New methods for characterizing solid acidity

 Title Page
 Table of Contents
 List of Tables
 List of Figures
 1. Introduction to the characterization...
 2. A new solid acidity scale
 3. The effects of proton exchange...
 4. The activity and acidity of...
 5. A new solid super acid: Silica...
 6. General conclusions
 Appendix I. The basic theory of...
 Appendix II. Operating the solid...
 Biographical sketch

Material Information

New methods for characterizing solid acidity
Physical Description:
xii, 130 leaves : ill. ; 29 cm.
Osegovic, John Philip, 1972-
Publication Date:


Subjects / Keywords:
Solids   ( lcsh )
Organic acids   ( lcsh )
Catalysts   ( lcsh )
Chemistry thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1999.
Includes bibliographical references (leaves 120-129).
Statement of Responsibility:
by John Philip Osegovic.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030366345
oclc - 41881639
System ID:


Material Information

New methods for characterizing solid acidity
Physical Description:
xii, 130 leaves : ill. ; 29 cm.
Osegovic, John Philip, 1972-
Publication Date:


Subjects / Keywords:
Solids   ( lcsh )
Organic acids   ( lcsh )
Catalysts   ( lcsh )
Chemistry thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1999.
Includes bibliographical references (leaves 120-129).
Statement of Responsibility:
by John Philip Osegovic.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030366345
oclc - 41881639
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
        Page vii
    List of Figures
        Page viii
        Page ix
        Page x
        Page xi
        Page xii
    1. Introduction to the characterization of solid acidity
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
    2. A new solid acidity scale
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
    3. The effects of proton exchange on 12-tungstophosphoric acid and its derivatives
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    4. The activity and acidity of two sulfated silica gels
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
    5. A new solid super acid: Silica supported antimony pentachloride
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
    6. General conclusions
        Page 99
        Page 100
        Page 101
    Appendix I. The basic theory of solid state magic angle spinning NMR
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
    Appendix II. Operating the solid state Bruker Avance 400 MHz NMR spectrometer
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
    Biographical sketch
        Page 130
        Page 131
        Page 132
        Page 133
Full Text








I would be strongly remiss if I did not try to mention all of those who contributed

to the writing of this work. To start with I would like to acknowledge the faculty and

staff of the University of Florida Department of Chemistry for allowing me to continue

my research after the passing of Professor Russell S. Drago. I would especially like to

single out Professors Russ Bowers and Mike Scott for dealing with and supporting me

during a difficult time.

I would not have been able to finish my work if not for the support of all of the

members of the Drago, Bowers and Scott groups. Every list must have a start so this one

might as well begin with the Burrito Brothers Lunch Brigade (Karen "Captain" Frank,

Steve Joerg, Alfredo Mateus, and Ben Gordon) which made lunch interesting during

those first formative years. Those Crazy Micropourous Guys (William Scott Kassell, J.

Michael McGilvray, C. Edwin Webster "Esquire," and Andrew Cottone) who seem to

have references for everything. The Former Inmates of room CLB 402 (Nick "The Tick"

Kob, Siliva Claudia Dias, Jos6 A. Dias, Danilo "The Quiet Man" Ortillo, and Krzysztof

"The Mad Pol" Jurczyk) who dealt with Emperor John of the Ever Expanding Desk

Regime. And finally the Bowers NMR Crew Team and Gymnastics Squad (Vincent

Storhaug, Anil Patel, Tony Zook, Gail Fanucci, and Charity Brockman), what more can

be said about them that is not already in the title.

I would like to thank Dr. Ion Ghiviriga for all of his help (and patience) in

teaching me to use the departmental spectrometers and Randy Duran for giving me the

opportunity to travel to the Advanced Photon Source at Argonne National Laboratory.

The true movers in the chemistry department are the secretarial staff. Without

them it seems that everything would grind to a halt. I know from experience that if you

have a problem with ordering chemicals, formatting text, or just figuring out the right

thing to say you go to visit Maribel Lisk. For those days when the University of Florida

Faceless Bureaucracy Thugs seems to be in every nook and cranny, Donna Balkom is the

woman to call. Her ability to drive the Thugs mercilessly to ground while at the same

time soothing the graduate student serfs has saved the sanity of countless individuals

(including me several times) over the years. Finally, Amanda Garrigues has been a

friendly and supportive face every since she arrived not too long ago.

My most heartfelt thanks are reserved for two people: Professor Russell Drago,

and my wife, Karen.

The loss of Doc was like losing a father for the second time. I deeply miss his


Even though she will not believe it, the most important thing in the world to me is

the love and friendship of my wife. I attribute the greatest part of my successes to her.



ACKNOWLEDGMENTS.................. ............................ ii

LIST OF TABLES.................. .. ............................... vi

LIST OF FIGURES..................................................... ..... ............. viii

A B ST R A C T ......................................... .................................................... xi

SOLID ACIDITY...................... ..... .............................. 1

A Definition of Solid Acidity................................................3
Using Probe Molecules of Measure Acidity..................................... 6
Calorimetric Measures of Acidity.................... ........ .......... ... 14

CHAPTER 2: A NEW SOLID ACIDITY SCALE................................ .......... 23

E xperim ental................ ................................................................ 28
Results and Discussion........................ ..... .....................31
C onclusions............................................................... ..................... 47


Experimental................................... ...................... 53
Results and Discussion.................... ...... ..................... 54
C onclusions...................................................................... ........ 67

SIL IC A G ELS........................ .... ............... ................................... 69

Experimental...................... .. .................. ........................... 73
Results and Discussion........................... .................... 75
Conclusions....................... ... ......... ..................... 81

ANTIMONY PENTACHLORIDE........................................................ 83

Experimental..................................... ................... .. 87
Results and D iscussion................ .................. ..... .............. .90
C onclusions.................... ... .. ............................................. 98

CHAPTER 6: GENERAL CONCLUSIONS..................... .................. 99

SPIN N IN G N M R............................................................................................. 102

AVANCE 400 MHz NMR SPECTROMETER............................................ 109

LIST OF REFERENCES.............................. .................... ........... .. 120

BIO G RA PH ICA L SKETCH ..................... ...................................................... 130


Table page

Table 1-1 An example of intermolecular interactions that lead to bonding....... 4

Table 1-2 The types of forces measured in calorimetric experiments............. 15

Table 1-3 The Cal-Ad results for 12-tungstophosphoric acid (HPW) in
acetonitrile solution and as a solid slurried in cyclohexane............ 18

T able 2-1 T he A 6 Scale........................................ ........................................ 32

Table 2-2 The inductive effect in Lewis haloacids..........................................39

Table 2-3 The Cal-Ad results for HZSM -5........................... ............... ...... 43

Table 2-4 The A6 value of TEPO on several bases.......................................... 44

Table 2-5 The coupling constants (J 31P-77Se) of TMPSe physisorbed on
solids.................................. ......................... 48

Table 3-1 Comparison of the Cal-Ad site populations to the position of the
protons in the crystal structure of HPW ............................. ....... 60

Table 3-2 The chemical shift of cesium salts of HPW..................................... 67

Table 4-1 Retention time of several alcohols and ethers.................................. 76

Table 4-2 The acidity of two sulfated catalysts measured by calorimetry......... 77

Table 5-1 Retention times of various hydrocarbon families.......................... 89

Table 5-2 The ECW model can be used to predict the strength of donor-
acceptor interactions.......................................... ..................... 91

Table 5-3 Ea and Ca parameters for a variety of Lewis acids........................... 92

Table 5-4 Assignment of the FTIR peaks of adsorbed pyridine on
(silica gel) Sb C 3................................................ ................... 93

Table 5-5 The results of alkylation reactions with silica supported antimony
pentachloride and aluminum chloride...................................... 96



Figure 1-1

Figure 1-2

Figure 1-3

Figure 1-4

Figure 1-5

Figure 1-6

Figure 1-7

Figure 2-1

Figure 2-2

Figure 2-3

Figure 2-4

Figure 2-5

Figure 2-6

Figure 2-7

Figure 2-8

Figure 2-9

Figure 2-10


An exam ple of site density............................................... ................... 5

The effect of magic angle spinning (MAS) on the NMR spectra of
solids........ ................ ...... ... ................... 7

Weakening of the phosphorous-oxygen bond............................... 9

Resonance forms of mesityl oxide in Br0nsted acid solutions................. 12

Energy diagrams for slurry and gas phase calorimetries...................... 16

Only Bronsted sites can be probed by 2,6-di(2-methylpropyl) pyridine.. 18

A Born-Haber cycle relating Cal-Ad to gas phase calorimetry............. 21

The three chemical shifts of adsorbed trimethylphosphine.....................25

Transfer of electron density from a base to an acid................................ 27

The correlation between calorimetry and AS..................................... 34

Phosphorus-31 MAS NMR of TEPO on (silica gel)nSbC13.................... 37

Phosphorus-31 MAS NMR of TEPO adsorbed on TaCl......................... 38

A comparison of two different methods of measuring of solid acidity.....41

Phosphorus-31 MAS NMR of trimethylphosphine oxide on zeolite
T S-1 ................. ............ .... ................... 43

The chemisorbed chemical shift of trimethylphosphine oxide compared
to calorimetric measures of acidity............ ............... .................. 45

An MO diagram for the acid TEPO interacting with KOH.................... 46

Phosphorus-31 MAS NMR of TEPO on potassium hydroxide............. 46

Figure 2-11 The 3P MAS spectrum of TMPSe on tungstic acid (H2WO4)..................48

Figure 3-1 The structure of the Keggin anion................................................... 51

Figure 3-2 Delocalization of charge in 12-tungstophosphoric acid............................ 52

Figure 3-3 The 3P MAS NMR of 12-tungstophosphoric acid................................. 55

Figure 3-4 The 3"P MAS NMR of 0.24 mmol pyridine on 1 gram of
12-tungstophosphoric acid..................... ............... ..................... 56

Figure 3-5 The chemical shift of the central phosphate changes with cesium
loading..... ...... .... ..... .... ... ............... .................. ..... ..... 57

Figure 3-6 The XRD data for 12-tungstophosphoric acid...................................... 58

Figure 3-7 The XRD data for 12-tungstophosphoric acid with one pyridine
molecule adsorbed per acid site...................................................... 58

Figure 3-8 A representation of the unit cell of 12-tungstophosphoric acid............. 61

Figure 3-9 A representation of site one for pyridine titration of HPW..................... 62

Figure 3-10 A representation of fully titrated HPW ............................................... 62

Figure 3-11 The 31P MAS NMR spectrum of CsH2PW.......................................... 65

Figure 3-12 The 31P MAS NMR spectrum of Cs2HPW......................................... 65

Figure 3-13 The 31P MAS NMR spectrum of Cs2.5H.5PW.................... ............ .. 66

Figure 3-14 The 3P MAS NMR spectrum of Cs3PW........................... ........... 66

Figure 4-1 The effect of solid acidity on the reaction of tertiary butanol with
m ethanol.......................... ... ....................................... 70

Figure 4-2 The mechanism for the catalytic production of MTBE from methanol
and t-butanol..................................................................................... 70

Figure 4-3 The reaction mechanism for the production of dimethyl ether from
m ethanol......................................................................................... 72

Figure 4-4 The transition states for ether production......................................72

Figure 4-5 Hydrolysis of a tethered sulfate group from a surface............................ 74

Figure 4-6

Figure 4-7

Figure 4-8

Figure 5-1

Figure 5-2

Figure 5-3

Figure 5-4

Figure 5-5

Amount of MTBE and t-butanol in the product stream from TSC over
the life span of the catalyst.................................. .............. 78

Amount of MTBE and t-butanol in the product stream from BSC over
the life span of the catalyst...................... ... ........................ 79

The products of the reaction change as the acidity of the catalyst
deteriorates....................... ..... ........................... 80

Examples of super acids............................................. 84

The reaction mechanism for the alkylation of isobutane to octane........... 85

The FTIR of pyridine on (silca gel)nSbvCl3..................................93

Titration of (silica gel)nSbVC13 with pyridine in cyclohexane.................. 95

The proposed acid sites of (silica gel)nSb Cl........................................ 97

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



John Philip Osegovic

May 1999

Chairman: Dr. Clifford R. Bowers
Major Department: Chemistry

Liquid acids are very important catalysts that are used to make many types of

products. However, the waste from industrial facilities that use liquid acids has

contributed to many environmental problems such as ozone depletion. Replacement of

these hazardous liquids with reusable solid acids is the key to developing green industrial

processes. Unfortunately, the complex interactions of a molecule with a solid acid site is

not fully understood, preventing the rapid and wide spread replacement of liquid acids

with new solid acids.

Current methods for characterizing solid acidity include many different types of

spectroscopic and calorimetric techniques. Many of these methods rely on observing the

changes that occur in probe molecules interacting with the acid site. Generally,

adsorption of the probe molecule onto the acid site results in the formation of an adduct

bond between the probe and the acid site. The change in the electronic state of

triethylphosphine oxide upon adsorption to an acid site can be measured with

Phosphorus-31 MAS NMR.

The position of the spectral lines) of the adduct can be used to identify the

relative acidity of multiple acid sites on a surface. The combination of this new method

with activity tests and calorimetry can be used to describe the total acidity of a solid.

Additional methods can be used to determine how the probe molecule reacts with

the surface of a solid. In the case of the important catalyst 12-tungstophosphoric acid and

its cesium derivatives, several techniques were used to characterize the surfaces of the

solids. Penetration of the lattice by small molecules is apparently limited by the energy

cost associated with opening the crystal lattice to include the molecule. The result is that

only a small fraction of the available acid sites of these solids are available for interaction

with the reactant.

The development of new solid acid characterization techniques will help to

increase the level of understanding of these important catalysts. The new insights gained

will lead to an increase in the pool of solid acids available to replace liquid acids,

reducing the burden placed on the environment by industrial processes.


The replacement of hazardous liquid acid catalysts with solid analogs is one of the

most important goals for the modernization of industrial processes. This substitution is

driven by economic and environmental factors. Removal of acidified waste from effluent

streams would lower the amount of sulfate and other waste products that are dumped into

watercourses by millions of tons per year.' Solids are also easier to transport, store,

handle, and are (generally) recyclable, resulting in a significant reduction in cost.

Finding a solid that is capable of catalyzing a current industrial process under laboratory

conditions is only the first step in the development of an industrially viable replacement

for liquid acids. The subsequent steps include up scaling catalyst synthesis and reaction

size and retooling or constructing production facilities.

The replacement of a liquid acid with a solid is the result of years of testing and

retesting. Much of this time is spent in the initial steps of identifying which reactions) a

candidate solid is capable of catalyzing. These initial steps include activity tests and

various measures of acidity, usually beginning with a screening process.23 One method

of screening catalysts starts with measuring the activity of a series of solids toward a

certain reaction. The catalysts that do not perform well are discarded, and the successful

solids are further tested to determine their life span or to isolate the components that

make them successful to help in the generation of another set of catalysts for testing. A

second screening method initially tests the solid acid to predict the type of reactions it

will be capable of catalyzing. The solid is then tested for its ability to catalyze reactions

in this range of acidity. Such a test is often used to estimate acidity through calorimetric

or spectroscopic techniques. This second method is more common when only a small

number of solids have been synthesized and require testing.

The use of solid acids has been driven by the need to separate products from

catalysts and to replace toxic liquids with less hazardous solids. A reaction driven by a

liquid acid has three major shortcomings. The first problem is how to store large

volumes of dangerous liquids. For example, handling hydrofluoric acid is not a simple,

inexpensive, or safe operation. The acid is produced on site and piped to the reaction

vessel as a gas where it is condensed. A small leak of HF can stop the production of a

large plant, as well as pose a significant health risk to the local community and

environment. The second problem is how to separate the product from the spent acid.

This is often performed by distillation or neutralization procedures.4 Finally, the cost to

regenerate acidified waste is more than the cost to produce fresh acid, resulting in a large

increase of waste. For example, production of methylmethacrylate results in several tons

of ammonium sulfate that must be disposed of for every ton of product.' This salt is

often dissolved and dumped into the environment.

Solid acids address each of these problems in a favorable way.2 Storage and

handling of solid acids is a simple task. They can be stored in bags or containers, and

require only a mask to avoid getting the particles in the eyes and lungs. Separation of the

catalysts from reaction products is an easy procedure: filtration or reactive distillation,

where a chemical reaction occurs simultaneously and concurrently with fractionalization

of products. Finally, many solid acids can be regenerated, extending their lifetime


A Definition of Solid Acidity

Solid acids are invaluable as catalysts in industrial reactions. The interaction of a

solid acid with a base can be described using Lewis acid-base theory--the sharing of an

electron pair.7-9 The intermolecular interactions involved can be distinguished from other

forms of intermolecular bonding through the use of some examples. If we allow two

non-acidic, non-basic molecules to interact, such as two helium atoms, the strength of the

interaction can be described using only van der Waals forces.'0 The same forces can be

used to describe the interaction of a basic molecule, such as pyridine, interacting with a

non-acidic solid such as silicalite. The magnitude of this interaction, measured through

the enthalpy of adsorption, has been determined to be -19.1 kcal mol'1." However, if we

exchange the silicalite with its acidic isomorph, HZSM-5, we must describe the bonding

as containing both a van der Waals and an acid-base interaction."2 The total enthalpy of

this interaction has been measured to be -47.8 kcal mol 'that can be separated into a van

der Waals component of -19.1 kcal mol1 and an acid-base component of -28.7 kcal mol'.

To truly describe a solid acid we must also rule out interactions that oxidatively lead to

the formation of a covalent or ionic bond (Table 1-1).

Describing Solid Acids

Solid acids are described in three ways: (1) the number of types of acid sites and

their site density, (2) the acid strength possessed by each type of site, and (3) the activity

of the solid toward acid catalyzed reactions. The overall strength of a solid acid (points 1

and 2) is perhaps its most significant characteristic. The maximum strength tells

Table 1-1. An example of intermolecular interactions that lead to bonding.

Reaction Type of Interaction leading to the Product

'/2 H- + Ar -) HAr + e Redox reaction

H' + Ar -- HAr Bronsted Acid-Base

Argon + Water -) Argon
n + ate A n Dipole-Induced Dipole

NH3 + HCI -- NH4CI Bronsted/Lewis Acid-Base

C2H4 + H20 CH3CH20H Addition Reaction

nCl2 + 2nNa Sodium Chloride Redox to a Ionic Bond

NH3 + BF3 -4 H3N:BF3 Lewis Acid-Base

researchers and engineers what kind of reactions the solid is capable of catalyzing. It

may also give the researcher some idea of how well the catalyst will carry out that

reaction.3 Site density is the second consideration (Figure 1-1).13 Site density is different

from normal density, p, in that it is a measure of the number of acid sites available for

interaction with a specific probe molecule for a given mass of solid. Site density is probe

dependent and is usually reported in units of mmol of acid site per gram of solid acid, or

simply mmol g'.

It is important to stress that site density is probe dependent.14 For example, the

two solids shown in Figure 1-1 might report the results of a single solid that was titrated

with two different probe molecules, one of which could not interact with the Type 1 sites.

Such differences can be related to the size, or steric requirement, and basicity of the

probe molecule.

2 Site, High Density Solid Acid

-:)--Type I sites.

*-'* y .2--s-e '

.. ,,- , .

-- --Tpe 2 sites.
---^- O. *-:. A-;-,

One Site, Low Density Solid Acid

-- S -'. -
.-,,.S o.-- -,. .-- ,,. .-- .,

,, .4.i.. . 4.
"I I Sf

,-^ .i, ^> ^.
\ \ \
r '*' It ^^ I
S .-\- S-\ ^'\-

, I r
-.4. .4.----:**--. -.

.*-< .-. ... .. .


'. S S o L t
S i

Figure 1-1. An example of site density. The diagram on the left shows a high-density
solid acid with two different types of acid sites. The diagram on the right shows a one
site, low-density solid acid. A reaction that can be catalyzed by either type of acid site
will be better performed by the high-density, two site solid. Tests of acidity that can only
detect the Type 2 sites cannot be used to distinguish one solid from the other--they will
have identical results.

The combination of site density and acidity are the dominant factors in

determining the usefulness of the catalyst. Both a low density solid with very strong

acid sites and a high density solid with very weak acid sites will be poor catalysts. They

may be capable of catalyzing a reaction, but only at a very limited rate. Additionally, the

stability of the acid sites is also very important.'5 If the acid sites are easily destroyed, for

example, by water hydrolysis, the catalyst will be of limited use in reactions where

complete removal of water is difficult.

Using Probe Molecules to Measure Acidity

Acidity is often characterized by measuring spectral line shifts of an adsorbed

probe molecule.6 Techniques based on IR,1721 UV,22 EPR23 and NMR (to be discussed

later) have been developed. The shift of the spectral lines) is compared to the shift of the

spectral lines) of the same probe molecule on other solid acids with different acidity to

generate a scale of relative acidity. A relatively larger shift of the spectral line indicates

higher acidity. When several different spectral lines from a probe on one solid acid are

observed, the additional spectral lines are taken to be additional acid sites with different

strengths.24 By monitoring the changes in intensity of the different spectral lines as the

total amount of adsorbed probe is varied, the amount of each acid site present in the solid

can be determined.

The technique for obtaining a useful NMR spectrum of a solid is called magic

angle spinning (MAS). (The basic theory behind MAS NMR can be found in the

Appendix.) Without MAS, the NMR signal of a solid is typically broad and the

determination of the position of the isotropic spectral lines) is inaccurate. The line width

of a solid sample can be in the tens of kilohertz.25 For example, a 10 kHz line for

phosphous-31 on a 400 MHz spectrometer will be approximately 62 ppm wide, while in

solution the line representing the same feature might only be 0.1 ppm in width (16 Hz).

The extreme broadening is due to the large number of rigid dipolar couplings in the solid

state. The use of MAS allows the acquisition of moderately high resolution NMR

spectrum of a solid, where the spectral line or lines are resolved and their chemical shifts

can be determined. Figure 1-2 shows a graphical representation of the effects of MAS.

Figure 1-2a is an example of the NMR of solid adamantane without MAS. Adamantane

Figure 1-2. The effect of magic angle spinning (MAS) on the NMR spectra of solids.
Figure 1-2a represents the NMR spectra of a solid adamantane. Figure 1-2b represents
the NMR spectra of the same solid while spinning on the magic angle. The line width
has been significantly decreased allowing for the observation of two spectral lines.
(Adapted from "Modern NMR Spectroscopy 2ed" by J. K. M. Sanders and Brian K.

is a symmetrical bicyclic organic compound with two types of carbon atoms in a 2:1

ratio. However, the static NMR spectrum shows only one featureless peak. Figure 1-2b

is of the same sample but acquired under MAS conditions. The line width has been

greatly reduced and two spectral lines of roughly 2:1 intensity can now be resolved.26

Phosphorus-31 MAS NMR is well suited for solid acid characterization. The

large gyromagnetic ratio and 100% natural abundance of phosphorus-31 allows detection

of very small amounts of probe molecules, which is advantageous when attempting to

observe the limited number of acidic sites present on the surface of a solid acid.

Previously reported phosphorus-31 MAS-NMR studies over the past decade have focused

primarily on trimethylphosphine as the adsorbate probe of choice."

Trimethylphosphine has been used to identify the number of different sites on the solid,

the density of each site, and the nature of each site as either Lewis or Bronsted.34

Unfortunately, the small variation of the phosphorus-31 chemical shift of

trimethylphosphine in solids of wide ranging acidity limits the utility of this probe for the


determination of acidity. For Bronsted acids the range of acidity correlates with chemical

shifts over a 5 ppm range, with an error in measurement of 0.5 ppm. For Lewis acids, the

range is covered in only 3 ppm, with an error of measurement of 1 ppm.

The most promising use of trimethylphosphine is for the determination of

densities of the Lewis and Br0nsted sites on the surface. Work performed using 15N

pyridine has shown s'5N of that probe can be used to give similar information to 83'P

from trimethylphosphine.3539

Liquid acids have been previously characterized using triethylphosphine oxide

(TEPO), upon which the acceptor number scale is based.19'30.40-42 The phosphorus-31

isotropic chemical shift of TEPO can identify the relative acidity of solutions. However,

while this and similar probe molecules have been used to distinguish multiple types of

acid sites on the surface of several solids,28'4344 no attempt has been made to use TEPO to

form a scale of solid acidity. A scale based on the change in chemical shift of this probe

would be advantageous due to the speed at which the characterization can be performed.

The origin of the phosphorus-31 chemical shift of chemisorbed TEPO is the shift

of electron density from the basic probe to the acid site.42 The interaction of the

phosphoryl oxygen atom with the acid site gives rise to a weakening of the phosphorus-

oxygen bond (Figure 1-3). Evidence for the weakening of this bond can be obtained from

the vibrational, or infrared, spectroscopy investigation of solvent acidity.17'1920 As the

acidity of the solution is increased, the phosphorus-oxygen stretching frequency can be

seen to decrease, indicating a weakening of the bond. Positive charge develops at the

phosphorus center while negative charge builds up at the acid site. The stronger the acid

site, the greater the shift of electron density and the greater the charge build up. This

Et- O Acid Site Et --- Acid Site Et-_ O- -Acid Site

Et Et Et EtE
Et Et Et

a) Coordinated b) Weak Acid Site c) Strong Acid Site

Figure 1-3. Weakening of the phosphorus-oxygen bond. The phosphorus-oxygen bond
of uncoordinated triethylphosphine oxide lengthens upon coordination to an acid site.
Figure 1-3b represents coordination to a weak acid site such as a hydrogen bonding site.
Figure 1-3c represents bonding to an extremely powerful acid site.

Note: The nature of the phosphorus-oxygen bond in trialkylphosphine oxides is not
understood.45 Due to the short length of this bond measured through crystallography and
other means, this bond has been characterized as a double bond. However, calculations
have shown that it is possible to represent the bonding as a single bond with a full
positive charge at the phosphorus and a full negative charge at the oxygen. The work
here can not add to this debate and the choice to show the phosphorus-oxygen bond as a
double bond throughout this dissertation is for clarity only.

shift in electron density causes a change in the chemical shielding factor, o, at the

phosphorus nucleus. The resulting change in the chemical shift can be readily observed

using MAS NMR.

It has been reported that the 31P chemical shift of triphenylphosphine oxide is

dependent upon the AI-O(P) distance in aluminum coordination complexes.46 This bond

distance was noted to decrease with increasing electrophilicity (acidity) of the aluminum

center. While the P-O bond distance was observed to increase slightly, the correlation to

831P was poor.

Since each acid catalyzed reaction requires a certain range of acidity, a researcher

can quickly determine what reactions a new solid acid will be suitable for. If the catalyst

causes TEPO to have a small shift in its 3'P MAS NMR signal, it will be suited to

perform reactions that require little acidity, such as the dehydration of tertiary butanol to

isobutene, an important chemical building block. If the shift is large, the catalyst can be

tested with reactions that require much greater acidity, such as the conversion of butane

to octanes, the major component of gasoline. Using the methods described in this

dissertation, a researcher can focus on optimizing the catalyst for a catalytic process, or

attempt to synthesize a new catalyst to perform the desired reaction.

Carbon-13 Probes of Solid Acids.

Several studies of solid acids using carbon-13 labeled ketones have been reported.

The work of Biaglow et al.47-50 and others51 focuses on the perturbation of the chemical

shift of 2-13C acetone to simultaneously determine the relative acidity and Br0nsted or

Lewis nature of acids sites in HZSM-5 and Faujasites.48 A chemical shift of about 220

ppm (relative to TMS) was reported to represent acetone adsorbed on the Bronsted acid

sites of the solids, and slight variations about this shift were taken to indicate changes in

the strength of the hydrogen bond. The stoichiometric adsorption complexes of antimony

pentachloride with the Bronsted acid sites of zeolite HZSM-5 were studied separately.50

The magnitude of the shift in this solid was larger than in clean HZSM-5. The shift was

enhanced due to the coordination of the antimony halide to the oxygen atom at the acid

site similar to the large increase in acidity observed in two component super acids over

either component alone. Treating the solid at 700 K resulted in dealumination of the

zeolite and a larger chemical shift for acetone, approximately the same as measured in a

magic acid solution (249.5 ppm). Additional work on other molecular sieve solids

showed that the chemical shift of acetone in these structures was largely invariant and

that the shift might even be correlated to the size of the cavity of the solids studied.47 The

change in chemical shift of acetone on silica supported aluminum chloride was used to

assert that the acidity of that solid was greater than HZSM-5.39 The overall picture for

characterizing acidity with acetone is a cloudy one. Acetone appears to be sensitive to

the strength of the acid site, but the relationship is not well known. It may also be

sensitive to size of the pore the acid site is located within and will have different,

overlapping chemical shift ranges depending on whether it is bound to a Br0nsted or

Lewis site.

Farcasiu et al. and others have developed a scale for Br0nsted acids based on the

difference in shift of the carbons of mesityl oxide (4-methyl-3-petene-2-one) in

solution.52-56 The P carbon of Mesityl oxide (Figure 1-4) showed a large dependence on

the acidity and polarity of solutions,5255 but it was less sensitive to the acidity of solid


Using 27Al and 2Si MAS NMR to Determine the Structure of Molecular Sieves.

Aluminum-27 and Silicon-29 MAS NMR are invaluable tools for studying the

structure of silica/alumina molecular sieve catalysts.57 The structures of these catalysts

are a series of tetrahedral silicon atoms linked into regular repeating ring structures of

varying sizes. The substitution of aluminum for a silicon atom (up to a certain limit: there

can be no AI-O-AI linkages in zeolites) results in a change in the chemical shift of the

nearest neighbors and results in the potential for acid sites to exist. A single chemical

shift can be measured for silicon atoms with zero, one, two, three or four adjoining

aluminum centers in the framework and the population of each type of silicon atom can

be obtained from integration.58-59 The bulk silicon/aluminum ratio includes non-

framework aluminum atoms (present predominantly as alumina, A1203) and a comparison

of the two ratios can be used to determine the purity of the sample. Additionally, the

0/ CH3 0 CH3
II i I+
H3C a d C P CH3 H3C a p CH3

Figure 1-4. Resonance forms of mesityl oxide in Bronsted acid solutions. The chemical
shift of the a and P of mesityl oxide carbons are both sensitive to protonic media.

extent of dealumination can be measured with 27A1 NMR.6066 Aluminum-27 (I = 5/2,

natural abundance = 100%) is a quadrapolar nucleus, so the spectral lines observed are a

mixture of both quadrapolar coupling and chemical shift of the aluminum centers.67

Essentially, only two peaks will be observed in the MAS spectra: one due to tetrahedral

framework aluminum and one due to octahedral extra framework alumina (virtually all

non-spherically symmetric 27Al signals will be lost due to rapid quadrapolar relaxation).60

The amount of dealuminated material can be determined by comparing the intensity of

each signal. By combining both the 27Al and 29Si MAS NMR of a molecular sieve the

possible ordering schemes of the solid can be derived that show the possible location of

the aluminum atoms in the unit cell.57 Additionally, Huggins and Ellis showed that

lateral surface diffusion of surface protons on aluminas causes an electric field gradient

that provides an efficient quadrapolar relaxation mechanism resulting in a loss of signal

intensity that increases with surface area.67

Methods Using 'H MAS NMR

Proton NMR has been used to label the chemical shift of acidic protons on

surfaces.3967'71 The chemical shift of protons varies tremendously with the type of solid

and whether or not the proton is acidic. A scale based on simulating the static 'H NMR

of H30O relative to the total number of protons at 4K was proposed to distinguish

between the acidic protons from hydrogen bonding protons and non-acidic protons and

rank zeolite acidity.70-71 The scale that was developed, however, ranked the acidity of

zeolite HZSM-5 lower than both HY and H-Mordenite in contradiction to all other scales

of acidity. Pearson68 used MAS NMR to identify two different types of protons,

chemisorbed and phyisorbed, on the surface of various transition-aluminas (the different

transitions of alumina can be obtained through heat and pressure treatments). It was also

reported that 83% of the acidic protons can be found in defect sites in the first two layers

of the surface.

In addition to labeling the acidic protons in zeolites, Anderson et al.69 utilized the

change in 'H chemical shift of CD30H to measure the proton donating ability (rather than

the acid strength) of molecular sieves. The origin of the shift is due to the formation of a

strong Bronsted-alcohol hydrogen bond or the formation of the oxonium ion (CD30H2 ).

This change in shift was greatest in HZSM-5 (9.4 ppm) while in silicalite (a non-acidic

isomorph of HZSM-5) the resonance was at 3.8 ppm. Solids with acidities between these

two extremes generally resonated between 4 and 6 ppm. It was clearly demonstrated that

the observed chemical shift was dependent on synthesis technique and silica/aluminum

ratios by using two separate methods for preparing HZSM-5. The two different solids

differed in the 'H chemical shift of adsorbed CD3OH by almost 2.4 ppm!


A basic description of NMR and the theory behind magic angle spinning can be

found in Appendix I. Chapters 2 and 3 will use MAS NMR to help characterize solid

acidity. Chapter 2 will focus on the use of triethylphosphine oxide to develop a new

scale to characterize solid acidity. A brief description of the use of other phosphine

chalcogenide probes will also be discussed. The scale developed in Chapter 2 will be

applied to the characterization of a new solid acid in Chapter 5. The location of the acid

sites in the crystalline solid 12-tungstophosphoric acid will be determined with the help

of MAS NMR in Chapter 3.

Calorimetric Measures of Acidity

The measurement and determination of the acidic properties of solids is of great

interest. Calorimetry is a common method used in the characterization of solid acidity.

As it applies to solid acid characterization, calorimetry can be described as a measure of

the initial and final enthalpy of a free or solvated probe molecule upon adsorbtion onto a

surface.72 The enthalpy change of the probe molecule as it adsorbs or desorbs from a

surface causes a measurable change in the temperature of the surroundings. The degree

to which the temperature is changed is dependent upon the strength and type of

interaction of the probe with the surface (Table 1-2). In the case of a basic probe

molecule adsorbing onto a clean (unsolvated) non-acidic surface, only the energy due to

van der Waals forces will be measured. For the same system in the presence of a solvent

the difference in the van der Waals forces due to probe-solid and solvent-solid

interactions can be obtained. However, if the solid is acidic, then the clean solid will

have a enthalpy equal to the sum of the van der Waals forces and the donor-acceptor

forces. In the case of a solvated, acidic solid, the extent to which the van der Waals

forces are measured depends on the difference between the van der Waals forces for the

solvent and probe interacting with the surface.

Table 1-2. The types of forces measured in calorimetric experiments.

Forces Present

Type of Surface van der Waals Donor-Acceptor Example

Clean Non-Acidic Yes No Carbonaceous Solid

Solvated Non-Acidic Reduced No Wet Graphite

Clean Acidic Yes Yes Aluminum Trichloried

Solvated Acidic Reduced Yes AI2C16 in alkanes

The measured energetic of adsorption of a probe molecule onto a surface varies

with the technique used. In gas phase adsorption calorimetry the energy measured is the

total of the non-specific van der Waals interaction of the probe with the surface plus the

donor-acceptor interaction of the probe with the acid site.72 The magnitude of the van der

Waals probe-surface interaction can be as high as 40% of the measured enthalpy."

A second type of calorimetry is performed in the presence of a non-participating solvent.

This method of calorimetry is called solution calorimetry if the compound under study is

dissolved in the solvent or slurry calorimetry if the compound is suspended in the solvent.

For slurry calorimetry, the non-specific van der Waals contribution to the enthalpy

measured in gas phase calorimetry is reduced or canceled by the energy required to

displace a solvent molecule from the surface (Figure 1-5).73 The complete cancellation of

the van der Waals component relies on the selection of a solvent molecule that is

approximately the same size as the probe molecule, therefore having essentially the same

van der Waals contribution.

Acid Site--Probe

Acid Site--Probe i
4 Probe onto
< g Surface

a RProbe onto
r Acid + Probe Surface Acid + Probe


a) Slurry Calorimetry b) Gas Phase Calorimetry

Figure 1-5. Energy diagrams for slurry and gas phase calorimetries. The energetic of
adsorption for a) slurry and b) gas phase calorimetries. The enthalpy observed during
slurry calorimetry is due only to the donor-acceptor interaction of the probe with the
surface. The observed enthalpy from gas phase calorimetry contains both the specific
and non-specific enthalpies of adsorption.

An additional step is often added that measures the amount of base adsorbed.

This addition is especially useful for solids where adsorption is far from complete. It

allows a more accurate description of the acid sites. The method of tracking the amount

of adsorbed probe is generally determined by taking the difference between the amount

of free probe and the total amount of probe in the system as measured through UV or IR

spectroscopy. The Cal-Ad method is an example of such a method.

Cal-Ad combines the data obtained from slurry adsorption calorimetry with a UV

adsorption isotherm.39'73-83 A least square fit to the data allows a determination of the

number of discrete acid sites and the population, acid strength, and equilibrium constant

for adsorption of the probe for each site. The probe typically used to perform Cal-Ad is

pyridine dissolved in either n-hexane, cyclohexane, or acetonitrile. Additional

information about the solid can be obtained by varying the probe molecule. For example,

comparing the results of a titration using pyridine with the results of an experiment using

2,6-di(2-methylpropyl) pyridine, a site can be designated as Lewis or Bronsted in

nature.7'80 The presence of the 2 methylpropyl arms will prevent the 2,6-di(2-

methylpropyl) pyridine from interacting with a Lewis acid site (Figure 1-6). Many solids

have been successfully characterized by this method, and a Cal-Ad scale of solid acidity

has been proposed. Table (1-3) shows the results from the Cal-Ad experiment for 12-

tungstophosphric acid dissolved in acetonitrile77 and slurried in cyclohexane.83 In both

cases the titrant was pyridine.

The following description of a typical solution or slurry calorimetric experiment

is adapted in part from an article authored by Jose A. Dias, John P. Osegovic and Russell

S. Drago.8 It is the method followed for all of the calorimetry reported in this

dissertation, and is the general method practiced by the Drago laboratory. As such,

similar descriptions can be found in many publications.7383 The repetition here is for


Samples are weighed (typical sample mass: 1.0 g) and transferred to an insulated

calorimetric cell with an internal mirrored coating containing a stir bar. For each

titration, 100ml of solvent is added to the cell. A calibrated syringe, filled with a solution

of known concentration of base (e.g., 0.1 mol L-') is inserted into the cell along with a

thermistor and a heater coil. The thermistor and heater coil are connected to an electronic

bridge and a computer.3 The completed cell is allowed to reach thermal equilibrium

with the environment. After thermal equilibrium is met, the experiment is begun. The

Figure 1-6. Only Bronsted sites can be probed by 2,6-di(2-methylpropyl) pyridine. The
alkyl arms on 2,6-di(2-methylpropyl) pyridine (bottom), a slightly stronger base than
pyridine, prevent the formation of an acid-base bond at the Lewis site in this diagram.
Pyridine (top) can freely interact with either site. The Cal-Ad data for each solid would
show one fewer site when titrated with 2,6-di(2-methylpropyl) pyridine.

Table 1-3. The Cal-Ad results for 12-tungstophophoric acid (HPW) in
acetonitrile solution77 and as a solid slurried in cyclohexane.83

HPW Dissolved in Acetonitrile Solid HPW in Cyclohexane Slurry

Site -AH Densitya Kb -AH Density Kb
(kcal mol') (mmol g1') (kcal mol-') (mmol g-')
1 21.0 0.33 2.1*103 32.7 0.079 3.7*105

2 11.8 0.33 2.4*102 19.6 0.16 2.9*103

3 18.6 0.33 4.1*10'

a All of the protons are available for titration in solution; however, only about 25% of the
protons of solid HPW are available for reaction.
Equilibrium Constant for adsorption.


0 N


thermistor is calibrated with the heater coil prior to or immediately after each titration. A

set of 12 calibrated brass stops is used for the additions of the titrant. After each addition

the heat evolved from reaction of the base with the sample is measured. The calorimetric

data generates an isotherm of total heat evolved versus the total moles of base added.

A Current Example of the Application of Calorimetric Methods

In a recent paper,' the analysis of the interaction of pyridine with HZSM-5 by the

Cal-Ad method73 was called into question. A discrepancy of 22.9 kcal mol'1 between the

value derived from gas phase adsorption calorimetry and Cal-Ad was believed to be due

to a flaw in the Cal-Ad technique, and Savitz et al." concluded that Cal-Ad was

unsuitable for the characterization of the acidity of HZSM-5. Gas-solid calorimetry

(GSC) results were directly compared to the average enthalpy of the two acid sites

described by Cal-Ad without insuring that the two HZSM-5 samples were prepared

identically. The authors had also arrived at the conclusion that there is only one type of

acid site present in HZSM-5 despite a variety of papers that disagree.73'8490 In their

response to the Savitz article, Webster et al.91 pointed out that they failed to accurately

describe the differences in the experimental procedures and sample preparation methods.

It was shown91 that several considerations could reduce the discrepancy between

the Cal-Ad and GSC results. The first reduction made originated when accounting for

the change in temperature between the two experiments. The heat absorbed by pyridine

gas, as the temperature of the experiment is changed from 298 K (Cal-Ad is performed at

ambient temperatures) to 473 K (the gas phase calorimetry experiment is performed at

elevated temperatures to ensure homogenous adsorption of the pyridine),11.92 is 4.06 kcal

mol'. Webster et al. assumed that the heat absorbed by the zeolite during the

temperature increase from 298 K to 473 K canceled the heat released by the zeolite-

pyridine adduct during the temperature reduction from 473 K to 298 K and left this step

out of their consideration.

Reports have shown that there are n-hexane molecules specifically bound to the

acid sites of HZSM-5.93 Every pyridine that interacts with the acid site must displace a

coordinated n-hexane.91 The desorption of a sufficient amount of n-hexane to allow

pyridine to occupy all of the acid sites resulted in a calculated enthalpy of 19.6 kcal mol1.

Sample preparation procedures for zeolite HZSM-5 directly affect the amount of

heat observed. The sample procedure used by Savitz et al. was not reported; however,

other preparation procedures used by the same group are very mild.95 Drago et al.73 used

a very harsh treatment condition designed to emulate the procedures used to prepare

HZSM-5 for industrial use. In their survey of the literature Webster et al. discovered gas

phase heats of adsorption for both n-hexane and pyridine onto HZSM-5 prepared in a

similar manner to the sample used in the Cal-Ad method.8493 When both the enthalpies

for n-hexane and pyridine adsorption were used in place of the values reported in

Reference 11, the difference between the two measures was reduced to -2.7 kcal mol'1.

The discrepancy between Savitz et al.1c and Drago et al.73 is almost certainly due to

differences in the preparation of the solid. A Born-Haber cycle helps to illustrate the

thermodynamic arguments used (Figure 1-7).

Calorimetric methods of characterization are very useful tools for exploring solid

acidity. There are a wide variety of procedures that are available to suit the researchers'

needs. However, due to the extreme dependence on sample preparation and experimental

parameters, care must be taken when comparing the results of two different types of

HZSM-5 +Pyg,473 +Hexg



AHCal-Ad=l 1.0 kcal/mol
Hcorrected gas= 13.7 kcal/mol

HZSM-5 +Py, +Hex,

HZSM-5hex +PY +Hexg

HZSM-5hex +Phex AHunmix

HZSM-5hex +PyI


AH py abs

HZSM-5-Py + Hex,

AHhex abs

HZSM-5-Pyhex + Hexg

AHliq, hex

IHZSM-5-Pyhex +Hex1 I

Figure 1-7. A Bom-Haber cycle relating Cal-Ad to gas phase calorimetry. The Born-
Haber cycle used to relate the gas phase adsorption of pyridine onto HZSM-5 to the Cal-
Ad experiment. The on an enthalpy denotes that the value was changed by Webster et
al. from those used by Savitz et al. These values were changed because the new values
were obtained from HZSM-5 samples that were treated in a very similar manner to the
one described by the Cal-Ad method.

experiments. An understanding of slurry calorimetry is all that is necessary to interpret

the enthalpies reported throughout this dissertation.

Characterizing solid acidity is important for the further development of

environmentally friendly processes that employ acid catalyst. The work in this

dissertation describes several new solid acids and a new method for their

characterization. The first part of the dissertation will focus on the use of MAS NMR to

describe solid acidity in two ways by 1) introducing a new method for using probe

molecules to measure the acidity of a solid and 2) by assigning locations to the acid sites

in a crystalline solid.

Chapter 2 will describe a new method for qualifying the acidity of solid acids.

This chapter will focus on the use of the phosphorus-31 solid state NMR to measure the

acidity of surfaces. The behavior of the probe molecule on solid acids of all strengths as

well as bases and neutrals will be discussed.

The third chapter will detail the efforts made to characterize the acidity of an

important series of solid acid catalysts. Spectroscopic methods will be combined with

calorimetry to help to understand the behavior of these solids. With the aid of a crystal

structure diagram, the location of the acid sites within the structure will be determined.

In Chapter 4 the relationship between structure and activity will be explored by

comparing the activity of two very similar sulfate catalysts for the production of methyl

t-butyl ether from methanol and t-butanol. This reaction will be shown to be an ideal

model reaction system for monitoring the changes in acidity of a catalyst throughout its


Chapter 5 will focus on the characterization of a new solid acid, silica supported

antimony pentachloride. Calorimetry, activity, and 31P MAS NMR will be used to

describe this catalyst as a new solid super acid.


Identifying the total acidity of a solid acid is a complex process, involving the use

of activity tests, calorimetric titrations, and many spectroscopic methods. Magic angle

spinning NMR of chemisorbed basic probe molecules is a very important step in this

characterization process. Probe molecules can be used to classify the acid sites of a solid

as Bronsted or Lewis in nature and measure relative acidity. To this end, a new scale

based on the chemical shift of the spectral lines of chemisorbed triethylphosphine oxide is

reported here.

Several NMR probes that have been used to study solid acids including 2-13C

acetone,47-51 4-13C mesityl oxide,53-56 and 1N pyridine.35-39 Unfortunately, due to the

low population of acid sites on the surface, the low sensitivity of these nuclei requires

either extensive signal averaging or the use of labeled materials. To avoid both of these

problems, a probe molecule including a more sensitive nucleus at or near the basic site is

highly desirable. An ideal replacement probe would contain phosphorus-31. The

gyromagnetic ratio and natural abundance of phosphorus-31 is much greater than either

13C or 15N resulting in an order of magnitude increase in the NMR signal that can be

observed. The relative sensitivity of 31P compared to 3C and '5N at natural abundance is

approximately 700 and 2400 respectively.96 The very high sensitivity allows for the

detection of small amounts of phosphorus containing compounds adsorbed on the acid

sites of a solid without the use of expensive enriched samples.

Over the past few years, the spectral probe molecule of choice for the study of

solid acids by MAS NMR has been trimethylphosphine.2734 Coordination of

trimethylphosphine to an acid site results in a characteristic shift of the spectral line that

depends primarily on the Br0nsted or Lewis nature of the acid site (Figure 2-1). The

similar chemical shift of Lewis bound and physisorbed trimethylphosphine results in

difficulty for using this probe to unambiguously determine the presence and population of

Lewis sites. Despite this shortcoming, trimethylphosphine has been widely used to

determine the total population of Bronsted sites on a surface. The strong basicity of

phosphines results in the formation of a protonated base upon coordination to any

Bronsted site. The chemical shift of this characteristic system is about -4 ppm (all 3'P

chemical shifts in this dissertation are referenced to 85% phosphoric acid) and is largely

invariant with the strength of the acid site.

The basicity of trialkylphosphine oxides is on the same order of magnitude as

trimethylphosphine. However, the removal of the phosphorus atom from the base site

leads to a wide range of chemical shifts that vary with the strength of the acid site.40 The

most prominently used trialkylphosphine oxides are triethylphosphine oxide (TEPO) and

trimethylphosphine oxide. Both have seen limited use as probes of solid acids. 28,29.31,44

The most well known use of TEPO is as a probe of solvent acidity. Meyer and Guttman

proposed the Acceptor Number scale based on the chemical shift of TEPO extrapolated to

infinite dilution in an acidic solvent (831Pinfinite dilution).40 Normal hexane was assigned an




a -55 ppm b -49 ppm c -5 ppm

Figure 2-1. The three chemical shifts of adsorbed trimethylphosphine. The chemical
shift of trimethylphosphine a) physisorbed (van der Waals forces only) b) Lewis
chemisorbed and c) Br0nsted chemisorbed.

Acceptor Number value of 0.0 and SbC15 was assigned a value of 100. The acceptor

number of any solvent can be determined by using Equation 2-1:68

Acceptor Number = 2.348*31Pinfinite dilution (2-1)

This scale has been criticized for the observation that even neutral and basic solvents can

have significant Acceptor Number values that have been attributed to van der Waals


Recently, Rakiewicz et al.44 characterized the acid sites of amorphous silica-

alumina; zeolties HY, dealuminated HY, and USY; and y-alumina with

trimethylphosphine oxide. They were able to distinguish several acid sites on these

materials, and then measure the population of those sites through spin counting

experiments. It was also reported that this probe could differentiate Lewis and Bronsted

sites. A small negative shift of the spectral line (compared to physisorbed

trimethylphosphine oxide) on dry y-alumina was observed. This feature disappeared

upon exposure to water. It is known that water destroys the Lewis acid sites of

y-alumina. From these observations, the negative shift was taken to indicate the presence

of Lewis acid sites. This trend of assigning shift ranges for Lewis or Bronsted bound

phosphine oxides includes work by Lunsford31 and Baltusis.28 However, each author has

assigned a different region to the chemical shifts for the Lewis chemisorbed probe.

Triethylphosphine oxide has been previously used only twice for the characterization of

solid acids: to identify two different acid sites on y-alumina and to identify several acid

sites on zeolite HY.29'31'44

The change in position of a spectral line through the chemisorption of a basic

probe molecule onto an acidic site is primarily due to electronic changes in the probe

molecule leading to a change in the local magnetic field about the nucleus.42.68 Electron

density flows from the base to the acid site (Figure 2-2) upon formation of an adduct

bond with the acid site. Coordination of the phosphoryl oxygen of a trialkylphosphine

oxide creates overlap between the electron deficient orbital on the acid site (LUMO), be it

Bronsted or Lewis, and one of the orbitals on the oxygen that contains a lone pair of

electrons (HOMO). The resulting adduct bond causes a loss of electron density from the

probe that can be observed by a change in the phosphorus-31 NMR. As the strength of a

Br0nsted acid site is increased, the acid site-probe adduct will begin to resemble an ion

pair. However, once the base is fully protonated (acid site:H-probe') no further change

in chemical shift will be observed by increasing the acidity of the site. Instead, the same

chemical shift will be observed for any stronger acid sites. At a Lewis site, coordination

of a probe will result in the formation of an adduct bond. In the case of a phosphine

oxide, the oxygen atom acts as a bridge between the Lewis site and the phosphorus

center. The strength of the acid site will be directly related to the strength of the adduct



Figure 2-2. Transfer of electron density from a base to an acid. Overlap of the LUMO on
A (the acid site) with the HOMO of B (the base) results in the formation of a molecular
orbital (BA). The component of B in BA is reduced from B alone and as the strength of
A increases so will its share of the electrons in BA.

bond. At very high acid strength electron transfer may occur forming an ion pair. Again,

no further change in the chemical shift is expected, unless oxygen atom transfer occurs.

Transfer of the oxygen atom will result in an oxidized acid site and the reduction of the

trialkylphosphine oxide to the corresponding trialkylphosphine.

By comparing the shift of a phyisorbed base to that of a chemisorbed base, the

contribution of the van der Waals forces to the chemical shift is reduced or removed.28

The chemical shift of physisorbed TEPO has been observed to be approximately 50 ppm.

By referencing the chemisorbed chemical shift to this value, the relative acidity of a solid

acid site can be measured. The change in the chemical shift of TEPO has an excellent

correlation with calorimetric measurements of acidity except in the case of

12-tungstophosphoric acid videe infra).

Currently there is debate as to the acid strength of the crystalline solid acid

12-tungstophosphoric acid (HPW). Many reports have been issued with conflicting

measurements of its acidity.97-112 HPW is composed of large anions called Keggin units

with a secondary structure composed of H502' counter ions. 13-115 The Keggin unit is

composed of four W3012 groups called W3 triplets. 16 These triplets are linked by shared

corners about a central P04 tetrahedron. The acidity of this solid is believed to come

from the large size of this anion and from the high proton conductivities that spread any

buildup of anionic charge throughout the solid.

For a probe to penetrate the crystalline HPW structure, the lattice must expand.

Evidence of this can be seen by comparing the lattice constants of HPW samples obtained

from X-ray diffraction analysis before and after loading with the base pyridine.83'117

Expansion of the lattice to incorporate a guest molecule requires energy, reducing the

enthalpy observed by methods based on differential measurements of heats of adsorption

and resulting in a lower estimation of acidity.

Reported in this chapter is new a method for the determination of the global

acidity of a solid acid. Chemisorption of the probe TEPO results in a chemical shift that

is only dependent on the strength of the acid site and not the Bronsted or Lewis nature of

the site. We discuss the observed static acidity of HPW compared to acidity determined

by calorimetric titration of this solid. The chemical shift behavior of TEPO on solid

neutrals and bases is also discussed.


The solid acids and bases were loaded with TEPO by combining a known mass of

1.0 M TEPO in anhydrous pentane solution with a known mass of the solid. The loading

is determined by the ratio of the moles of TEPO, determined by using a solution density

of 0.679 g ml' and the molar mass of TEPO of 134.16 g mol ', to the mass of the solid.

For example, a TEPO loading of 0.74 mmol TEPO g-' solid was obtained by combining

approximately 100 mg of the TEPO solution to 200 mg of the solid. A volume of 2 ml of

anhydrous pentane was also added to the samples to facilitate mixing. The slurries were

allowed to equilibrate for twenty to thirty minutes with additional time for the zeolite

samples. The samples were then dried in a vacuum oven maintained at 500C. Our

preferred method, however, was to individually mix a small mass of TEPO dissolved in

4-5 ml anhydrous pentane with the solid, allowing twenty to thirty minutes for

equilibrium to be established, and then treating the sample in the vacuum oven.

The MAS spectrum of each sample was acquired on either a Varian Unity 500

MHz spectrometer operating at 202.270 MHz or a Bruker Avance 400 MHz spectrometer

operating at 161.976 MHz. Both spectrometers were equipped with Jakobsen style MAS

probes operated at spinning speeds of 4-6 kHz (Unity) or 10-12 kHz (Avance). A simple

one pulse-acquire sequence was employed with a pulse corresponding to the 900 flip

angle of 22 lts (Unity) or 3.0 gs (Avance). A recycle delay of 30 s was found to allow

complete spin lattice relaxation between pulses. The spectra were referenced to an 85%

phosphoric acid external standard.

Preparation of the Solids

Many of the solid acids used in this report were supported on a treated silica gel.

The preparation of this silica gel can be found in Chronister et al.74 The preparation

includes washing the silica with aqueous HCI, 30% hydrogen peroxide, and deionized

water followed by a heat treatment. This preparation procedure is believed to increase the

number and strength of silanol sites on the silica surface.74

Silica supported aluminum chloride,ll8"119 (silica-gel)nAl'Cl12, is a strong Bronsted

acid that was prepared by refluxing the silica gel described above with anhydrous

aluminum chloride in carbon tetrachloride. This solid was prepared and stored under an

inert atmosphere. The antimony (V) analog, (silica-gel)nSbVCl3, was prepared in an

identical manner, except trichloromethane was used as the solvent.

Silica supported sulfated tungsten oxide, (silica-gel)nWO3SO3, is also a strong

Bronsted acid.78 The preparation of this solid was performed in two steps. The first step

was refluxing tungsten hexachloride with silica gel for 24 hours. The solid was then

collected and dried in a 140 OC oven for one day. The supported tungsten oxide was

cooled to room temperature and then washed with 1 M sulfuric acid. It was activated at

200 C in flowing air prior to use.

The sulfated silica gel solid was prepared by washing silica gel with 1 M sulfuric

acid. This solid was activated at 200 C in flowing air prior to use.

The sol-gel sample was prepared by hydrolyzing tetraethyl ortho silicate in the

presence of a catalytic amount of HCI at 80 C. The resulting gel was crushed and

calcined at 400 C. This solid was found to have a very small number of weak acid sites.

It had a surface area of approximately 400 m2 g-.2

Zeolites HZSM-5 and H-Mordenite were prepared by ion exchanging the solids

with 3 M ammonium chloride in deionized water. Both solids were then treated at 400 C

in flowing air for four hours and then for 12 hours under vacuum. The furnace was then

cooled and filled with dry nitrogen and the samples were taken to an inert atmosphere

glove box where they were stored until used. Zeolite TS-1 was prepared by heating it to

200 OC in flowing air for two hours, and then in vacuum for twelve. The furnace was

cooled and transferred to the glove box as described above.

The preparation of 12-tungstophosphoric acid will be described in detail in

Chapter 3.

The method of preparing the 1:1 adduct of TEPO with SbCI5 was similar to one

found in the literature for the preparation of a 1:1 adduct of SbCl5 with

triphenylphosphine oxide.120 In separate vials 0.124 g of TEPO and 0.276 g SbCI5 were

dissolved in hot anhydrous CC14. After full dissolution, the hot CCl4/TEPO solution was

transferred by syringe to the vial containing the hot CCl4/SbCl5 solution. A clear white

solution formed that contained a brown oil. The oil was extracted by syringe, allowed to

cool, and the CCI4 was evaporated. Both the solution and the oil yielded a white solid

that was stable in air. The oil yielded the bulk of the solid, and this portion was used for

the NMR studies. The same procedure was followed to produce the 1:1 adduct of TEPO

with TaCI5 (0.128 g TEPO and 0.309 g TaCIs) and the 2:1 adduct with A12C16 (0.305 g

TEPO and 0.155 g A12C16). The 'H MAS NMR of 2:1 TEPO:Al2C16 showed only the

ethyl protons.

Results and Discussion

The Acidity Scale

A new solid acidity scale based on the change in the chemical shift of

chemisorbed TEPO has been discovered.121 This scale is based on the change in 31p

MAS NMR chemical shift of TEPO from the shift of physisorbed TEPO, a A5 value. The

A5 value is calculated by subtracting the chemical shift of physisorbed TEPO, estimated

as 50.0 ppm, from the chemisorbed chemical shift. The A6 values for all of the solid

acids measured are reported in Table 2-1. By referencing all the shifts to physisorbed

TEPO the van der Waals forces component of the chemical shift is significantly reduced,

making this an excellent, one parameter measure of the donor-acceptor interaction

between TEPO and the acid site.

Table 2-1. The A8 Scale.

Acid or Adduct A8
(silica-gel)nSb C13 Site 1 66.6
Site 2 48.0
Site 3 36.4
Site 4 22.6
(silica-gel)nAlm'Cl2 52.0
1:1 TEPO:TaCI5 51.3a
HPW 45.8
1:1 TEPO:SbCl5 41.7c
Supported Sulfated Tungsten Oxide 41.7
HZSM-5 Site 1 39
Site 2 7
2:1 TEPO:A12C16 36.6
Sulfated Silica Gel 35.9
Tungstic Acid Hydrate Site 1 27.6
Site 2 13.4
H-Mordenited 25.6
Silica Gel 6.0
Sol-Gel 4.5
NaCl*6H20 2.7
NaOH -1.2

a Various solids with different acidities.
b Identified through the spinning side bands.
c This is a minor peak with less than 10% of the total intensity.
d TEPO can probably only access the large cavities where the weaker of
two acid sites can be found.

The comparison of the A6 scale to calorimetric measures of acidity results in a

linear relationship (excluding 12-tungstophosphoric acid vide infra). The correlation of

A5 to the Cal-Ad and slurry calorimetry scales in not fortuitous (Figure 2-3). Both

experiments measure the extent of interaction of a basic probe molecule with a solid acid

site. Stronger acid sites will cause greater interactions with the probe, decreasing the

electron density about the central phosphorus atom in TEPO or increasing the enthalpy of

interaction with pyridine. The chemical shift of TEPO will consistently increase with the

acid strength until oxygen atom transfer or full protonation occurs. In the case of oxygen

atom transfer, chemical shifts similar to that expected for triethylphosphine should be

observed. In the case of full protonation, no further increase in the chemical shift will

occur, as the conjugate acid of TEPO, HTEPO+, can no longer easily donate electron


Fitting the points where one acid site was detected resulted in a straight line with

an excellent r2 value of 0.9855. The correlation of these single points allows us to

assume with a degree of certainty that both the pyridine and TEPO are probing the same

acid sites (i.e., the strongest site detected by pyridine on a particular solid is the same as

the strongest site detected by TEPO). The correlation considering all of the acid sites is

improved slightly over the previous trend: r2 = .9862. The use of a second order (or

higher) polynomial results in a better fit (r2 = .9879); however, within experimental error

it cannot be distinguished from a first order fit, leading to a choice of the linear fit.

The derivation of a relationship between the change in chemical shift and enthalpy

of adsorption is not known. However, there are two empirical methods to justify a

relationship between the two values. The first comes from Drago's E and C method.122



60-: -1
HPW Sj -2 AC13
50 :

40 a-3iS HZSM3-5 one
30 S -4 803-SG

20 B- HZSRI-5 two
Ssol- A 1.03(-AH) -0.045 ppm
S^silicagel R=0.98
0 10 20 30 40 50 60 70

-A H (kcalmol)

Figure 2-3. The correlation between calorimetry and AS. The acids are: sol-gel, a low
acidity silica solid; silica gel, a typical silica gel; HZSM-5 one and two, the acid sites of
zeolite HZSM-5; Sb-1 through Sb-4, the four acid sites of silica supported antimony
pentachloride; S03-SG, sulfuric acid washed silica gel; SgWO3SO3, silica supported
sulfated tungsten oxide; HPW, 12-tungstophosphoric acid; AIC13, silica supported
aluminum trichloride. The *points are enthalpies derived from the Cal-Ad method,
while the U are derived from slurry calorimetry alone. The A represents 12-
tungstophosporic acid. The line was fit without 12-tungstophosphoric.

E and C uses two parameters to describe the donor-acceptor interaction between two

molecules (Equation 2-2): 123

A = EaEb + CaCb W (2-2)

The nature of AX can be enthalpies of reaction, spectral shifts, or even rate constants. The

first set of parameters, Ea and Eb, represent the electrostatic component of the interaction,

while the second set, Ca and Cb, represent the covalent interaction. The W term accounts

for any barrier that must be overcome to allow the acid and base to interact, such as the

breaking of a dimer. The E and C model indicates that the same factors that lead to an

evolution of an enthalpy of interaction will lead to a proportional change in the spectral


Literature precedence has established a second link between acidity measured by

enthalpies of interaction and changes in chemical shift. It can be shown that acidity is

proportional to strength of the acid-base bond formed (Equation 2-3): 1172-73,77.94

-AHacid-base oc acid-base bond strength (2-3)

And that the change of chemical shift of TEPO or 2-'3C acetone in solution is

proportional to the bond strength of the donor-acceptor pair in the solution (Equation 2-


A831P, AI' C c acid-base bond strength (2-4)

These two equations allow the relationship between chemical shift, a spectral parameter,

and enthalpy of interaction, a thermodynamic parameter to be written:

A~31P, A5'3C oc acid-base bond strength oc -AHacid-base (2-5)

This relationship allows us to look for a correlation between A6 and calorimetry.

Evidence for the Measurement of Global Acidity

The A5 scale is a global acidity scale34--it measures acidity without regard for the

type of acid site the probe is bound to. The total donor-acceptor interaction is the primary

contribution to the observed change in chemical shift. The 3P MAS NMR spectra of

TEPO on the solid acid (silica gel)nSbVC13 (Figure 2-4) demonstrates this phenomenon.

This solid contains four acid sites, at least one of which is a Lewis site as determined by

pyridine adsorption IR.124 Additionally, the A8 of the four acid sites observed all

correlate with differential heats of pyridine adsorption obtained from slurry calorimetry

(reported in Chapter 5).

Additional evidence for a global measurement of acidity comes from TEPO

adsorbed on a series of Lewis acids. The spectrum of the adducts formed between TEPO

and tantalum pentachloride (TEPO:TaCls) is presented in Figure 2-5. Each adduct

measured (TaCls, SbCI5, Al2C16) had a positive A5 at a different chemical shift. The

position of this peak indicates that the Lewis region predicted by Rakiewicz et al.44 (a

small, negative A6) for trimethylphosphine oxide cannot be extended to TEPO. Such

negative A6 values can be shown to be the very weak Lewis acid TEPO interacting with

strong basic sites on the surface videe infra).

By changing the loading of TEPO on TaCIs, a new of shift can be observed. The

variation in shift is attributed to the nearest neighbor effect. In samples with less than a

1:1 TEPO: TaC15 ratio, induction effects that will increase the acidity of the acid site are

likely to be observed. Two different loadings of TEPO on TaCl5 demonstrate this effect

(Table 2-2).

/ /
O---H--O CI Si
Et--.P Clo--Sb--0

EtEt ci 0----

k' IN
1 44i\ j ,

160 140 120 100 80 60 40 20

Figure 2-4. Phosphorus-31 MAS NMR of TEPO on (silica gel)nSbVCl3. The four acid
sites observed correspond to the four acid sites detected by slurry calorimetry.
Additionally, this solid contains both Lewis and Br|nsted acid sites. One of the proposed
Bronsted acid sites appears in the inset.

i f i


1 i i i

S1 '
I i'

14 131 3IZ Ill 1lt o 8ii 7l It 5 4I 30 24 1t I pI p

Figure 2-5. Phosphorus-31 MAS NMR of TEPO adsorbed on TaCls. The four lines
observed are each a different solid form of the adduct with different bonding between the
metal and the phosphine oxide. A fifth spectral line can be identified through the
spinning side bands.

Table 2-2. The inductive effect in Lewis haloacids.

Compound Compound Observed Chemical Shift A6
(ppm) (ppm)

1:1 TEPO:SbCl5 TEPO:SbCl5 87.5 37.5

91.7 41.7

2:1 TEPO:A12CI6 TEPO:AIC13 86.6 36.6

88.0 38

91.8 41.8
Various TEPO:TaCI5
1:1 TEPO:TaCl5 -92.2 -42.2
96.3 46.3

101.3 51.3

TEPO:TaCI5 101.6 51.6
1:10 TEPO:TaCl5
TEPO:TaCl5-TaCl5 110.2 60.2

Static and Differential Measures of Acidity

Chemisorption of a probe molecule results in a static measurement of acidity.

This static interaction is a measurement of the acidity of the site that a molecule

experiences while bound to it. Differential heats of adsorption may indicate a different

acidity than such a static measure depending on the acid under study. The acidity of

12-tungstophosphoric acid has been measured calorimetrically both in solution77 and in

the solid state (Table 1-3).83 Most results indicate that HPW is a powerful solid acid;

however, there is much dispute over the true strength of this solid. The Cal-ad acidity of

the strongest site is 32.7 kcal mol ', the expected A6 is therefore 33.6 ppm. The observed

A5 of HPW, however, is 45.8 ppm (predicted enthalpy of 44.5 kcal mol'), much greater

than the expected value. The A6 scale predicts that HPW is a much stronger acid than

measured by the Cal-ad techniques.

The source of this discrepancy between calorimetry and A8 can be readily

accounted for with an understanding of the nature of the two experiments. The heat

measured from slurry calorimetry is the heat of interaction of the base with the acid site

less the heat required to displace a solvent molecule from the acid site, less the heat

required, in this case, to open up the lattice for the basic probe to enter (Figure 2-6). The

heat of interaction of the base with the acid site contains contribution from both donor-

acceptor interactions and van der Waals forces. It can be shown that with the correct

choice of basic probe and solvent, the van der Waals component can be removed from the

measured enthalpy by the displacement of the solvent molecule.73123

The reduction of the measured enthalpy by opening the lattice cannot be

accounted for without a measure of the lattice energy of the initial and final states. The

reduction of the exothermic energy by the opening of the lattice will lower the estimate of

the acidity of the solid by the amount of energy required to open the lattice. The initial

adsorption of TEPO contains all of these energetic steps, but the measure of the acidity by

31P MAS NMR occurs after all of these processes have been completed. The A8 is a

measure of the static interaction of TEPO with the acid site after the lattice is opened and

the molecule is fully adsorbed at the site. The difference between the enthalpy estimated

from the change in chemical shift of TEPO and the enthalpy measured by the Cal-Ad

method leads to an estimated energy of 11.8 kcal mol'1 required to open the HPW lattice

to include a pyridine sized molecule

Pyridine Adsorption (Differential)
Total Energy Measured

I Lattice
S ^(Void

TEPO (Static Measure)
K Void

Keggin Unit



Figure 2-6. A comparison of two different methods of measuring solid acidity. The top
scheme demonstrates the energy changes measured during differential calorimetry of 12-
tunsgstophosporic acid (HPW) with pyridine. The first step is the endothermic opening of
the lattice to accommodate the probe molecule. The second step is the exothermic
enthalpy for adsorption of the probe. The total energy measured is the sum of both these
steps. The lower scheme is for the measurement of the chemical shift of chemisorbed
triethylphosphine oxide. Only the interaction between the probe and the acid site is


( 1

Triethylphosphine Oxide in Zeolites

Previous measurements of trimethylphosphine oxide in zeolites HY and

dealuminated HY have been carried out by Zalewski et al.32 and Rakiewicz et al.44and of

zeolite USY by Rakiewicz et al. They report that the shift observed is very sensitive to

the type of acid site to which the probe is adsorbed. We extend their observations of

phosphine oxide behavior to include the measurement of zeolites HZSM-5 and H-

Mordenite with TEPO and zeolite TS-1 with trimethylphosphine oxide.

The 31P MAS NMR spectrum of 0.1 mmol g' TEPO on zeolite HZSM-5

demonstrates the presence of two acid sites. The structure of HZSM-5 is large enough to

allow complete penetration by the probe into the large chambers and channels;125

however, a significant amount of time is required to allow TEPO to penetrate the pores of

the zeolite. The first acid site is a low population strong site that has been assigned as a

Bronsted site through pyridine IR adsorption and Cal-Ad.73 The second site has been

shown to be a weaker hydrogen bonding site of significantly larger population. The

acidity determined for these two sites are in good agreement with results from

calorimetric measurements (Table 2-3).

The spectra of TEPO adsorbed on H-Mordenite shows the presence of only one

acid site. The A8 value for this moderate strength site is 25 ppm. It is probable that

TEPO cannot penetrate the small channels of H-Mordenite where additional acid sites are

believed to exist.

Zeolite TS-1 is a very weakly acidic zeolite.79 The acidity of TS-1 arises from

defect sites resulting in tricoordinate Ti4+. The acidity of this solid was measured with

trimethylphosphine oxide. Two overlapping spectral lines were observed (Figure 2-7).

Table 2-3. The Cal-Ad and A6 results for HZSM-5.80

Site 1

Site 2

-AHave (kcal mol") 42.1 0.8 8.6 3.8

Population (mmol g-') 0.0415 + 0.0057 0.53 0.01

K 4.9 + 2.3*106 2.3 + 2.0*106

A5 (ppm) 39 7

Figure 2-7. Phosphorus-31 MAS NMR of trimethylphosphine oxide on zeolite TS-1. The
peak at about 30 ppm corresponds to physisorbed trimethylphosphine oxide.

1 I


2 6

140 20 00 0 4 \0
< -;------"~-;' I I\

l^O 120 100 8C 60 40 20 C ppmn

From limited studies of the behavior of trimethylphosphine oxide on solid acids (Figure

2-8), the change in the chemical shift of trimethylphosphine oxide can be roughly

compared to the A8 values for TEPO but with a chemical shift for the physisorbed probe

of -30-34 ppm. This assumption leads to the conclusion that the acidity of TS-1 is very

mild, and arises from titanium sites scattered throughout the structure (there are twelve

different types of titanium centers in TS-1) and possibly from surface silanols.

Interaction of Triethylphosphine Oxide with Solid Bases

The adsorption of a large amount of TEPO onto a strong, basic solid results in a

small negative A8 shift (Table 2-4). The shift is only observed at very high loadings. It is

Table 2-4. The A6 value of TEPO on several bases.

System A8 (ppm)

NaOH -1.36



y-alumina -1.35

believed that the cause of the shift is an acid base interaction between the TEPO and the

basic sites of the solid. In this case, TEPO is acting as a very weak acid and accepting a

small amount of electron density from the basic site (Figure 2-9). A similar result is

80 -

70 -


50 -

40 31P = 0.95(-AH) + 34.892 ppm
R2 = 0.9521
30 -----
0 10 20 30 40 50

-AH (kcal mol')

Figure 2-8. The chemisorbed chemical shift of trimethylphosphine oxide compared to
calorimetric measures of acidity. The two chemical shifts of trimethylphosphine oxide
interacting with zeolite TS-1 (represented by I) were averaged together to estimate the
shift of the acid site found by Cal-Ad for this solid. The Cal-Ad method cannot separate
two sites that have similar acidity. The chemical shifts reported by Rakiewicz et al. (54)
for zeolite HY were used for this solid.

reported for trimethylphosphine oxide adsorbed on y-alumina,44 although the shift was

attributed to adsorption onto a Lewis acid site. Upon exposure to water, the resonance

observed in Rakiewicz et al. disappeared. Exposure to water destroys the Lewis acid

sites of y-alumina, as reported, but will also greatly diminish the basicity of the solid.

Figure 2-10 is the spectrum of TEPO adsorbed on potassium hydroxide.

The Use of Trimethylphosphine Selenide to Explore Solid Acidity

The use of trimethylphosphine selenide (TMPSe) to measure solid acidity was

explored. It was hoped that both the 3'P and 77Se chemical shift information could be

Phosphine Oxide



Figure 2-9. A MO diagram for the acid TEPO interacting with KOH. The negative A5
observed for TEPO adsorbed on the solids y-alumina, potassium hydroxide, calcium
hydroxide, and sodium hydroxide to be representative of an interaction with a basic site.

Figure 2-10. Phosphorus-31 MAS NMR of TEPO on potassium hydroxide. The small
peak at around 48.5 ppm represents TEPO interacting with a basic site.


urS~nri5, .1-.. J. I'A


obtained from one probe molecule. However, the solid was unstable, decomposing at

various rates depending on the components of the solid it was adsorbed on producing an

insidious fine red powder polymericc selenium?) and a stench (trimethylphosphine). All

attempts to measure solids with any sulfur content failed due to exchange of the selenium

with the sulfur moiety of the acid. The use of TMPSe must be restricted to solids without

sulfur, and serves only to detect the presence of acidity, not as a measure of acid strength.

The coupling constants of physisorbed TMPSe may be a measure of the surface polarity,

and further research into this phenomenon would help to improve understanding of the

physiochemical properties of surfaces.

The chemical shift of TMPSe chemisorbed on a solid appears to be limited to one

shift at about 30 ppm. However, chemical shifts from the decomposition products

(trimethylphosphine and trimethylphosphine oxide) can be observed (Figure 2-11).

Information can be gained from the study of the coupling constant, J(3'P-77Se), of

physisorbed TMPSe. The coupling constant is very sensitive to the distance between the

coupled nuclei. Polarization of the phosphorus-selenium bond will lead to small changes

in the bond length that will be observable in the change of J("P-77Se). Since the coupling

constants varied with the adsorbent (Table 2-5), they may prove useful as a measure of

the polarity of a surface.


The basic probe molecule triethylphosphine oxide is capable of measuring the

acidity of solid acids using simple one pulse 31P MAS NMR. The shift of TEPO has been

shown to correlate well with acidities measured with Cal-ad or slurry calorimetry. The

nature of the acid site, as Lewis or Bronsted, does not affect the shift of TEPO. The value

90 --- 70--- V7_0_0_0_2___0______-__-__0_p.

Figure 2-11. The 31P MAS spectrum of TMPSe on tungstic acid (H2WO4). The peaks at
51, 58 and 72 ppm are decomposition products. The spinning side bands are at -60 ppm
and 88 ppm. The peak at 14 ppm is assigned to physisorbed TMPSe and shows coupling
to 7Se (J = 662 Hz).

Table 2-5. The coupling constants (J 31P-77Se) of TMPSe physisorbed on solids.

Solid J (3P-7Se)

Crystallized TMPSe 638 Hz

Silica Gel 644 Hz

HZSM-5 646 Hz

Tungstic acid (H2W04) 662 Hz

10 0 -10 -20 0 -0 -40 -50 ppm

90 80 70 60 50 40 10 2()

of measuring acidity through several techniques has been demonstrated by a comparison

of the acidity measured by differential heats of adsorption to the static method involving

3'P NMR to determine a "lattice penetration" energy for pyridine into HPW. Reports of a

Lewis acid region observed for TMPO have been shown to be inappropriate for TEPO.

Small, negative A8 values have been demonstrated to be interactions with basic sites.

Overall, this is a powerful, accurate and fast tool for characterizing and exploring solid


Triethylphosphine oxide is an ideal probe for the measurement of the global

acidity of solid acids. It is capable of identifying different strength acid sites on the

surface. However, simple one pulse experiments are incapable of identifying the nature

of the acid site. TEPO lacks the chemical shift ranges dependent upon the nature of the

acid site observed for TMP. It has been demonstrated elsewhere44 that the population of

acid sites can be determined with phosphine oxides. The combination of all of these

traits makes phosphine oxides ideal for the characterization of solid acids.


Interest in heteropoly acid catalysts has been raised by their interesting results in

laboratory scale work and their industrial applications. Their use as homogeneous and

heterogeneous catalysts for redox and acid catalyzed reactions have been extensively

reviewed.16'126-132 Promising use of heteropoly acids in the synthesis of fine chemicals

and for biological applications has extended the range of use of these acids.133-134

The crystal structure of 12-tungstophosphoric acid (HPW) hexahydrate has been

determined by X-ray113-"4 and neutron diffraction115 analysis. The Keggin anion is

highly symmetrical and can be considered a tetrahedron about the central P04 group.

The central phosphate is surrounded by twelve W06 octahedra. These octahedra are

arranged in four groups of three edge-shared W3013 groups called W3 triads (Figure 3-

1).116 The Keggin anions are close packed on a body centered lattice with the central

anion turned 90 degrees to its eight neighbors. The counter ions to the -3 charge on each

Keggin ion of HPW are H502+ located in the tetrahedral vertices. Extensive dehydration

will leave a proton in the same position as the H502+ ion.127

An additional feature of the solid HPW hydrates are their high proton

conductivities. The proton conduction in solid HPW is comparable to the proton

conduction of water. The low activation energy for conduction is attributed to the

location of waters of crystallization and hydrated protons in the interstices of the large,

Figure 3-1. The structure of the Keggin anion. This particular Keggin ion, PWi2040 has
a negative 3 charge.

globular anions.135-137 The high mobility of the proton is thought to increase the acidity

of this compound by distributing any anionic charge built up throughout the solid (Figure

3-2). The catalytic properties of this solid can be tailored at atomic and molecular levels

by exchanging the hydrated protons with other cations. Changing the counter cation

modifies the properties of the heteropoly acid and can result in a change in both structure

and activity.

Recently, CsxH3.xPWi2040 derivatives (abbreviated as CsxH3.xPW) have gained

attention because of their claimed super acidity and shape-selectivity in alkylations

reactions.138-140 Unlike HPW, the CsxH3.xPW salts have limited solubility in water or

organic solvents and are used exclusively as heterogeneous catalysts. The cesium salts

Figure 3-2. Delocalization of charge in 12-tungstophosphoric acid. Abstraction of a
proton by a base results in a negative charge that is distributed throughout the solid by
proton conduction.

show higher catalytic activity and higher thermal stability in some reactions than

HPW.126 The cesium salts are thought to have a similar structural motif as HPW.

The structure of Cs3PW is the same as HPW but with Cs in place of the HsO2+

groups and larger lattice constants from the larger spaces between Keggin units. The

structure of the other CsxH3.xPW compounds is not completely established. The

proposed structures of cesium complexes with x<3 are reported to be mixtures of HPW

and CsxH3-xPW. A study of the precipitates from the titration of HPW with Cs2CO3

concluded that in the range of 0
Cs2HPW.142 Compounds prepared in the range of 2
HPW, although Cs2.5Ho.5PW is considered to be a mixture of Cs2HPW and Cs3PW.

Reported XRD patterns for CsxH3.xPW (x = 0, 1, 2, and 3)143 claim that CsH2PW is a

mixture of HPW and Cs2HPW which is transformed into a nearly homogeneous acidic

salt after thermal treatment of the precipitate at 300 C. The lattice constants for those

cesium salts indicate all of them belong to cubic systems.141

The cesium salts were studied by 31P NMR.39 Each loading of cesium has been

found to have a distinct MAS NMR spectrum. Literature reports that the 31P spectrum of

Cs2HPW is composed of four peaks, which were interpreted as a mixture of four species

(x = 0, 1, 2, and 3). Results for Cs25Ho.5PW also claim that this compound is a mixture

of Cs2HPW and Cs3PW. Supported HPW on Cs3PW presents the same 3P spectrum as

Cs25Ho.5PW,139 indicating in homogenous exchange between H' and Cs' during the

heating process. In those studies'41-143 the cesium derivatives were prepared similarly,

although the thermal treatment before analysis of the solids was not always the same.

The location of the pyridine binding sites in HPW will be identified. Pyridine

will be shown to be incapable of completely penetrating the solid, but that it can cause

the widening of certain crystalline layers allowing for an interaction with various protons

in the structure. The effect of cation exchange on the phosphorus-31 chemical shift of the

cesium salts will be reported.


Materials Preparation

The solids studies were prepared by J6se A. Dias,83 but since the method of

preparation is of critical importance the relevant experimental steps will be summarized

here. Elemental analysis of the HPW (Aldrich) revealed 16 moles of water per mole of

HPW, therefore a drying procedure is required to obtain the hexahydrate. The drying

procedure was to heat the solid at 160 C for 4 hours under vacuum. It has been shown

that heating HPW at that temperature yields the anhydrous acid.77

Spectral Analysis

The 3P MAS NMR spectra of the solids were taken on a Varian Unity 500 NMR

operating at 202 264 MHz. The magic angle was tuned using K79Br. A recycle delay of

2 seconds was used. The acquisition time was set to 0.02 ms. Signals were indirectly

referenced to 85% phosphoric acid. At least ten minutes of signal averaging were

allowed for all spectra.

Calorimetric Titrations and Adsorption Measurements

J6se A. Dias performed all calorimetric titrations. The methods he used can be

found in his dissertation81 or in Reference 83, a paper co-written by this author. The

calorimetric data is instrumental in the data analysis presented in this chapter, particularly

for HPW. The exact experimental procedure does not differ significantly from the

calorimetric procedure presented in Chapter 1. Cyclohexane was chosen as the solvent to

make the slurry because it is a non-interacting solvent in which the solids are insoluble.

Time dependent UV adsorption experiments were performed to insure that equilibrium of

the pyridine with HPW was achieved in the time scale of the calorimetric titration.

Results and Discussion

Spectroscopic Results of 31P MAS NMR and XRD for HPW and CsH3,PW

The 31P MAS NMR spectra of HPW and several Py-HPW species with different

ratios of pyridine to the total number of acid sites were obtained. The spectrum of

anhydrous HPW shows a single peak at -12.0 ppm (Figure 3-3), corresponding to

literature values of -10.9 ppm,141 -12.4 ppm,'44 and -11.1 ppm.98 The product spectrum

of the first two sites titrated with pyridine, 1.04:0.24 mol ratio of HPW to pyridine, gives

only a single resonance at -17.5 ppm (Figure 3-4), which changes to -16.5 ppm upon

Figure 3-3. The 31P MAS NMR of 12-tungstophosphoric acid. There is only one peak
(-12.0 ppm), the other peaks are spinning side bands.

exposure to water. Proton conduction acts to distribute the anionic charge throughout the

structure, making the electron density on all the Keggin units similar on the time scale of

the NMR experiment. Spectra of 1.04:0.08 and 1.04:0.12 mole ratio of HPW to pyridine

also show a single peak at -17.5 ppm providing further evidence that the protons are

mobile in the structure.

Exchange of protons for less acidic cesium atoms causes a small increase in the

electronic density at the phosphorus core. The resulting increase in electron density can

be observed in the 31P MAS NMR of the solids CsxH3.xPW (x=0, 0.9, 1.8, 2.4, 2.7). Each

40 20 0 -20 -40 -60 -80

Figure 3-4. The 1P MAS NMR of 0.24 mmol pyridine on 1 gram of
12-tungstophosphoric acid.

spectrum shows a single peak that shifts to lower field going from HPW to Cs3PW. The

presence of only one narrow peak in the NMR, unlike the reported literature results,141

demonstrates that the method of preparation produces homogenous compounds.

The chemical shift of the cesium salts changes regularly with increasing cesium

substitution (Figure 3-5). The shift is due to the increasing electron density on the

Keggin units due to the greater charge separation afforded by the larger cesium cations.

In essence, the cesium cations hold a larger portion of a positive charge than the solvated

protons, allowing for a greater negative charge to build up on the Keggin units. The

result of this buildup can be observed through the shielding of the phosphorus atoms.

Increasing the amount of cesium in the structure results in a gradual decrease in the

chemical shift.


15 *

10 \
0 0.5 1 1.5 2 2.5 3
Cesium Loading (x-= to 3)

Figure 3-5. The chemical shift of the central phosphate changes with cesium loading.
The data can be fit by a second order (or higher) polynomial indicating that there are at
least two factors which influence the chemical shift videe infra).

Locating the Acid Sites in the HPW Crystal Structure

The XRD spectra of HPW gives a pattern similar to that reported in the

literature.' 7 The main peaks, related to the strongest reflections, are present at 28 =

10.40, 25.40, and 33.50 (Figure 3-6). The broader peaks indicate loss of crystallinity, as

reported for HPW dried at different conditions.105'145 The XRD from the reaction of 1.04

mmol of HPW with 0.24 mmol of pyridine shows a few new peaks (Figure 3-7). The

most intense peak at 20 = 10.40 splits up into two less intense shoulders at 20 = 10.300

and 10.520. Two new reflections at 20 = 9.270 and 9.980 are present in the powder

pattern. These smaller angle reflections indicate a larger d-spacing in the unit cell to

accommodate the pyridine molecules attached to the protons. Other reflections shifted

slightly with significant changes in intensity. No large-scale change was observed in the

The XRD data for 12-tungstophosphoric acid.

400 -


100 -

! I

25 30 35

i i i 1

5 10 15 20


Figure 3-7. The XRD data for 12-tungstophosphoric acid with one pyridine molecule
adsorbed per acid site. The smaller 20 values compared to pure HPW indicate an
expansion of the lattice.


20 25 30

Figure 3-6.

_ __

XRD pattern when enough pyridine was adsorbed onto HPW to account for all of the

strong acid sites determined by Cal-Ad.83

The Cal-Ad results for HPW83 (reported previously in Table 1-3) demonstrate that

the heat evolved from adsorption of pyridine indicates that only a fraction of the 1.04

mmol of protons in one gram of HPW are titrated. After 0.24 mmol of pyridine was

added (about 24% of the moles of protons in a gram of solid), heat evolution essentially

stopped along with the adsorption of pyridine by the solid. The number of surface

protons reported for solid HPW is 0.008 mmol g-'.146 This value can also be calculated

by assuming that the Keggin ions are spherical and that the surface area is 5 m2 g-1

(calculated 0.0079 mmol surface sites per gram HPW). Thus, in addition to reacting with

the surface protons, pyridine also penetrates into regions of the solid. This result is

consistent with Misono131 who reports that the adsorption of polar molecules by HPW

leads to catalytic reactions in the bulk of the crystal as well as on the surface. This

behavior is analogous to a concentrated solution and is described as a pseudoliquid

phase131,147 where non-polar molecules are not absorbed.

In order to make sure that the incomplete reaction was not caused by forming a

layer of product on the surface that could not be penetrated by pyridine, the sample was

ground to finer particles inside an inert atmosphere box and titrated with pyridine.8 An

identical isotherm was produced demonstrating that diffusion due to restrictions of

particle size is not a problem, and suggesting that pyridine was not able to access certain


In the tertiary structure, the protons are located at the center of each edge and

face, while the Keggin anions are arranged in a cubic body centered manner. The unit

cell (Figure 3-8) consists of a total of two anions (8 anions centered on the vertices and 1

anion in the center of the cube) and six protons (6 positioned on the center of the faces

and 12 at the edges). The ratio of anions to protons (2:6) corresponds to the molecular

formula. The Cal-Ad results suggest that the total number of protons titrated (0.24 mmol

per gram) could be assigned to the number of protons in every even plane of the lattice

(EOO). The Cal-Ad determined first site would be the face-centered proton,

corresponding to about 8% of the total protons in the cell. Summing over all the face

protons in the even planes in one gram of HPW would yield 0.08 mmol of protons for

this site (Figure 3-9). The second site closely corresponds to the edge protons in this

plane (Figure 3-10). These edge protons represent 16% of the protons of each unit cell.

Again, summing over all the edge protons in the even planes in one gram of HPW would

yield 0.16 mmol of sites. Both the individual components and the total number of moles

determined in this manner correspond with the Cal-ad results of nl= 0.08 mmol g~' and

nz= 0.16 mmol g~' (Table 3-1).

Table 3-1. Comparison of the Cal-ad site populations to the position of the
protons in the crystal structure of HPW.

Bulk Crystala Even Planesa (EOO) Cal-ad PopulationC
Site One (mmol g-')b 0.35 0.08 0.079 0.002
Site Two (mmol g-I)b 0.69 0.16 0.16 0.05

a. Site one for the Bulk Crystal and the Even Planes are the face centered protons. Site
two represents the edge protons.
b. To convert from mmol site (gram HPW)l1 to mmol site (mmol HPW)-' multiply each
value by 2.880 g mmoln' HPW.
c. The units for Cal-ad Population are the number of acid site of a particular type that
interact with the probe molecule pyridine for an average gram of material (from
Reference 83).

H + H 02+
H 2 H 2


HH502 2+ H

Figure 3-8. A representation of the unit cell of 12-tungstophosphoric acid. The shaded
spheres represent the Keggin anions.

Figure 3-9. A representation of site one for pyridine titration of HPW. Pyridine can only
interact with the face protons of every even lattice plane.

Figure 3-10. A representation of fully titrated HPW. The site two protons are all the
edge protons in the even planes.

The assignment of the protons that can react with pyridine to the faces and edges

of the unit cell account for the two sites found in the titration of HPW. The unit cell of

this preparation is thought to be identical to that of the hexahydrate, bcc. Pyridine

initially reacts with one of the six available face protons, producing site 1 of the titration.

With one equivalent of pyridine attached to the structure, additional pyridine is in effect

titrating (CsHsNH)+ H2PW-. The second site has a lower enthalpy, AH2= -19.6 kcal

mol-l but a higher site density. Considering the fraction that each proton shares in the

unit cell, the ratio of face protons to edge protons (1:2) titrated by pyridine gives the Cal-

Ad ratio of n\ to n2 (1:2). The lower enthalpy for site two is probably due to a loss or

lowering of the proton conduction due to the nearby pyridinium cation. The sum of the

nl and n2 value from Cal-Ad, 0.24 mmol g' confirms that protons have reacted and are

located inside the solid as well as on the surface.

The chemical interaction of the probe molecule with the solid is of great

importance. In studies conducted using microcalorimetry of gaseous ammonia

adsorption,107109 all of the protons in HPW were observed to react. The complete

penetration of ammonia into HPW can be attributed to its small size. The HPW lattice

will not have to be distorted for an ammonia molecule to enter and displace a water

molecule from the H502+ cations. Additionally, the stronger basicity of ammonia will

help to drive the reaction to completion.

Homogeneity of Cesium Distribution in Cesium-HPW Salts

Literature reports have shown that the cesium salts of HPW are mixtures of HPW

with Cs3PW. Phosphorus-31 MAS NMR has shown multiple peaks indicating the

complex nature of the precipitated, uncalcined solids. However, a thermal treatment

process developed in the Drago lab can be shown to produce a homogenous distribution

of the cesium ions throughout the structure.

For CsH2PW only 3% of the available protons in the solid are titrated which is

lower than the 24% titrated for HPW. Coupled with the 3P NMR (Figure 3-11) results

which lacks a peak for HPW, it can be concluded that the surface is coated with cesium

ions which prevent access to most of the protons in the solid. However, the bulk solid

has a regular substitution of cesium ions for protons. The location of the pyridine binding

sites in this solid, nor for any of the other cesium salts, can be determined as for HPW.

The blocking of the surface by the cesium ions prevents even penetration of the pyridine

into the structure.

The 31P MAS NMR spectrum of Cs2HPW (Figure 3-12) has only one signal that

is at a different chemical shift from pure HPW. This indicates that the bulk solid has a

homogenous distribution of cesium ions. A previous report found that Cs2HPW was a

mixture of four different species of CsxH3-.PW (x= 0, 1, 2, and 3).141 The preparation of

this salt in aqueous solution might generate a series of equilibria in solution involving up

to four different species, and when the solid precipitates all four species are present in the

solid. After thermal treatment the solvated protons and the cesium ions are completely

homogenized into the solid Cs2HPW structure.

Like Cs2HPW, solid Cs2.5Ho.5PW is reported to be a mixture of two compounds:

Cs2HPW and Cs3PW.141 Our results from 3P MAS NMR (Figure 3-13) again

demonstrate that a thermal treatment results in a homogenous solid. Neither a spectral

line for Cs2HPW or Cs3PW (Figure 3-14) was observed.

.C~r~~----. .-/ \/' \ *'' -r*
1s0 89 i 41 21 1 2 -41 -6 -a -1jz -1 l ppM

Figure 3-11. The 31P MAS NMR spectrum of CsH2PW. The peak at 14.9 ppm is the
isotropic peak while the other peaks are spinning side bands.

L30 82 HI 40 20 B

I ^- i-~-;- V;" -----
L~0 -41 -04 -80 -II -120 ,.

Figure 3-12. The 31P MAS NMR spectrum of Cs2HPW. The signal near -20 ppm is the
transmitter signal.

S* 1

1 -20 -4& -"4 -1It -124 pp.

Figure 3-13. The 31P MAS NMR spectrum of Cs2.5Ho.5PW.

It aAsl S 5 241 -2I -4l -i8 -II -UI -1i9 ppI

Figure 13-14. The 31 MAS NMR spectrum of Cs3PW. The line at -20 ppm is the
transmitter spike.

:II Io

r) II II

The single chemical shift observed for each solid indicates that thermal treatment

results in a homogenous distribution of the cesium units throughout the solid. The

change in chemical shift is probably due to a combination of electronic effects

concomitant with cesium substitution and changes in the surface area, and hence the

space between the Keggin anions (Table 3-2).

Table 3-2. The chemical shift of cesium salts of HPW.

Number of Cesiumsa Surface Area (m" g ) 831P (ppm)

HPW 0 5 12.05

CsHPW .9 1 12.5

Cs2HPW 1.8 63 14.9

Cs2.5Ho.5PW 2.2 119 15.5

Cs3PW 2.7 72 17.0

a Determined from ICP-MS.


Spectroscopic methods have led to a better understanding of the behavior of HPW

and its cesium salts. For HPW, the large number of sites (0.24 mmol g-1) found by

titration with pyridine compared to the number of surface protons (0.008 mmol g-1)

demonstrates penetration of the solid by pyridine. XRD results confirm the opening of

the lattice as pyridine absorbs on sites 1 and 2 of solid HPW. By comparing the Cal-Ad

results to the crystal structure of HPW, the limitation that pyridine can only interact with

every other plane of the HPW lattice was discovered. The proper thermal treatment of


the cesium derivatives was found to result in a homogenous solid as determined by 31P

MAS NMR, in conflict with literature reports. The change in the isotropic 31P chemical

shift with increasing cesium loading is probably due to a combination of electronic and

structural changes due to the large size of this counter cation.


The activity of a solid acid catalyst is proportional to its acidity.3 148-150 A more

acidic solid is generally able to catalyze a reaction faster than a weaker solid or at a lower

temperature. However, there is a minimum acidity that is necessary to drive any catalytic

pathway. For the production of methyl t-butyl ether (MTBE) from methanol (MeOH)

and t-butanol an acid which is both capable of dehydrating t-butanol to isobutene and

then alkylating it with methanol is necessary.149 Until, recently, MTBE was considered a

safe oxygenate to help increase the octane number of automobile fuel.5 's0 The

dehydration of tertiary butanol occurs on fairly weak acids with the coupling reaction to

MTBE occurring under more rigorous conditions. The dehydrative coupling of two

methanols to form dimethyl ether occurs under very acidic conditions. This reaction

system gives an ideal handle for an activity test measure of acidity. A stream of

methanol and t-butanol can be exposed to a surface and the products can be utilized to

measure the current acidity of the solid (Figure 4-1). If the products are monitored as a

function of time, then the acidic lifecycle of a catalyst can be determined.

The reaction mechanism for the production of MTBE (Figure 4-2) starts with the

dehydration of t-butanol to isobutene.149 The isobutene can then be protonated to form the

t-butyl cation, and the electrophilic central carbon is then attacked by a MeOH. Loss of a

proton returns the acid site to the initial condition and allows the MTBE to diffuse

+ H-Acid

SUnique Products
Strong Acid
o CH3-O-CH3 + H20
Reactor / Moderate Acid
CH30H + (CH3)3COH O-- (CH3)3COCH3 + H20
Weak Acid
i (CH3)2C=CH2 + H20

Figure 4-1. The effect of solid acidity on the reaction of tertiary butanol with methanol.
Strong acids can catalyze the formation of dimethyl ether, while moderate strength solids
produce MTBE, and the weakest solid can only dehydrate t-butanol.


+ H20


> + Acid



/H H /CH3
Net Reaction: *0 + H3 / O- CH + H20

Figure 4-2. The mechanism for the catalytic production of MTBE from methanol and

away. The production of dimethyl ether is the most prominent side reaction.4 The

mechanism for this dehydration reaction is very similar to the production of MTBE;

however, there is no initial dehydration step. The first step is the protonation of the

methanol to form an oxonium ion (Figure 4-3). The oxonium ion is intimately associated

with the acid site. A second methanol can then approach and in a concerted reaction a

water molecule is produced and the ether linkage is formed. Transfer of a proton back to

the acid site allows the dimethyl ether to diffuse away from the surface.

The current process for the synthesis of MTBE is the alkylation of isobutene by

methanol carried out over a sulfonic acid resin, Amberlyst 15.150 Tertiary butanol cannot

be used in place of isobutene with this system as the water produced reduces the acidity

of the solid (requiring higher temperatures) and eventually leads to deactivation of the

surface and corrosion of the reactor.5 An efficient process for producing MTBE from

alcohol feed stock was devised by Nicolaides et al.149 Nicolaides et al. used isobutyl

alcohol in their feed in stead of t-butanol or isobutene. Due to the association of the

protonated isobutyl alcohol with the acid site, the methanol was forced to attack the

primary carbon producing methyl isobutyl ether (MIBE), an undesirable product. It was

discovered that two separate catalyst beds were necessary to produce MTBE from this

system. If the isobutanol is first dehydrated to isobutene over one catalyst and then

alkylated with methanol by a second, the primary product was MTBE. The shift in

product specificity is due to the location of the carbocation of isobutene at the central

carbon. Methanol attack at the central carbon results in production of MTBE (Figure 4-


H3C-O/ + H-Acid + H3C- 0

H3C .0O-
0---- H ----Acid- + H20- HC
H3C o0--I


H3cO cH

Net Reaction: 2 H3C-O --- H3C / CH3 + H20

Figure 4-3. The reaction mechanism for the production of dimethyl ether from methanol.


Isobutyl Alcohol

Acid-- --H -


H3C -CH3

Methyl Isobutyl Ether

t-Butyl Cation

\ e I
----.C --- Acid'




Methyl t-Butyl Ether

Figure 4-4. The transition states for ether production. The position of the cation in the
isobutanol transition state results in the formation of methyl isobutyl ether. When the
transition state is the t-butyl cation, the ether produced is MTBE.

+ H3C--0


Deactivation of a catalyst can occur in many ways. In the presence of acid sites,

alkenes can polymerize and block the surface from interaction with reactants. The

common method used to remove the polymer and other carbonaceous deactivators

include dissolving them in organic solvents or burning them off in an oxygen stream at

elevated temperatures. A second, more disastrous deactivation pathway occurs through

the loss of the acid site. Sintering of a mixed metal oxide results in loss of surface

homogeneity and the properties that resulted in the acidity of the solid, metal A in close

proximity to metal B, are lost. Dealumination of a zeolite catalyst is a special type of

sintering that results in the loss of the internal acid sites of the solid and the creation of

weaker alumina sites. For catalysts which rely upon sulfate groups hydrolysis of the

sulfate linkage to the surface by water (Figure 4-5) results in the production of sulfuric

acid which is lost from the surface and is a corrosion problem.15 In the MTBE reaction,

all of the reaction pathways result in the production of water, which makes this system

ideal for studying the loss of sulfate functionality from an acid surface.

Two solid acids that derive their acidity from surface sulfate groups have been

developed and characterized. These solids have had their acidity tested by both

calorimetric and activity tests. They both have been found to be of similar acid strength.

It was discovered that inclusion of the sulfate group into the silica lattice increases the

life span of the catalyst without sacrificing acidity.


Preparation of the Catalysts

The bridging sulfate catalyst (BSC) was prepared by mixing 23 ml diethyl sulfate

with 30 ml tetraethyl ortho silicate in a Teflon container (The synthesis procedure was

0 0-H

0 0 + H2,0 O + H2SO4

Figure 4-5. Hydrolysis of a tethered sulfate group from a surface. The sulfuric acid
produced is swept out of the reactor by the carrier gas.

devised by John Michael McGilvray'15 who also supplied the bulk of the BSC sample).

Ethanol was added and the solution was stirred for twenty minutes. The solution was

diluted with 20 ml of deionized water, followed by the addition of ~0.1 ml of sulfuric

acid. The Teflon container was sealed in an autoclave and heated at 80 OC for five days.

The resulting gel was crushed and heated in a tube furnace, under flowing air, at 400 C

for three days.'51 These conditions were reported to remove any residual organic

compounds, water, and sulfuric acid.

The tethered sulfate catalyst (TSC) was prepared by washing acid and peroxide

treated silica gel with 1M sulfuric acid. The catalyst was heated to 200 C before use.

Activity Tests

Activity tests were performed in a gas flow reactor with 0.5 g of catalyst. Glass

beads were packed over the catalyst bed to vaporize the incoming liquid reactants. Dry

nitrogen was used as the carrier gas with flow rates of 5 ml min1'. The addition of

reactants was accomplished with the use of a syringe pump. The liquid alcohols were

vaporized on the hot glass beads then carried through the catalyst bed. Temperature

control was accomplished by monitoring the temperature of the catalyst bed with a K

type thermocouple. The thermocouple was attached to an Omega Temperature Control

unit that regulated the bed temperature to 150 OC unless otherwise noted.

Additional tests were performed in a glass bomb reactor. These were prepared by

loading 0.5 g of catalyst into a 250 ml bomb reactor. A solution of methanol and

t-butanol was then poured into the bomb. The bomb was then sealed and placed into an

oil bath operating at either room temperature or 50 oC.


Calorimetry was performed on TSC by slurrying 1.0 g of the solid into 100 ml of

anhydrous cyclohexane. The solution was titrated with a 0.10 M or 0.18 M pyridine in

anhydrous cyclohexane solution. Cal-Ad experiments were performed on BSC by John

Michael McGilvray.151

All analyses were performed on a Hewlett Packard Model 5890 gas

chromatograph operating at 70 OC equipped with an FID detector and a Hewlett Packard

HP3394 Integrator. The retention times of several compounds under these conditions can

be found in Table 4-1. The effluent gases were sampled with a gas-tight syringe. Infrared

spectra of the catalysts were obtained using a KBr pellet on a Nicolet 5PC FTIR.

Surface area and pore volume measurements were performed on a Micromeritics

ASAP 2000.151

Results and Discussion

The Acidity of the Solids

The TSC solid was characterized by pyridine adsorption slurry calorimetry. This

solid can be classified as a moderate strength solid acid, with acidity that is slightly less

than zeolite HZSM-5. The Cal-Ad procedure was used to characterize BSC (Table 4-

Table 4-1. Retention time of several alcohols and ethers.

Compound Retention Time (min)

dimethyl ether 3.42

Methanol 5.12

t-butanol 7.07

MTBE 10.42

MIBE 11.5

2).'51 BSC was found be slightly less acidic than TSC. However, the small difference in

acidity (4 kcal mol ') is not likely to be of great significance unless the acidic threshold--

the minimum acidity required to drive a reaction--occurs in this range. Porosimetry was

performed, with the TSC having a surface area of 240 m2/g and BSC having a surface

area of 540 m2/g.

Results of the Activity Tests

Further testing of the catalysts began with a simple test to determine their

stability. Three containers containing identical t-butanol/methanol solutions were set up.

TSC was placed in one of the containers, BSC in the second, and a small amount of

sulfuric acid was placed in the last. One hour was allowed for the reactions to progress,

and then the solutions were analyzed by GC. All three catalysts had the same products

and product distributions, notably an excess of methyl iso-butyl ether. Leaching of the

sulfate functionality by water in the solvents to form sulfuric acid caused the identical

results for all three systems. More reactions were undertaken using dry solvents and

sealed from the atmosphere by using a bomb reactor. Results were significantly different

Table 4-2. The acidity of two sulfated catalysts measured by calorimetry.

Catalyst nl (mmol g') -AH, (kcal mol ) n2 (mmol g-1) -AH2 (kcal mol')

TSCa 0.2 32 0.7 17

BSCb 0.1 28.5 0.9 17

aPyridine adsorption slurry calorimetry in cyclohexane.
b Cal-Ad results from J. M. McGilvray (151).

from the wet reactions or the sulfuric acid standard, primarily producing MTBE and

isobutene with maximum conversion of methanol.

More ambitious testing of the catalysts was performed in the gas phase. The

product of the reaction of a methanol and t-butanol stream is strongly dependent on the

acidity of the catalyst used. Strong acids are capable of dehydratively coupling methanol

to form dimethyl ether and water. Acids of fairly weak to moderate strength dehydrate

the alcohols to form MTBE. Weak acids are capable of dehydrating t-butanol to form


Tests using this alcohol system were performed using various ratios of

methanol/t-butanol at 1500C in a flow reactor unless otherwise noted. Initial testing

began with 10% (mol/mol) methanol in t-butanol. TSC converted all of the methanol to

MTBE and MIBE and 95% of the remaining t-butanol to isobutene. The catalyst lasted

for twenty hours before losing activity and was dead by fifty hours (Figure 4-6). The

rapid deactivation of this catalyst was probably due to loss of the sulfate moiety as

sulfuric acid. BSC showed complete conversion of the methanol to MTBE with high

selectivity (MTBE/MIBE>20) and nearly complete conversion of t-butanol to isobutene.

25 ----

E 20 ...

15 .----- ..
t 0I o MTBE

0 I
0 10 20 30 40 50 60

Time (hrs.)

Figure 4-6. Amount of MTBE and t-butanol in the product stream from TSC over the life
span of the catalyst.

The initially high activity dropped off slowly starting at fifteen hours but was still

producing a significant amount of MTBE at 120 hours (Figure 4-7).

MTBE forms an undesirable azeotrope with both t-butanol and methanol. In an

attempt to find out if we could avoid formation of the azeotrope, we tried to determine if

BSC could stoichometrically convert methanol and t-butanol to MTBE and/or isobutene.

A 50% (mol/mol) methanol in t-butanol solution was prepared from solvents dried over

4A molecular sieves (only excess water was eliminated: water is produced by the

reaction). Initial results showed 100% conversion of t-butanol to isobutene, and 25%

conversion of methanol to dimethyl ether. After four hours, the reactants began to be

converted to MIBE. After seven hours, MTBE production was observed. MTBE

production plateued at 10% of the products. Greater than 20% conversion of methanol

was achieved with 100% conversion of t-butanol (>80% selectivity to isobutene). Small


30 ----- -. ---
S* t-butanol
G 20- -0- *-.

S15 -


5 '
(J ----- ---^------- --r-- -------*<--- -----ii

0 10 20 30 40 50 60 70 80 90 100 110 120 130

Time (hrs.)

Figure 4-7. Amount of MTBE and t-butanol in the product stream from BSC over the
life span of the catalyst.

amounts of water were added to attempt to force the t-butanol dehydration equilibrium

towards t-butanol, but little effect was observed. A similar activation pattern was

observed for TSC.

The products produced by the catalyst allow us insight into how the catalyst

changes during the reaction (Figure 4-8). Both catalysts start very acidic, producing

dimethyl ether for the first few hours of reaction. The catalysts then begin to lose acidity

as can be seen by the production of methyl isobutyl ether. The production of methyl

isobutyl ether from r-butanol or isobutene requires a strong enough acid to stabilize a

primary carbocation. As the acidity continues to decrease, we see MTBE production

begin. Isobutene and MTBE are the primary products of both catalysts during the middle

DME Threshold
I- ---------------

S MIBE Threshold

SMTBE Threshold
S--. ----------
< -

.Isobutene Threshold
7 i

Time '

Figure 4-8. The products of the reaction change as the acidity of the catalyst deteriorates.
At the start of the reaction the catalysts are strong enough to produce dimethyl ether.
However, the acidity quickly deteriorates to the point where MIBE production begins.
As the reaction continues, MTBE production begins and MIBE production fades.
Finally, the solid can no longer catalyze MTBE formation and only isobutene is

portion of their lifetimes. The TSC, however, quickly begins to lose activity. Soon this

solid can no longer catalyze the production of MTBE. The BSC catalyst loses its activity

toward this reaction much more slowly.

Threshold Acidity for Heptane Isomerization

Both catalysts were used to isomerize n-heptane. The small difference in acidity

between the two solids resulted in different product streams. The TSC catalyst was found

to have no activity towards isomerization, while the BSC was found to have a small

activity toward n-heptane isomerization. From these two observations the threshold

acidity for heptane isomerization can be said to be slightly less than the BSC catalyst, or

-32.5 kcal mol" measured by pyridine adsorption slurry calorimetry.

Structure of the Solids

The structure of the acid site of the two solids is very different. Fourier transform

IR of both catalysts shows a distinct difference in the nature of the bonding about the

sulfate group. The BSC solid was shown to be a bridging group by the presence of the

characteristic band at 1057 cm ',151'52 while the TSC lacks this band. The stability of

BSC over TSC is due to this structural difference. The bridging sulfate group of the BSC

catalyst results in a more stable sulfate moiety. A double hydrolysis reaction must occur

to remove this sulfate from the solid, and a longer life span is to be expected.


Further research into obtaining a wide range of solids of varying acidity is

required. They are important for the further discovery of threshold and optimum

acidities. The methanol/t-butanol system is offered as a new system for characterization

of solid acidity. Monitoring the products is a good method for determining the overall


strength of the catalyst and may give insight on how that catalyst deactivates. Further

study with other acids may prove this system to be a general method for broad

classification of acids.

The structure of solid acids has been shown to be of great importance. The

stability of the bridging sulfate catalyst (BSC) when compared to the very similar

tethered sulfate catalyst (TSC) is impressive. Stabilization of future catalysts may use

this incorporation system to improve lifetime and increase resilience.


Conventional strong acids can be used as inspiration for new strong solid acids.

For example, Drago et al. supported aluminum chloride, a conventional strong Lewis

acid, on a silanol rich silica gel, a Bronsted acid, to produce a solid that was significantly

stronger than either component alone.154'155 However, their goal was not to form a solid

with Lewis acid sites, but instead to form an even more powerful Bronsted acid. The

interaction of a powerful Lewis acid with a Bronsted acid leads to the enhancement of the

acidity of the Bronsted site.156'157 In the case of Drago's catalyst, the interaction between

the unsaturated aluminum center with the oxygen of a silanol promoted the acidity of the

solid such that it can be labeled as a solid super acid (Figure 5-1).39

The ability of a Lewis acid to enhance the acidity of a Bronsted acid is due to the

inductive effect. The coordination of the Lewis acid to the Br0nsted site leads to a

further reduction in electron density at the proton. The net effect is the weakening of the

O-H bond, and in extreme cases, the protonation of the Bronsted acid occurs. In the case

of Magic Acid, antimony pentafluoride in fluorosulfonic acid, the super acid formed is

capable of protonating fluorosulfonic acid (Equation 5-1):

SbF5 + FSO3H -) SbF5-FSO3-H + FSO3H SbFs-FSO3- + H20SO2F+ (5-1)

The protonated fluorosulfonic acid is then capable of forming carbonium ions from


F _FH H Al
F-Sb- -F-S--
4j II I
F F O Si SSi

a) The classic super b) The solid super
acid Magic Acid. acid silica supported
aluminum chloride.
Figure 5-1. Examples of super acids. A super acid is formed, primarily, through the
interaction of a strong Br0nsted acid with a strong Lewis acid. a) Antimony
pentafluoride in fluorosulfonic acid, a classic super acid. b) The solid super acid silica
supported aluminum chloride.

Since super acids are capable of protonating alkanes to form carbonium ions, they

are ideal for performing certain petroleum reforming reactions.148 The alkylation of

isobutane and butane to higher molecular weight hydrocarbons is important for creating a

value-added product from these materials. The mechanism (Figure 5-2) for this reaction

begins with an initiation step. The isobutene present in the stream is protonated by a

strong Br0nsted acid forming the t-butyl cation. The t-butyl cation can then interact with

another isobutene to form a variety C8 carbocations. These carbocations then abstract a

hydride from isobutane to produce another t-butyl cation continuing the catalytic cycle.

Side reactions that occur include isomerization, polymerization, cracking, and

disproportionation. Isomerization of the C8 carbocations to a more stable form may occur

before hydride abstraction. Polymerization occurs if the C8 carbocations comes into

contact with an alkene, thus the concentration of the alkene must be kept high enough to

initiate and sustain the alkylation reaction, but low enough to prevent polymerization

Other Intiators

Additional Reactant










' (Isomerization)



Figure 5-2. The reaction mechanism for the alkylation of isobutane to octane. The
catalyst is deactivated by the build up of polymer on its surface.





which leads to catalyst deactivation. Disproportionation is a bimolecular interaction that

results in one hydrocarbon abstracting a methyl group from another hydrocarbon

(Equation 5-2). Cracking of a hydrocarbon results in two smaller fragments, one of

which is an alkene (Equation 5-3).

2C8Hi8 C7HI6 + C9H20 (5-2)

CnH2n+2 Cn-xH2(n-x)+2 + CxH2x (5-3)

The activity of a catalyst towards each of these reactions is related to its acidity.

The more acidic the solid the lower the temperature it can carry out each reaction. By

comparing two (or more) catalysts at the same temperature, the relative acidity can be

found by comparing the activity. Strong catalysts, such as Drago's aluminum chloride

catalyst, can perform alkylations (little or no cracking, isomerization, or

disproportionation) with high selectivity at 0 OC.

The growing importance of solid acid catalysis and the correlation of activity with

acidity has lead to a surge in the development of new, strong solid acid catalysts. A new

solid acid that uses antimony pentachloride as an acidity promoter has been developed.

The acidity of this catalyst has been measured through a new method (reported in Chapter

2). The synthesis and characterization of this new solid super acid will be reported. The

acidity and activity of this solid will be shown to be greater than that of Drago's silica

supported aluminum chloride catalyst.



Davisil silica gel (mesh 100-200, surface area 250 m2 g ') was purchased from

Aldrich. It was washed with 1M HCI three times (about 20 ml 1M HCI per 10 grams

silica gel), then with deionized water three times, followed by a 35% hydrogen peroxide

wash three times, and finally several washes with deionized water before drying it in an

oven at 140 C overnight. The silica gel was removed from the oven and allowed to

adsorb water from the atmosphere for 24 hrs.

Synthesis of the Catalyst

The catalyst is prepared by slurrying silica gel in dry chloroform (carbon

tetrachloride or methylene chloride can be used, but the best results were obtained with

chloroform). The slurry is heated to 350C with stirring under a slowly flowing N2

atmosphere. Enough SbC5l (Aldrich) was added to make 0.5 mmol SbCI5 (g' SiO2).

This was accomplished by loading a syringe with SbC15 (98%, Aldrich) in a dry box. The

syringe was capped and removed from the N2 glove box. SbCI5 was added dropwise into

the slurry by injection through a septum. The solution immediately turned brown.

Passing the N2 purge gas through a water bubbler collects the HCI gas produced during

the reaction. Titration of the water trap with 0.0469M NaOH shows that 1.8 0.1 mol

HCI was produced for every mol of SbCl5 reacted. Substituting PCI5 for SbCl5 produced

the phosphorus analog.

Activity Tests

Alkylation reactions were performed in a glass batch reactor by loading, in an

inert atmosphere glove box, one gram of each solid into a separate 250 ml glass reactor

bomb. The pressure head was attached before removing the assembly from the box. The

bomb was cooled in a dry ice/acetone bath (after sealing and removing from the glove

box). Butane was condensed (10 ml) into the bomb, followed by 0.2 ml of isobutene.

The isobutene is necessary as an initiator, however excess isobutene will polymerize and

deactivate the catalyst by covering the surface with a polymer. The system was removed

from the dry ice/acetone bath and placed in an ice water bath. The reaction was then

allowed to proceed for thirty minutes. At that point the products were filtered off, 2 ml of

n-pentane was added as an internal standard, and the residual butane and isobutene were

allowed to boil off.


Analysis of the reaction products were performed on a Hewlett Packard 5890 Gas

Chromatograph equipped with a FID detector and a Hewlett Packard HP3394 Integrator.

A temperature ramp was used to separate the products of the reaction:

1) Initial Temperature 350C.

2) Initial time 10 minutes.

3) Increase by 5 C min' to 2000C

4) Hold at 200 oC forl0 minutes.

The retention time of various standards can be found in Table 5-1.

The solid state NMR experiments were described in Chapter 2. The sample was

prepared and loaded into the rotor in an inert atmosphere glove box.


The procedure used to perform calorimetry is standard and was described in

Chapter 1. The highest heats of interaction were observed when a small amount of