The cal-ad method

MISSING IMAGE

Material Information

Title:
The cal-ad method a technique for the determination of surface aciditybasicity
Physical Description:
viii, 150 leaves : ill. ; 29 cm.
Language:
English
Creator:
Chronister, Chris W., 1967-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 146-149).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Chris W. Chronister.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001973121
notis - AKE9913
oclc - 31743884
System ID:
AA00003209:00001

Full Text








THE CAL-AD METHOD: A TECHNIQUE FOR THE DETERMINATION
OF SURFACE ACIDITY/BASICITY
















By

CHRIS W. CHRONISTER


DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA












ACKNOWLEDGEMENTS


It would be impossible to thank everyone that made this possible, so I will try to


thank the people who have had the greatest influence on my life.


has been the ideal research director and friend.


Dr. Russell S. Drago


He has a great knowledge of chemistry


and the ability to share it with those he comes in contact with.


great host for the numerous group parties.


Ruth has also been a


I would be remiss if I did not mention the


tremendous


done


secretaries.


Maribel


Lisk,


April


Kirch,


Diana


Williamson have always been a great help and have provided interesting conversation.

While attending the Florida Institute of Technology, I received a solid foundation


in chemistry.


I would like to thank the entire faculty there including Dr. Clayton Baum,


Dr. Alan Brown, and Dr. Richard Mounts.


The time I spent doing research for each of


them was time well spent.

I would also like to thank all the members of the Drago group from the time I


arrived


until


particular,


have developed


great


friendship


Mike


Naughton and Don Ferris "Bueller" and I hope the good times will continue as we go our


separate ways.


Ngai Wong and Larry Chamusco deserve a special thanks for the writing


of the computer programs necessary to complete this work. It was also very enjoyable

to help direct the research of Melissa Hirsch and Steve Joerg. Dr. Doug Bums, Andy


Dadmun, and Mike McGilvrey are another group that must be acknowledged as they are






my fellow "cal-aders"


. I would also like to thank Dr. Robert Beer, Dr. David Singh,


and Dr. Phil Kaufman for their helpful advice.


Another group of people that have had a profound effect on

arrived in Gainesville are the members of my billiards league team. Pat


life since I


rick Koenigstein,


Valerie Brown,


Terry Schmittendorf, Kevin


White, Jack Nugent, and Bill Keen have


been very good friends and have provided me with a lot of great memories of my time

in Gainesville.


final


member of my


league


team,


Hoffmann,


also doubles


as my


girlfriend.


She has had a great influence on my life.


The time always seems to pass to


quickly when she is around.


I couldn't ask for a better match.


Last, but certainly not least, I would like to thank my family.


They have always


been a stable force in my life from which I can draw unending love.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS


ABSTRACT


CHAPTER 1
INTRODUCTION.....
Background .
Cal-Ad Method
Calorimetry
Adsorption
Method


* . . a a a a a a a a a a .
* a a . a a a a a a a a a a a S a a a a a a
* S S I S S 5 0 a a a a a a a a a a a a a a a a a a a a a a S
* S S S S a a S S S S S S S S S I S S 0 a a a a a a a a a *
* S S S S S S S S S S S S S S S S S a S 0 5 5 a a a S a S S a a
* a a a a S a a a a a a a a a a a a a S a a a a a a a a S S


CHAPTER 2
ACIDITY OF SILICA GEL . . . .
Introduction ......... .... ..
Experimental . . . .
Purification of Materials .
Calorimetry and Adsorption .
Ultraviolet/Visible Spectroscopy
Fourier Transform Infrared (FTIR)
Results and Discussion
Acidity of Hydrated Silica Gel
Acidity of Dehydrated Silica Gel
Characterization of the Silica Gels
Gas-Solid Equilibria . .
E and C Analysis
Conclusions . . . .


13
14


Studies . . . 15


23

with Betaine . . . 29
a32
33
36


CHAPTER 3
ACIDITY OF A1Cl2(SG)n .
Introduction . .
Experimental
Reagents . .
Instrumentation


* a a a a a S S S a S a
- - -* *
. . . .
. . . .


* a a a a a a a a a a a a a a a S
* S a a S *S* S *S* *

* a a a a a a a a a a a a a a a
* . . a a .







Calorimetric


Titrations


Results and Discussion


Infrared Examination of Adsorbed Pyridine
Infrared Titration .. . . .


Calorimetry and Adsorption


Conclusions


S S t U S S S 5 0 ft ft f ft ft ft S f ft f ft S S S S S 5 5 5 5 t /5


CHAPTER 4


ACIDITY OF PHOSPHOTUNGSTIC ACID


Introduction
Experimental


Purification of Materials
NMR Spectroscopy
Calorimetry


S S . . 4 S . 56
.* . S S f t S S 5 S 56
* . . S S S S S S S . 5II7


Fourier Transform Infrared (FTIR) Spectroscopy


Results and Discussion
NMR Spectroscopy
Calorimetry
FTIR Spectroscopy
Conclusions .


* f f f S S S S S S S S S S 5 5 7

a a a a a a a a a S S S f f. S S ft S S S f S a S ft IFt 62


Sf S S f ft ft 70


CHAPTER 5


ACIDITY OF HY ZEOLITE . . . . . . . 73

Experimental 77
Purification of Materials . . . . . . 77


Calorimetry . . . .
Fourier Transform Infrared (FTIR)


Results and Discussion
Calorimetry .
FTIR Spectroscopy
E and C Analysis


Conclusions


CHAPTER 6
SUMMARY


APPENDIX I

APPENDIX II

APPENDIX III


pectroscopy


91
ft f ft t f ft ft S ft ft t S ft t S S S S S f S S S 5 8
S S S S S ft S t S S S S S S ft S ft S S S S ft S S S S S S S I 91


93


94


SCHEMATIC OF ELECTRONICS


PASCAL PROGRAM FOR DATA ACQUISITION


MANUAL FOR DATA ACQUISITION


128


52


52
56








REFERENCE LIST


BIOGRAPHICAL SKETCH .


. 146


. p p 5 S S . * . . 150












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


THE CAL-AD METHOD:


TECHNIQUE FOR THE DETERMINATION


OF SURFACE ACIDITY/BASICITY

By


Chris W


Chronister


April 1994

Chairperson: Dr. Russell S. Drago
Major Department: Chemistry

A novel calorimetric-adsorption (cal-ad) method has been developed to determine


the surface thermodynamic properties of solid acids.


This method combines calorimetric


and adsorption titrations of a slurry of the solid acid in a non-polar solvent with various


basic probe molecules.


The adsorption experiment is a spectroscopic determination of


the amount of base remaining in solution throughout the titration.


Using this information


along with heats evolved over the same concentration range yields a thermodynamic

description of the solid acid surface (i.e. enthalpy, equilibrium constant of binding, and


number of sites for each of the different types of binding sites).


The application of this


method to several solid acids leads to a number of interesting conclusions.


Investigating


various hydration levels of silica gel enables the separation in reactivity of three different

types of hydrogen-bonding sites (indicating that this method is more sensitive than all


-- -






investigated was a silica supported aluminum chloride superacid (A1Cl2(SG) ) useful for


hydrocarbon


cracking


reactions.


This


catalyst


found


to contain


highly


acidic


BrOnsted sites with an absence of any Lewis acidity.


The Bronsted and Lewis acidity of


HY zeolites was dependant on the pretreatment temperature of the solid.


A heteropoly


tungstate (phosphotungstic acid) was also investigated and its acidity was also found to


be dependent on the pretreatment temperature.


It was also shown to contain Brbnsted


acidity exclusively.











CHAPTER 1
INTRODUCTION


Background


The field of catalysis is of great economic importance to the chemical industry.

In 1989 it is estimated that over $1 trillion worth of goods were produced using catalytic


technology.'


Recently,


there


been


greater


emphasis


placed


on developing


heterogeneous catalytic systems.


The greatest advantage of heterogeneous systems is the


ease of separation of catalyst from the reactants and products.


More than 400 million


barrels of oil ($8 billion dollars) per year have been saved by the use of zeolite solid acid


catalysts.2


Another advantage


steric


selectivity


which


can be


forced


on the


reactants and/or the products of a reaction with a proper shape selective catalyst structure

(e.g. zeolites).

The characterization of solid acids and bases has become an area of increasing


interest in recent years because of their applications in heterogeneous catalysis.3


characterization of acid/base sites on


solid surfaces is recognized as a difficult task.


Shifts in infrared bands of coordinated molecules were measured and proposed to reflect


acid strengths.4,5


More recently,


nuclear magnetic resonance spectroscopy


has been


used


to study


the coordination


donor molecules


to the


surface.6'7


Temperature


Programmed Desorption (TPD) is another technique which has gained increasing use as








an indicator of the acidity of solid acids."


The use of calorimetry in the past has been


limited


mostly


to the


adsorption


gaseous


bases onto


solids.9


Recently,


solution


calorimetry has been used to measure the total heat evolved when donors are added to


solid


acids. 10,11


This provides an


enthalpy


units of kcal


mole-'


donor added.


These measurements are difficult to interpret because the heat evolved has contributions


from the amount of base completed (i.e.

the number of sites involved. A full cha


the equilibrium constant), the enthalpy, and


iracterization of the solid acid would involve


measuring the equilibrium constant and enthalpy of adduct formation for each of the


different acceptor


solid.


None


above


approaches


provide


information.


Thus, it is not surprising to find that the measured estimates of acidity from


most of these procedures fail to parallel catalytic activity.

In our laboratory, a recent approach to the characterization of solid acids, the cal-

ad method, has been used to successfully distinguish two different sites for coordination


of donors to a Pd/Carbon heterogeneous catalyst.'2


Conventional differential scanning


calorimetry and desorption (thermal gravimetric analysis) techniques could not distinguish


sites


provided


average


values


interaction


sites.


dissertation, the cal-ad method has been improved and used to characterize silica gel,


zeolites,


phosphotungstic


acid,


silica


supported


aluminum


chloride


catalyst


To aid


this characterization,


other physical


methods (e.g.


Fourier


transform infrared spectroscopy and UV/vis spectrophotometry) were used.


the different acids with a series of bases, an E and C analysis13


By studying


of the enthalpies of


(AICI,(SG)~.








in the E and C database.


This analysis is demonstrated for the solid acids and for strong


acids soluble in weakly basic solvents.


Cal-Ad Method


Calorimetry


The calorimeter to be described consists of the calorimeter cell, a thermistor, a

heater coil, an injection syringe, an electronic circuit bridge, a chart recorder, and a

computer with supporting software.


Cell.


The calorimetry cell is shown in Figure 1-1.


It is a silvered dewer flask


fitted with three joints that are used to accommodate a thermistor, a heater coil and an


injection syringe.


There is also another port to allow the cell to be purged with an inert


The maximum volume of the cell is


certain precautions are taken,


Thermistor.


75 mL but smaller volumes can be used if


vide infra.


The thermistor (manufactured by Fenwal Electronics, part # 121-


402-EAJ-Q01, and distributed by Newark Electronics),


is fitted in a teflon plug at the


end of a glass probe.


Heater coil.


The heater coil is a hand-wrapped 0.5 mm diameter wire with a


resistance of about 100 ohms.


This coil is placed in a glass probe and mineral oil is


added to a level just above the top of the coil.


When the probe is inserted in the cell,


it is important that the level of the oil is below the level of the solvent in the calorimeter


cell. This is necessary


efficient heat


transfer to


the solution


to obtain a


valid


















ringe


Heating


coil


Thermistor


Magneti


tirrer








Injection syringe.


The syringe is a Hamilton Gastight Series 1000 2.5 mL syringe


with a luer tip.


A glass tip is then made from a luer type outer joint with capillary wall


thickness (Ace Glass, Inc.) by heating the glass and drawing it out to a tip length of


about 30 cm.


To accompany the syringe, there is a set of 12 brass stops which can be


placed on


the outside of the plunger.


These stops are used,


in order of decreasing


length, to supply constant volume injections. The stops are calibrated by determining the

injection weight of water delivered for each injection. The weight is then converted to


a volume using the density of the water.


Electronics. A schematic of the electronic circuit bridge is supplied in Appendix



Chart recorder. The chart recorder is a Fisher Recordall Series 5000 and the


chart paper used is EC-146.


Chart speed is set at


cm/min and the voltage scale is


set at 1.0 volt full scale.


Computer.


An analog to digital board is used to link the electronic circuit bridge


to an IBM Model XT


personal computer.


The program for data acquisition is supplied


in Appendix II and a manual for the use of the program can be found in Appendix III.

The Quattro (Borland) spreadsheet software was also used for data analysis.


Procedure.


The solid is added to the cell in the dry box.


The cell is sealed and


transferred to the dry bag. Solvent is added to the cell and the syringe is filled with the

injection solution in the dry bag. All the probes are placed in the cell and it is removed


from the dry bag to be attached to the electronic bridge.


About ten minutes is allowed




































0 200 400 600 800 1000


1200


time (arbitrary units)






7

For each calorimetric titration, the heater coil and the thermistor are calibrated.


A digital


multimeter is used


to calibrate the heater coil


during


the timed electronic


injections of a constant voltage (5 V).


By measuring the voltage over the heater coil and


the voltage over a known constant resistor the power output of the heater coil can be

calculated as follows:


VV
p= V R
R,


(1-1)


The units of power for this calculation result in J sec'e


Dividing this number by 4.184


gives


power in


The calibration


thermistor is


a linear


regression of the magnitudes of the deflections caused by the timed electronic injections.

A typical calogram is shown in Figure 1-2 to illustrate the deflections observed.

The experimental injections are then made by depressing the injection syringe to


the next smallest brass stop.


Time is allowed between each injection to allow the system


to come to thermal equilibrium (a flat baseline).


It is important to note that if large


concentrations of base remain in solution, the heat capacity of the solvent may change


significantly.


This may require the recalibration of the thermistor.


To test if this is the


case, a timed electronic injection can be done at the end of the experimental injections

to determine if it still agrees with the original calibration curve.


Adsorption


The adsorption apparatus (see Figure


1-3) consists of a 3-neck flask, injection











Syringes.


The injection syringe is the same as the one used in the calorimetry


titration but in the adsorption titration, a stainless steel tip is used instead of the glass tip.


The sampling syringe is a Hamilton Gastight Series


1000


1.0 mL syringe with a luer


joint and a stainless steel tip.


UV/vis


spectrophotometry.


spectra


were obtained


with a Perkin


Elmer


Lambda 6 UV/vis spectrophotometer.


The cells used were Suprasil quartz with a 1 or


0.1 cm pathlength.


Procedure.


The solid is added to the 3-neck flask in the dry box.


The flask is


sealed and transferred to the dry bag.


Solvent is added to the flask and the syringe is


filled with the injection solution in the dry bag.


While the solution is stirring, an injection is made.


About three to four minutes


is allowed for the system


to equilibrate (same time as in


the calorimetric titration).


Stirring is stopped to allow the solid to settle.


A sample of the solution is withdrawn and


the sample volume is replaced


neat solvent.


Stirring is resumed and the next


injection


is made.


During


the equilibration


time


new


injection,


UV/vis


spectrum


can be obtained


on the sample that


just removed.


This sequence


repeated until all the injections have been made.


Method


The purpose of the adsorption titration in the cal-ad method is to determine the

equilibrium concentration of pyridine in solution for each added increment of pyridine








calorimetric


titration,


problems


can anse


making


direct


comparison


equilibrium position in the two experiments.


The direct comparison is necessary in order


to solve


the adsorption


calorimetric


data simultaneously


for the


This


problem is overcome by using a constant ratio for the volume of solution/mass of solid


for both


measurements.


An adsorption


isotherm


is constructed


from the adsorption


titration data


which covers the range of base concentrations used in the calorimetric


titration.


The best-fit isotherm


can be calculated


using


a modified simplex routine


(Appendix


designed


to solve


following


multiple


Langmuir


type14


equilibrium equation for a series of base concentrations.


nK [B]
1+K,[B]


(1-2)


For i sites, STB is the total number of moles of base adsorbed per gram of solid, nq is the

number of moles of site i per gram of solid, K1 is the equilibrium constant of binding at


site i, and [B] is the equilibrium concentration of base in solution.


In the calorimetric


experiment, only the total concentration of base added is known for each calorimetric


injection.


The total number of moles of base on the solid (SrB) is expressed as follows:


S([TJ [B])V
-


(1-3)


where [T] is the total molar concentration of injected base and V is the volume of the


experiment.


By substituting Equation (1-2) into Equation (1-3) the following relationship


is obtained:


S~B = I:









([7]-[B])V


n1 +K,[B]
1+K([B]


(1-4)


The equilibrium concentration of base is measured in the adsorption experiment.


order to obtain the equilibrium base concentration for a base addition in the calorimetric

experiment a preliminary Langmuir analysis of the adsorption data is carried out to give


Ki and ni.


These values of K, and ni are used to calculate the [B] corresponding to the


calorimetric measurement.


This allows substitution of these [B] values calculated for


each base addition along with the heats obtained from the calorimetric titration into the

following equation:


hs
- g
g


n1,K[B]
1 +K.[B]


(1-5)


Where h is the heat evolved and g is the mass of the solid.


Two sets of simultaneous


equations are written. (

form of Equation (1-2).


3ne set is of the form of Equation (1-5).


The second set is of the


Each set includes an equation for each base addition.


The two


sets are then solved simultaneously, using the modified simplex routine in Appendix IV,


for the best value of K,, ni and AH. that reproduce the experimental quantities.


important to note that the ratio of the volume of solution to the mass of solid (V/g) must


identical


the calorimetric


adsorption


experiments.


This


is necessary


simultaneously solve the combined calorimetric and adsorption data sets for nr, K,, and

AH. values (i.e. the base concentration in solution relative to the amount of solid must


be the same in both experiments).


If this condition is not met,


the data sets may be








higher V/g ratio a more reactive site could be averaged in with a less reactive site which

would give rise to a faulty analysis if this data set was then used in combination with a


data set obtained at a lower V/g ratio in which the two sites were separable).


It should


also be noted that the concentration range used in the calorimetric titration should be


within


the concentration


range


studied


the adsorption


titration


to avoid


uncertain


extrapolation of the best-fit adsorption isotherm outside of the measured range.












CHAPTER 2
ACIDITY OF SILICA GEL


Introduction


Silica was chosen as the solid acid for this study because of its extensive use as

a support in heterogeneous catalysis, in separations and in microelectronic fabrication.


It has


been


studied


extensively


infrared


spectroscopy'5


nuclear


magnetic


resonance spectroscopy.'6


Previous calorimetric investigations10"' of silica bonding to


donor molecules assumed only a single binding site for a bases binding to the silica


surface and in one study," complete complexation to this site was also assumed.


In the


absence of equilibrium data, the units for the resulting enthalpies are kcal mole' of donor


added.


If more than one site is involved


, the enthalpy corresponds to a mole fraction


average


pyridine


coordinated


sites.


Model


studies


silsesquioxanes'7 raised suspicions that sites of differing reactivity should be present on

the solid depending on the extent of hydrogen bonding between neighboring hydroxyls


on the surface.


Isolated silanols are shown to be less acidic than clusters possessing at


least three mutually hydrogen bonded hydroxyl groups in the soluble silsesquioxanes.


Using the cal-ad


method we can determine equilibrium and enthalpy data for


different sites on Fisher silica gel S-679 coordinating to pyridine.


Furthermore, we have


been able to demonstrate the dramatic influence that pretreatment of the silica has on its









reactivity.


Comparison of this data with gas phase-solid data demonstrates the magnitude


of the non-specific interactions that are involved when gas phase solid equilibria are

measured.


Experimental


Purification of Materials


Fisher silica gel S-679 (LOT NO. 903424) was used as supplied.


Studies were


carried out on samples that were


"dried" by evacuating at


C, and 200 C.


BET


surface areas of 590 m2/g were obtained after both pretreatment temperatures.

Pyridine (Fisher) was stored over BaO and redistilled over CaHll2 using a 12 in.


Vigreaux column.


Cyclohexane (Aldrich) was treated with activated charcoal three times


in order to remove traces of benzene.


It was then distilled over P205 and stored over 4A


molecular sieves at least 24 hours prior to use.

Reichardt's Dye (2,6-diphenyl-4-(2,4,6-triphenyl-pyridinio)phenolate), 95 %, was

used as supplied by Aldrich.


Calorimetry and Adsorption


The calorimetric and adsorption titrations were carried out as described in Chapter


1 of this dissertation.


A series of pyridine additions was made to provide a similar range


in pyridine concentration (3.5 pM


- 50 mM) for the two sets of titrations.


After each


fl A *P r. 4


mt r' 1 r' _r11.


I










Ultraviolet/Visible Spectroscopy


UV/vis


spectra


were


obtained


using


Perkin


Elmer


Lambda


UV/vis


Spectrophotometer. Suprasil quartz cells of 1 and 10 mm pathlength were employed in

the adsorption studies. For spectra of the solid silica, a mineral oil mull of silica was


made on a piece of filter paper and referenced against mineral oil on filter paper.


Fourier Transform Infrared (FTIR) Studies


All FTIR spectra were taken on a Nicolet 5DXB FTIR.

in these studies were prepared by forming a slurry in cycle

addition of varying amounts of pyridine. The pyridine conce


correlate with those used in the calorimetry and adsorption titrations.

filtered after five minutes was allowed for the solutions to equilibrate.


The silica samples used


ohexane followed by the


:ntrations were chosen to


The solid was then

The spectrum was


then obtained as a fluorolube grease GR-362 (Fisher) mull or mineral oil (Aldrich) mull

on sodium chloride plates.


Results and Discussion


Acidity of Hydrated Silica Gel


The thermodynamic data from the cal-ad analysis of the sample evacuated at 28


C (VAC28) are shown in Table 2-1.


Two different reactive sites were found with the


-.r t


. .


--.. -J -- ------- .. .---------. 1-- L- L -- t -AUL.. L- --. t


11~1- r- I


A
















O 00


OB ?



o o
Q 0


The fit of these results to the experimental data points is shown in Figure 2-1


illustrates both the adsorption and calorimetric data.

In the cal-ad analysis, the pyridine concentration ranged from the smallest amount


detectable to near saturation of the silica surface.


The different reactive sites were


discovered because both the amount of pyridine adsorbed and the amount of heat evolved


were measured over this large donor concentration range.


A further decided advantage


of the cal-ad method is the determination of the number of active sites as 0.86 mmol g'

for site A and 0.86 mmol g-' for site B.

It is interesting to note that a simple Langmuir analysis of this adsorption data yields


a different result than that obtained by the cal-ad method.


The Langmuir equation:


= +1
n^K 7il


is solved by plotting [B]/SB vs. [B].


(2-1)


The adsorption data produces a straight line (see


Figure


indicating


there


is 1.6


mmol/g


single


equilibrium constant of 2700 M'


n


This is clearly inconsistent with the calorimetric data.


- S. -1


I-


Al


A


trn rr amnnr n ri rn r.t nfl t.,.r YIt f In flrn flint~ i*n ~ rr nn i r* Plf annr r a C ar n


which


*


rl:CSa*at\) C~n\aa


r, ,














Table 2-1.


Thermodynamic


parameters obtained


cyclohexane for silica-


pyridine hydrogen bond formation.


Evacuated @ 28C


Evacuated @ 200 C


n, (mmol/g)


0.86


K, (M-')

1H, (kcal/mol)


17600


25400


-AG, (kcal/mol)


-22.8


n, (mmol/g)


0.86


1.30


K2 (M-')


-AH2 (kcal/mol)

-AG2 (kcal/mol)


AS, (eu)



































0 5 10 15 20 25 30 35 40 45


[py] total (mmol/L)


Figure


Adsorption isotherm and heat of adsorption


vs. total pyridine injected

















rU


0.005


0.015


0.02


0.025


0.03






20
the line is dependent solely on the number of sites, there will be no change in the slope


of the line if the two sites are present in equal quantities.


By dividing Equation (2-5) by


[B], another linear equation is obtained in which a plot of 1/(SiB) vs. 1/[B] should yield


another


straight


line (see


Figure


2-3).


This emphasizes data


points taken at lower


concentrations and, in the pyridine on silica case, shows curvature consistent with the


sites


found by


calorimetry.


This serves


to illustrate the difficulty in extracting


meaningful data solely from adsorption isotherms.

In this system, the possibility exists that the first site involves formation of the

hydrogen bond adduct and the second site involves reaction of this adduct with more


pyridine to form (CsHsN)2H +. Accordingly, the reaction between silica and pyridine,

was probed using FTIR spectroscopy. It has been reported that bands at 1447 cm'1 and


1599


pyridine


adsorbed


on silica


are indicative


of hydrogen-bonded


pyridinei4


Figure 2-4 shows the spectrum obtained for a silica sample at the completion


of a calorimetric titration as well as the silica surface before pyridine addition.


These


spectra clearly show that the interaction between pyridine and the silica surface involves


a hydrogen bonded adduct. Pyridinium ion is not formed, for it would be indicated by

a characteristic band at 1540 cm1'. The second site which forms in excess pyridine is

also a hydrogen bonded adduct. It is significant to note that the infrared spectrum does


not resolve


separate shifts


for the two adducts


these samples.


Furthermore,


pyridine shifts upon hydrogen bonding to the strong site of VAC28 are the same as those


for hydrogen bonding to the weaker sites of VAC200,


vide infra,


within experimental
















2500


2000


1500


1000


500


0 2 4 6 8 10 12 14 16


1/[py]
(Thousands)































































WA V ENUMBER


Figure 2-4.


PTTR


cn~rvtn t


nf VAP g?


in a minhnrinha


niilli


1,,,\.


cnaFntrin


AI bZL 'L LSAZLI 11 .JA *LL AI *t t AZ*I a1L1I *111 .LLt~A tfl k1'. A IUtl.~ *III






23

Arnett et al." report an enthalpy of -12.37 kcal per mole of pyridine added to a


silica slurry in


hexane.


This is not the enthalpy of adduct formation unless, at the


pyridine concentrations used, all the donor is coordinated and only one site is involved.

Comparison of these values with those determined for site (A) in this work indicates the


extent to which these assumptions are correct.


Base concentrations remaining in solution


are not specified in the Arnett study," but at the low base concentrations used in their


work most of the donor is coordinated to the first site.


Solution enthalpies for pyridine


bonding to silica gel reported previously by Fowkes et. al.'0 vary from 11.3 to 12.9 kcal


mole-'


These


values'0


were


calculated


correcting


base


concentration


remaining in solution.


However, at the concentrations studied, our results show that this


data has contributions from the second site averaged into the reported enthalpy.


could account for the slightly lower enthalpy values found in some systems.


Since these


studies were carried out on different types of silica than employed in our study,


similarity in the enthalpies of interaction indicates that similar surface sites give rise to

the strong acid centers of the different samples.


Acidity of Dehydrated Silica Gel


The combined cal-ad


measurements on a sample of silica gel S-679 that was


pretreated by heating to 200 C under vacuum (VAC200) were carried out next.


results of the best-fit analysis show that two different sites exist.

data and the values of n have been compiled in Table 2-1. Thb


The thermodynamic


e fit of the calculated




































0 20 40 60 80 100


[py] (mmol/L)


Fi pure


Adsomtion isotherm and heat of adsomtion vs. total nvridine infected








Figure


The enthalpy calculated for pyridine hydrogen bonding to the first site on


VAC200 is similar to that found for the second site on VAC28 and is therefore attributed


to a (B) type site.


The difference in the entropy of adduct formation between the two


different samples is very interesting and may be due to an increased propensity of the


VAC200 surface to physisorb cyclohexane.

hydrophobic surface from loss of hydroxyl g


The more stringent drying results in a more

;roups. Dispersion interactions which are a


function of mass, are expected to be larger with silicon and oxygen atoms than with


hydrogen atoms.


As the number of hydroxyl groups are decreased, hydrophobicity and


dispersion interactions increase.


When the pyridine hydrogen bonds to the remaining


hydroxyl


groups


hydrophobic


solid,


surface


expulsion


cyclohexane


molecules into the bulk solution would lead to a positive entropy.


The above proposal is supported by infrared measurements.


The FTIR spectra


of VAC200 was the same as observed for VAC28 in the 1400-1700 wavenumber region.

An interesting difference in the two spectra was observed in the region above 3000 cm'.


O-H


stretching


frequencies


silica,


above


3000


wavenumbers,


have


been


thoroughly


studied


are quite


informative.


isolated


hydroxyl


stretching


frequencies occur at 3747 cm'1 when studied as pellets in evacuated cells.'


Figure 2-6


shows a spectrum in a mineral oil mull of a silica sample that was heated to 450 C


under flowing 02 and then cooled to room temperature.


The peak appearing at 3696 cm


assigned


an isolated


silanol


to its


sharpness.


Hydrogen


bonding


neighboring silanols leads to a significant broadening of the O-H stretching vibration.











shift


from 3747


in vacuum


to 3696 cm-1


in a


mineral


mull is


attributed to dispersion forces from physisorbed mineral oil.


Absorbances from both the


hydrogen-bonded


isolated


silanols


were


present


in the


spectrum


obtained


VAC200.


The spectrum of VAC28 was less interesting in this region.


Only a large


broad peak was observed for the hydrogen-bonded silanols and no sharp absorbance from

isolated silanols was seen.


In order to determine which silanols react first when pyridine is added,


VAC200


was


titrated


FTIR


spectrum


recorded


after


each


addition


(Figure 2-7).


frequency decrease and broadening of the absorbance is expected when hydrogen-bond


acceptors coordinate to donors.


From these spectra, it is apparent that the first silanols


to react are those in the hydrogen bonded region around 3500 cm-'


Beginning with the


addition of the second increment of pyridine, a slight broadening of the isolated silanol


absorbance occurs indicating that they have begun to react.


Reaction continues as more


pyridine


added


isolated


silanol


vibration


broadens


extensively.


This


substantiates our earlier proposal that the hydrogen-bonded silanols are more acidic than

the isolated silanols.

An explanation for the difference in reactivity of silanol sites is suggested by solution


reactivity


studies on silsesquioxanes.17


The most acidic silanol is found when three


adjacent silanols are capable of hydrogen bonding to each other.


Our infrared results


with incremental pyridine addition indicate that the hydrogen-bonded silanols react first.


The strongest site in


VAC28 is as,


or more,


extensively hydrogen bonded than the






























































000


WAVENUMBER


- -v..c,7o n~f 7rh1PA t~ Sit 1fl0C


"rTf enanet. af T4 char


f*\ nn nvrtdine


sloo


3600


3600


i








the first, more acidic, site on VAC200.

can be assigned to the second site of VA(


The isolated silanols (C) are weaker acids and

C200. The structures for sites A, B and C are


same


types


surface


species


proposed


to arise


from


different


pretreatment


temperatures by Hench.'8


There


is precedence


for a hydrogen


bonding interaction


leading to increased


acidity.


example,


ionization


constant"9


maleic


(cis-1,2-


ethylenedicarboxylic


acid)


x 10-2


fumaric


(trans-1


ethylenedicarboxylic acid) is 9.3 x 104


. Intramolecular hydrogen bonding stabilizes the


conjugate base and increases the acidity of the cis derivative.


A similar explanation is


offered to account for the higher acidity of silanol clusters.


Characterization of the Silica Gels with Betaine





































300 400 500 600


wavelength (nm)








from a methylene chloride solution.


This organic dye has been


used as a probe of


solvent


polarity


ultraviolet/visible


is reported


region


ever


to have


largest


observed.20


shift


solvatochromic


protonic


shift


solvents


contributions


from


a specific


hydrogen


bonding


interaction


from


non-specific


solvation.20


With


silica


different


types


should


shift


longest


wavelength band of the dye by different amounts.


This should give rise to two different


absorption maxima for the lowest energy transition in the dye.


Figure 2-8 illustrates the


UV/vis spectra obtained for VAC28 before and after adsorption of the dye.


maximum is observed at 492 nm.


Only one


In view of the broad adsorption bands, this probe is


not able to distinguish separate sites.


Furthermore, the spectrum of Reichardt'


dye on


VAC200 is the same as that of VAC28 within experimental error.


Spectra were taken


at different loadings of the dye, and though the color intensified with increased loading,


the maximum in the spectrum of each sample occurred at the same wavelength.

lowest loading, the silica was a faint red with a lot of the white silica particles.


At the

The red


color continued to darken with increased loading until the surface would no longer adsorb

any more dye, at which point, the red color was so intense that the silica appeared a deep


purple.


The observed shift of 492 nm corresponds to an Er(30) value of 58.1 kcal/mol.


Since the shift of VAC28 and VAC200 are the same, betaine must be oriented on the

surface in a manner that does not permit the specific hydrogen bonding of its carbonyl


group to the silanol.


Thus, the non-specific interaction with the silica surface is more


than the combined


"polar" interaction with methanol (Er(30)


= 55.4 kcal


mol') but








mol-).20


A larger shift occurs for the silica surface than for any polar, non-hydrogen


bonding solvent e.g. CH3NO2 (46.3 kcal mole').


Gas-Solid Equilibria


Extensive literature is available describing the use of temperature programmed

desorption, TPD, and differential scanning calorimetry, DSC, to characterize the acidity


of solids.


As shown in an earlier report,12 these methods are not as effective as the cal-


ad method in distinguishing sites of different acidity.


Thus, data from these studies often


provide values representing a complex averaging of the different interactions.


In trying


to correlate reactivity, in an application that involves only the stronger sites on a series

of solids, the average value provides little insight concerning the number or strength of


the reactive site.


A second complicating factor in ascertaining the acid-base component


reactivity


from


TPD and


DSC


involves


the added contribution


to enthalpies


measured with these techniques from non-specific dispersion interactions.


Even in the


absence


donor-acceptor


interactions,


a negative


enthalpy


accompany


condensation of a gaseous molecule on a solid surface.


This can be larger than the heat


condensation


gaseous


molecule.


indication


magnitude


contribution


from


effect


can be


appreciated


comparing


recently


reported


enthalpy for a gas phase-solid equilibrium with that reported here for the solution-solid

equilibrium.


Dumesic21


reported


an enthalpy


7 kcal


binding


gaseous








pretreated as reported for the Cab-O-Sil sample.


This sample preparation leads mainly


to isolated silanol


groups on the surface of both silica gel samples.


A calorimetric


titration


was


cyclohexane


room


temperature


employing


pyridine


concentrations.


An enthalpy of -3 kcal mol


was measured indicating an interaction with


an isolated silanol.


The large difference of nearly 20 kcal mol' in the cyclohexane and


gas-solid


enthalpy


must


associated


contribution


from


non-specific


dispersion interactions of gaseous pyridine with the solid surface in the gas phase-solid


equilibrium.


When


the equilibrium is studied in a cyclohexane solution of pyridine


instead of gaseous pyridine, these interactions are canceled out.


A cyclohexane molecule


on the solid surface of the reactants is displaced by a pyridine molecule in the adduct


roughly canceling the non-specific dispersion interactions.


The non-specific pyridine-


cyclohexane


interactions


reactant


leaving


solution


replaced


cyclohexane-cyclohexane interactions from the displaced cyclohexane entering solution.

These solution interactions correspond to a small fraction of the heat of solution because

the pyridine molecules on the surface are interacting with cyclohexane molecules at the

interface.


E and C Analysis


Enthalpy values for a series of different donors bonding to silica gel have been


measured


by a


calorimetric procedure


using


dilute


base


hexane."


The reported


enthalpy for the reaction of pyridine with silica gel correlated with that determined by







34
enthalpies of adduct formation reported for the other bases can be assumed to involve

only the same site and the donors can be assumed to be fully coordinated to the extent


of a +


1 kcal mol' error limit.


The E and C model13 (Equation (2-2)):


(2-2)


-- BEAE


+ CACB


been


used


to interpret


enthalpies


reactions


solution


to determine


when


contributions other than donor-acceptor interactions are involved.


Accordingly, it was


of interest to determine if the enthalpies measured calorimetrically for coordination to

silica gel were dominated by the same factors that influence bond strengths in solution.


The enthalpies and reported"13 EB and


values for the donors are substituted into


Equation (2-2) to produce a series of equations that are solved for EA and CA.


and CA


The EA


values obtained for the strongest site on silica gel are 2.30 + 0.88 and 2.48 +


0.15 respectively.


This C/E ratio (1.08) is similar to the C/E ratio for the hydrogen


bonding acceptor octanol (C/E


= 1.02).


It is interesting that both the EA and CA


values


calculated for silica are larger than the corresponding values for phenol (EA


= 2.27, CA


= 1.07) or trifluoroethanol (EA


= 2.07, CA


= 1.06).


This indicates that this site on the


silica is more acidic than the above organic hydrogen bonding acceptors.


The data


fit is


shown


Table


good


results


indicating


that the


enthalpies


for the adsorption


of these donors by


silica


are dominated by


donor


acceptor interactions.


These EA and CA


the ER and C1 values reported13c


for over


values can be combined in Equation (2-2) with

70 different donor molecules to predict their


+ W



































Q U2




-C


0~
4.2.

.00


aCS
4-
wE!
S'-4.
eq.)


4-
E
,I4


1 .0


N


0 T3








be calculated


from


the differences


the enthalpies of


adsorption


to determine


enthalpic contribution to competitive binding.


The practical applications of competition


binding include separations, adhesion, removal of bound molecules and solvent selection

for acid catalyzed reactions.


Conclusions


Upon evacuation of


silica,


there is a change in


the types of hydroxyl groups


available


reaction.


With


mild


pretreatment,


there


large


number


polyhydrogen-bonded silanols which give rise to the strongest binding.


Also present are


monohydrogen-bonded silanols which are weaker acids than the polyhydrogen-bonded


silanols.

200 C.


A noticeable change in the strength of the two sites occurs upon evacuation at

At this point the polyhydrogen-bonded silanols have been effectively eliminated


leaving the monohydrogen-bonded silanol as the strongest site while a new,


weaker, site


has appeared that can be attributed to the isolated silanol on the silica surface.


enthalpy of adsorption of donor molecules by the strongest site are dominated by donor-


acceptor interactions.


Gas phase-solid equilibria have shown substantial contributions to


the enthalpy of adsorption from non-specific, dispersion interactions.












CHAPTER 3


ACIDITY OF


A1CI,(SG),


Introduction


Recent efforts from our laboratory have produced a novel solid acid, denoted as


22,23,24


material


is synthesized


condensation


reaction


between


aluminum chloride and silica gel in CCl4.


Characterization by


NMR spectroscopy23


indicates the presence of surface tetrahedral aluminum sites.

these surface sites are illustrated in Figure 3-1. The very


Two possible structures of


strong acidity of AlC12(SG),


been


shown


tivity


cracking23


and


dehydrohalogenation/hydrodehalogenation


reactions2s


under


mild


conditions


where


typical zeolite or silica/alumina catalysts are inactive.

understanding of the nature of the acidity was needed.


With a species this acidic, a better

Hopefully this investigation would


determine the cause of the catalyst deactivation and lead to the synthesis of a more

durable catalyst.


Previous efforts to characterize the surface of the catalyst included


FTIR spectroscopy.23


Al NMR and


Al NMR investigation revealed the presence of tetrahedral


aluminum centers in the active catalyst and octahedral aluminum in catalysts that were


inactive.


One of the most common methods of examining acid sites on solids utilizes the


infrared


spectrum


adsorbed


pyridine.4'26


Pyridine


bound


to Lewis


AIC1,(SG),.






















CI




0









(including


hydrogen


bonded


pyridine)


exhibits different spectral


shifts


pyridinium ion arising from protonation by Br6nsted sites.4


The shift in the pyridine


frequency


bound


to Lewis


relative


to that


free pyridine


gives a


qualitative


assessment of acid site strengths.


AICl2(SG), prepared in a series


This procedure has been used to evaluate the acidity

s of solvents23'27 and is found to correlate with the


reactivity of the solid in acid cracking reactions.


"spectroscopic


titration"


catalyst


pyridine


would


important tool for the characterization of the catalyst surface.


Since the pyridinium ion


and Lewis bound pyridine sites have two distinctly different absorbances in the 1400-


1700 cm


region, incremental additions of low concentrations of pyridine can indicate


the nature of the most acidic site.

Calorimetry is the most direct method of determining acidity because it yields a

value for the enthalpy of interaction which is a direct measure of the bond strength of the


solid


acid-donor


determine


molecule adduct.


surface


acidity


The cal-ad


a heterogeneous


method has been


Pd/carbon


used


catalyst"2


previously


silica.2'


Correlation


enthalpy


changes


with


spectral


changes


various


donor


concentrations would indicate if there are different types of acid sites and which sites are


most


reactive.


comparing


results


obtained


from


calorimetric


titrations


AICl2(SG),, using pyridine and 2,6-lutidine as basic probe molecules, the contribution

to the measured enthalpy from coordination to protonic and aluminum centers can be


determined.


Steric effects with the latter donor are expected to lead to smaller enthalpies








Experimental


Reagents


AIC12(SG)n


catalyst


used


as prepared


previously.24


Pyridine


cyclohexane were purified as described earlier.


used as supplied by Aldrich.


2,6-Lutidine (redistilled, 99+%) was


Carbon tetrachloride was distilled over P205.


Instrumentation


FTIR was performed on a Nicolet 5PC FTIR spectrometer.


UV/vis spectra were


obtained using a Perkin Elmer Lambda 6 UV/vis Spectrophotometer.


Infrared Analysis


A sample of AlCl2(SG), is placed in a vacuum desiccator with a pyridine reservoir


and evacuated to generate a pyridine atmosphere inside.


The catalyst is kept in the


pyridine environment at room


temperature


hours,


during which time the


catalyst changes to an off-white color.


The catalyst is then removed from the desiccator


and placed under vacuum


for at least 2 hours to remove excess pyridine.


Infrared


spectra are taken as fluorolube mulls of the catalyst on NaCi plates prepared in an inert

atmosphere.

The infrared adsorption titration of pyridine was conducted in an inert atmosphere


of dry N2.


Injections of pyridine were made into a slurry of 0.10 g of catalyst in 10 mL









to dry in the inert atmosphere.

mull on NaCI plates. Pyridin


Calorimetric


The infrared spectrum was then obtained as a fluorolube


e loadings ranged from 1.2 mmol/g to 11 mmol/g.


Titrations


The calorimeter and the procedure to measure enthalpies was described in Chapter

The cell was loaded with 0.50 g of catalyst and 50 mL solvent (CCl4 or cyclohexane)


in a dry N2 atmosphere.


The concentrations of pyridine and 2,6-lutidine ranged from 0.2


- 50 mmol/L.


Results and Discussion


Infrared Examination of Adsorbed Pvridine


order


to obtain


qualitative


assessment


different


AIC12(SG),, the infrared spectrum of adsorbed pyridine can be used.


A band at


- 1540


involves a C-N+-H bending vibration and is used as a fingerprint for pyridinium

A band at 1452 cm-1 is assigned to Lewis-bound pyridine and this includes pyridine


that is hydrogen bonded or bound to aluminum.

both Lewis and pyridinium species. Previous Fl


solely on the vapor deposition of pyridine technique23

which the catalyst was saturated with pyridine. This ai


A third band at 1485 cm' results from


:IR characterization of this catalyst relied


This provided a spectrum in


analysis indicated the presence of


Brinsted and Lewis acid sites but was unable to determine which site was stronger and








Infrared Titration


Figure 3-2 illustrates the results of the infrared titration in carbon tetrachloride.

Upon addition of 1.2 mmoles of pyridine per gram of catalyst, the pyridinium ion is the


only


species present absorbancee at


1540 cm').


As excess pyridine is added to the


catalyst (11 mmol/g), absorbance from another species is observed.


This is assigned to


a hydrogen-bonded adduct since coordination to an aluminum center can be ruled out

with the results obtained from the calorimetric titrations, vide infra.


Calorimetry and Adsorption


A non-polar solvent is utilized to elucidate the enthalpy of interaction between the


AICl2(SG), catalyst and pyridine.


Cyclohexane is the best choice for this except the


catalyst has been shown to be a good hydrocarbon cracking and isomerization catalyst.?

Carbon tetrachloride is another good solvent except it is known to form a complex with


strong electron donor amines (e.g. pyridine).


The results of titrations in both solvents


yielded the same heats within experimental error (the value in carbon tetrachloride was


corrected by adding 1 kcal mole`' for the heat of complex formation with pyridine).


remaining titrations were done in cyclohexane since there was no contribution to the heat

from isomerization or cracking at the conditions of the calorimetric titration.

To determine the types and strengths of acid sites present on the catalyst surface,


a calorimetric titration of the catalyst at low base concentrations was done.


Pyridine and


- -- -




































C
C
U
ED
U
C
UI
C










CLC


ED

C1
U'


[Base] (mM)









slightly stronger


hinderance


donor than


acceptors


pyridine.


larger than


However,


proton


methyl


resulting


groups result


in a


in stenc


weaker interaction.


Figure 3-3 shows the similarity in reactivity between the two bases.


This titration rules


out the possibility of any Lewis acidity other than hydrogen bonding, since the enthalpy

of adduct formation for 2,6-lutidine is slightly higher than the enthalpy for pyridine.

Steric constraints are expected with 2,6-lutidine toward an aluminum acceptor center and

would lead to a reduced enthalpy for this donor compared to pyridine if such centers


were involved.


It is obvious from this experiment that the acid sites on the catalyst are


protonic and are very strong.

By assuming that the pyridine is completely completed (a necessary assumption

due to the adsorption results, vide infra), an enthalpy can be calculated for each addition


of base.


This enthalpy is calculated by dividing the amount of heat for each injection by


the number of moles of base added for that injection, yielding a value in kcal (mol base


added)-1


Figure 3-4 is a plot of the calculated enthalpies vs. the total amount of base


added.


We see from this graph that the first site has a AH


-40 kcal (mol pyridine


added) '


This interaction is considerably stronger than the -28 kcal molr' enthalpy of


adduct formation between trimethylaluminum and pyridine and much larger than the


-12.6 kcal mol'1 interaction of pyridine with silica gel.

there is a second plateau in the calculated enthalpies.

type of acid site on the catalyst with a AH = -20 kcal


It is also interesting to note that

This is an indication of a second


I (mol pyridine added)-'


In order to quantify the acid sites present on the catalyst, the cal-ad method was





























x x
X x
X
X'

'C

X'
X'

)C '
'C'


0.15


pyridine added (mmol)
































A A


A





A C
O

O







AC
A

IU
C l
,A
I


(mmol/L)








over the entire concentration range in which the catalyst adsorbs the base.


Figure 3-5


shows the results of the calorimetric titration of the active catalyst over a large pyridine


concentration range.


A catalyst which has been deactivated in a flow reactor by allowing


a constant flow of N2 to pass over it for 24 hours at room temperature (this treatment


resulted in the loss of HCI and reduced the activity of the catalyst) is also shown.


deactivated catalyst has undergone a loss in the number of acidic sites resulting in the


reduction of the total heat evolved in the titration.


A small number of the active sites


remain


the deactivated


catalyst


as indicated


evolved at


low pyridine


concentration.


The adsorption titration produced an interesting result.


An initial addition of 0.14


mmoles of pyridine to 0.5 grams of catalyst in 50 mL of cyclohexane was made.


was allowed to come to equilibrium and the UV/vis spectrum was taken of a sample of


the solution above the solid (see Figure 3-6).


This spectrum indicated that there was no


pyridine in solution but there was a small amount of a new species present at longer


wavelength (267 nm).


Appearance of this band was accompanied by an increase in base


line adsorbance at all wave lengths most likely due to small solid particles in the solvent.

At this amount of added pyridine, (nearly 0.3 mmol/g), the FTIR spectrum of the solid


catalyst shows that the pyridinium ion is formed at this concentration (Figure 3-2).


successive additions of pyridine are made to the solid in cyclohexane, the absorbance at


nm increases.


When


the amount


of pyridine added


reached


1.6 mmol/g,


pyridine was found in solution absorbancee at 251 nm).


At this and higher pyridine






































240 260


wavelength (nm)






50

To identify the species responsible for absorbing at 267 nm, small amounts of a


cyclohexane


solution


saturated


were


added


solution


pyridine


cyclohexane.


The protonation of pyridine was followed spectroscopically.


A species


was formed which absorbed at 267 nm and an increase in base line absorbance at all


wavelengths is again observed.


From this result, it can be concluded that in the titration


of AlC12(SG),, at concentrations between 0.3 and 1.6 mmol/g of pyridine, pyridinium


chloride is formed from dehydrohalogenation of the catalyst.


The dehydrohalogenation


of AIC12(SG), is also the principal reaction leading to loss of activity when the catalyst


used in


cracking.


As reported


previously,23


addition of HCI


to the feed


greatly


enhances catalyst lifetime for the cracking reactions.


Conclusions


The calorimetric titrations indicate that the most acidic site on the surface of the


AICl2(SG), catalyst produces a heat of -40 kcal mol' with pyridine.


shown,


It has also been


using the pyridine/2,6-lutidine comparison, that this was a Brinsted acid site.


In addition, Figure 3-5 shows that no significant heat is evolved after the loading of

pyridine reaches about 1.5 mmol/g.


It is


clear


from


infrared


titration


catalytic


activity


A1C12(SG)n solid acid results from a Brdnsted type acid site (initial addition of pyridine


produced a spectrum in which the pyridinium ion was the only species present).


the loading of pyridine reaches about


After


1.3 mmol/g, a second absorbance in the FTIR






51
In attempting to use UV/vis spectrophotometry as the method to quantify the

amount of pyridine adsorbed by the active catalyst, is was discovered that the pyridinium


ion was being produced.


When the amount of pyridine added reached 1.6 mmol/g, no


more pyridinium ion was formed and free pyridine was found in solution.

In conclusion, the AlCl2(SG), catalyst is a very strong Br6nsted acid with about

1.5 mmol/g of active sites and it has been confirmed by the UV/vis adsorption study that

the catalyst is deactivated by the loss of HC1.












CHAPTER 4
ACIDITY OF PHOSPHOTUNGSTIC ACID


Introduction


Phosphotungstic acid


(H3PW,2040)is a member of the


heteropoly acid (HPA)


family with a Keggin unit structure (see Figure 4-1).


A phosphorus(V) ion occupies the


central atom position and is surrounded by a tetrahedron of oxygen atoms.


Each of these


oxygens is shared by three tungsten(VI) ions and occupies an octahedral position about


each tungsten ion.


The secondary structure of the bulk solid is illustrated in Figure 4-2.


It can be seen here that the Keggin anion exhibits body center cubic packing with the


cation being a protonated form of a cluster of water molecules.


The number of water


molecules depends upon the degree of hydration.

The acidity of heteropoly acids has been investigated by comparing activities of


liquid


phase30


supported31


HPA


to conventional


acids


model


catalyzed


reactions.


Attempts have also been


made to quantify the acidity of HPA


using the


Hammett acidity indicator method.32'33


It has been reported that phosphotungstic acid


supported on activated carbon is a stronger acid than a sulfonated ion exchange resin34


In an attempt to obtain more information about the nature of the acidity of HPA, an in-

depth calorimetric and spectroscopic investigation of phosphotungstic acid is reported

here.















w?


0-






















HPA


HPA


HPA


HPA








Nuclear magnetic resonance (NMR)


has been a


useful


for the structural


characterization of a large number of different heteropoly acids.35


was directed toward the heteroatoms and addenda nuclei.


Most of this work


Attempts to correlate shifts in


the NMR due to changes in acidity (as a result of varying hydration conditions) have not


been reported.


It is not expected that the nuclei mentioned above will undergo significant


changes upon loss of water from the structure, but the proton NMR of the HPA in an

aprotic solvent may lead to an interesting result.

Calorimetry is an important tool that has not been utilized for the determination


of thermodynamic


for phosphotungstic


acid.


The cal-ad


method'1228 has been


designed specifically for the investigation of solid acids/bases.


There is a great deal of


interest in being able to quantify the acidity of strong acids and in being able to compare


the reactivity to other strong acids.


By determining enthalpies of interaction of the


phosphotungstic acid with a series of bases, E and C values36 can be calculated for the


acid.


This then allows for the calculation of an enthalpy of interaction for the acid and


any base in the E and C database.

Another important function of calorimetry is its ability to distinguish between


Br6nsted and Lewis acidity.


By comparing the heats evolved using a stericly hindered


base (e.g. 2,6-lutidine), coordination to a Lewis acid would be unfavorable and would

give less heat than coordination by a slightly weaker base (e.g. pyridine) which has an


accessible base site.


For a Br6nsted base, the sterics would not play a role and the heat


for the lutidine would be slightly greater than that for the pyridine.









Infrared


spectroscopy


is another


method


which


been


used


to distinguish


between Brinsted and Lewis acid sites.


Separate absorbances for the pyridinium ion and


Lewis bound pyridine can be found in the


1400-1600 cm-'


region of the spectrum37


photoacoustic-infrared


study,


addition


pyridine


to phosphotungstic acid


resulted in the appearance of exclusively pyridinium ion38 indicating that this HPA is

a Brbnsted type acid.


Experimental


Purification of Materials


The phosphotungstic acid was obtained from Fisher.


after various dehydration pretreatments.


It was used as supplied and


These pretreatments included evacuation (VAC)


and calcination under a flow of N2 (CALC) at different temperatures.


The hydrated


phosphotungstic acid was white while all heat treated samples of the acid had a faint gray


color.


The pyridine was distilled over CaHll2 using a 12 in.


Vigreaux condenser.


other solvents were purified using literature methods39.


NMR Spectroscopy


All 'H and 31P NMR spectra were obtained on a Varian VXR300 300 MHz NMR


spectrometer


using


distilled


water


or d4-DMSO


as the


solvent.


proton


phosphorus


spectra


are referenced


to tetramethylsilane


(internal


standard)









Calorimetry


The procedure for the calorimetric titrations has been described in Chapter 1.


A volume


either


or 75


used


calorimetric


experiments.


gram


phosphotungstic acid was used for a typical titration.


Additions of dilute solutions of


base to the acid solutions above resulted in final base concentrations ranging from 0.6

to 50 mM.


Fourier Transform Infrared (FTIR) Spectroscopy


All FTIR spectra were obtained on a Nicolet 5PC FTIR spectrometer as Fluorolube or

Nujol mulls on NaCI windows and were prepared under a N2 atmosphere.


Results and Discussion


NMR Spectroscopy


No solvent effect was observed in changing from water to perduetero dimethyl


sulfoxide (d6-DMSO) for the 3"P NMR of the untreated phosphotungstic acid.


chemical shift of the singlet in either solvent was -14.5 ppm (see Figure 4-3).


is in good agreement with previous reports"


drying procedure.


This value


The resonance was also insensitive to the


From the untreated sample to the most stringent drying condition


(evacuation at 250


there


was no change in


the 31P NMR.


The fact that this


resonance was not affected by the drying procedure is not surprising since the phosphorus






























A
_________A


Temperature (degrees C)


R x j








Since the protons on


phosphotungstic acid are acidic, an aprotic solvent


DMSO was chosen) was necessary to observe the 'H NMR of the compound.


untreated acid exhibited a singlet chemical shift of 4.6 ppm (see Figure 4-4).


Adding a


drop of water to the NMR tube containing the acid in d6-DMSO, the chemical shift was


reduced to 4.3 ppm.


The 'H NMR of all the dried samples exhibited singlet chemical


shifts of 7.8 + 0.2 ppm (see Figure 4-5).


Upon addition of a drop of water to each of


these samples, the chemical shift was reduced to 4.4 + 0.2 ppm.

chemical shifts obtained at different evacuation temperatures. Ev


Figure 4-6 shows the


en though this physical


method reveals a change in the structure with a change in the pretreatment, it is not a

good measure of acidity since the spectral change is not proportional to the change in

acidity, vide infra.


Calorimetry


The initial


titrations were conducted on a 25


w/w%


phosphotungstic acid on


Ambersorb 572 carbon molecular sieve.


Due to difficulties in obtaining heats from the


bead form of the support4


, the doped carbons were powdered.


To determine the solid


acidity of the supported acid, the cal-ad method was used.


This method requires the use


a poorly


solvating


solvent


order


to eliminate


enthalpic


contributions


dispersion interactions.


When a pyridine titration was conducted in cyclohexane, there


no noticeable difference


evolved


between


undoped


doped


supports.






63

In a further attempt to determine the solid acidity of the phosphotungstic acid, a


similar pyridine titration was conducted on the bulk solid in cyclohexane.


There was no


detectable heat evolved but an infrared spectrum of the solid after the titration provided


a promising result,


vide infra.


The difficulty in obtaining any heat from the solid arises from its low surface area

(surface area for the phosphotungstic acid is generally found to be less than 10 m2 g-).

It has been reported31'41 that the heteropoly acids display "pseudo liquid phase" behavior


with respect to polar solvents. That is, the presence of some polar solvent molecules

allow transport into the bulk of the solid. This allows access to sites which are not on


the external surface of the bulk solid.

Acetonitrile was then chosen as the solvent for the calorimetric titrations (it should

be noted that the phosphotungstic acid, with no pretreatment, was soluble in acetonitrile).

It is polar and weakly basic, so the heat evolved for the addition of a stronger base will

include the displacement of the acetonitrile from the acid site as well as enthalpy for the


coordination of the stronger base to that site. When pyridine was used as the titrant, a

significant amount of heat was observed (see Figure 4-7). A calculation of the enthalpy


of interaction for each injection, assuming complete complexation of the pyridine, gives


rise to an interesting result. Figure 4-8 shows the graph of calculated enthalpies as a

function of total moles of pyridine added. This graph shows that a constant enthalpy (-


6.8 0.3 kcal mol-') is obtained for this titration and heat is evolved until the amount


of pyridine reaches about one mmol.


This corresponds to three times the number of































X )
X)
'C


Pyridine (mmol)







65










8-

7 X
X
X
x
6
0
E 5-
ro x


(4


2-


1


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Pyridine added (mmol)






66

moles of phosphotungstic acid in solution which suggests all three protons of the acid

have been completely completed.

In an attempt to determine how the acidity of the phosphotungstic acid is affected


dehydration,


several


samples


were


prepared


under


various


drying


conditions.


Figure 4-9 and Figure 4-10 illustrate the effects of evacuation and calcination at various


temperatures respectively.


It is interesting to note that all the dehydrated samples were


only slightly soluble in acetonitrile but the endpoint of each titration continued to occur


at the same place as in the hydrated sample (about one mmol of base added).


None of


the enthalpies obtained from all of the samples that were dried under vacuum remained

constant although it is interesting that the samples evacuated at 150 OC and 200 C had

a similar site distribution but the sample evacuated at 250 C contained much stronger


sites.


The decrease in enthalpy with each injection suggests that the conversion of the


weaker to stronger acid sites is not complete and the enthalpies calculated are an average


of the strong and weak sites.


Calcination at elevated temperatures seemed to convert all


the acid sites to the same type of site, but the acidity of that site was dependent upon the


amount of time the sample was exposed to the elevated temperature.


Figure 4-10 shows


that an enthalpy of 17.4 + 0.4 kcal mol'1 is obtained when the acid is calcined at 300


C for three hours and an enthalpy of 31.4


+ 0.9 kcal mol-' is obtained after twelve


hours at 300 C.

It is important to determine if the dehydration procedure changes the nature of the


acidic site.


To determine if the Br6nsted sites were converted to Lewis sites, titrations


























X
X
X









O X


A X



A. A0 S C


Pyridine added (mmol)




































0 0.2 0.4 0.6 0.8 1 1.2 1.4


Pyridine added (mmol)








69











16-

14

12x
A A

i 10--- ------


.0 A
e8-

4 x
X

2-&
Sx
X

0-
0 0.5 1 1.5 2 2.5
Base added (mmol)








as described in Chapter


Figure 4-11 illustrates the results of these titrations.


heats obtained for the additions of 2,6-lutidine to the dried phosphotungstic acid yielded


heats slightly higher than those obtained for pyridine at the same concentration.


This


indicates that the drying procedure merely enhances the Br6nsted acidity and does not

convert the Brinsted sites to Lewis sites.


FTIR Spectroscopy


The FTIR spectrum of the bulk phosphotungstic acid after the initial titration in


cyclohexane (which did not yield any detectable heat) is shown in Figure 4-12.


spectrum shows the presence of pyridinium ion absorbancee at 1540 cm1) indicating that


there are sites on the acid which are strong enough to be completely ionized.


apparent from this and the calorimetry data that there are strong acid sites present in the


phosphotungstic


a majority


them


are not accessible


non-polar


cyclohexane.


Infrared spectra taken of the phosphotungstic acid samples recovered at the


conclusion of the calorimetric titrations in acetonitrile also contained the same bands

indicating Br6nsted acidity.


Conclusions


The acidity of the phosphotungstic acid


nature.


has been shown


Upon dehydration of the acid, the sites become stronger ac


to be of a Br6nsted

:ids. The conversion


of the sites is dependent upon the method and length of pretreatment.


When the acid is






































022139








0.2850~





0.2800~


022700


0.27001





0.2550


0.25270O


WA;tt S








which averages the strengths of the sites.


Calcination of the acid converts all the sites


to the same strength and longer calcination times at the same temperature yield stronger


sites.


"pseudo liquid phase" behavior reported previously31'41 for the heteropoly


acids has also been confirmed here.


The calorimetric experiments in cyclohexane didn't


yield measurable heats because of the small number of sites accessible, but the strong


sites were found to be present in the FTIR spectrum.


When acetonitrile was used as the


solvent, its polarity allowed access to all the sites of the acid.












CHAPTER 5
ACIDITY OF HY ZEOLITE


Introduction


Research in the field of zeolites encompasses a very large range of materials and


applications.


Novel


zeolites are continually being synthesized


for a great variety of


chemical needs (a number of these have been reviewed by Davis42).


The basic structure


of a zeolite consists of a framework of neutral silicon dioxide moieties covalently linked


to negatively


charged


aluminum(III)


moieties43


Figure


5-1).


This


creates


negatively charged framework and requires a cation to maintain charge balance.


ability of the zeolite to act as a cation exchanger is one of the properties responsible for


its popularity.


The amount of counter ion needed to balance the framework charge can


be tailored by incorporating the desired amount of aluminum dioxide moieties. Another

reason for the zeolites popularity arises from the fact that the framework is porous. This


porosity is another variable which can be synthetically tailored.


A good example of this


is the shape selective catalytic isomerization of hydrocarbons


using


the channel-like


structure of H-ZSM-5.

Out of this vast array of zeolites, HY was chosen as the first to be studied using

the novel cal-ad method because the acid form of this zeolite has been of interest in acid


catalyzed reactions in industry.


There has been a Rreat deal of work reported on the












structural characterization of Y


zeolites43 including an X-ray crystal structure" and a


pulsed-neutron powder diffraction45 as well as many methods for the determination of

the acidity of the acid form.

A physical method that has been utilized extensively is infrared spectroscopy.


Early work in this area included


investigations of the hydroxyl


region to distinguish


between different types of hydroxyls on the surface.46'47


The results of these studies


concluded that there are three band


in the hydroxyl region.


A band at 3742 cmn' results


from absorbance by isolated silanols (similar to those found in silica).


Two bands


present at 3643 cm-' and 3540 cm-' result from absorbance by bridging hydroxyls in the


large cage (a-cage) and small cage (3-cage) respectively.


Many of these studies also


included adsorption of bases (ranging from ethylene to water) from the gas phase to

determine the effect of coordination on the infrared spectrum.

Another method which has received a great deal of attention more recently is


nuclear magnetic resonance.


Although the aluminum and silicon centers are active NMR


nuclei, they have not been very useful for the determination of the acidity of zeolites.


The most informative work has involved 'H NMR.


Resonances corresponding to the


same hydroxyl species described above have been separated and assigned.48


The use


of basic probes with NMR active nuclei has also been an area of increasing interest.


great deal has been done with nitrogen49 and phosphorus50 NMR studies but they have

not led to a direct quantification of the zeolites acidity.

Other methods that have been used (temperature programmed desorption51 and








been


microcalorimetric


investigations into


the acidity


of HY.53


These


have studies


involved the gas phase adsorption of pyridine or ammonia and concluded that there is a

uniform distribution of strong acid sites with heats of adsorption of 113-125 kJ/mol for


ammonia.


It is interesting to note that other zeolite samples (e.g. H-ZSM-5) examined


in these studies did not contain a uniform distribution of strong acid sites, but did exhibit

slightly higher heats of adsorption.

Another important aspect of the acidic zeolite is its ability to convert its Brinsted


acidity to Lewis acidity.


It has been well established that this can be accomplished by


calcination of the zeolite at temperatures above 500 C.47


It is also well established54


that the mechanism for this conversion consists of the release of aluminum from the


framework


concomitant with dehydroxylation creating


hydroxide species.


This species


an extraframework aluminum


is further dehydroxylated to yield a coordinatively


unsaturated A13' Lewis acid center.


In this study,


the cal-ad method was used on the HY


zeolite in an attempt to


distinguish the different acid sites calorimetrically.

was used in addition to the calorimetric results. The


Concurrently, FTIR spectroscopy


;se results were also correlated using


the E and C analysis'3 to determine if the interactions of the bases with the solid acid

could be predicted using the solution data for the bases currently available in the E and

C database.








Experimental


Purification of Materials


zeolite was obtained


from Aldrich in


the ammonium exchanged form


(molecular sieves catalyst support, Linde LZ-Y62, Lot # 00220BZ).


This was converted


to the acid form by calcination under flowing N2 for 24 hours at either 350 C (HY350)


or 650 C (HY650).


The cyclohexane was filtered three times over activated carbon


followed by distillation over P205.


Vigreaux column.


The pyridine was distilled over CaH2 using a 12 in.


All other solvents were purified using literature methods39


Calorimetry


The procedure for the calorimetric titrations has been described in Chapter 1.


volume


mass


zeolite


were


used


calorimetric


experiments.


Additions of dilute solutions of base to the slurries above resulted in final


base concentrations ranging from 0.8 to 50 mM.


Fourier Transform Infrared (FTIR) Spectroscopy


FTIR


spectra


were


obtained


on a


Nicolet


FTIR


spectrometer


Fluorolube on NaCI windows and were prepared under a N2 atmosphere.


titrations were done in 10 mL of CCL on 0.1 g of zeolite.


The infrared


The zeolite was handled


under an inert atmosphere at all times.









Results and Discussion


Calorimetry


A non-polar solvent (cyclohexane) was chosen for the calorimetric titrations in

order to obtain an enthalpy of interaction, between the solid acid and the base probe


molecule, in which the dispersion interactions are cancelled out.


Figure 5-2 shows the


results of


several


titrations of HY350 with


pyridine.


There was


some difficulty in


obtaining results that were as reproducible as the previous calorimetric results.


be attributed to the zeolites great affinity for water.


This may


The difficulty may also arise from


a difficulty in diffusing to the interior sites of the zeolite on a reasonable timescale for

the calorimetric experiment (in almost every case, higher concentration injections yielded

broader deflections than at lower concentrations, indicating heat evolution over a longer


time frame).


In order to use the cal-ad method,


samples were then obtained.


the adsorption data for the HY350


For every addition of pyridine, up to


1.32 mmol g-'


zeolite


sample


completely


adsorbed


probe.


This


result


leads


one of


conclusions.


The first is that the number of acid sites in the zeolite is larger than the


amount of pyridine that has been added (for the formula H56[(A102)56(SiO)136)], the


maximum number of protonic sites would be 0.49 mmol/g).


If this is the case,


equilibrium constant for coordination to these sites would have to be very large so that


complete complexation of the base was occurring.


The second possible conclusion is that


physisorption


of the base


is facilitated by the porous structure of the zeolite which




































,i,





dr
A


















t
A A

A A


A A

AA A





1A
A A
AA


AA







A


pyridine added (mmol)







80

The difficulty with either of these interpretations is that the base concentration in solution


cannot


measured


over


range


in concentration


calorimetric


experiments.


This makes it impossible to determine a value for the equilibrium constant


of binding for the cal-ad analysis.


Since the adsorption data are not helpful in


the determination of the zeolites


surface thermodynamics, a strictly calorimetric approach has been taken.


This approach


is susceptible to the same scrutinies as previous calorimetric investigations of solids were


as described


Chapter


Figure 5-3


illustrates


the calculated enthalpies for each


addition of pyridine assuming complete complexation of the base on HY350.


shows an initial enthalpy of -24.0 +


This graph


1.9 kcal mol' up to 0.05 mmol of pyridine added.


From 0.05 to 0.12 mmol of pyridine a constant enthalpy of -20.4 +


1.8 kcal mol


obtained.


Between 0.12 and 0.3 mmol of pyridine added,


there is a decrease in the


enthalpy calculated for each addition.

sites to the much weaker sites. The


This is caused by the transition from the strong


: calculated enthalpies in this range are the mole


fraction average of the few strong sites still remaining along with the abundance of weak


sites.


From 0.3 to 0.52 mmol of pyridine added a third enthalpy of -9.3 2.4 kcal mol-


calculated.


Even


though


enthalpy


calculated


have


overlapping error limits, it appears more clearly in Figure 5-3 that they are two distinct

sites.

Since the mass of zeolite for the calorimetric titrations was 0.5 g, the theoretical

maYimlm number of strnn nrotnnic sites corresnonds to 0.245 mmol of nvridine added.





































































































0 0.1 0.2 0.3 0.4 0.5


pyridine added (mmol)


r~ tr rIr % .nrn*


"


rrl rn


'"'


----~-'


i








weak sites.

(note in Cha


The weak sites in the zeolite may be due to remaining silanol functionality


pter


that at a calcination temperature of 450 C for silica there were still


isolated silanols remaining).


comparison


Figure 5-4.


heats evolved for pyridine and 2,6-lutidine is given


The first addition of 2,6-lutidine yields an enthalpy of -26.4 kcal mol-' but


it is obvious that the number of sites accessible to this base are greatly reduced compared


to those accessible to pyridine.


of Lewis acidity,


a-cage43).


In this case this reduced number of sites is not the result


vide infra, but is due to the small pore openings (7.1 A opening into


The distance across the methyl groups of the 2,6-lutidine molecule is 6.99 A.


This denies accessibility of the probe to the internal pore structure.


The conclusion from


this is that the number of sites on the external surface (surface accessible from the bulk

solution) of the zeolite is slightly less than 0.1 mmol g-' (since the first addition of 2,6-

lutidine was 0.049 mmol onto a 0.5 g sample of HY350).

The heats obtained for the calorimetric titration of HY650 using pyridine as the


probe are shown in Figure


throughout the entire concentration range.


A constant enthalpy was not obtained at any point


All the heats were less than those obtained


for HY350 indicating that the Lewis sites that are produced by this calcination procedure

are slightly less acidic than the corresponding Bronsted site.


FTIR Spectroscopy


The 1400-1700 cm-' region of the infrared spectrum has been used extensively47'4







83










9

8

7

6-
C 5-
(0 A

4





Sa A
OA



0 0.1 0.2 0.3 0.4 0.5 0.6
base added (mmol)




























U
U
U
U
U
U
U


pyridine added (mmol)


1 ~





































0.2894


0.290





0.260





0.240





0.220





0.200





0.~11

0.1746


SAMPLE



















































0.240


0.223





0.230





0.1625









for the pyridinium ion (Br6nsted site) occurs around

absorbance can be found between 1440 and 1455 cm-1


1540 cm' while the Lewis site

Figure 5-6 shows the presence


exclusively


Bronsted acidity


on HY350 whereas


Figure 5-7 illustrates


Lewis


acidity of HY650.


Infrared spectra for the pre-calcined (NH4Y),


shown in Figure 5-8.


HY350, and HY650 zeolites are


It is easy to see, from these spectra, the loss of ammonia in going


from


NH4Y


to HY350.


acidic


hydroxyls


apparent


from


comparison of the HY350 and HY650 spectra.


The spectrum of HY350 is curious in that


there appears to be only one absorbance in the acidic hydroxyl region.


to all previous reports43,46'47 for this type of zeolite.

in which the spectra were obtained. In the previot


This is contrary


The reason for this is the technique


is reports, all the spectra were taken


as self-supporting pellets of the zeolite, whereas here the spectra are taken as Fluorolube


mulls of the zeolite.


The peak position of the single absorbance in the Fluorolube mull


is nearly in the middle (3588 cm-1) of the two absorbances previously reported (3643 and


3540 cm-1).


It appears from this that the presence of Fluorolube broadens these bands


to the point that they are no longer separable.


An infrared


titration


of HY350 with


pyridine is shown in Figure 5-9.


asymmetry of the hydroxyl peak becomes more apparent after the addition of pyridine


indicating the presence of the two different hydroxyl absorbances.


Another point that is


interesting is that instead of a shift in


the O-H stretch,


as would be expected upon


coordination,


it seems as


though the proton has been completely removed from




























































WAVENUY.3ER.eO 3 I p I 1 51f


330








































































1400.0











pyridinium ion is formed in this reaction.


In order to see a shift in the O-H stretching


frequency,


an infrared


titration


HY350


benzonitrile


was


performed


Figure 5-10).


Again the asymmetric nature of the hydroxyl absorbance becomes more


apparent with the addition of base.

be observed (about 700 cm-'). Thi


In this case, the O-H stretching frequency shift can


e shift in the cyanide stretch can also be seen (from


2232 cm1' for neat benzonitrile to about 2425 cm"').


The absorbance at 2257 cm' is


assigned to physisorbed benzonitrile as this absorbance is removed upon evacuation at

25 C.


E and C Analysis


A determination of the acidity of HY zeolite can be obtained by determining heats

of interactions from calorimetric data for the interaction of a series of bases with the


zeolite.


The E and C model13 (see Equation 2-2) has been used to interpret enthalpies


if interaction


large database of


acids and bases.


substituting


the enthalpies


obtained for the low concentration additions of each base and the reported13' EB and CB


values for each respective base into


Equation


a series of


equations is obtained.


These equations are then solved simultaneously for the best EA and CA for the zeolite.


The EA and CA


mol-)')05 and


values obtained for the strongest site on HY350 are


74 0.52 (kcal mol-')05 respectively.


7.15 + 0.96 (kcal


The data fit is shown in Table 5-1.


A good fit results indicating that the enthalpies of adsorption for these donors by HY350


are dominated by donor-acceptor interactions.


The EA and CA


values obtained here for





















Table 5-1


Results of E and C fit for the most acidic site on HY350.


Base" -AHx -AHa Deviation
(kcal/mol) (kcal/mol) (kcal/mol)

34 Benzonitrile 13.6 13.9 0.3
79 THTPb 12.0 13.0 1.0
49 DMAb 19.5 20.4 0.9
16 Pyridine 24.0 22.4 1.6


a Numbers to the left of the base indicate their call number in
database. 3c


the E and C


b THTP is tetrahydrothiophene and DMA is N,N-dimethylacetamide