THS

{AL RE: M? THC} E ' fig {fig-1%.? Q?-
\‘A {V‘R 1A? EGH ERR; "" ASJ “EQGETD
GREG MAKE (ERA? H ‘f

T593493 éw {€19 08:31:36 0‘; M. 5‘
34.! Risk“ Shim .IVE’E‘. ET?
Mseph E“ Schiiier
3958

 
    
    

  

I L {8 RA RY
Michigan .551“
3A Um

‘fiES'S

”(gm/gm, WNW/WW

947970

, o Hj .(fim
N...” L Li" J L

n-mmm ‘7“) 1m? 'fi (*7? "WP“IZu-UPTOII

l ‘ - lit. -4. V _,; \31'.

|
. r) T ' 7'. -- 1 7 "7'qu ‘f‘? #31717
"*o-JJL'JL) Cni‘I-H'Cufld‘ )

-M.~---L..L --..-.4..

\J

‘hnmJ- n.” Vq-..~,n~.i r'-~ 1‘ 3 m)“. .'\* n nr"vv)"-~.'\1nq(" *r 13') (I #3101”? r: l an
1 .. ~ 4 ' / f ‘- g; '- L! x.‘ - .A J‘ .-. ‘J :x 7- v L ‘. Apt; 9“ 9 . .z‘ ‘ {-1 1'. I '.

frwm *Ph PF2“*O in rhffiffi on find as ? fvr0140n O

_o clromatorraphlc Cflwun.

tQMWGratHrn in a was—Jirn*

~

The r0,orfiiOD time-tmfoW35tan bnhavior Of thq com—

- r

fiovnfls studind woro contflrod with that of a rpfocnnca

comhnunfi frog tho ano hom)1owous series for wbic

tdé bait of vaW0?1fiat10n was known. From the ratio
A” .1.“ r“ .z. ~ .2 a ,. ,

of 93))os of pho lo; vnucntzcn t7JQ—T0CLTTOC41 tam—
firvaffro fiiots n” w c¢3flmnnfi ¢flfl fhe PQFOWOWCQ, the

x.-. 4.

fififit of Vdnarifiqtion of t%o firhp conhounfi was cairn—

4.

I 1tr i. 1% SFWTQ]¢)rc> thlnirq Wrzs 113cw3, :wwfi (flair) VERS
‘“Lv5rvw3 frw’1?~“°fi“*??iflfi?, finrflTi+? ’“Ttdws

v-~-— my — 4‘1« j fo 1\ - - np-t 11 .‘L flux
FiloQWTWWPC L»d F“ uflfi H.8 shown to r;vn RJTuUulQ

. . 4—1-,‘0 -. 4—711. m D “a,“ 1 .30
V “T L: “or? #3,";L an 0.x. C‘V=.;..’rm»lrLL«oo

A DIRECT METHOD FOR HEAT OF VAPORIZATION
BY GAS-LIQUID CHROMATOGRAPHY

BY

‘1
1‘
l

Joseph E? wSchiller

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry
1968

ACKNOWLEDGMENTS

Deepest thanks must be given to my advisor,
Dr. MacDonald, for his leadership and help through-
out the course of this work. His interest and
enthusiasm in this project were an inspiration.
Mr. Paul Handy and my wife, Ruth, gave much per-
sonal support. Thanks must also be given to
Michigan State University Department of Chemistry
for financial support and facilities, which made

this work possible.

Table of Contents

Title Page
Acknowledgments

Table of Contents

List of Tables

List of Figures
History

Theory

EXperimental

Results and Discussion
Conclusion

Bibliography

iii

iv

ll
15
28
29

 

Table

List of Tables

Results for Heat of Vaporization
Results for n-Alcohols

AH of Vaporization and AH of Mixing
of n-Alcohols

Page

21
23
24

Table

List of Figures

Title

Retention Time-Temperature Data
n-Taraffins

Retention Time-Temperature Data
Acetate Esters

Retention Time—Temperature Data
Ethers

Retention Time-Temperature Data
Haloalkanes

Retention Time-Temperature Data
n—Alcohols

Variation of Retention Time With
Sample Size

for

for

for

for

for

22

26

History:

Gas chromatography has developed into one of the
most powerful analytical tools available, with the
separation and determination of mixtures as its major
application. Gas chromatography, however, has been
used to determine thermodynamic constants and physical

10’11 In the

preperties by two distinct approaches.
first case, the phenomena under study takes place out-
side the chromatographic system, and the chromatograph
is used as a convenient analytical tool. Secondly,
with the development of the theory of gas chromato-
graphy, it became evident that the direct measurement

of thermodynamic constants and physical properties wa

possible.

The composition of the vapor over a solution as a
function of solute concentration and temperature has

25

been determined by gas chromatography. A sample of
the vapor above the solution was introduced into a gas
chromatograph where the components were separated and
determined. The partition isotherms of the systems

studied have been determined by repeating the experiment

with varied parameters.

The solubility of hydrocarbons in water has been
12
determined by passing helium carrier gas through a

known volume of a hydrocarbon saturated aqueous solution

directly into the chromatographic column. water was
removed from the gas stream by a drying tube, and the
remaining hydrocarbon passed to the detector. The re—
sults obtained were in good agreement with spectrophoto-

metric measurements.

The heat of vaporization of gasses and liquids

8’14 A bulb

has been measured in the following way.
containing the sample was connected to a bypass samp-
ling system of known volume. The bulb and sampling
system were thermostatted, and the sample was allowed
to equilibrate with the previously evacuated bypass
system. The sample in the fixed vapor volume was then
passed through the column and determined quantitatively
at the detector. The heat of vaporization was calcu-
lated using the Clausius—Clapeyeron equation. For

this technique, samples must be pure, and the compound

studied must obey the Clausius-Clapeyeron equation.

The direct determination of thermodynamic and
physical quantities by gas-liquid chromatography has
been done by treating the interaction of the sample and
stationary liquid in the column as the solution of a
gas in a liquid. The equations which describe the dis—
tribution of a vapor between the gas phase and the solu-

tion are thus applicable. These are combined with the

equation which relates retention volume and the parti—
tion coefficient for a chromatographic system.9 The
heat of solution, heat of mixing, activity coefficient,
partition coefficient, and several other quantities

may be calculated directly from gas-liquid chromato-

, 1
graphic data.

The partition coefficient was first determined by
Porter, Deal, and Stross.23 A number of columns were
prepared, and the parameters such as the flow rate of
carrier gas, weight of liquid phase in the column, dead
volume, etc. were known accurately. The retention
volumes were corrected for pressure drop across the
column. No comparison between the results published
in this paper and previous measurements were made, but
values for the partition coefficient were of a reason-
able value. Further investigation has shown the approach

of Porter, Deal, and Stross to be valid.lo’ll

The partition coefficient is directly proportional
to the Henry's Law constant, so solubilities may be
calculated from gas-liquid chromatographic data. The
solubility of benzene and cyclohexane in di-n-nonyl
phthalate was determined by gas—liquid chromatography
and found to be in excellent agreement with conventional
measurements.18 More recently, the solubilities of
group III, IV, and V hydrides were measured by Devyatykh,

24
et.al. using water as the stationary liquid -hase.

The parameters of the system were known accurately,
and the results for the solubilities were in very good

agreement with previously published values.

It can be seen from equation 15 that the molecu-
lar weight of the stationary liquid may be measured
if all of the other values in the equation are known.
Martire and Purnell16 determined the molecular weights
of three liquids of molecular weight approximately
BOO, 400, and 1200 respectively with good results and

demonstrated the possibility of using gas-liquid

chromatography to make this measurement.

Rearrangement of equation 15 yields an expression
for the activity coefficient in terms of measurable
quantities. The measurement of very large or very
small activity coefficients may be made conveniently
by this technique. A very large activity coefficient
would be difficult to measure by conventional techni-
ques because the solubility would probably be extremely
low. In cases where results from gas-liquid chromato-
graphy may be compared with previously published re-
sults, there is excellent agreementlg"22 The measure-
ment of activity coefficients by gas-liquid chromato-
graphy is sufficiently reliable that it has been
13

suggested as an undergraduate eXperiment.

Muhs and Weiss determined the formation constants
of silver-olefin complexes in ethylene glycol using
gas chromatography.15 The partition coefficient of the
olefin with pure ethylene glycol was first determined.
A solution of silver nitrate in ethylene glycol was
made and used as the stationary liquid in a chromato—
graphic column to determine the partition coefficient
between the solution and the vapor phase. Since the
concentration of silver nitrate in the ethylene glycol
was known, the formation constant for the complex

could be calculated.

The change in partition coefficient or retention
volume with temperature for a given system allows the
calculation of additional constants. It has been pre-
dicted and found experimentally that when the log of
retention volume for a solute is plotted against recep-
rocal temperature that a straight line is obtained.9
The slope of this line is the heat of solution divided
by 2.303R. By determining experimentally the retention
volume as a function of temperature for a given solute-
solvent system, the heat of solution of the solute may
be calculated. The heat of solution is composed of
two terms, the heat of condensation and heat of mixing.
If the heat of vaporization of the solute is known, the
heat of mixing may be obtained by difference. However,

if the solute and solvent form an ideal solution, the

heat of mixing is zero, whereupon the heat of solution
equals the heat of vaporization.l This possibility
for measuring heat of vaporization, however, is essen-

tially restricted to hydrocarbons.

Vapor pressure as a function of temperature and
the boiling point of the solute may be determined by

1 found

an extention of the above. Hoare and Purnell
that under the preper conditions, some families of com—
pounds have a common log retention volume-log vapor
pressure plot. The retention volume as a function of
vapor pressure was measured for one member of such a
series, and the retention volume as a function of
temperature was measured for another compound in the
series. The vapor pressure as a function of tempera-
ture and the boiling point were calculated for the
second compound. Data were obtained for n—paraffins

and n-olefins, and results were in good agreement with

previously published results.

Theory:

When a vapor is dissolved in a liquid, the partial
pressure of the vapor over the solution, PA, is given
by Roults Law:

0
(1) PA = PA B XA

P2 is the vapor pressure of the pure solute, B the

activity coefficient, and XA the mole fraction of the

solute in the solution. The mole fraction of the solute

may be expressed by the equation.
(2)X=_£A___
ANN
A B
If the solution is very dilute in component A, then

(3) XA ; NA
B

If v: is the molar volume of the solvent, equation 3

may then be written:

m l m
(4) X = NA ' = C V
A -v— B A B

B

The partial pressure of A in the vapor phase may be
expressed by the ideal gas equation:

v
(5) PA =fl = CXRT
v

VA

Substituting equations 4 and 5 into 1,
V o .1
(5) CART = PA sag":l

Rearranging:

 

KA is the distribution coefficient of component A
between the solution and the vapor space over the
solution. A Clausius—Clapeyeron type equation for the
vaporization of component A from solution is:

ABS

eC

 

(8) BP0 = e -
A RT‘

AHs is the heat of vaporization of component A from

solution. A constant (a) is defined as:

(9) a = AHs
V

AHV is the heat of vaporization of pure A. For com-
ponent A the Clausius-Clapeyeron equation is:
o -— e {ARV e
A " ' "RT"
Combining equations 9 and 10:

(10) P C'

_o __ o a
(11) PA — (PA)
Equation 7 may then be written:

(1) K RT
2 :._______
(PXWVE

For a gas-liquid chromatographic system, the
interaction of the sample vapor with the stationary
liquid is simply the solution of‘a gas in a liquid. The
equation

(13) VR 2 KV +v

l d

gives a relationship between the retention volume and

the distribution coefficient.23 V1 is the volume of the

liquid phase, and Va is the dead volume in the chroma-

tographic system. Combining equations 12 and 13:

 

( 4) RTVi + HT“N1
1 v == V‘ = + v
R 05 m d o a d

N1 is the number of moles of the stationary liquid in
the column. If the retention volume is corrected for
dead space,

RT‘Nl

(15) VR = W

T'is the temperature when the solute passes through the
column, and Po is the vapor pressure of the pure solute
at temperature T. If F is the flow rate of the carrier

gas and tR the retention time, then

RT N1

(16) FtR = TEUj‘E“

The flow rate measured at ambient temperature must be

corrected to give the flow rate at the column tempera-
3

ture. Then:

. T” RT N1
(17) tRFa, Ta " (PO) a

Ta is the ambient temperature and Fa is the flow rate

measured at this temperature. Cancelling like terms and

rearranging:
= a
(18) ta EFT???—

The usual correction of retention volume for pressure

drop across the column4 need not be made. The reason

10

for this will be discussed shortly. Taking the log—

arithm of both sides of equation 18:

O Brawl
(19) log tR’ = -a log P -+ log-—E;—-
Substituting equation 10 into 19:
a AHV RTéNl
(20) log tR == ._____. + C + log
2.3RT‘ F

a
RTéNl

The term C + log —F;—— may be easily held constant.
A plot of log retention time versus reciprocal

temperature should yield a straight line of slope

:73égl' if data are obtained for retention time as a

function of column temperature. The constant (a) has

been found experimentally to be the same for each member

of a homologous series.1 If the heat of vaporization

is known for one member of the series, the heat of

vaporization of other members may be calculated from

retention time-temperature data.

As indicated previously, the retention time was
not corrected for pressure drop across the column.
Since the change in retention time with temperature is
of interest, multiplying the right side of equation 18

by a constant will not change the result.

11

Experimental:

To verify the equations derived above, retention
time-temperature data were found for a number of homo-
logous series of organic liquids. To obtain reliable
data, the flow rate of carrier gas through the column
and the column temperature had to be controlled pre-
cisely. The F&M Model 810 gas chromatograph (Hewlett-
Packard, F & M Div., Avondale, Pa.) was found to be a
suitable instrument for the study. The temperature in
the column compartment was measured with a mercury
thermometer accurate to .100. After temperature equi-
librium had been established in the column compartment,
the temperature was found to vary less than .loC. A
flow rate of approximately 40 ml/minute of Helium
carrier gas was used, and the change in flow rate with
temperature was found to be less than .1% for a 20°C
change in column temperature. At constant temperature,
no measurable change in the flow rate was observed.

The flow rate was measured with a soap bubble flowmeter.

The column packing was prepared in the following
way. Gas Chrom Q (Applied Science Labs, State College,
Pa.), an acid washed, silanized, diatomaceous earth
solid support was weighed accurately. The desired

quantity of liquid phase was weighed and dissolved in a

l2

suitable solvent. The solid support and the solution
were put into a ribbed round bottom flask, and the
flask was rotated as the solvent was removed under
vacuum. The sides of the flask were washed down with
the solvent, and the additional solvent was removed
in the same way. The solid was then assumed to be
coated uniformly with the liquid.

The Gas Chrom Q was 100-120 mesh, and 1/4 inch cop-
per tubing was used for the column. A 4-foot column

packed with 25m squalane (Eastman Organic Chemicals) on

the solid support and a 6-foot column with 15% diisode-

cyl phthalate (Eastman Organic Chemicals) packing were
made. The dead space in the instrument and column had

to be determined since the column dead volume is not con-

stant with temperature, and erroneous results would be

obtained if these terms were neglected. The usual pro-

cedure for dead time correction is to measure the time

between the emergence from the column of a compound not

retained in the column and the emergence of the compound

of interest. Methane was introduced with each sample

since a flame ionization detector was used. This is the

lowest boiling compound which will give a response with

the detector used, and was found to be suitable, based on
the following. From the retention time—temperature be-
havior of pentane, it was calculated that methane would
be retained only 2.5 seconds at 40°C. {ore importantly,

the change in retention time of methane with temperature

13

in the range of interest would be of the order of one
second for a 10000 temperature change. No significant
error was introduced into the results by assuming the

retention time of methane to be essentially zero.

Retention time as a function of temperature was
measured for n-paraffins, haloalkanes, ethers, and
acetate esters using the column containing squalane.
The n—alcohols were studied with the diisodecyl phthal-
ate column because the retention time of the alcohols
on the squalane column was too short for accurate data
to be obtained. It also was of interest to see if the
equations derived above could be applied when the

stationary liquid was somewhat polar.

For each compound, the retention time was measured
at at least four temperatures from approximately 10
degrees below the boiling point to approximately 10
degrees above the boiling point. The heat of vaporiza-
tion at the boiling point was then calculated, and

could be compared with previously published values.

The liquids to be measured were put in small vials
and stoppered with a rubber septum. After standing for
several hours, five microliters of the vapor above the
liquid were withdrawn into a Hamilton ten microliter
syringe. Approximately one microliter of natural gas

was withdrawn from another container, and the contents

14

of the syringe were injected into the chromatographic
column. A stopwatch was used to measure the time
between the peak recorder response for methane and

the sample. All measurements were made in duplicate.

15

Results and Discussion:

Data obtained for retention time as a function
of temperature for the n-paraffins, acetate esters,
ethers, and haloalkanes are plotted in Figures 1, 2,
3, and 4 respectively. The plots show very little
scatter of the experimental points, and the slope of

each line was determined by the method of least-squares.

The calculation of the heat of vaporization of
each compound was done in the following way. Cne com-
pound from each series was taken as a reference, and the
slope of the retention time-temperature plot was calcu-
lated for each member of the series. Since the lepe is
proportional to heat of vaporization within a series,
the ratio of the slopes for the unknown and reference
compound was multiplied by the heat of vaporization of
the reference compound. The result is the heat of vapor-

ization of the unknown.

An alternate method of treating the data is to first
determine the constant (a) for each series. This would
be done by choosing a reference compound from each
series and determining the retention time as a function
of vapor pressure. Based on equation 19, a plot of the
data would yield a straight line of slope —(a). For

other members of the series, a plot of the logarithm of

16

 

 

 

3.35

2.6 _
(2)
(1)
2.5 _
2.4 a
L .
og TR
2.3 _
l. Heptane
2.2 _ 2. Hexane
3. Pentane
2.1
J I l l I
2.55 2.60 2.80 3 3.00 3.20
10 /T
Figure l. Retention Time-Temperature Data for n-Paraffins

l7

 

 

 

(1) (2)
2.3
" (3)
2.2 p
2.1 _
Log TR
2.0 -
l. Butyl Acetate
2. Propyl Acetate
3. Ethyl Acetate
1.9 _
1.8 1 I J I
2.4 2.5 2.7 2.9 3.1
103/T

Figure 2. Retention Time-Temperature Data for Acetate
Esters

18

 

 

(1)

(2)

2.6 ..

('5)
2.5 -
Log TR
2.4 l. Butyl Ether
- 2. Propyl Ether
3. Ethyl Ether
2.3
I I I I 1 I I
2.3 2.4 2.6 2.8 3.0 3.2 3.4 3.6

103Ar

Figure 3. Retention Time-Temperature Data for Ethers

19
2.6

(l) Butyl Bromide
Butyl Chloride
Propyl Bromide

Propyl Chloride

2.5 r. (2)
(3) (4)
2.4 ,.
Log Th
2.3
3.1 3.3

4:"me
000

2.2

 

J L

2.5 207 209

103/1r

Figure 4. Retention Time-Temperature Data for Haloalkanes

20

the retention time versus reciprocal temperature has
a lepe of the value a AHV/2.3RT1 The heat of vapori-
zation of the other members of the series could then

be calculated, once (a) had been determined.

The first method is preferred for two reasons.
First, very accurate vapor pressure data over the
range desired are unavailable except for a few com-
pounds which have been studied extensively. An addi-
tional error is also introduced by plotting vapor
pressure at a given temperature rather then tempera-

ture directly.

The results for heat of vaporization are shown
in Table I. There is excellent agreement between
values obtained by the method described above and those
obtained by conventional techniques. In fact, because
assumptions were made in the derivation, (that the
sample obeys the ideal gas law and the Clausius-
Clapeyeron equation) the results indeed are much better
than would be expected. However, one would expect
deviation from the ideal gas law and the Clausius-
Clapeyeron equation to occur approximately to the same
degree within a series. The effect of these devia-
tions is essentially eliminated, since a compound from

the series is used as a reference.

Compound

Pentane
*Hexane

Heptane

*Propyl Chloride
Propyl Bromide
Butyl Chloride
Butyl Bromide

Ethyl Acetate
*Propyl Acetate

Butyl Acetate

*Ethyl Ether
Propyl Ether
BUtyl Ether

21

 

* Reference compounds

 

Table I
AH AH

313.. 33:32:33.1 Ltiz‘iflti:
1.28 6.05 6.16
1.46 ‘____ 6.90
1.62 7.66 7.66
1.30 __ 6.51
1.39 6.96 6.98
1.45 7.26 7.17
1.55 7.76 7.61
1.47 7.74 7.76
1.56 -____ 8.21
1.64 8.63 8.59
1.32 .____ 6.22
1.61 7.58 7.77
1.94 9.14 9.01

Reference

 

O\O\O\O\

Retention time as a function of temperature was

also determined for n-alcohols eluted from diisodecyl

phthalate. A plot of the data is shown in Figure 5.

Using ethyl alcohol as the reference compound, the

heat of vaporization of the other alcohols was calcu~

1ated. The results are

shown in Table

In this case

 

 

2.5 " 22
(3)
2.4 _
2.3
(2)
2.2 b
(l)
2.1
1. Butyl Alcohol
2. Propyl Alcohol
3. Ethyl Alcohol
2.0..
LA 1 1 1
2.4 2.5 2.7 2.9 3.1
103/1r

Figure 5. Retention Time-Temperature Data for Alcohols

23

there seems to be no correlation between experimental

and accepted values for the heat of vaporization.

 

 

 

Table II
AH' AH‘
Experimental Literature
Compound Slope keel/mole kcal/mole Reference
Ethyl Alcohol 1.25 9.41 5
Propyl Alcohol 1.47 11.1 9.88 5
Butyl Alcohol 1.25 9.41 10.5 5

(D

This, however, may b explained in the following
way. Consider the solution of a rather polar gas in a
nonpolar liquid. The heat of solution is composed of

two parts, the heat of condensation and the heat of

mixing. The constant (a) previously defined is of the

 

form:
a = AHS = AHV AHm
AER, ‘AHV

Since the solvent can at most interact only slightly
with the solute, the termziHh arises mainly from removing
the attractive forces between the solute molecules of a
liquid which must be broken for vaporization to occur.
Therefore, the termlifim should be approximately a linear
function of the heat of vaporization of the solute.

Then (a) will be very nearly a constant over a wide

temperature range.

24

In the case where both the solute and solvent are
somewhat polar, the heat of mixing then is affected by
considerable solute—solvent interaction. The heat of
mixing is then not a simple function of ARV of the solute,
and therefore the constant (a) would vary with tempera—
ture. Reliable results then could not be expected for
heat of vaporization by the method described above when

a polar solvent is used.

To support the conclusions stated above, no accur-
ate data from the literature could be found for the
heat of mixing of polar compounds in hydrocarbons.
For the heat of mixing of n-alcohols in diisodecyl phthal-
ate at 100°C, Porter, Deal, and Stross obtained the

results shown in Table III.23

 

Table III
AH’ AH
Compound Vaporization Mixing
Ethyl Alcohol 9.41 4.61
Propyl Alcohol 9.88 1.41
Butyl Alcohol 10.5 1.80

Again, there is no correlation between heat of mixing
and heat of vaporization. This is probably due to solute-

solvent interactions.

25

There is an additional point worthy of note which
became evident during the course of this work. As
indicated in the derivation, the sample size taken to
determine retention time must be small so that a very
dilute solution will result. The method used in this work ‘
was to introduce a very small amount of the sample into
the chromatographic column. The alternate, and more
widely used procedure, is to determine the retention
time for a number of rather large sample sizes. The
retention time as a function of sample size is found
and a plot of the data is extrapolated to zero sample
size. The retention time at zero sample size is

assumed to be that for an infinitely dilute solution.

It was found, however, that there is no simple
relationship between sample size and retention time
when expremely small samples are used. This was observed
also by Urone and Parcher.l7 Data are shown in figure 6
for retention time as a function of sample size for pro-
pyl alcohol and chloroform eluted from squalane on Gas
Chrom Q. As indicated by the plot, the extrapolation
does not give the retention time for an infinitely small

sample.

A solute in a chromatographic column is retained

by both the solid support and the stationary liquid.

200

150

T" (sec
R )

100

50

Figure 6.

25

 

 

 

 

+E>E:-__._ (1)
“h
f-
l. Chloroform
2. 2-Pr0panol
2
L...— = ()
L, L J
O 10 2O 25

Moles of Sample x 10

6

Variation of Retention Time With Sample Size

 

27

Even a very inert solid support has some active sites
which may adsorb sample molecules. When a rather large
sample is eluted from a column in which the solid sup—
port is quite inert, the solid support appears to have
no effect on the retention time. This occurs because
the number of molecules of sample is large with respect
to the number of active adsorption sites. As the sample
size is decreased, retention by the solid support be—
comes more significant since the relative abundance of
active adsorption sites is increased. The extrapola-
tion technique would, however, give accurate data if
one were interested in absolute retention volumes for
the liquid phase, since the effect of the solid support
has been removed. In this work, satisfactory results
were obtained using a constant small sample size.
Retention by the solid support was either constant with
temperature, or varied in such a way that the results

were not affected.

28

Conclusion:

The method of gas liquid chromatography has been
shown to be a general technique for determining the
heat of vaporization of volatile liquids, and for many
cases may be the only practical technique. Samples
which are impure may be used since data for the com-
pound of interest in the mixture may be determined.
Since the sample size may be extremely small, toxic
compounds and those not available in large quantities

may be measured easily.

In addition to the work previously described,
some additional related experiments should yield use-
ful information. First, it was concluded from the data
presented above that the heat of mixing of a polar
solute in a non-polar solvent is a linear function of
the heat of vaporization of the solute. This should be
investigated either by conventional means or by gas

chromatography.

The concept of a "homologous series" should also
be defined more clearly. EXperiments should be con-
ducted which determine whether certain compounds are
homologs. For example, 2 heptanol and 5 heptanol are
both secondary alcohols, but one may not be a suitable
reference for the other in determining heat of

vaporization.

10.

ll.

12.
13.

14.

29

Bibliography

Heare, M.R., and Purnell, J.H., Trans. Farad.
Soc., 52, 222 (1955)

Driesbach, R.R., "Pressure-Volume-Temperature
Relationships of Organic Compounds", Handbook
Publishers Inc., Sandusky, Ohio, 1952

Ambrose, D., Keulemans, A.I.M., and Purnell,
J.H., Anal. Chem., 29, 1582 (1958)

James, A.T., and Martin, A.J.P., Biochem. J.,
:9. 679 (1952)

International Critical Tables, E.W. Washburn,
editor, Vol. 5, p. 137, McGraw—Hill Book Co.,
New York, 1929

Driesbach, R.R., "Advances in Chemistry Series",
Vol. 22, Am. Chem. 800., Washington, D.C., 1959

Rose, H.C., Stern, R.L., and Karger, B.L., Anal.
Chem., 2E, 459 (1965)

Kackle, H., Hayrick, R.G., and Rooney, J.J.,
Trans. Farad. Soc. 56 115 (1960)

"_’

Littlewood, A.B., Phillips, C.S.G., and Price,
D.T., J. Chem. Soc., 1480 (1955)

Purnell, H., "Gas Chromatography", John Wiley
and Sons, Inc., New York, London, 1962

Dal Nogare, S. and Juvet, R.S., "Gas Chromatography",
Interscience Publishers, New York, 1952

Kusy, V., Anal. Chem., :1, 1748 (1965)

Kenworthy, 8., Miller, J., and fiartire, D.E.,
J. Chem. Ed., 52, 541 (1963)

Mackle, H., and McClean, R.T.B., Trans. Farad. Soc.,
£9. 817 (1964)

50

Bibliography -- Cont.

15.

l5.

17.

18.

20.

21.

22.

25.

24.

25.

Muhs, M.A., and Weiss, F.T., J. Am. Chem. 800.,
§&. 4697 (1962)

Martire, D.E., and Purnell, J.H., Trans. Farad
Soc., fig, 710 (1965)

Urone, P. and Parcher, J.I., Anal. Chem.,‘2§,
27c (1966)

Adlard, E.R., Khan, M.A., and Whitham, B.T.,
Gas Chromatography 1960, Butterworths, London,
P. 251

Bradford, E.W., Harvey, D., and Chalkley, D.E.,
J. Inst. Petrol., 3;, 80 (1955)

Kwantes, A., and Rijnders, G.W.A., Gas Chromato-
graphy, 1958, Butterworths, London, P. 125

Anderson, J;B., and Nepier, K.H., Australian J.
Chem. 19, 250 (1957)

Ashworth, A.J., and Everett, D.H., Trans. Farad.
800., 5g, 1609 (1960)

Porter, P.E., Deal, C.H., and Stross, F.H., J. Am.
Chem. Soc., IQ, 2999 (1956)

Devyatykh, G.G., Ezheleva, A.E., Zarin, A.D., and
Zueva, M.V., Russ. J. Inorg. Chem., §, 678, (1963)

Gianetto, A. and Panetti, M., Am. Chim. (Rome)
0, 1713. (1960)

U

"Tififilliliu‘ulfiliiflfmu

1111111155

31293010