• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Literature review and introduc...
 Experimental procedure
 Experimental results
 Appendix
 Reference
 Biography
 Copyright














Title: Properties of polar molecules adsorbed on metal oxides.
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091619/00001
 Material Information
Title: Properties of polar molecules adsorbed on metal oxides.
Series Title: Properties of polar molecules adsorbed on metal oxides.
Physical Description: Book
Creator: Neikam, William Charles,
 Record Information
Bibliographic ID: UF00091619
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000565729
oclc - 13577988

Downloads

This item has the following downloads:

Binder1 ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
    List of Figures
        Page vii
        Page viii
    Literature review and introduction
        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
    Experimental procedure
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    Experimental results
        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
        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
        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
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
    Appendix
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
    Reference
        Page 102
        Page 103
        Page 104
    Biography
        Page 105
        Page 106
    Copyright
        Copyright
Full Text










PROPERTIES OF POLAR MOLECULES

ADSORBED ON METAL OXIDES












By
WILLIAM CHARLES NEIKAM


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










UNIVERSITY OF FLORIDA
June, 1961
















ACKNOWLEDGEMENTS


The author wishes to express his appreciation to

his research director and teacher Dr. W. S. Brey, Jr.,

for his expert instruction in the classroom, and for

his enthusiastic assistance in all the phases of this

research.

To Miss Akemi Saji the author is indebted for her

construction of the figures herein, for her assistance

in the measurement of line widths in the n.m.r. experi-

ments, and for her excellent translation of articles

from the Japanese language.

Thanks are due Messers K. D. Lawson and C. C. Shera

who have assisted in various stages of this research.

The National Institutes of Health who sponsored

much of this work and who granted the author a fellow-

ship during his first year of study are especially ap-

preciated.

The n.m.r. portion of this work was supported by

the University of Florida through their nuclear budget;

their help is gratefully acknowledged.

Finally the author wishes to thank the Department

of Chemistry for the assistantships granted him and for

the use of their facilities.
















TABLE OF CONTENTS


ACKNOWLEDGEMENTS

LIST OF TABLES

LIST OF FIGURES

Chapter

I LITERATURE REVIEW AND INTRODUCTION

A. Review of Literature on Dielectric
Properties

B. Review of Literature on Nuclear
Magnetic Resonance Line Width

C. Introduction to the Present Work

II EXPERIMENTAL PROCEDURE

A. Vacuum Apparatus Used in Dielectric
Experiments

B. Vacuum System Used in Nuclear
Magnetic Resonance Experiments

C. Alumina Sample

D. Capacitance Measuring Apparatus

E. Procedures for the Nuclear Magnetic
Resonance Studies

F. Measurement of Heat of Adsorption

III EXPERIMENTAL RESULTS

A. Surface Properties of the Alumina
Samples

B. Discussion of Dielectric Experimental
Results


iii


Page

ii

V
vii


1


1


11

15

18


18


20

20

22


26

28

29


29


40













C. Discussion of Nuclear Magnetic
Resonance Experimental Results

D. Comparison of Dielectric and Nuclear
Magnetic Resonance Results

E. Summary

APPENDIX A

I Calibration of Volume of Vacuum Line Used
in Dielectric Measurements

II Calibration of Vacuum Line Used in Nuclear
Magnetic Resonance Experiments

APPENDIX B

I Theory of Dielectric Studies

II Theory of Nuclear Magnetic Resonance
Studies

LIST OF REFERENCES

BIOGRAPHY


Page


56


66

70




95


97

98

98


99

102

105















LIST OF TABLES


Table Page

1 Description of Samples Used in Dielec-
tric Experiments 30

2 Surface Areas of Samples Used in N.M.R.
Experiments 31

3 Surface Areas of Samples Used in Dielec-
tric Experiments 33

4 Tabulated Heats of Adsorption Sample
Activated at 500C in Vacuum 37

5 Slopes of Temperature Dependent Line
Width Curves 63

6 Dielectric Studies
Sample Activated at 4000C for 4 Hours
in Vacuum 73

7 Dielectric Studies
Sample Activated at 5000C for 6 Hours
in Vacuum 76

8 Dielectric Studies
Samples Activated at 800C for 6 Hours
in Vacuum 78

9 Frequency Dependence of Tan 9
Sample Activated at 4000C for 4 Hours
in Vacuum 81

10. Dependence of Tan & on Frequency and
Water Content
Sample Activated at 400C for 4 Hours
in Vacuum 82

11 Measurement of Heat of Adsorption on
Sample Activated at 5000C 84

12 Dependence of Line Width on Amount of
Water Adsorbed
Sample Activated at 4000C for 4 Hours
in Vacuum 86












Page


Table


13 Dependence of Line Width on Amount of
Water Adsorbed
Sample Activated at 5000C for 6 Hours
in Vacuum 87

14 Dependence of Line Width on Amount of
Water Adsorbed
Sample Activated at 8000C for 6 Hours
in Vacuum 88

15 Temperature Dependence of Line Width
Sample Activated at 4000C for 4 Hours
in Vacuum 89

16 Temperature Dependence of Line Width
Sample Activated at 5000C for 6 Hours
in Vacuum 91

17 Rate of Desorption 93















LIST OF FIGURES


Figure Page

1 Vacuum Apparatus Used in Dielectric
Experiments 19

2 Vacuum Apparatus Used in N.M.R, Ex-
periments 21

3 Dielectric Cell and Leads 23

4 Water Adsorption Isotherm 35

5 Desorption Isotherms 38

6 Clapeyron Plot for Determining the Heat
of Adsorption 39

7 Dielectric Properties of the Alumina-
Water System for the Sample Activated at
4000C 41

8 Dielectric Properties of the Alumina-
Water System for the Sample Activated at
5000C 42

9 Dielectric Properties of the Alumina-
Water System for the Sample Activated at
8000C 43

10 Capacitance of the Alumina-Water System
for the Sample Activated at 4000C 47

11 Capacitance of the Alumina-Water System
for the Sample Activated at 5000C 48

12 Capacitance of the Alumina-Water System
for the Sample Activated at 8000C 49

13 Dielectric Properties of the Alumina-
Water System at 8, 16, 20 mc. 53

14 Frequency Dependence of the Loss Tangent
at Water Contents of 0, 11.9, 15.1 mg.
H20/g. A1203 54


vii












Figure


15 N.M.R. Line Width of the Alumina-Water
System for the Sample Activated at 4000C

16 N.M.R. Line Width of the Alumina-Water
System for the Sample Activated at 5000C

17 N.M.R. Line Width of the Alumina-Water
System for the Sample Activated at 8000C

18 Temperature Dependence of the N.M.R. Line
Width at Water Contents of 20.24, 37.72
and 46.80 mg. H20/g. A1203

19 Temperature Dependence of the N.M.R. Line
Width at Water Contents of 20.40, 35.01,
46.34, and 53.14 mg. H20/g. A1203


20 Comparison of the Loss
Width, 4000C Sample

21 Comparison of the Loss
Width, 5000C Sample

22 Comparison of the Loss
Width, 8000C Sample


Tangent and Line


Tangent and Line


Tangent and Line


viii


Page
















CHAPTER I

LITERATURE REVIEW AND INTRODUCTION


A. Review of Literature on Dielectric Properties.

The first extensive attempts to interpret the di-

electric properties of solids with molecules adsorbed

on their surfaces were made by R. McIntosh and his co-

workers (1 through 11).

Using the systems ethylene oxide-porous silica gel,

butane-porous silica gel, and ethyl chloride-porous sili-

ca gel, McIntosh (2) found that the curves obtained by

plotting change in electrical capacitance versus the

volume of material adsorbed per gram of solid consisted

of linear segments with different slopes. This type of

curve will be referred to as the capacitance curve and

the linear segments will be called the first region,

second region, and so on. The points where two linear

regions intersect will be designated as the first break

point, second break point, and so on.

With ethylene oxide and butane two linear regions

were observed while with ethyl chloride there were three

linear regions. Calculation of the apparent dielectric

constant of the adsorbed phase indicated that nonpolar

butane was adsorbed in a state similar in mobility to












that of the liquid, while polar ethyl chloride was in a

state less mobile than the liquid. The break points did

not agree with the value of the B.E.T. monolayer (12).

In Waldman's (4) study of n-butane and ethyl chloride

on nonporous titanium dioxide some interesting results

were obtained. Again for ethyl chloride two linear re-

gions were found in the capacitance curves. The inter-

section of these two lines agreed quite well with the

value for the B.E.T. monolayer. When butane was the ad-

sorbate, no correlation could be found between the break

point and the B.E.T. surface area. The temperature de-

pendence of the capacitance curves was investigated and

with ethyl chloride the temperature coefficient below

the monolayer was essentially zero. This suggested rigid

bonding in the first layer. After completion of a mono-

layer the presence of a temperature coefficient indicated

that a less rigid bond was present.

Additional investigation by Channen (5) and by Wald-

man (8) of polar molecules (NH3, SO2, ethyl chloride) ad-

sorbed on silica gel and titanium dioxide showed that

with silica gel no correlation existed between the ob-

served break points and the B.E.T. surface area. With ti-

tanium dioxide, however, the first break point always

agreed fairly well with the completion of the B.E.T.

monolayer.













Petrie (11) extended these studies on rutile (Ti02)

to include the adsorbates dichlorofluoromethane and

1,1,1,-difluorochloroethane. As in the earlier work an

increase in the rate of change of capacitance was ob-

served upon the completion of a monolayer. An analysis

of the temperature dependence of the capacitance ac-

cording to the methods developed by Kurbatov (14), Snel-

grove (10), and Channen (5) indicated that the molecules

behaved as rotational oscillators in the first layer;

that is, they were free to rotate but were held to their

positions by a considerable restoring force. With some

adsorbates an inverse temperature effect was observed

in the second linear region; this was interpreted as

being due to molecules constrained to rotate within a

given cone angle or in the plane of the surface.

The adsorption of water, ethyl chloride, ethyl chlor-

ide-water, and butane-water on silica gel was studied by

Snelgrove (9). Several break points were noted on the

capacitance curves; results with water and ethyl chloride

were similar to earlier work (2, 5, 8). With the ethyl

chloride-water mixture it was concluded that ethyl chlor-

ide and water were adsorbed on different types of sites.

No satisfactory explanation could be found for the re-

sults of experiments using butane-water mixtures.












The investigation of the dielectric properties of

porous Vycor glass was undertaken by Petrie (7). Using

an ethyl chloride-water mixture and butane as adsorbates,

he observed two linear sections; the second linear re-

gion had a lower slope than the first. It was proposed

that the lower slope in the second region might be due

to the existence of pores within which the electric field

was different from that at the surface.

Heukelom (15) has studied the adsorption of water,

methyl alcohol, methyl-ethylketone, chlorobenzene, diethyl

ether and benzene on silica gel in the frequency range

20 to 20,000 cycles. High values of the real part of the

dielectric constant (W') were observed at low frequency

(see Appendix B for a discussion of dielectric theory).

The value of E' did not depend on the quantity adsorbed

or on the nature of the adsorbed species. It was con-

cluded that ( in the low frequency region was only de-

pendent on the shape of the pores and on the general

structure and arrangement of particles of silica gel.

A maximum in E" dielectricc constant due to loss) was

observed. The Cole-Cole plot of E' versus (" was in-

dependent of the adsorbate but characteristic of the

adsorbent. The conductivity of the sample increased

with the number of monolayers of adsorbed liquid. This

phenomenon was attributed to the pore structure which was












believed to consist of pores passing entirely through

each grain of sample.

In order to aid other investigators who may have

difficulty obtaining translations of articles from the

Japanese language, the work of Kirosaki and Shimizu

will be discussed in some detail. Kirosaki (16) studied

the system silica gel-water in an attempt to find where

and how capillary condensation occurred. The adsorption

isotherms of water on silica gel were found to be S-shaped.

From the adsorption isotherm the quantity of water neces-

sary to complete a monolayer (Vmo) was determined by the

B.E.T. method. Vmo was found to be 38.6 mg. H20/g. Si02.

Kirosaki assumed that the adsorbed phase had the density

of ice and calculated the surface area occupied by one

water molecule from the equation proposed by Emmett and

Brunauer (18). He obtained a value of 11.1 A2 which

corresponds to a surface area of 143 M2/g. if water mole-

cules are in contact with each other in the monolayer.

Measurements of the heat of adsorption (LHA) of water on

silica gel showed that in the early stages of adsorption

it was approximately 10 kcal./mole greater than the heat

of condensation (6HC) of water, which is 9.7 kcal./mole.

This result was of course given as evidence for tight

bonding to the surface in the early stages of adsorption.












At a water content of 100 mg. H20/g. Si02, LHA was ap-

proximately equal to AHC; it was proposed that water

adsorbed in this region exists in a condition similar

to liquid water.

A plot of the apparent dielectric constant (('A) of

the adsorbate versus mg. H20/g. Si02 yielded three linear

regions of different slope. The first region extended

from zero to 15 mg. H20/g, Si02, the second region from

15 to 94 mg. H20/g. Si02, and the third region above

94 mg. H20/g. Si02, A plot of ('A and ("A versus fre-

quency in the frequency range 2 to 100 kc. showed a

maximum in ("A (E"A max.) at 10 kc. + 2. As the amount

of water increased, E" A max. shifted to higher frequency.

In the first adsorption region (0 15 mg. H20/g. Si02)

("A was essentially zero. This, considered along with

the heat of adsorption, suggested restricted motion of

OH groups in this region. E"A increased at low fre-

quency in the second region (15 94 mg. H20/g. SiO2) and

it was suggested that the water was in a situation simi-

lar to water dissolved in nonpolar solvents. In the

third region E"A max. appeared and shifted to higher

frequency as the amount of water increased. It was

concluded from the finite relaxation time in this region

that the OH groups turn with the field in bunches. The

situation is believed to be a state similar to ice, with












strongly developed hydrogen bonds. The explanation of

the shift in ("A max. with increasing water content was

that a state similar to liquid water is approached

where there is more freedom of motion than in ice. An

alternate explanation was offered for the appearance

of ("A max. in the third region. It was suggested that

capillary condensation begins there and that, as water

condenses into the pores and thus changes state from

vapor to liquid, the peak in ("A appears. Upon examin-

ing the gel with an electron microscope it was concluded

that the capillaries were cylindrical in shape. By pre-

paring samples rich in NaCl impurities it was determined

that the shift in ("A max. was not due to the solution

of impurity ions into the capillary condensed water.

The investigation of ethyl alcohol adsorbed on silicon

gel was undertaken by Shimizu (17). The heat of ad-

sorption was determined from the Clapeyron plot of the

logarithm of equilibrium pressure against the inverse

temperature. Two linear regions were obtained for low

alcohol coverage, the break point occurring at 1200C.

Below 120C, and for low coverage, the heat of adsorp-

tion was about 12 kcal./mole; above 1200C it was approxi-

mately 15 kcal./mole. For high alcohol coverage the heat

of adsorption was constant over the entire temperature

range and had a value of approximately 10.5 kcal./mole.











Plots illustrating the effect of temperature on capa-

citance showed a gentle curve toward lower values of capa-

citance with increasing temperature. At high water con-

tent the change in capacitance was most pronounced. It

was concluded that, since the capacitance shifts to lower

values with increasing temperature, the OH groups couple

in the direction of minimum polarization. Desorption

occurred readily over the entire range studied, and it

was concluded that no irreversible chemical change oc-

curred during adsorption.

Calculation of the B.E.T. surface area, from the

ethyl alcohol adsorption isotherms, gave values which

range from 170 mg. EtOH/g. Si02 at 200C, to 124 mg.

EtOH/g. Si02 at 1500C. At approximately 180 mg. EtOH/g.

SiO2 the heat of adsorption approached the heat of con-

densation of ethyl alcohol; this was cited as additional

evidence for the completion of a monolayer at approxi-

mately 180 mg. EtOH/g. Si02.

The polarizability of the adsorbate was calculated

according to the Kirkwood-Frohlich equation (19, 20) for

polar mixtures. Plots of the specific polarization of

the adsorbate against the weight fraction adsorbed gave

several linear sections at low temperature. The last

break point occurred at a water content 15% beyond the

completion of a monolayer. The linear sections before











the monolayer point did not vary with temperature and

were considered regions where no condensed liquid exists.

No explanation was offered as to the nature of adsorp-

tion in these regions.

Ebert (13) studied the adsorption of water on alumina.

He found three linear regions on the curves obtained

by plotting tan 5 versus mg. H20/g. A1203 (hereafter re-

ferred to as tan 9 curves). The temperature coefficient

of the first region is small compared with that for

other regions. Along with the other authors whose work

has been discussed above, Ebert attributed this phenomenon

to rigid bonding in the first layer and less rigid bond-

ing in successive regions. It was assumed that the first

linear region corresponds to rigid bonding before the

completion of a monolayer, and that the second linear

region is due to mobility within the monolayer. The as-

signment of the second break point to the monolayer is

consistent with a surface area per water molecule of

10.6 A2. He suggests that the heat of adsorption in

the first region is at least 14.6 kcal./mole.

Thorp (21) investigated the dielectric properties

of alumina with methyl and ethyl alcohols adsorbed on

the surface. The first break point in his curves of

capacitance versus mg. alc./g. A1203 agrees fairly well

with the value of the B.E.T. monolayer as calculated












from point B of the adsorption isotherms. Capillary

condensation was observed to occur, as indicated by a

hysteresis loop in the adsorption isotherm and in the

capacitance curve after the completion of a monolayer.

The desorption capacitance points lie at higher capa-

citances than the adsorption points. Apparently during

desorption the condensed liquid in the pores contributes

more to the capacitance than alcohol adsorbed in multi-

layers. Thus, during pore filling only capacitance

due to multilayers was believed to exist, but upon

completion of filling of the pores a contribution of

high capacitance resulted from liquid in the pores and

persisted during desorption until the pores were empty.

In his dissertation Baldwin (22) hns reported some

dielectric studies of water vapor adsorbed on alumina.

He obtained curves of the same shape as did other in-

vestigators when he plotted ("A versus mg. H20/g. Al203.

The first break point in his curves is quite constant

with change in frequency, and he attributes it to the

completion of a monolayer. However, this break point

occurs at 40 mg. H20/g. A1203 while the surface area

from the B.E.T. nitrogen isotherm is 270 M2/g. A1203.

If the first break point were the completion of a mono-

layer, it would require that the effective area occupied

per water molecule be 20.2 A2, which seems a bit high.

Point B of his water adsorption isotherm suggests that











the true value of the monolayer occurs nearer 55 mg.

H20/g. Al203, which would give a surface area per water

molecule of 16.2 A2, a more reasonable value in view of

the activation procedure employed.

Below the first break point Baldwin found values of

the relaxation time that approach infinity, and he there-

fore concluded that molecules in this region do not orient

in the applied field. No temperature dependence was found

below the first break point, but one did exist after the

break point. The heat of adsorption at high water content

is 12.7 kcal./mole.

B. Review of Literature on Nuclear Magnetic Resonance

Line Width.

Due to the relative infancy of n.m.r. little work

has been done to investigate adsorption phenomena by

n.m.r. methods. In 1950 Shaw (23) measured the peak in-

tensities of samples of apple, potato, and wood of various

water contents. A linear relationship between peak in-

tensity and water content was obtained. The line width

could not be measured because only crude apparatus was

available (see Appendix B for a discussion of n.m.r.

theory). Later Shaw (24) was able to determine the line

width of water adsorbed on starch and egg albumin. A

narrow line superimposed on a broad line was observed.

These two lines were attributed to "sorbed water (narrow











line) and water in the solid (broad line)." Tanaka (25)

also observed this sharp line superimposed on a broad

line with water adsorbed on carbon. He attributed the

narrow line to water on the monolayer and the broad line

to water in the monolayer.

The spin lattice relaxation time T1 was measured,

using spin echo technique, for water, methyl and ethyl

alcohols, and hexane on gamma alumina by Hickmott (26).

A linear relationship between T1 and the amount of liquid

adsorbed was observed. Tj moved to higher values as more

liquid was adsorbed. It was proposed that the first

liquid adsorbed was tightly bound, with proton exchange

occurring between lattice vacancies and residual water.

This exchange modifies the relative orientation of nu-

clear dipoles, with the result that an increasing number

of spin transitions occur and consequently a short re-

laxation time is observed. With increasing amounts of

liquid, T1 approaches the longer time characteristic of

bulk liquid.

Zimmerman (27, 28, 29) obtained the relaxation

times T1 and T2 of water adsorbed on silica gel using

the spin echo method. He found three regions on the

plot of T2 versus water content. At water contents

greater than two layers a distinct resolution of relaxa-

tion times due to water molecules was obtained. This












was attributed to water in and on the monolayer. Between

one half a monolayer and two layers, single-phase be-

havior existed. At water coverages of less than one half

a monolayer T2 was large; this fact was considered to be

due to a decrease in effective inter-molecular interac-

tions caused by the low coverage of water molecules.

A comparison was made between the apparent dielec-

tric constant as observed by Kirosaki (16) and the T2

values obtained on the silica gel-water system. The

changes in slope of the curves obtained in the dielectric

experiment were at approximately the same water coverage

as the observed changes in T2. It is to be noted in this

connection that Kirosaki assumed that adsorbed water had

the density of ice and therefore occupied an area of

11.1 A2, while Zimmerman, for a supposedly similar sample,

assumed an area of 10.6 A2 for a water molecule.

In later experiments with water on silica gel Zim-

merman (29) found two values for T2. One value, which

was low and constant over the entire range of water con-

tent, was attributed to water adsorbed on high energy

sites. The other was large and constant until the com-

pletion of two layers, where it appeared to break to higher

values, probably because of capillary condensation. Two-

phase behavior was also observed for Tl. It was concluded

that the phases responsible for T1 and T2 were identical.












Odajima (30) compared the dielectric properties of

water on mulberry paper with the n.m.r. line width.

Three sections had previously been reported (31) on

the curve of dielectric constant versus water content.

The first section due to localized water could not be

detected with the n.m.r. The second break point on the

dielectric curves was found to occur at approximately

the same water content as the break point in the n.m.r.

experiment.

The adsorption of water on alumina was investigated

by Winkler (32). T1 was found to increase linearly with

increasing water coverage. T2, however, showed a sharp

break at high water content. This change of slope in

T2 was attributed to the completion of filling of micro-

pores and the start of filling of macropores.

O'Reilly (33) has attempted a measurement of the

chemical shift of residual protons in a silica-alumin,

catalyst, and in silica gel. He has concluded on the

basis of a shift from water of -0.3 + 0.2 parts per

million that protons are bound as OH groups on both

catalysts.

Aston (34) has measured the line width of water and

methane on rutile. At a water content of 0.1 monolayers,

the line width was 1.1 gauss as compared to 16 gauss for

ice and 10-4 gauss for liquid water. It was concluded












that, due to the sparse surface coverage, there was little

magnetic interaction between water molecules in this re-

gion. With increasing water content the line width de-

creased. Using methane, the line was much narrower and

it was concluded that reorientation with respect to trans-

lation and rotation occurs readily.

Line width measurements were also made by Mays (35)

on titanium dioxide with water adsorbed. At low water

content a narrow line was observed which persisted even

down to 770K. At water contents approaching the mono-

layer value a wider line was observed. It was concluded

that the narrow line was due to the sparsely covered

surface where magnetic interactions are small between

neighboring protons.

C. Introduction to the Present Work.

All the authors previously referred to are agreed

that during the initial stages of adsorption, molecules

are bound rather tightly to the surface when compared

with molecules adsorbed after the completion of about one

monolayer. In some cases, however, an uncertainty exists

as to where a monolayer is completed because of a lack

of knowledge of the surface area occupied by an adsorbed

molecule. In the present paper this difficulty has been

avoided by determining the amount of water necessary to

fill one layer, according to the theory by Brunauer,












Emmett, and Teller, by determining the adsorption isotherm

of water on alumina and applying the B.E.T. equation to

it. No previous investigation has been made into the ef-

fects of sample activation upon the dielectric and n.m.r.

properties of the adsorbate. In this paper an attempt

has been made to correlate the properties of three dif-

ferent alumina samples with the dielectric and n.m.r.

properties of the adsorbate. While there has been a great

deal of work done on the temperature dependence of the

apparent dielectric constant of adsorbate-adsorbent sys-

tems, very little has been done to investigate the tem-

perature dependence of n.m.r. line width in these systems.

Some interesting phenomena are reported here in connec-

tion with the temperature dependence of n.m.r. line width.

In addition an interesting adsorption isotherm will be

discussed in relation to the dielectric properties of

the system. Heats of adsorption have been measured

for various water contents in order to give a quantita-

tive estimate of the strength of bonding to the surface.

Finally, certain correlations are pointed out between

the dielectric behavior and the n.m.r. line width of the

samples.

In addition to the intrinsic interest of these

studies they have been pursued with the purpose of ob-

taining information about adsorption in connection with









17


catalysis studies on alumina being carried out in these

laboratories. It is hoped that these investigations will

prove valuable in connection with studies into the nature

of the freezing process in biological materials (36), in

which there is interest at the present time.















CHAPTER II

EXPERIMENTAL PROCEDURE


A. Vacuum Apparatus Used in Dielectric Experiments.

The apparatus is shown in Figure 1. Pressures of

less than 10-5 mm. of mercury as read from the McLeod

gauge were attained by using a mercury diffusion pump

in series with a Cenco-Hyvac vacuum pump. The McLeod

gauge and mercury diffusion pump were purchased from

the Scientific Glass Apparatus Company.

The procedure for adsorbing the water is as follows:

Section 7 was filled with distilled water; the water was

frozen to ice with a Dry Ice-acetone bath. The Dry Ice-

acetone bath was removed and replaced by a water bath;

the ice was then pumped on as it melted to remove dis-

solved gases. This procedure was repeated until no gas

bubbles were visible in the water. Stopcock S2 was

closed and the entire system was brought to a pressure

of less than 10-5 mm. of mercury to insure that no appre-

ciable amounts of oxygen or nitrogen remained and that

there were no leaks in the system. Stopcock S2 was then

opened and water vapor admitted into section 6, 3, and

4. The pressure was recorded and stopcock S, opened and

the water vapor allowed to adsorb on the sample in the


























-TO
VACUUM
PUMP


II I I'
II I I
KLIJ~ j II
II I'
A B C D E F G


Fig. 1. Vacuum Apparatus Used in Dielectric Experiments












cell A. The final pressure was recorded and the weight

of water adsorbed was calculated from a knowledge of the

volume of the system. The calculation of the amount of

water adsorbed is described in Appendix A.

B. Vacuum System Used in Nuclear Magnetic Resonance

Experiments.

The apparatus is shown in Figure 2. It was evacuated

by means of a Cenco-Hyvac vacuum pump and the pressure

measured by means of the manometer D. The system was

calibrated in terms of weight of water at a particular

pressure. This calibration is described in Appendix A.

G is a drying tube filled with Drierite through which

nitrogen could be admitted to the system. F is a safety

valve to insure that pressures greater than one atmos-

phere would not be built up during nitrogen addition. E

is an eudiometer, D the manometer, C the water sample.

B is an auxiliary tube for admitting vapor, and A the

sample tube which fits between the magnet poles of the

n.m.r. instrument. The diameter of the sample tube is

12 millimeters.

C. Alumina Sample

The alumina was Alcoa Activated Alumina, grade F-10;

obtained in granule form of mesh 8-14. In the dielectric

studies it was used without further reduction of particle

size. For the n.m.r. studies it was ground and the portion
































MAGNET G


TO
SVACUUI
A B C D E F PUMP


Fig. 2. Vacuum Apparatus Used in N.M.R. Experiments












which passed through a 16 mesh screen was used. Three

portions of the alumina were activated under different

conditions. The first portion was activated at 400C

for 4 hours, the second portion at 500C for 6 hours nnd

the third at 800C for 6 hours. The temperature was con-

trolled by a chromel-alumel thermocouple and Wheelco pro-

portioning controller. The properties of the samples

are listed in Table 1. The residual water was determined

by heating to constant weight at 20000F.

D. Capacitance Measuring Apparatus.

The dielectric cell was constructed of copper cy-

linders which had tungsten leads and which were enclosed

in a glass cylinder. It is shown in Figure 3. The outer

and inner copper cylinders are connected and are at elec-

trical ground potential. The middle cylinder is the high

side of the electrical circuit. The weighed sample was

placed between the high potential cylinder and the two

grounded cylinders. This was done in a dry box filled

with nitrogen. The cell was then attached to the system

while nitrogen flowed through the system. This was to

insure that the sample adsorbed no water from the air in

the laboratory.

After each increment of water was adsorbed by the

sample, measurements of capacitance and conductance were

made with a General Radio Twin T Impedance-Measuring

Circuit Type No. 821-A. The Twin T Impedance-Measuring







































BANANA
PLUG


TOP VIEW


Fig. 3. Dielectric Cell and Leads












Circuit is a null instrument with a frequency range of

460 kc. to 30 mc. It is used basically with a parallel-

substitution method for measuring unknown impedances in

terms of their parallel admittance components, namely

susceptance, B, and conductance, C. The susceptance

is related to the capacitance, which is read directly

from a dial calibrated in micromicrofarads (Auf) by

the equation:

B W AC

The conductance is read directly from a dial cali-

brated in micromhos (pmho). From these quantities it

is possible to calculate tan S. Tan is defined as

the ratio of the dielectric constant due to loss divided

by the real dielectric constant. That is, tan S equals

7, and for this bridge is given by G divided by B.

G is related to the conductance dial reading by the equa-

tion:

G =-2o

where C is the conductance dial reading C converted to

the units of mho, f is the operating frequency in cycles

per second and fo is 106 times the position of the fre-

quency range switch, on the bridge. B has been defined

before; it is equal to 2nf&C, where f has the same mean-

ing as above and OC is the difference, expressed in











farads, before and after the dielectric cell was attached

to the bridge.

The oscillator used was a General Radio Bridge Os-

cillator Type No. 1330-A and the detector was a Halli-

crafter's Radio Receiver Model SK-62A. These combina-

tions permitted the frequency to be varied from 460 kc.

to 20 mc. Measurements above 20 mc. could not be made

because of the lack of stability in the detector at

these frequencies.

It is to be noted that the capacitance and conduc-

tance values obtained are the combined values of the

cell and its leads, the alumina and adsorbed water, and

the water vapor in the dead space of the cell. Since

these contributions to the capacitance and conductance

are either constant or very small they will not affect

the interpretation of the data as regards the physical

situation at the alumina surface.

A word is in order about the leads from the dielec-

tric cell to the bridge. Figure 3 shows the banana

plugs used to make connection between the cell and the

bridge. It was found that this method of connection

insured complete reproducibility of the readings. Other

methods such as making electrical contact through a wire

dipped into a mercury pool were found to give different

results depending on the time the wire and mercury were











exposed to the air or on the depth to which the wire was

dipped into the mercury pool.

Most of the measurements were made at a frequency of

750 kc. This frequency was chosen because of the lack

of noise in the Hallicrafter's receiver at this point,

which thus facilitated measurement.

E. Procedures for the Nuclear Magnetic Resonance Studies.

The alumina samples previously described and with

their dielectric properties known as a function of water

adsorbed were used. The nuclear magnetic resonance spec-

trometer was purchased from Varian Associates and was

Model V-4300-2 operating at a frequency of 60 mc.

The sample, resting in the probe holder between the

magnet poles as in Figure 2, was allowed to adsorb known

increments of water, as previously described, and the

line width was measured from the derivative of the ad-

sorption signal obtained by the wide line technique.

From 6 to 10 measurements of line width were made after

each addition of water and the results were averaged.

Sideband calibration was done on a water-saturated alu-

mina sample to obtain accurate line widths. It was not

possible, because of the weakness of the n.m.r. signal,

to obtain line widths when the water content was less

than about 25 milligrams per gram of alumina.












All line width measurements were at a response set-

ting of 3. A setting of 3 selects a time constant of

3 seconds in an R-C integrator which controls the band

width of the signal from the synchroverter-phase de-

tector. It was found that a response setting of 3 in-

tegrated out much of the noise without distorting the

signal. Response settings above 3 distorted the signal

too much for measurement purposes.

Since weak nuclear signals were recorded on a graphic

recorder it was necessary to limit the sweep field to

small values (low-3) in order to minimize artificial

sweep broadening effects. These effects arise when the

slope of the resonance curve is not constant over the

area swept by the modulation field. The result is that

a signal appears before the true resonance absorption

is reached and thus the signal is too broad. It is

necessary, therefore, to use a small sweep field and to

keep it constant for the duration of the experiments.

In order to study the variation of line width with

temperature a Varian Variable Temperature Probe Model

V4340 was used. Through this was passed a stream of

dry nitrogen previously cooled by passing it through

a copper coil immersed in a Dry Ice-acetone bath. The

temperature was varied by regulation of the flow rate

by means of a reduction valve in series with the copper












coil. An estimate of the flow rate was obtained by the

use of a Predictability Type Flow Meter manufactured by

the Matheson Company.

F. Measurement of Heat of Adsorption.

A sample of alumina of known water content was at-

tached to the vacuum system previously used in the di-

electric experiments and shown in Figure 1. The sample

tube was surrounded by a water bath for measurements at

low temperatures and by a furnace for measurements at

high temperatures.

The temperature of the water bath was controlled by

means of a Fisher Transistor Relay in conjunction with

a mercury-tungsten temperature controller. The tempera-

ture of the furnace was controlled by a chromel-alumel

thermocouple and Wheelco proportioning controller.

Pressure readings were taken at a series of tem-

peratures and the log pressure versus the inverse tem-

perature plotted according to the Clapeyron equation.

From the slope of the straight line obtained, the heat

of adsorption was calculated.















CHAPTER III

EXPERIMENTAL RESULTS


A. Surface Properties of the Alumina Samples.

As was pointed out earlier, samples of alumina were

prepared by utilizing different conditions of activation.

One sample was activated at a temperature of 400C for

4 hours under vacuum. This sample will always be referred

to as the 4000C sample. Two other samples were prepared

by activation for 6 hours at 5000C and 8000C respectively

and will be referred to as the 5000C and 800oC samples.

The surface areas of these samples, as determined from

the nitrogen adsorption isotherms by the use of the B.E.T.

equations, and values for the residual water in the solid

may be found in Table 1 for samples used in the dielec-

tric experiments. The properties of the samples used

in n.m.r. experiments may be found in Table 2.

The symbol Vmo in all the tables and figures is the

amount of water expressed in mg. H20/g. A1203 necessary

to complete the monolayer. For the dielectric experi-

ments this value was calculated from the water adsorp-

tion isotherms by application of the B.E.T. equation.

The nitrogen adsorption isotherm was determined for

each sample investigated. From the value of Vmo and

















TABLE 1



Description of Samples

Used in Dielectric Experiments


Temperature of
Activation
(oC)



400


500


Time of
Activation
(hr.)



4


6


Residual
H20
(mg.


18.5


10.7


B.E.T.
Surface Area
Meterss)
g.A -A

120


129.6


3 105.5


800
















TABLE 2

Surface Areas of Samples Used in N.M.R. Experiments


Activation
Temperature
(oC)


800

400

500


Surface
Area
(B.E.T. N
Isotherm)


95.96

122.0

128.6


Molecular Area
of H20 (A2)





13.2

14.2

14.9


Vmo

mg. H20)




21.7

25.7

25.8












the surface area of the sample as determined from the

nitrogen adsorption isotherm the surface area occupied

per water molecule was calculated for each sample used.

The value obtained is different for each sample used

(13.2, 14.2, 14.9 A2). These values differ rather

greatly from the value of 11.1 A2 assumed by Kirosaki

(16) for adsorption on silica gel, and the value of 10.6

A2 assumed by Zimmerman (28) for water on silica gel.

The surface areas obtained on each sample are listed

in Table 3.

Also listed in Table 3 are "Vm" values calculated

from the nitrogen surface areas assuming each of two

different values found in the literature (18, 19) for

the area occupied by one water molecule. It will be

noted that these values for the amount of water neces-

sary to complete one monolayer differ rather greatly.

It is believed that Table 3 indicates the imprudence

of assuming a knowledge of the area occupied by a water

molecule adsorbed on a particular surface. It is sug-

gested that the area covered by a water molecule depends

upon the nature of the surface on which it is adsorbed

and therefore must be determined from the water adsorp-

tion isotherm on that surface.

For the samples used in the n.m.r. experiments no

water isotherms were determined and therefore Vmowas












TABLE 3


Surface Areas of Samples Used in Dielectric Experiments


Activation
Temperature
(oC)


800

400

500


Surface
Area (-.)
(B.EoT. N2
Isotherm)


105.5

120.2

129.6


v /mg. H20)
m g.Al203
H20 14.8 A2


21.3

24.3

26.2


Vm Mg. H20)
Vm g.Al203;
H20 = 10.5 A2


30.5

34.3

37.0


/mg. H20
Vmo g.Al203-

(B.E.T. H20
Isotherms)

24.0

25.4

26.2


Molecular
Area of
H20(A2)


13.2

14.2

14.9












calculated from the nitrogen surface area assuming that

the surface area occupied by a water molecule was the

same as on the similarly prepared samples used in the

dielectric experiments. This assumption should intro-

duce no significant error. These Vmo values may be

found in Table 2.

The isotherm for adsorption of water on the 8000C

sample, which is shown in Figure 4, is rather unusual.

Two regions of constant vapor pressure exist on the iso-

therm; the first begins at a water content of 26 mg.

H20/g. A1203 and the second at 32 mg. H20/g. A1203.

These two sharp breaks are interpreted to mean that

molecules here begin to be adsorbed in states with

lower free energies of adsorption.

An adsorption isotherm has been run on a sample ac-

tivated at 5000C under very careful conditions. Times

from a few hours to two days were allowed for equilibrium

to be attained between successive points on the isotherm.

The same general shape was observed as in Figure 4, al-

though the break points were not nearly so sharp. This

seems to indicate that, with time, molecules take posi-

tions of lower energy and that the energy of adsorption

does not change so drastically.

At this point it is appropriate to point out that

curves of the same general shape as that in Figure 4









-MG H 20
G Al203


36h


32h


28k


24 VMo


20-


L 1 1 II, P/P0


0.1


0.2


0.3


0.4


Fig. 4. Water Adsorption Isotherm


80 0 C


40












were obtained on all the samples used in the dielectric

experiments. However, not as much care was taken as re-

gards the time interval between readings of "equilibrium"

pressure and consequently they do not form quite as con-

vincing a plot. They were sufficiently accurate in the

low pressure region, however, to permit an excellent plot

according to the B.E.T. equation.

Figure 5 shows the amount of water remaining on the

surface of the three samples, after they had adsorbed

large quantities of water, studied as a function of time

of constant pumping with a Cenco-Hyvac vacuum pump.

Initially the water came off the surface very easily,

indicating that rather loose bonds held it to the sur-

face. Additional pumping for as long as 16 hours did

not remove water in the monolayer region. This illus-

trates the rigid nature of the bond at water contents

lower than one layer.

In order to obtain a quantitative measure of the

strength of bonding of water molecules to the surface,

heats of adsorption were measured at various water con-

tents. The Clapeyron plots are shown in Figure 6. The

heat of adsorption, calculated from the slopes of these

plots for each water content, is listed in Table 4. It

will be noted that the heat of adsorption is approxi-

mately 17 kcal./mole before the monolayer is complete















TABLE 4


Tabulated Heats of Adsorption

Sample Activated 5000C in Vacuum


H20 /mg. H20
Content \g. A1203/


16.39

29.00

34.34

48.35

71.37


/kcal.
H mole




17.30

17.30

11.74

11.7

10.7









o= 8 0 0 C
, = 4 0C

S= 5 0 0C


50




45




40




35




30


MI N.)


Fig. 5. Desorption Isotherms


uG mH20
G A12 03


o 16HR.




* IOHR.
S5HR.


10 20 30 40 50









50 0C


1 39






I A
2.2 2.4


29.00


2.6


2.8


3.0


.2
.2


__ /T(K)X 10
34 3.6


Fig. 6. Clapeyron Plot for Determining the Heat of Adsorption


L
0
G
P


~\


1.0


0.6


0.4 -


71.73


34.34


0.2


0.0


3.


I I


I


!


!











and decreases with addition of water until it is very

nearly equal to the heat of condensation of water.

B. Discussion of Dielectric Experimental Results.

Several terms and conventions must be defined be-

fore proceeding with a discussion of the data. Plots of

tan E versus mg. 20/g. A1203 will be referred to as

tan S curves. The words mobility and motion will be

used to denote rotation, vibration, or reorientation of

a molecule. When movement of a molecule over the sur-

face occurs it will be termed migration. It will be as-

sumed that similar slopes on the tan S curves correspond

to similar situations on the surface of the alumina.

The plots of tan E versus mg. H20/g. A1203 at 750 kc.

are shown in Figures 7, 8, and 9. V'm, the monolayer

value obtained by assuming an area per water molecule

of 14.8 A2, is indicated along with Vmo for purposes of

comparison.

For each sample the tan S curves in Figures 7, 8,

and 9 show four rather distinct linear regions followed

by a region where tan S begins to level off. It would

be expected, from a consideration of the results ob-

tained by Baldwin (22), that tan S will decrease at

higher water contents, thus exhibiting a maximum, and

then begin to rise again.





T
A o
14-N 4000 C
6 75 0 KC
x
103
12-



I0-



8-



6-



4-



2-
0
SI I MG HO/G AlIO
4 8 12 16 20 24 28 32 36 40

Fig. 7. Dielectric Properties of the Alumina-Water System for the Sample
Activated at 4000C








12-T
A
N


5000C
750KC


x
03


MG H


Fig. 8. Dielectric Properties
Activated at 5000C


20 24 28 32 36 40


of the Alumina-Water System for the Sample






14-
T 800 C
A
N 750 KC
12 -6
x
103

I0-



8



6 -



4-



2 -


I I I I MG H2OG A O
4 8 12 16 20 24 28 32 36 40


Fig. 9. Dielectric Properties of the Alumina-Water System for the Sample
Activated at 8000C ^












The first adsorption region has a value for tan 5

very close to that of the solid and presumably the ad-

sorbed water is in a state of similar mobility to that

of residual water in the solid. The point Vmo (cal-

culated from the adsorption isotherm by the use of the

B.E.T. equation) in all cases falls at about the middle

of the second linear region. Addition of the weight of

residual water in the solid to Vmo produces no correla-

tion with respect to break points on the tan 6 curve.

Thus it appears that, although water in the first region

is in a state of mobility similar to that of residual

solid water, it differs from the water in the solid in

that the latter is apparently part of the surface while

water in the first region is adsorbed on the surface.

In this connection it should be recalled that McIntosh

(2) could find no correlation between break points and

the monolayer value for adsorption of polar molecules

on porous silica gel. The possibility does exist, how-

ever, that residual water is not part of the surface and

therefore would have to be added to the monolayer value

as determined from adsorption isotherms.

Because of the good agreement between point B of

the adsorption isotherm of Figure 4 and the end of the

first linear region of Figure 9, the first break point

on the tan S curve is chosen as marking the completion











of the monolayer. This choice, rather than the second

break point at 26 mg. H20/g. A1203 on Figure 9, seems

reasonable in view of the fact that there is also a

break on the adsorption isotherm at 26 mg. H20/g. A1203

which would logically be associated with the second

break point of Figure 9.

Upon choosing the first break point as marking the

completion of a monolayer three reasonable possibilities

exist for the explanation of linear regions and break

points on the tan 9 curve.

In the first explanation the second linear region

is imagined as the beginning of the second layer of water

molecules. At the second break point capillary filling

is imagined to begin in small pores. This is consistent

with the shape of the adsorption isotherm of Figure 4,

where a sharp break is observed at 26 mg. H20/g. A1203.

The vapor pressure remains constant along the third

linear region, as would be true if a small pore were

filling. At the third break point on the tan S curve

the pore is imagined filled; the radius of the meniscus

would then change from a negative to a positive value and

the vapor pressure would increase sharply, as is observed.

In this model the second region of constant vapor pres-

sure on the adsorption isotherm indicates that larger

pores begin to be filled. Along the fourth linear region












of Figure 9 pores would again be filling. The water con-

densed in these pores would be more mobile than the water

in smaller pores as indicated by the steeper slope of the

tan S curve. The sharp increase of vapor pressure in

Figure 4 at 37 mg. H20/g. A1203 would be evidence for the

completion of pore filling.

This explanation has the difficulty of supposing that

pore filling begins without appreciable addition of water

to other regions of the monolayer. In addition there ex-

ists the fact that after the completion of the monolayer

the capacitance shifts to lower values with time, and from

this it would be expected, that during desorption capaci-

tance curves could be obtained which fall below those of

Figures 10, 11, and 12. In Figures 9 and 12 the discon-

tinuity in the curves is due to a time lapse of approxi-

mately 15 hours between the measurement of the two points

at 35 mg. H20/g. A1203. If pore filling were the mechanism

the capacitance would be expected to be at higher values

during desorption than adsorption because of the emptying

of filled high-capacitance pores. Thorp (21) has ob-

served this high capacitance hysteresis. Another dif-

ficulty is to be found in the very low vapor pressure

which this picture demands of liquid in the pores; this

of course suggests that rather narrow pores exist.







40 00 C

750 KC


MVMo


MG H 2 A1203


N~ V I I I


Fig. 10.


Capacitance of the Alumina-Water System for the Sample Activated
at 4000C


AC
-~


10



8



6



4



2



0


20


36


40


0

^ 3o"-- 0







50 O0C

7 5 0 KC


1 AC
10 .


8 12 16 20


M 0

28


MG H 20/G A1203
36 40
36 40


Fig. 11.


Capacitance of the Alumina-Water System for the Sample Activated
at 5000C


4


------ I -


I









10-A 8 0 0 C

,7 750 K C
8 -f



6



4 0



20 0








V IMO MG HO GAI O
4 8 12 16 20 24 28 32 36 40


Fig. 12. Capacitance of the Alumina-Water System for the Sample Activated
at 8000C












A second explanation, which has a serious difficulty

also, involves the formation of lenses of water on the

monolayer. In order for this explanation to be consis-

tent with Figure 7 it would require that during the forma-

tion of lenses the vapor pressure of the lens remain con-

stant. It is difficult to imagine how this situation

could arise since presumably addition of water to a lens

will either increase the radius of the lens, in which

case a decrease in vapor pressure would be observed, or

it would decrease the radius of the lens, resulting in

an increase in the vapor pressure.

The third explanation is considered to be the most

reasonable. It is thought that molecules causing separ-

ate linear regions, after the completion of the mono-

layer, are adsorbed on the monolayer with different mo-

bilities and different energies of adsorption. It is

supposed that the first molecules adsorbed after the

monolayer are lying flat on the monolayer and can only

move or rotate in the plane of the monolayer.

Because of the increase in slope of tan S after the

second break point, this break point must be interpreted

as the beginning of adsorption of molecules which are

attached more weakly than those in the second linear re-

gion. That these molecules exist in states of lower

energy is suggested by the coincidence of the second












break point of Figure 9 with the break at 26 mg. H20/g.

A1203 in Figure 4. These molecules may be thought of as

being attached to the monolayer by two bonds rather than

three as those in the second linear region of Figure 9

were.

Again at 32 mg. H20/g. A1203 a sharp break is ob-

served in Figures 4 and 9. Water molecules can be thought

of as being attached by only one bond or by rather weak

hydrogen bonds and as having considerable freedom to

orient in the applied field. This region of constant

vapor pressure in Figure 4 indicates another change in

the energy of adsorption.

The leveling observed in the tan 9 curves at high

water contents is due to the rather high capacitance

in the second layer as the second layer nears completion.

The conductance in this region and at this frequency of

750 kc. approaches a constant value while the capacitance

continues to increase sharply. The water is considered

to be somewhat more mobile than water in ice but not

nearly as mobile as liquid water. The reasons for this

belief will be given with the nm.r. results.

This model is consistent with Figires 4 and 9 and

will explain the observed decrease of capacitance with

time. As pointed out in section A the breaks in the

water adsorption isotherm were not nearly as sharp if












long periods of time were allowed between successive

additions of water. This suggests that with time mole-

cules migrate to states of lower energy and form tighter

bonds to the monolayer such that their contribution to

the total capacitance diminishes.

To explain the different kinds of adsorption on the

monolayer it is necessary to assume that either some

surface forces extend through the monolayer or that the

monolayer itself has active sites.

A model based on the existence of active sites on

the monolayer would consider situations existing in the

monolayer where all the potential bonds that a water

molecule might use in attaching to the alumina, as being

completely saturated in some cases and not saturated in

others. It is not possible to decide between the two

possibilities and both may contribute to the true explana-

tion.

Preliminary experiments were begun on the frequency

dependence of tan S Figure 13 shows the dependence of

tan S on frequency and on water content at three frequen-

cies. The curve at 16 mc. is higher than those at the

other frequencies, This is shown more vividly in Figure

14. As the water content increases in Figure 13 the

maximum is no longer observed in the frequency range

studied. This is in accord with the previous assignment








7- T 4 00C
A
N
6-5
x
103

5



4



3 i



2




8
II 0 0M MG H 2 9 /G At 0
4 8 12 16 20 24 28 32 36 40
Fig. 13. Dielectric Properties of the Alumina-Water System at 8, 16, 20 mc.









2.0- T4 00 C
N
6
.8-x
103

1.6 -






1.2




1.0 -



0.8-
!5.1


0.6- 1.9



0.4 -



0.2 -


FREQUENCY IN MC
1 I I I I I I
4 8 12 16 20 24 28

Fig. 14. Frequency Dependence of the Loss Tangent at
Water Contents of 0, 11.9, 15.1 mg. H20/g.
A1203












of rigid bonds to water in the first region. The water

in the first region is rigidly held and cannot keep up

with the alternating field and a maximum in tan 8 is

therefore observed. As more water is added the bonding

is less rigid, the molecules keep up with the field more

readily and the maximum is no longer observable in this

frequency range.

Figures 10, 11, and 12,show the change in capacitance

with increasing water content. An increase in the rate

of change of capacitance is observed at about the com-

pletion of one layer of water molecules. No detailed

structure as observed on the tan curves could be found.

McIntosh (1 through 11) has observed linear sections on

similar plots of capacitance versus quantity of material

adsorbed. The total change in capacitance was of the

order of 100 micromicrofarads, over a surface coverage

of about one and a half layers, with a precision of

+ 0.2 micromicrofarads. With the curves reported here

the total change in capacitance is about 10 micromicro-

farads, over a surface coverage of about one and a half

layers, with a precision of + 0.2 micromicrofarads, so

that it is difficult to determine whether linear sections

exist on these curves.











C. Discussion of Nuclear Magnetic Resonance Experimental
Results.

Figures 15, 16, and 17 show the nuclear magnetic re-

sonance spectral line width plotted as a function of

water content for samples activated in the same manner

as those used in the dielectric experiments. Table 2

lists the properties of the samples. The method used

to calculate the quantity of water equivalent to one

monolayer has been discussed previously.

With all the samples studied no accurate line widths

could be measured before the completion of a monolayer

due to the extreme broadness of the derivative signal.

However, at very low water contents a narrow line was

observed superimposed on the wide line. This narrow

line was approximately 500 c.p.s. wide and did not change

in intensity or width with addition of water. It was

finally masked by the broader line which increased in

intensity and narrowed with increasing water content.

This narrow line was much narrower than line widths

measured at very high water contents. It is about 1000

times wider than liquid water and about 100 times nar-

rower than ice (34).

This narrow line might be attributed to protons on

the alumina surface that are far enough away from other

protons so that magnetic interactions are small and T2










UN
WIDH o
(C.P S.
\0


400C


o =ADSORPTION
* DESORPTION


2VL MGH20/GAl203


44


Fig. 15.


N.M.R. Line Width of the Alumina-Water System for the Sample
Activated at 4000C


2000


1750K


I 50C


I 250F


I 000k


o


750


V, VMO
,I1 I


24


40









LI NE
WIDTH
(C.P. S.)


5 00 C


2000




1750





1500




1250




1000





750


VMO


Fig. 1


28 32 36 40 44 48 52 56 60

6. N.M.R. Line Width of the Alumina-Water System for the Sample
Activated at 500oC


o = ADSORPTION
* = DESORPTION


















o









8 000 C


LI NE
WiDTH
-(C.P. S)

0


2VM
S !


24 28 32
Fig. 17. N.M.R. Line Width of
Activated at 800C


40


2 1VMo


44


*

MG HO/ A 1203
18 52


the Alumina-Water System for the Sample


o =ADSORPTION
* =DESORPTION


2000


1750


1 500


I 250




1 000


750












relaxation is very slow. If this is true, adsorption

must not occur around these protons since the line does

not change in intensity or width as adsorption proceeds.

It is difficult to imagine a surface upon which such a

proposed situation could exist.

The line width in Figures 15, 16, and 17 is found

to decrease linearly with increasing amounts of water

until about one and a half layers are complete. At

this point a flattening occurs on Figures 15 and 16

after which a straight line section exists, of lower

slope than the first.

The narrowing of the line in the first region must

be due to the increased freedom of motion of water mole-

cules after the completion of a monolayer.

The flat middle sections of Figures 15 and 16 are

difficult to explain but must be due to adsorption

where the thermal motions of the molecules are not par-

ticularly increased by further addition of water. A

way in which this might arise would be by water mole-

cules being adsorbed with their protons toward the mono-

layers, rather than away from it where they would presum-

ably have more thermal motion.

Figure 17 does not exhibit the flat transition

region observed in Figures 15 and 16 and the break

point is very near to the completion of two layers.












The second linear section must be due to rather

loosely bound molecules adsorbed shortly before and

after two layers are complete.

It should be noted that even at the highest water

coverages the width of the line is still about 800

c.p.s. If the assumption is true that line narrowing

is due only to thermal motions of adsorbed molecules

then comparison with ice, whose line width is 64,000

c p.s., and liquid water, whose line width is 0.5

c.p.s.i, would suggest that even at these high coverages

water is in a state of mobility nearer to that in ice

than liquid water.

Points determined during desorption of water fall

reasonably well on the adsorption curve. It will be

recalled that the capacitance decreased with time in

the same region, suggesting migration of molecules to

states of lower energy. There is no contradiction

between these two results since the decrease in capa-

citance at 750 kc. means that water molecules are no

longer capable of orienting with a field which is chang-

ing polarity at the rate of 750 kc. per second; while the

line width of about 800 c.p.s. means that molecules are

participating in motions up to a frequency of 800 cycles.

This amounts to saying that, while an increase in bond

strength may hinder the molecules from orientating at












frequencies of 750 kc. it need not restrict them from

participating in thermal motions of a frequency of 800

cycles.

Figures 18 and 19 show the temperature dependence

of line width at various water contents. The slopes

of the high temperature portion of these curves are

listed in Table 5. The slopes above a water content of

one monolayer are generally quite similar and small as

compared with the slopes at water contents of 20.2 and

20.4 mg. H20/g. A1203. This verifies the existence of

rather rigid bonds in the monolayer region.

At about -280C a sharp change is observed in the

slope of the lines. This change must correspond to

the "freezing out" of some thermal motions which have

contributed to line narrowing. In this connection it

should be pointed out that the transition temperature

is lower, for most of the curves, at higher water con-

tents where thermal motions are greater, as might be

expected.

,The energy of activation has been calculated, for

the two regions on Figures 18 and 19, from the slope

of the line obtained by plotting log (line width) versus

the inverse temperature. The slope of this line is

equal to AE/2.3 R. The energy of activation below -28C

is 1.7 kcal. For the high temperature region, at a water















TABLE 5

Slopes of Temperature Dependent

Line Width Curves



Sample Activated at 400C


mg. H20
g. Al203


Slope (c.p.s.
(LC /


- 1.96

- 1.07

- 1.66


Sample Activated at 5000C


- 3.57

- 1.21

- 0.83

- 1.00


20.24

37.72

46.82


26.46

35.01

46.30

53.14



















4 0 0C


20.24


0

0 .-37.72

46. 8 2
___L I T (C)


20


40


Temperature Dependence of the N.M.R.
Line Width at Water Contents of 20.24,
37.72, and 46.80 mg. H20/g. A1203


LINE
WIDTH
(C.P.S)


1300o


1200h


1100



1000


900


800


-20


Fig. 18.













5 00C


0 -'.( 0


o-O 20.40


5.01


I 0
0 0
0~~ 53.14


-20 0


Fig. 19.


T ,(o C)


20 40


Temperature Dependence of the N.M.R.
Line Width at Water Contents of 20.40,
35.01, 46.34, and 53.14 mg. H20/g.
Al203


LINE
WIDTH
(C. P. S.)


1400 h


1300 f


1200


I 100k


lOOO0


900o


8ob












content of 20.24 mg. H20/g. A1203, the energy of ac-

tivation is 350 calories, while at a water content of

37.72 mg. H20/g. A1203, the energy of activation is 200

calories. These figures confirm the existence of a

more rigid bond below the monolayer and suggest that

the motion "frozen out" at -280C is one of rotation.

D. Comparison of Dielectric and Nuclear Magnetic Reson-

ance Results.

A comparison of tan 9 and the n.m.r. line width is

made in Figures 20, 21, and 22 for the three samples

studied. The abscissa in these figures is mg. H20 per

square meter of surface.

The fact that line width measurements could not be

made until after the completion of a monolayer is ap-

parent.

The middle section of the line width curve seems to

begin at about the start of the fourth linear region on

the tan 6 curve and extends until the tan S curve flat-

tens out. This suggests that the orientation of dipoles

in the alternating field is not restricted by the addition

of further water molecules to the second layer. However

the thermal motion of the molecules must not increase in

this region, presumably because of the closeness of neigh-

boring molecules. This situation could exist if molecules

in this region were adsorbed perpendicular to the plane

of the monolayer and attached to the mololayer through







LINE
WIDTH
(C.P S)

20 ,O-L




1750-




1500 -



1250








750 -


40 0C


/

/
/


0.17 0.21


0.25 0.29 0.33 0.37 0.41


Fig. 20.


Comparison of
Sample


the Loss Tangent and Line Width, 4000C


T
A
N 14
6
x
103
-12



I0


8



-6



4


2


0.45










5 0OOC


1750




1500




1250




1000




750


0.25 0.29 0.33 0.37


A- 12
N


x 10
103


8



6



4



2



SMG0.45 0.49
0.41 0.45 0.49


Comparison of the Loss Tangent and Line Width,


0.17 0.21


5000C Sample


Fig. 21.







800C T

A 14


200




175'




150




1 25,


Fig. 22. Comparison of the Loss Tangent and Line Width,


LINE
WIDTH
(C.P S)


103- 12



10



8



6



4



.2


IGH20 M2
0.49 0.53


0.17 0.21 0.25 0.29 0.33 0.37 0.41 0.45


8000C Sample












their protons. A situation such as this would allow the

molecule to orient in an external electric field, while

the protons would not have to participate in thermal

motions. This may also be the reason that the detailed

structures observed in the tan i curves, does not ap-

pear on the plots of line width.

E. Summary.

Plots of tan versus mg. H20/g. A1203 exhibited

four linear regions. The first linear region had a con-

stant value of tan S and ended at the completion of one

monolayer. The heat of adsorption in this region was ap-

proximately 17 kcal./mole. The adsorbed water in this

region is held to the surface by rather strong bonds.

Upon further addition of water to the sample, ad-

sorption on the monolayer begins. In this second layer

water exists in four different states.

The first state, corresponding to the second linear

region or the tan S curve, is believed to be due to

water adsorbed in the plane of the surface and probably

attached to the monolayer by three bonds which restrict

it from moving out of the surface plane.

The second state, corresponding to the third linear

region, is more mobile and of lower adsorption energy.

Molecules in it may be attached to the monolayer by two

bonds such that they orient only partially with the ap-

plied electric field.












The next state is the most mobile of the three and

the one with lowest adsorption energy. It corresponds

to water along the fourth linear region. Molecules may

be attached to the monolayer by one bond or by weak hy-

drogen bonds and may be free to move on the surface.

The fourth state resulting from subsequent addition

of water is believed to be somewhat more mobile than ice

with respect to an alternating electric field.

The possibility of active sites within the monolayer

is proposed as an explanation for the various types of

adsorption observed.

The heat of adsorption after the first region ap-

proaches the heat of condensation of water, indicates

the rather weak nature of bonding after the monolayer.

It is pointed out that the surface area occupied by

an adsorbed molecule is dependent on the nature of the

surface upon which it is adsorbed.

Plots of line width versus mg. H20/g. A1203 show

that after the completion of a monolayer water is ad-

sorbed in two different ways separated by a short trans-

ition region. The first water adsorbed on the monolayer

is considered to be more mobile than water adsorbed after

the completion of one and a half layers.

The transition region is thought to be due to the ad-

sorption of molecules in a manner which restricts their












thermal motion without altering their motion in

an alternating field.

Studies of the temperature dependence of line width

showed that the temperature coefficient was larger be-

fore the completion of the monolayer than after, there-

by indicating a more rigid bond in the monolayer than

in the second layer. A sharp change in slope occurred

at about -280C, indicating that some motional freedom

was "frozen out" at that temperature.

A comparison of n.m.r. line width with tan 9 was

made and it was pointed out that no contradiction be-

tween results was involved, since tnn $ was a measure

of motion in an electric field, while line width re-

flected the thermal motion of adsorbed molecules.













TABLE 6
Dielectric Studies
Sample Activated at 4000C for 4 Hours in Vacuum

Measurements at 750 KC


mg.H20 Capacitance Conductance Tan x 03
g.A203 yuf) A(pmho)


0 57 1.6 3.34

0.71 57.3 1.8 3.74

1.49 57.5 1.8 3.73

2.11 57.5 1.9 3.94

2.88 57.4 2.0 4.01

3.63 57.5 2.0 4.01

4.97 57.7 2.1 4.34

7.03 58.1 2.2 4.52

7.73 58.0 2.2 4.52

8.99 57.9 2.5 5.14

9.49 58.0 2.7 5.55

10.06 58.1 2.9 5.95

11.88 58.1 2.6 5.33

15.09 58.3 3.0 6.13

15.98 58.5 3.1 6.32

16.57 58.6 3.3 6.71

17.24 58.8 3.8 7.70

18.15 58.9 3.6 7.28










TABLE 6 continued


mg.H 20
gAl203


19.15

19.81

21.80

22.58

23.30

24.02

24.78

25.53

26.33

27.00

27.64

28.28

29.00

29.66

30.31

31.01

31.81

32.43

33.02

33.69

34.32

34.97


Capacitance



58.9

59.0

60.1

59.6

59.9

60.1

60.3

60.7

61. 1

61.4

62.0

62.3

62.9

63.1

63.7

64.0

65. 1

65.3

66.0

66.6

66.7

68.3


Conductance
(ymho)


4.0

4.2

7.5

6.0

6.8

7.8

9.5

10.8

13.0

15.4

18.3

19.5

23.5

26.5

30.4

34.0

40.0

40.4

46.0

51.0

57.0

61.0


Tan x 103


8.09

8.48

14.8

12.0

13.5

15.5

18.8

21.2

25.3

29.9

35.2

37.3

44.6

50.0

56.9

63.3

73.3

73.9

83.2

91.3

101.8

106.3












TABLE 6 continued


Capacitance



69.4

70.0

70.9

71.9

73.3

74.3

75.6

76.6

77.7


Conductance
A mho)


67.0

70.0

74.0

78.0

84.0

88.0

92.0

95.0

97.0


mg. H20
&1A203

35.50

36.04

36.59

37.14

37.69

38.31

38.98

39.44

39.82


Tan 6 x 103



115.1

119.2

124.2

129.2

136.5

141.2

145.0

148.0

149.1














TABLE 7
Dielectric Studies
Samples Activated at 5000C for 6 Hours in Vacuum

Measurements at 750 KC


mg.H20 Capacitance Conductance Tan 9 x 103
g.A1203 (qpf) (tmho)

0 54.0 1.5 3.31

4.67 54.1 2.0 4.41

10.21 54.2 1.9 4.18

15.23 54.8 2.3 5.00

20.16 55.1 3.5 7.57

22.26 55.4 3.9 8.40

23.26 55.6 4.5 9.64

24.23 55.6 5.0 10.7

25.02 55.9 6.4 13.7

25.95 56.2 7.9 16.8

26.87 56.4 9.5 20.1

27.76 56.9 11.9 24.9

28.63 57.3 14.2 29.6

29.57 57.7 17.5 36.2

30.37 58.2 20.0 41.0

31.19 58.6 24.0 48.8

31.90 59.1 27.8 56.1

32.67 59.5 30.5 61.1











TABLE 7 continued


mg.H20



33.46

34.18

34.91

35.58

36.29

36.99

37.57


Capacitance
(P$Af)


60.1

60.8

61.6

62.4

62.8

63.8

64.5

63.5*

64.6

65.4

66.0

67.0

68.1

69.1

69.8

70.8

71.7

72.7

73.2


Conductance
(amho)


34.0

38.0

42.2

45.9

48.2

53.0

56.0

51.9*

56.0

58.0

60.0

62.5

64.2

66.0

67.0

68.0

68.0

68.0

68.0


*12 hours elapsed between this reading and the previous one.


Tan 8 x 103



67.3

74.5

81.5

87.6

91.2

99.0

103.0

97.3*

103.0

106.0

108.0

111.0

112.0

114.0

114.0

114.0

113.0

112.0

111.0


38.10

38.95

39.58

40.18

40.78

41.31

41.90

43.07

43.63

44.14

45.11














TABLE
Dielectric
Sample Activated at 800C

Measurements


8
Studies
for 6 Hours in Vacuum

at 750 KC


mg.H20



0

3.29

7.78

12.64

14.87

16.53

18.14

19.71

20.36

21.43

22.69

23.89

24.51

25.02

25.52

26.04

26.51

27.02


Capacitance
(5uf)

55.0

55.0

55.2

55.3

55.8

55.6

55.8

55.9

56.1

56.0

56.4

56.8

56.9

57.0

57.3

57.5

57.7

57.8


Conductance
umho)


1.5

1.5

1.8

2.0

2.4

2.5

2.8

3.0

3.3

3.0

3.5

5.0

6.0

6.5

7.5

8.5

9.5

11.0


Tan x 103



3.25

3.25

3.89

4.31

5.13

5.36

5.76

6.38

7.03

6.38

7.40

10.5

12.6

13.6

15.6

17.6

19.6

22.7











TABLE 9 continued


mg. H20
gA1203


27.69

28.25

28.75

29.26

29.76

30.22

30.62

31.10

31.76

32.24

32.73

33.17

33.82

34.26

34.79


Capacitance
(/ f)

58.2

58.6

58.9

58.9

59.3

59.5

59.8

60.2

60.6

61.1

61.5

62.2

62.7

63.5

63.0

62.1*

63.2

64.2

65. 5


Conductance
(pmho)


13.0

15.0

17.0

19.0

21.5

23.5

26.0

29.0

32.0

36.0

39.0

41.0

45.0

48.0

52.0

41.0*

49.0

55.0

60.8


*15 hours elapsed between this reading and the previous one.


Tan J x 103



26.6

30.5

34.4

38.4

42.2

47.1

51.8

57.5

63.0

70.3

75.6

78.6

85.5

90.2

98.4

78.7*

92.5

102.0

110.5


35.51

36.13

36.79











TABLE 8 continued


Capacitance
(/Pf)

66.8

67.9

69.1

70.6

71.7


Conductance
wumho)

66.0

70.0

74.0

78.0

78.0


Tan & x 103


117.8

123.0

127.5

131.5

133.0


mg. H20
g.A1203

37.36

37.93

38. 58

39.34

40.04














TABLE 9

Frequency Dependence of Tan

Sample Activated 4000C 4 Hours in Vacuum


Frequency
(mc)


0.750

4

8

12

14

15

16

17

18

20

24

28


0.00 mg.H20


3.34

4.35

4.97

5.05

3.90

8.00

11.7

9.27

9.9

8.0

12.0

14.0


Tan S x 103
11.9 mg.H20


5.33



7.40



5.93

11.6

12.8



10.7

10.5


15.1 mg.H20
g.A1203

6.13



7.77



6.52



14.5





11.2



20.2














TABLE 10
Dependence of Tan S on Frequency and Water Content
Sample Activated at 4000C for 4 Hours in Vacuum


mg.H20 Tan S x 103
gA03 8 mc 16 mc 20 mc

0.00 4.97 8.67 8.00

7.71 5.80 12.90 11.10

11.88 7.40 12.80 10.50

15.09 7.77 14.50 11.20

19.15 8.80 16.95 14.50

21.80 10.25 19.30 16.40

23.30 10.47 18.35 16.90

24.02 11.90 20.10 17.80

24.79 12.93 22.50 19.25

25.53 14.62 22.80 20.20

26.33 16.00 20.60 21.20

27.00 17.13 24.30 22.60

27.64 18.20 24.30 24.60

28.28 19.85 26.20 25.00

29.00 22.00 26.50 27.40

29.66 23.90 30.90 29.80

30.31 26.40 32.22 31.30

31.00 28.40 32.70 34.10

31.82 32.40 38.60 35.40











TABLE 10 continued

mg.H20 Tan S x 103

8 me 16 mc 20 mc

32.42 33.40 36.60 36.40

33.02 36.30 39.20 38.40

33.69 39.40 43.90 41.20

34.31 43.30 45.70 45.60

34.97 47.20 49.40 50.60

35.50 51.20 53.30 52.00

36.04 53.80 57.00 55.20

36.59 56.40 57.20 57.80

37.14 60.70 62.50 61.80

38.31 68.60 67.80 69.80















TABLE 11

Measurement of Heat of Adsorption on

Sample Activated at 500C


Pressure
(cm.Hg.)


Temperature
(C)


Water content, 16.39 to 18.70 mg. H20/g. A1203


0.06

0.24

0.60

1.01


139

175

200

212


Water content, 34.34 to 39.62 mg. H20/g. A1203


0.40

0.52

0.76

1.04

1.61

2.04


43.5

49.0

55.0

59.0

67.5

73.0












TABLE 11 continued


Pressure Temperature
(cm.Hg.) (0C)


Water content, 48.35 to 49.65 mg. H20/g. Al203

0.69 25

0.87 28

1.20 33

1.59 37.5

2.22 44.5

Water content, 29.00 mg. H20/g. Al203

0.18 66.0

0.49 79.0

0.73 88.5

1.21 99.5

1.55 100.5

Water content, 71.37 mg. H20/g. A1203

0.63 14.0

0.93 20.0

1.45 27.5

1.97 32.5

2.05 33.0















TABLE 12

Dependence of Line Width on Amount of Water Adsorbed

Sample Activated at 4000C for 4 Hours in Vacuum


Adsorption

mg. H20 Line Width
g. A1203 (c.p.s.)

21.6 1999

24.0 1870

26.1 1580

28.4 1473

30.2 1378

32.0 1260

33.9 1240

35.8 1164

37.5 1192

40.1 1040

42.0 990

53.5 873


Desorption

mg. H20 Line Width
g. A1203 (c.p.s.)

53.53 855

49.49 912

41.69 1018

38.95 1164

35.14 1193

33.61 1300

32.38 1271

31.00 1484
















TABLE 13

Dependence of Line Width on Amount of Water Adsorbed

Sample Activated at 5000C for 6 Hours in Vacuum


Adsorption

mg. H20 Line Width
g. Al203 (c.p.s.)


24.8

27.3

29.8

31.8

33.8

35.7

38.2

40.6

43.6

46.0

48.6

51.6

54.6

58.8

59.5

63.5


1840

1765

1552

1407

1321

1187

1158

1149

1032

936

897

897

810

762

782

733


Desorption

mg. H20 Line Width
g. A1203 (c.p.s.)


48.70

43.27

38.04

35.98

34.20

33. 11

32. 17

31.08


916

970

1125

1251

1266

1319

1261

1513












TABLE 14

Dependence of Line Width on Amount of Water Adsorbed

Sample Activated at 8000C for 6 Hours in Vacuum


Adsorption

mg. H20 Line Width
g. A]203 (c.p.s.)


21.5

23.4

25.4

27.4

29.3

31.3

33.2

36.6

38.6

41.3

44.0

45.9

46.2

51.6


2163

1872

2490

1989

1763

1600

1532

1319

1222

1140

1048

989

961

854


Desorption


mg. H20
g. A1203

53.53

51.08

45.74

40.94

39.02

37.35

35.88

32.93


Line Width
(c.p.s.)

815

824

1048

1096

1174

1324

1406

1833














TABLE 15

Temperature Dependence of Line Width

Sample Activated at 4000C for 4 Hours in Vacuum


20.24 mg. H20/g. A1203

Temperature Line Width
(oc) (c.p.s.)
32.8 934

27.6 949

16.0 991

9.3 996

-2.4 999

-19.0 1058

-25.8 1172

-30.7 1212

-34.2 1301


37.72 mg. H20/g. A1203

Temperature Line Width
(oC) (c.p.a.)
33.6 781

13.7 781

7.5 809

-2.9 814

-13.4 814

-18.4 827

-24.0 846

-30.2 862

-33.4 921












TABLE 15 continued


46.88 mg. H20/g. A1203


Temperature
(oC)


Line Width
(c.p.s.)


32.8

10.6

- 1.5

- 8.4

-17.7

-31.3

-33.8

-35.4

-37.0


708

761

768

781

797

819

822

830

841















TABLE 16

Temperature Dependence of Line Width

Sample Activated at 5000C for 6 Hours in Vacuum


20.40 mg. H20/g. A1203

Temperature Line Width
(C) (c.p.s.)

27.0 1071

21.5 1079

18.1 1065

5.6 1103

3.6 1135

2.6 1167

7.2 1167

-10.7 1167

-12.9 1202

-18.8 1199

-21.5 1439

-24.8 1247

-27.4 1279

-37.5 1455


35.01 mg. H20/g. A1203

Temperature Line Width
(C) (c.p.s.)


32.6

21.7

-14.5

-21.9

-28.9

-32.9

-34.6


824

846

879

895

970

993

1064











TABLE 16 continued


46.34 mg. H20/g. A1203

Temperature Line Width
( C) (c.p.s.)

33.4 783

20.9 785

6.5 799

-8.8 815

-14.4 817

-19.7 815

-25.9 831

-30.6 897

-35.8 911


53.14 mg. H20/g. A1203

Temperature Line Width
(C) (c.p.s.)

32.8 718

23.5 735

21.0 703

14.2 751

0.3 759

-8.1 785

-23.6 783

-33.3 815

-34.7 823

-39.2 927




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs