. Instead, the AS’ values for the 3 solvents
are the same (Table 19). The reactions in octane have a
smaller value of AH’ than those in toluene. 'The value of
AH’ in carbon tetrachloride is even smaller than for
toluene. These values point to the importance of a step
discussed in relation to the first step of the proposed
reaction mechanism. Recall that a solvent molecule must be
desorbed in order for a reactant molecule to adsorb. Since
the adsorption was exoergonic, the solid loses energy when
the solvent molecule is desorbed. The amount of energy the
solid loses is proportionate to the strength of interaction
between the solid and the adsorbate. Therefore, the solid
loses more energy in desorbing toluene and chlorobenzene
than in desorbing octane. Desorbing carbon tetrachloride
requires the least energy since it interacts more weakly
than the other solvents.‘ The energy that is lost by the
solid is energy that could have gone into the activation of
the iodide substitution reaction. Therefore, since less

additional energy is required to activate the reaction, the

124

solvent system which causes the smallest loss of energy
upon desorption should have the smallest value of AH’.

The above explanation accounts for the enthalpies of
activation observed in various solvents. It<does not ex-
plain the anomalous entropy of activation of the carbon
tetrachloride system. One would predict that the reaction
would proceed faster in carbon tetrachloride since the
energy lost upon desorption of the solvent would be smaller
than for any other solvent. The entropy of activation is
approximately 10 cal/mol °K more negative than for the
other solvents, indicating a smaller effective collision
frequency factor. My proposal is that too much alkyl
bromide is adsorbed for the reaction to proceed
efficiently. The concentration of alkyl bromide in the
gallery is very large because of a large amount of
adsorption. The bromide groups from the molecules are very
close to one another. When the C-Br bond is activated by
interaction with the hydroxyl groups of the gallery,
reaction with a nearby activated bromide group becomes the
preferred reaction under the (crowded conditions.
Therefore, even though the solid is activated to a greater
extent than in the other solvents, reaction of one alkyl
bromide with a nearby molecule becomes important enough to
inhibit formation of iodated products. The collision fre-
quency between iodide and alkyl bromide is thereby smaller,

resulting in a rate smaller than was predicted.

125

Conclusions.

(1). The swelling behavior of alkylsulfate-inter-
calated Zn-Cr LDH has been reproduced in Ca-Al and Li-Al
LDHs. The swelling is a general property of LDHs.

(2). Intercalation of the disulfate radical spin probe
into LDHs does not occur because of interactions between
the -SO3 groups and the hydroxyl groups of LDHs. This
interaction is similar to that observed between dodecyl-
sulfate ions and hydroxide surfaces.

(3). The structure of an unaged Li-Al LDH formed by
coprecipitation is variable. The Li+ may be removed from
the structure by extended washings, leaving mixed Al(OH)3
phases. The initial structure of the fresh precipitate is
probably intercalated gibbsite analagous to that of a com-
pound formed by reaction of a LiX solution and gibbsite.110

(4). Iodide intercalates of Zn-Cr LDH are active in
nucleophilic displacement reactions of alkyl bromides.
LDHs from which the water have been removed react 3-4 times
faster than air-dried materials. The reactions occur in
various organic solvents, but do not occur in the presence
of water. The rate of reaction is dependent on the nature
of the solvent utilized. The solvents affect the surface
properties of the solid and influence the adsorption of
alkyl bromides. Desorption of the solvent molecules, a
prelude to reaction, affects the energy content of the

solid and the activation energy of reaction. A mechanism

126

is proposed in which the bromide group is inserted into the
LDH gallery to undergo reaction. Activation of the mole-
cule is aided by the electrostatic attraction of the posi-
tive layers and the partially negative bromide group. The
anomalous activation parameters of reactions in CC14 sol-
vent are explained on the basis of a solvent intercalation
model which decreases the enthalpic requirements of
reaction, but increase the entropic requirements. Chlori-
nation reactions do not occur in solution phase because the
bromide group of the alkyl halides is larger than the
gallery height of the chloride intercalate. The enery
released upon adsorption of the bromide from solution is
not sufficient to expand the gallery enough to allow pene-
tration of the bromide group, a prerequisite for reaction.
131the vapor phase, the reaction does proceed because the
added heat of vaporization is sufficient to activate the

reaction.

REFERENCES

10.

llI'

12.

13.

14.

15.

16.

17.

18.

19.

REFERENCES

Allmann, R. Chimia 1970, 24, 99.

 

Miyata, S.; Kumura, T. Chem. Lett. 1973, 843.

Beck, C.W. Amer. Mineral. 1950, §_5, 985.

 

Wells, A.F. Acta Cryst. 1951, 4, 200.
Ghose, S. Acta Crystall. 1964, fl, 1051.
Brasseur, H. z. Krystall., 1932, §_2_, 195.

Pastor-Rodriguez, J.; Taylor, H.F.W. Mineral. Mag.
1971., §_8_, 286.

Nickel, E.H.; Clarke, R.M. Amer. Miner. 1976, l, 366.
Taylor, H.F.W. Mineral. Mag. 1973, §_9, 377.
Frondel, C. Amer. Miner. 1941 2_6_, 295.

Allmann, R.; Lohse, H.-H. N. Jahrb. Mineral. Monats.
1966, 161.

Ingram, L.; Taylor, B.E.W. Mineral. Mag. 1967, 36,
465. ‘

Allmann, R. Acta Cryst. 1968, 824, 972.

Turriziani, R. "The Chemistry of Cements"; Taylor,
H.F.W., ed.; Academic Press, New York, 1964, 233, and
references within.

Lerch, W.; Ashton, F.W.; Bogue, R.H. J. Research Natl.
Bur. Standards 1929, 3, 715.

Feitknecht, W.; Buser, H. Helv. Chim. Acta 1951, 34,
128.

Feitknecht, W.; Buser, H. Helv. Chim. Acta 1942, _2_5_,
106.

Feitknecht, W. z. Argew. Chem. 1936, 49, 24.

Feitknecht, W. Helv. Chim. Acta 1938, _2_1, 766.
127

 

20.
21.
22.
23.
24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.
37.

38.

39.
40.

128

Ross, S.J.;. Kodoma, H.1Amer. Mineral. 1967, 22, 1036.

 

Kuzel, H.-J. N. Jahrb. Mineral. Monats. 1965, 193.
Kuzel, H.-J. N. Jahrb. Mineral. Monats. 1966, 193.
Ahmed, S.J.; Taylor, H.F.W. Nature 1967, £12, 622.
Allmann, R. N. Jahrb. Mineral. Monats. 1968, 140.

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

Mascolo, G.; Marino, 0. Mineral. Mag. 1980, a, 619.

 

Hunsaker, V.E.; Pratt, P.F. Soil Sci. Soc. Amer. Proc.
1970, 34, 813.

Schalscha, E.B.; Pratt, P.F.; Gonzalez, C. Soil Sci.
Soc. Amer. Proc. 1972, g, 752.

 

Kinniburgh, D.G.; Jackson, M.L.; Syers, J.K. Soil Sci.

 

Soc. Amer. Proc. 1976, 40, 796.

Taylor, R.M. Clay Minerals 1984, l_9, 591.

 

Boehm, H.P.; Steinle, J.; Vieweger, C. Angew. Chem.
Int. Ed. Eng. 1977, 16, 265.

 

Lal, M.; Howe, A.T. J. Solid-State Chem. 1981, Q,
337.

 

Morrison, S.R.; Bonelle, J.-P. J. Catal., 1972, _2_5,
416.

 

Courty, P.; Marcilly, C. "Preparation of Catalysts
III"; Poncelet, G.; Grange, P.; Jacobs, P.A., eds.,
Elsevier, Amsterdam, 1983, 485.

Feitknecht, W.; Buser, H.W. Helv. Chim. Acta 1949, 33,
224.

 

Feitknecht, W. Fortschr. Chem. Forsch., 1953, 2, 670.
Feitknecht, W.; Buser, H.W. Chimia 1951, 4, 261.

Feitknecht, W.; Gerber, M. Helv. Chim. Acta 1942, 35,
121.

 

Feitknecht, W. Helv. Chim. Acta 1942, 22, 555.

 

Feitknecht, W.; Held, F. Helv. Chim. Acta 1944, _2_7,
1498.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.
52.

53.

54.

55.

56.

S7.

58.

129

Miyata, S.; Kumura, T.; Shimada, M. Ger. Offen.
2,061,094; Chem. Abstr. 1971, 75:153442u.

 

Miyata, S.; Kumura, T.; Shimada, M. Ger. Offen.
2,061,114; Chem. Abstr. 1971, 75:119620r.

 

Miyata, S.; Kumura, T.; Shimada, M. Ger. Offen.
2,061,136; Chem. Abstr. 1971, 75:131249j.

 

Miyata, S.; Kumura, T.; Shimada, M. Ger. Offen.
2,061,156; Chem. Abstr. 1971, 75:119619x.

 

Miyata, S.; Kumura, T.; Shimada, M. Japan Kokai
72 42,493; Chem. Abstr. 1973, 78:55527p.

 

Miyata, S.; Kumura, T.; Nosu, T. Japan Kokai
73 29,477; Chem. Abstr. 1974, 80:85256n.

 

Miyata, S.; Kumura, T. Japan Kokai 73 29,478; Chem_._
Abstr. 1974, 80:85257p. '

Gastuche, M.C.; Brown, G.; Mortland, M.M. Clay Miner.
1967, _7_, 177.

 

Brown, G.; Gastuche, M.C. Clay Miner. 1967, 1, 193.

 

Brindley, C.W.; Kikkawa, S. Amer. Minera1., 1979, 64,
836.

 

Miyata, S. Clays Clay Miner. 1975, 23, 369.

 

Miyata, S.; Okada, A. Clays Clay Miner. 1977, 22, 14.

 

Miyata, S.; Hirose, T. Clays Clay Miner. 1978, 2_6_,
441.

 

Miyata, S. Clays Clay Miner. 1980, _2__8_, 50.

 

Hashi, R.; Kikkawa, S.; Koizumi, M. Clays Clay Miner.
1983, 31, 152.

 

Kruissink, E.C.; Alzamora, L.E.; Orr, S.; Doesburg,
E.B.M.; van Reijen, L.L.; Ross, J.R.H.; van Veen, G.
"Preparation of Catalysts II"; Delmon, B.; Grange, P.;
Jacobs, P.A.; Poncelet, G., eds, Elsevier, Amsterdam,
1979, 143.

Kruissink, E.C.; van Reijen, L.L.; Ross, J.R.H. it
Chem. Soc.,Faraday Trans. I 1981, 21, 649.

 

Puxley, D.G.; Kitchener, I.J.; Komodromos, C.;
Parkyns, N.D. "Preparation of Catalysts III";
Poncelet, G.; Grange, P.; Jacobs, P.A., eds, Elsevier,
Amsterdam, 1983, 237.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.
71.
72.

73.

74.
75.

76.

77.

78.

130

Huheey, J.E. "Inorganic Chemistry", 2nd ed.; Harper
and Row, New York, 1978, 70.

Rouxhet, P.G.; Taylor, H.F.W. Chimia 1969, 23, 480.

Butler, F.G.; Dent-Glasser, L.S.; Taylor, H.F.W. J;
Amer. Ceram. Soc. 1959, 42, 121.

 

"Handbook of Chemistry and Physics", 58th ed., Weast,
R.C., ed., CRC Press, Cleveland, 1977.

 

Dosch, W.; zur Strassen, H. Zement—Kalk-Gips 1967, 2_0,
392.

Miyata, S.; Takata, M.; Okada, A. Ger. Offen.
2,222,473; Chem. Abstr. 1974, 81:54438q.

 

Miyata, S.; Kuroda, M. U.S. Patent 4,299,759.

Miyata, S.; Kuroda, M. Ger. Offen. 2,950,489; Chem;
Abstr. 19XX, 93:96250a.

Ross, J.R.H. "Specialist Periodical Reports.
Catalysis"; Royal Society of Chemistry, London, 1985,
1, l, and references within.

Alzamora, L.E., Ross, J.R.H.; Kruissink, E.C.; van
Reijen, L.L. J. Chem. Soc., Faraday Trans. I 1981, E,
665. '

 

Busetto, C; Del Piero, G.; Manara, G. J. Catal. 1984,
85, 260, and references within.

 

Reichle, W.T. U.S. Patent 4,458,026.

Reichle, W.T. J. Catal. 1985, _9_4, 547.

 

McKillop, A.; Young, D.W. Synthesis 1979, 401.

 

Akelah, A.; Sherrington, D.C. Chem. Rev. 1981, _8__l_,
557.

 

McKillop, A.; Young, D.W. synthesis 1979, 481.

 

Whitehurst, D.D. Chemtech 1980, 44.

Helfferich, F. "Ion Exchange"; McGraw-Hill, New York,
1962.

Gordon, M.; DePamphilis, M.L.; Griffin, C.E. J. Org.
Che_m_. 1963, 28, 698.

Borders, C.E.; MacDonell, D.L.; Chambers, J.E.. J. Org.
Chem_. 1972, _3__7_, 3549.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

131

Cainelli, G.; Manescalchi, F.; Panunzio, M. Synthesis
1976,472.

 

Miller, J.M.; So, K.H.; Clark, H. J. Chem. Soc., Chem.
~Comm_. 1978, 466.

Hodge, P.; Sherrington, D.C. "Polymer-Supported
Reactions in Organic Synthesis"; Wiley, New York,
1980, 92.

Cainelli, G.; Manescalchi, F. Synthesis 1979, 141.

 

Gelbard, G.; Colonna, S. Synthesis 1977, 113.

 

Cainelli, G.; Manescalchi, F. Synthesis 1975, 723.

 

Shimo, R.; Wakamatsu, S. J. Org. Chem. 1963, _2_8, 504.

Manescalchi, F.; Orena, M.; Savoia, D. Synthesis 1979,
445.

 

Regen, S.L. J. Amer. Chem. Soc. 1975, 15, 5956.

Regen, S.L. Angew. Chem. Int. Ed. Eng. 1979, 18, 421,
and references within.

Starks, C.M. J. Amer. Chem. Soc. 1971, 21, 195.

 

Starks, C.M.; Liotta, C. "Phase Transfer Catalysis:
Principles and Techniques"; Academic Press, New York,
1978.

Dehmlow, E.V. Angew. Chem. Int. Ed. Eng. 1977, _1_6_,
493, and references within.

Tundo, P.; Venturello, P. J. Amer. Chem. Soc. 1979,
101, 6606.

Tundo, P.; Venturello, P.; Angeletti, E. J. Amer.
Chem. Soc. 1982, 10__1,6551.

 

Kadkhodayan, A.; Pinnavaia, T.J. J. Mol. Catal. 1983
11, 109.

 

Cinquini, M.; Colonna, S.; Montanari, F.; Tundo, P.
J. Chem. Soc. Chem. Comm. 1976, 392.

Halasz, 1.; Vogtel, P. Angew. Chem. Int. Ed. Eni,
1980, 12, 24.

Tundo, P. J. Chem. Soc. Chem. Comm. 1977, 641.

 

Monsef-Mirzai, P.; McWhinnie, W.R. Inorg. Chim. Acta
1981, §_2_, 211.

 

99.

100.

101.

102.

103.

104.

109.

110.
111.

112.

113.

114.

115.

116.

Kadkhodayan,
sity, 1983.

132

A., Ph.D. Thesis, Michigan State Univer-

"Powder X-ray Diffraction File, Inorganic Phases",
JCPDS-International Centre for Diffraction Data,
Swarthmnore, PA, 1983.

Serna, C.J.; Rendon, J.L.; Iglesias, J.E. Clays Clay
Miner. 1982, 3_0_, 180.

 

Landau, S.D., Ph.D. Thesis, Michigan State Univer-
sity, 1984.

McBain, J.W.; Bakr, A.M. J. Amer. Chem. Soc. 1926,
18, 690.

Yoon, R.H.; Salman, T. in "Colloid and Interface
Science, Vol. III", Kuker, M., ed; Academic Press,
New York, 1976, 233.

Clementz, D.M.; Pinnavaia, T.J.; Mortland, M.M. 11
Phys. Chem. 1973, 11, 196.

 

McBride, M.B.; Pinnavaia, T.J.; Mortland, M.M. Am.
Mineral. 1975, 60, 66.

McBride, M.B. J. Phys. Chem. 1976, 80, 196.

 

Quayle, W.H.; Pinnavaia, T.J. Inorg. Chem. 1979, 18,
10.

 

Sachs, F.; LaTorre, R. Biophys. Journal 1974, 11,
316.

 

Burba, J.L. III, U.S. Patent 4,348,295.

Pinnavaia, T.J., private communication.

 

 

Atkins, raw. "Physical Chemistry",W.H. Freeman, San
Francisco, 1978.

Hine, J.; Thomas, C.H.; Ehrenson, S.J. J. Amer. Chem.
Soc. 1955, 11, 3886.

Evans, A.G.; Hamann, S.D. Trans. Farad. Soc. 1951,
47, 25.

Young, D.M.; Crowell, A.D.
Gases", Butterworths,
within.

"Physical Adsorption of
London, 1962, and references

Joyner, L.G.; P.H. Chem.

_7_9_, 2353.

Emmett, J. Amer. Soc. 1948,

117.

118.

119.

120.

121.

122.

133

Emmett, P.H.; Brunauer, S. J. Amer. Chem. Soc. 1937,
52, 1553.

 

Houtepen, C.J.M.; Stein, H.N. J. Coll. Int. Sci.
1976, 56, 370.

 

Clark, A. "The Chemisorptive Bond", Academic Press,
New York, 1974.

Hayward, D.G.; Trapnell, B.M.W. "Chemisorption", 2nd
ed., Butterworths, Washington, D.C., 1964, and
references within.

Cusumano, J.A.; Low, M.J.D. J. Catal. 1971, 11, 214.

 

Little, L.H.; Kiselev, A.V.; Lygin, V.I. “Infrared
Spectra of Adsorbed Species", Academic Press, London,
1966.