RAMAN STUDIES OF CONFORMATIONAL ENERGIES

AND HYDROGEN BONDING IN ALCOHOLS

THESIS

Presented to the Graduate Council of the

North Texas State University in Partial

Fulfillment of the Requirements

For the Degree of

MASTER OF SCIENCE

By

Simindokht Maleknia, B.S.

Denton, Texas

August, 1982

AlO, 5O

Maleknia, Simindokht, Raman Studies of Conformational

Energies and Hydrogen Bonding in Alcohols. Master of

Science (Chemistry), August, 1982, 52 pp., 8 tables,

9 illustrations, bibliography, 20 titles.

The conformational energy differences have been deter-

mined for ethylene glycol, 2-chloroethanol, and 2,2-

dichloroethanol in the neat liquid, DMSO, and H20 with

Raman spectroscopy. Spectra in the 0-H valence region were

utilized to determine the energy difference between inter-

and intramolecularly hydrogen bonded species. It was found

that the solvent effect on the relative stabilities of the

gauche and trans rotamers of the alcohols differ signifi-

cantly. The results also indicate that, unlike ethylene

glycol, there is significant intramolecular hydrogen bond

formation in the halogenated alcohols in the neat liquid

phase. Stronger intramolecular hydrogen bond formation

was observed in dichloroethanol than in 2-chloroethanol.

TABLE OF CONTENTS

PageLIST OF TABLES.. ............... . . . iv

LIST OF ILLUSTRATIONS .0... .... .......... v

Chapter

I. CONFORMATIONAL EQUILIBRIUM ANALYSIS WITHVIBRATIONAL SPECTROSCOPY . . .. .....

IntroductionConformational EnergiesExperimental

II. CONFORMATIONAL ENERGIES OF 2-CHLOROETHANOLAND ETHYLENE GLYCOL IN HYDROGEN BONDINGSOLVENTS . ..... . . .......... . 6

IntroductionExperimentalResults and Discussion

III. CONFORMATIONAL ENERGIES AND HYDROGEN BONDINGIN DICHLOROETHANOL.......... ......... 27

IntroductionExperimentalResults and Discussion

APPENDIX.................................. . .* . 47

BIBLIOGRAPHY .............. ..................... 51

iii

LIST OF TABLES

Table Page

I. Area Ratios for Neat 2-Chloroethanol . . . . . 8

II. Spectroscopic Parameters for 10 MolePercent 2-Chloroethanol in Water . . . . 9

III. Spectroscopic Parameters for 40 MolePercent 2-Chloroethanol in DMSO . . . . . 10

IV. Spectroscopic Parameters for 65 MolePercent 2-Chloroethanol in DMSO . . . . . 11

V. Calculated Equilibrium Constant for 10 MolePercent Ethylene Glycol in Water . . . . 13

VI. Solvent Dependence of AH (=HTrans-HGauche- 16

VII. Spectroscopic Parameters of the HydroxylStretching Modes ... .. ......... 34

VIII. Comparison of Energy Differences . . . . . . . 35

iv

LIST OF ILLUSTRATIONS

Figure Page

1. Temperature Dependence of Area Ratios forNeat 2-Chloroethanol (Squares), 40% inDMSO (Open Circles), 65% in DMSO (SolidCircles), and 10% in Water (Triangles) . . 15

2. High Frequency Raman Spectrum of2-Chloroethanol in the Neat LiquidPhase at Room Temperature ........ 20

3. High Frequency Raman Spectrum of 40%2--Chloroethanol in H20 . *... ....... 22

4. High Frequency Raman Spectrum of EthyleneGlycol in the Neat Liquid Phase . . . . . 24

5. High Frequency Raman Spectrum of DCE at-140C; Circles--Experimental Spectrum;Solid Line--Theoretical Curve . . . . . 31

6. High Frequency Raman Spectrum of DCE at+470C; Circles--Experimental Spectrum;Solid Line--Theoretical Curve......... 33

7. Temperature Dependence of Peak Frequenciesof DCE (Solid Circles) and Chloroethanol(Open Circles).................. . . . 39

8. Temperature Dependence of Line Widths ofDCE (Solid Circles) and Chloroethanol(Open Circles)............. . ........41

9. Temperature Dependence of RelativeScattering Intensities in DCE (SolidCircles) and Chloroethanol (OpenCircles) . . . . a . . . . . - -0 . . . . . 43

CHAPTER I

CONFORMATIONAL EQUILIBRIUM ANALYSIS WITH

VIBRATIONAL SPECTROSCOPY

Introduction

The techniques of infrared and Raman spectroscopy have

long provided a means for studying the conformational

behavior of various hydrocarbons and their derivatives in

different chemical environments. Distinct peaks arising

from the vibrations of the individual gauche and trans con-

formers can be observed because the time scale of molecular

vibrations (103-104 sec) is considerably shorter than

that for rotation about carbon-carbon single bonds. Thus,

variations in peak areas can be correlated to the influence

of temperature and environment (i.e. solvent) on the enthalpy

differences between the gauche and trans rotamers (1,2).

In this work, the conformational energies of some

alcohols have been investigated by Raman spectroscopy. The

ability of these molecules to form intermolecular and intra-

molecular hydrogen bonds influences the relative enthalpies

of the conformational states. The results of hydrogen

bonding effects in the liquid phase and in solution are

discussed. The conformational energy differences for

2

ethylene glycol, 2-chloroethanol and 2,2-dichloroethanol in

the neat liquid and some solvent systems with varying hydro-

gen bonding capacities are also presented.

Conformational Energies

The conformational energy differences (AH) are deter-

mined spectroscopically by measuring the temperature and

the solvent dependence of the relative intensities of

vibrational peaks assigned to the trans and gauche conformers

(3). In the gas phase this technique is best employed with

infrared spectroscopy because of the inherently greater

intensity of the IR absorptions. Raman spectroscopy is

preferred for the liquid phase and in solution due to the

small sample size and facile variation of temperature over

a broad range.

For the conformational equilibrium in disubstituted

ethanes, where the two gauche rotamers are equivalent, the

equilibrium constant can be expressed by:

2XT -AH/RT AS/RK =X G=e

XG

where XT and XG represent the mole fraction of the two

conformers.

In Raman spectroscopy, where the concentration is

directly proportional to the peak intensities, the equili-

brium constant is given by:

3

2NTAK- -A(2)

NGAGG G

where AT and AG are the integrated areas of the two bands;

NT and NG are the Raman scattering cross sections. By

substituting equation 2 into equation 1, the Van't Hoff

equation is obtained after taking the natural logarithm:

LN(AT/AG) = -AH/RT + C .

Enthalpy differences (AH) may be determined directly

-lfrom the slope of a plot of LN(AT/AG) vs T , where the

constant C is a function of both the entropy difference and

relative scattering cross sections of the two conformers.

Experimental

The samples of ethylene glycol, 2-chloroethanol, and

2,2-dichloroethanol were obtained commerically and were

purified by vacuum distillation prior to use. Solutions of

alcohols in the solvents DMSO and H20 were prepared gravi-

metrically. The samples were contained in sealed melting

point capillary tubes and were inserted in a standard

Harney-Miller cell. The temperature was regulated by either

heated air flow or liquid nitrogen boil-off through the

cell, and was measured with an iron-constantan thermo-couple.

The associated accuracy in temperature measurement was

determined to be +10C.

4

The excitation source was a Coherent Radiation CR-3

0

argon ion laser operated at 4800 A. Typical powers used

were in the range of 20 to 500 milliwatts. The scattered

light was collected at 900 to the incident radiation and

was analyzed through a Spex 14018 double monochromator,

followed by a RCA C31034 photomultiplier tube, Spex photon

counting electronics and a Linear Instruments 9011 recorder.

For determination of conformational energies, the

integrated areas of the peaks were calculated by multiplying

the height by the width at half height.

For hydrogen bonding studies, the Raman spectra in the

O-H valence region (3000-3700 cm 1 ) were acquired automa-

tically by an interfaced microprocessor (APPLE II+). In

this region, the two peaks were overlapping, and they were

resolved by a non-linear curve fitting procedure (4,5),

outlined in the Appendix, prior to area ratio calculations.

CHAPTER BIBLIOGRAPHY

1. Mizushima, S., Structure of Molecules and InternalRotation, New York, Academic Press, 1954.

2. Orville-Thomas, W. J., Internal Rotation in Molecules,New York, John Wiley and Sons, 1974.

3. Barnes, A. J., and Orville-Thomas, W. J., VibrationalSpectroscopy Modern Trends, New York, Elsevier,1977.

4. Horak, M., and Vitek, A., Interpretation and Processingof Vibrational Spectra, New York, John Wiley andSons, 1978.

5. Maddams, W. F., Appl. Spectrosc. 34, (1980), 245.

5

CHAPTER II

CONFORMATIONAL ENERGIES OF 2-CHLOROETHANOL AND

ETHYLENE GLYCOL IN HYDROGEN BONDING SOLVENTS

Introduction

Ethylene glycol and 2-chloroethanol differ from the

simple alcohols in their capacity to form intramolecular

hydrogen bonds; it is this common feature that is believed

responsible for their similar behavior in dilute carbon

tetrachloride solution (1) as well as in the vapor phase

(2,3). In both cases, the gauche molecular conformation is

observed to predominate, stabilized by this form of hydrogen

bonding.

Quantitative studies on the structural equilibrium of

ethylene glycol revealed the trans conformer to be greatly

stabilized by dilution in dimethylsulfoxide (4,5), a Lewis

base solvent. In contrast, a Raman spectrum of this mole-

cule in aqueous solution (4) appeared to indicate no signi-

ficant shift in the gauche-trans equilibrium; however,

quantitative experiments were not performed.

In order to determine whether the above mentioned

similarity between the two molecules exists in active,

hydrogen bonding solvents, the conformational energy

6

7

difference (AH=HT-HG) has been determined for 2-chloroethanol

at two concentrations in DMSO, and for both compounds in

dilute aqueous solution.

Experimental

Energy differences were calculated from the Van't Hoff

equation, utilizing 14 to 18 intensity (area) ratios mea-

sured for each sample at various temperatures throughout

the liquid range (typically from -100 C to +900 C).

Area ratios (AT/AG), for 2-chloroethanol (10 mole

percent) in water were obtained from the bands (7) at

750 cm~1 (trans C-Cl stretch) and 660 cm~1 (gauche C-Cl

stretch). Interference from solvent vibrations prohibited

the use of these bands in DMSO. However, it was found that,

in DMSO-d6 , area measurements on the C-C-0 bending modes

(7) at 395 cm~1 (trans) and 475 cm~1 (gauche) could be

obtained down to moderate concentration. The spectroscopic

parameters and area ratios are listed in Tables I to IV.

For ethylene glycol, the study of direct area ratios

(AT/AG) of the two conformers was limited due to solvent

interaction. However, the equilibrium constant (2XT/XG

was obtained from the bands (4) at 865 cm (C-C stretching

of both conformers) and 480 cm-1 (gauche, symmetric C-C-0

bending). At sufficiently low concentration in solvent,

the integrated intensity ratio, (A4 8 0/A86 5 ), approaches a

Y) (N O o % H ' LO H H(N qzr 0)(N HO Y 0000o

4H H*1 r4, IHHr4

o m r H r-001 m l (N ON M L)

r- r- r, r-01 kD N q:

H Ho o' e N H yH H I I

8

,40z

Pipq0

P

0

CN

z

O0

U)0

H-

E-1

4 U

ior4

OrdLH

rj

c4-)u

OH A0U

rd u

4

r.4 -)

z C)

y I)

rd UO4H

-I ci

w r-

H (I

UC)o

H- eV o C% tO tor o NO oq 0 0 r 0 -0 0

V LO Iu L IC) N N Om

N Ns Ce H oh w H- CO COS 0 0 0 - 0 . 0 0

rr) C0) CC m C?) N qr H qi'

N% CN N CN N N N N N

E-4

z

0z

0

0

P

zP4

U

E-Q

P4

0

P40

U2

U

H

P4

.

0P4

H

P4U)

co N N r-m LO

N N H H Hr-r- N

( co r% N CO 'ONll2 qq qcv N N 0 c o o

00HH40

0)N0N

rN D L0 )C 0 CO H COm) CO 0% L OmC N N H-

0- 1

C ~

SCH

9

H

-H

p4E-p

o C) C 0 0 0 0 0 0o o oo *- m o r,- r. -CO C LO N C) Co Hs Ns Nsco co rs o -i rsrHt

H

U

1d

H

~0-ICIC

'jx

H

H

0)

C-)

Cd

*-)

0-

i . . . - 4

0)-

E-

H

O4J

4JCd

rO

H

H-r--

0)a)

4

4-)(d

EaH0)41

-- -

.-

Cm n k. N m r-I r-I )mL0 Lf) CO N '.0 H (N O'I 00

4c4 o N (N (N N N H Hr-

- - 4H

(N '.Jo CC) 0 'V0 (N '0

o m c O O ) CO CONN H (N H H H H Hr- H

0U)

0

H

04

0

z

0

C)

0

LO

E-1U)

14

04

0ur4)0P4E-

u

H

H ( C) O CO CO CO

H -L N N r- H Lo LS r-I r- r- N N N (N

1

LO)

'.0L0

C- r-4 C4 r- CH H H riH

) C) Ci L0 C)

C O 10 l C) .0.0 .0 0lo '. .0

HH1

N 0'. .0

(N

r-")

0

--- 1- 1

O - Co oC >jI '.0 (N '.0O(n) 00 LO " (N - (N

I I

10

S4 )

C0

H

U'--'K

rI.

'Hr-

HH

F4

E-

4. 4-1

LO

,Cd

r10

C')

E-i

-,

x

H

H

E0

Q

H

faI

4-)

L

4J4J

4-)

4-)

H

i - I-V

>14

0'H41)

CE-i54

_____ "I

0U)

:iCl

zH

t4

0

0

P4

0

CN

E-

z

P4

0

1%0

0

0uU)0H

U)

H-

C,)

'K

x

H

NLC LO N O WCWO .

OC'n O M 00C0 N 00 e0 m 0m0 0 0 0 0H H o'- o 0 I H H

(N H H H (N HH H H

co ~ L H mL m m 0 0 mC4 0 0 0C N 0

H H HCN (N N C')CM m

___ __I_ _ _ _

NOH

H

0

'K'K

H

4)

a)O

E-q

LO LOr- H

(NH

Lf LCHH

0 .Ln 04*H- H I

(N

H

000

A _________-4_

LO 00 LOt,0 rl

00 F" a3 L

______ '1~

o 0 0O0 H O N (N U'0 C 0 C) m N H H (

I I

11

H

E-H

H

C)

C)

Cd

0

4-)

H

,C

4-)rco'H

H~H

'K

Na1)

4J

C

4

4-)4-)

Irri

>1

4

1 - -

4100

110r--

12

constant value (0.327) corresponding to a trans mole fraction

of unity (4), represented as:

A480A8 6 5T 0.327

XG T= -XT

and

K = 2XT/Xeq T G.

The area ratios and corresponding equilibrium constant are

listed in Table V.

Results and Discussion

The linear temperature dependence of the AT/AG ratio

for 2-chloroethanol in the neat liquid and at various concen-

trations in the two solvents, is displayed in Figure 1. The

calculated gauche-trans energy differences for ethylene

glycol and 2-chloroethanol are presented in Table VI.

Earlier studies (4,5) show that dilution of ethylene

glycol in DMSO has a most striking effect on the skeletal

equilibrium, stabilizing the trans conformer (or, alterna-

tively,- destabilizing the gauche form) by an amount greater

than 1.5 kcal/mole at only moderate dilution. The precise

opposite trend is observed in 2-chloroethanol, where dilution

in the same solvent actually induces a significant enhance-

ment of the relative stability of the gauche rotamer.

00 o 191, ) (

ON H H H H H

O 1000 0 m0 N%-0e LO MCIm n m10 0 OOCHHHHNH H H

0000000 o

E-1HwUP4

0

-40

HZOr4

E- Z

2H

zO

H 4

z

w

HE-

CI

E-

13

N oCO c\ H H

E-

E-H

4I-)

Co4-)

0 xUN

PH

PH

'H 0

01

0'H 10)(dO(d O

1) 00N 4t

0Go

-E-4Q-)H

14

Fig. 1--Temperature dependence of area ratios forneat 2-chloroethanol (squares), 40% in DMSO(open circles), 65% in DMSO (solid circles),and 10% in water (triangles).

15

1-0 1

03

10

3.6 4.6

10OO/T

0.0

40AP --,zo

-1-.o

-20

2.6: 1I a A

(k~ 1)

16

r-i0 ro ro o o

W 0 0 0 0C:) C)I 0 C

4)o 0 0 00

0 N +1 -H +1 +1

0 d -00 HLOr-O N 0 0 N

0 H H H

E1 r-V

m 00 C) ( D L0

UI+ .1 +1 + +1

0- -

H -r-I e00e

U>4<H H

u4 4 + + 1 +l +0 T0

E- 0 ri

O-O CtI N r- LIO

> 000 0 . 0. 0 0-0 C4 f4 P4P 4P H

v ++ I I 0

z +z a4 )r

[4 4

E-t a oa 0

D> c O O4-44-)4-4---

Hr ' o O o o~ O,

-a -r-* -i-

U) U 0\0W\0 0\0U)C01 LC)rnz -o l~

17

In aqueous solution, a marked difference in the behavior

of the two alcohols is observed. The conformational energy

difference of aqueous ethylene glycol is of the same magni-

tude as that in the neat liquid. (The small variation lies

within the range of experimental error in the measurements).

2-Chloroethanol, on the other hand, exhibits a 4 to 5 fold

increase in AH relative to the liquid phase value.

Although the complexity of these strongly hydrogen

bonded systems precludes any unique interpretation of the

above results, it is possible to furnish a tentative expla-

nation of the behavior of ethylene glycol in the two solvents.

In DMSO, the formation of two intermolecular hydrogen bonds

with solvent molecules is the likely cause for the shift in

equilibrium to the trans conformation (4). By contrast,

since the glycol contains two hydrogen bonding protons and

no hydrophobic groups, the relative invariance of the equili-

brium in water may reflect the ability of these molecules to

form water-like pseudolattice structures in the liquid as

well as in aqueous solution.

The observed stability of the gauche conformer of

2-chloroethanol in both solvents could be concomitant with

an increase in the degree of intramolecular hydrogen bonding,

similar to that observed in solutions of the haloethanols

in carbon tetrachloride (1). To investigate this possibility

further, the room temperature Raman spectrum of 2-chloro-

ethanol in the O-H stretching region (3200-3700 cm) as

18

obtained. In the neat alcohol (Fig. 2), the extremely

broad band centered around 3370 cm~1 is due to intermole-

cularly hydrogen bonded hydroxyl groups, while the narrower

peak in the vicinity of 3580 cm arises from intramolecu-

larly bonded species (8). Any increase in intramolecular

hydrogen bonding would be manifested clearly by a rise in

the intensity of this latter band. However, this peak is

diminished greatly, if present at all, in aqueous solution

(Fig. 3). A similar marked diminution in this band also

is exhibited in solution with DMSO. This indicates that

there is a decrease in intramolecular hydrogen bonding even

though the equilibrium is shifted towards the gauche confor-

mation. Thus, there is no simple relationship between the

conformational equilibrium and the hydrogen bonding properties

of 2-chloroethanol in complex solvents.

The major differences in conformational behavior of

the two alcohols extend also to the intra-intermolecular

hydrogen bonding equilibria. In stark contrast to neat

2-chloroethanol, the high frequency spectrum of ethylene

glycol (Fig. 4) reveals the virtually complete absence of

any intramolecular hydrogen bonds, both in the neat liquid

and in either solvent.

In conclusion, it may be seen that the apparent simi-

larities between 2-chloroethanol and ethylene glycol exhibited

in dilute CC 4 solution and in the vapor phase do not exist

in interactive solvent systems. From this, it may be

19

Fig. 2--High frequency Raman spectrum of2-chloroethanol in the neat liquidphase at room temperature.

20

3600 3400 3200

FREQUENCY (CM 1 )

"now

21

Fig. 3--High frequency Raman spectrum of40% 2-chloroethanol in H2 0.

22

3600 3400 3200

FREQUENCY (CM 1 )

23

Fig. 4--High frequency Raman spectrum ofethylene glycol in the neatliquid phase.

24

3600 3400 3200

FREQUENCY (CM )

25

inferred that differences in the properties of substituted

alcohols such as the two studied here from those of simple

alcohols are due, not directly to intramolecular hydrogen

bond formation, but rather arise from more complex factors.

CHAPTER BIBLIOGRAPHY

1. Krueger, P. J., and Mettee, H. D., J. Mol. Spectry. 18,(1965), 131.

2. Buckley, P., and Giguere, P. A., Can. J. Chem. 45,(1967), 397.

3. Buckley, P., Giguere, P. A., and Schneider, M., Can.J. Chem. 47, (1969), 901.

4. Schwartz, M., Spectrochim. Acta 33A, (1977), 1025.

5. Pruettiangkura, P., Ho, S., and Schwartz, M., Spectrosc.Letters 12, (1979), 679.

6. Davenport, D., and Schwartz, M., J. Molec. Struct. 1,(1978), 259.

7. Wyn-Jones, E., and Orville-Thomas, W. J., J. Molec.Struct. 1, (1967), 79.

8. Gupta, A., Davenport, D., and Schwartz, M., Spectrochim.Acta 36A, (1980), 601.

9. Kastha, G. S., Roy, S. B., and Nandy, S. K., Indian J.Phys. 46, (1972), 293.

26

CHAPTER III

CONFORMATIONAL ENERGIES AND HYDROGEN BONDING

IN DICHLOROETHANOL

Introduction

Vibrational spectroscopy has proven to be of consider-

able utility in the study of rotational isomerism about the

carbon-carbon bond in disubstituted ethanes (1,2). One

series of particular interest is the halogenated ethanols

(3-10) since they are among the simplest of molecules with

the capacity to form both intra and intermolecular hydrogen

bonds. Earlier studies have employed Raman spectroscopy to

determine the factors affecting conformational (gauche/trans)

energy differences in the 2-haloethanols (XCH2CH2 OH, X=F,

Cl, Br, I) in the neat liquid (7) and in hydrogen bonding

solvents (9), and also to analyze the inter/intramolecular

hydrogen bonding .equilibrium in these systems (10).

Analogous results have been obtained with 2,2-dichloro-

ethanol (DCE). Like 2-chloroethanol, this molecule has

three equilibrium skeletal conformations, of which two are

spectroscopically equilvalent (11):

27

28

OH

C1 C CI CI C C

OH HO

Gauche (G) Gauche (G) Gauche' (G')

However, in contrast to the former alcohol, the hydroxyl

group in dichloroethanol is always gauche to at least one

halogen atom. Another factor expected to influence the

conformational and hydrogen bonding' behavior in this mole-

cule is the increased interelectronic repulsion between the

oxygen and the two halogen atoms on the molecular skeleton.

Experimental

The calculation of conformational energy differences

from the Van't Hoff equation requires measurement of the

temperature dependence of relative intensities of modes

assigned to the two rotamers. In the neat liquid and in

aqueous solution, the 570 cm~1 C-C-0 deformation (G') and

790 cm~1 C-Cl asymmetric stretch (G) were used (6). This

latter band was obscured by solvent interference in DMSO.

Therefore, the 420 cm- C-C-0 deformation (6) was used in

its place in these solutions (12). As in 2-chloroethanol,

the presence of nearby DMSO vibrations prohibited accurate

29

measurements at concentrations below approximately 40 mole

percent.

The high frequency 0-H valence region in the Raman

spectrum of DCE is qualitatively similar to that of the

monohalogenated alcohols (10), containing a very broad band

at approximately 3400 cm and a higher frequency, narrower

peak, assigned to the O-H stretching modes of the inter-

and intramolecularly hydrogen bonded species, respectively.

In addition, there is a small residual intensity at the

lower frequencies (3000-3100 cm1 ), arising from the C-H

scattering modes. There was no evidence for non-hydrogen

bonded 0-H groups at any of the temperatures studied.

By the method described in the Appendix, a non-linear

least-squares curve fitting algorithm was used (13) to

resolve the band parameters of the inter- and intramolecu-

larly hydrogen bonded hydroxyl modes. Fits performed both

with and without convolution to the experimental slit

function gave essentially the same linewidths (14). The

experimental and theoretical spectra at two typical temper-

atures are displayed in Fig. 5 and 6. The numerical values

of the calculated band parameters in DCE at the various

temperatures are contained in Table VII.

Results and Discussion

The calculated conformational energy differences for

DCE and 2-chloroethanol (7,9) are contained in Table VIII.

A precise comparison of the two sets of results is, of

30

Fig. 5--High frequency Raman spectrum of DCE at-140 C; circles--experimental spectrum;solid line--theoretical curve.

31

Cc

IJ

3050 3240 3430 3620 381[FREQUENCY (CM-1)

32

Fig. 6--High frequency Raman spectrum of DCE at+470 C; circles--experimental spectrum;solid line--theoretical curve.

33

S

co

cE

-U):nLLJ

3050 3240 3430 3620 3810FREQUENCY (CM-1)

(N 0) C CN C) N O LL iOH r- H H 0 0 0 0

0 C; 0 0 0 0 0 0

-Hk.0 L LO o t o Li -O) N

S N (N ( (N (N ( N

tC

+i1C) N N C f-CY O V- N ci 0iLi L L ) % D O %O L C)- H H H H-r H-i H-A H r-

U)

0

z

H

E-)

E-i

x

0

E-i

U)

E-4

L4

0

H

A4

0

U)0

O4

E-uw)04Hn

(N

LO(Y

C)

+1ko

R:34(Y

N " 'rV C H ' (N) 00 00 00 .00- N N

Li L Lin LOO L ifn Lm Ci cY io cq Y

re Y H 0- to(N 00 N ' Y 0 mC4

Co4 m Ci ci c( )c (Ni ci i ci ci ci c

r- % 11V o r-- c o o) 0 or c00Hn O t -r N r- 0 m0O

I I

34

0)

dH.

Cd4

1H

H

H4

-H

C)(Y)+1

LO(Y

,.QH-

C)

Cd

03

Hj

H

0i)

H

C

4

40)

H

0)-4

4.3Cd

0)oH2

0

4-1

0

4r4

-)

(d

ri

rdro

o

4-Xor4-0)

~4H -H m

Cd-) 0

c ,. -H

-r-

x r4 r-IM ,.4 00)rd r- p H:1a) -H HP4 '-

-r- i -

.. I i

i

i

35

r-)

H HH C

o0 H0

CD 0CC)

C C) CD

CO+0 1 - 0 0

H ti o o o

z <r-4 r-

I H HN N

0000

N + 1 +

U4

zH U 0

H CHHH ,.C C

H 0 -iJ M> 0\0

C0 0 0 C- qO co

E-' 0 < +1 +1 -HI 00 H LOnz (N 0 (N

0 - - .U) 0 H H

H + + +

rd O 4-1 4-4 -r-

o (Ncr-)

36

course, not possible since DCE contains no skeletal confi-

guration directly analogous to the trans rotamer of the

former molecule (in which no intramolecular hydrogen bond

formation is possible). However, the molecules in the neat

liquid phase exhibit energy differences that are of the same

magnitude, but somewhat lower than typical values obtained

in other, non-hydrogen bonded substituted ethanes (1,2).

In solution, on the other hand, quite striking differences

between the two alcohols are observed. Whereas the gauche

form in chloroethanol is greatly stabilized (by approximately

a factor of four) in both DMSO and H20, the energy differ-

ences between the G and G' configuration remain low in DCE.

Although no unambiguous interpretation is possible at

present in these complex systems (10,15), one may still

offer a tentative explanation of the above results. It is

not unreasonable to speculate that the increased energy

differences of chloroethanol arise, at least in part, from

dielectric stabilization of the more polar gauche form in

these two solvents (E:(H20)=78.5 (16) and s(DMSO)=44.7 (17)

at 250C). One would at first expect a similar stabilization

of the G' conformation of DCE, which calculations have shown

to have a much greater dipole moment than the G rotamer (6).

However, it should be noted from the high frequency spectra

(18) that the hydrogen bonding in this alcohol in both sol-

vents is almost exclusively intermolecular in nature. As

pictured below, this would result in greater halogen-oxygen

37

repulsion in the G' configuration.

Cl 0- H -- O H O-H--OC1_ __C1I, \

/ H NH

H H Cl H

Gauche' (0') Gauche (G)

Thus, the relative insensitivity of AH to dilution in the

two solvents may result from a competition between dielectric

stabilization and interelectronic repulsion.

Comparisons of the temperature dependence of the cal-

culated peak frequencies and linewidths of the inter- and

intramolecularly hydrogen bonded 0-H modes in DCE and

chloroethanol (10) are displayed in Fig. 7 and 8, respec-

tively. The results are almost identical in the two alcohols.

Indeed, the only significant difference is that the line-

width of the intramolecularly bonded species are slightly

lower in DCE. It is interesting to note that the same phe-

nomenon was observed in early investigations of both alcohols

at low concentrations in CCl4 (3).

In contrast to the above results, one does observe

differences in the band area ratios between the two alcohols.

As shown in Fig. 9, 1I /I. is noticeably greater inintra inter

DCE than in chloroethanol. Furthermore, the calculated value

of AH(=Hintrainter) was found to be 2.4+0.13 kcal/mole,

38

Fig. 7--Temperature dependence of peak frequenciesof DCE (solid circles) and chloroethanol(open circles) .

39,

3580 -

INTRA

-F 3480:E

33800'0

3280 INTER

-30 0 30 60 90

Temperature (*C)

40

Fig. 8--Temperature dependence of line widths ofDCE (solid circles) and chloroethanol(open circles).

41

32 1

26

am"0O

201

1801

1601

14O

30

Temperature

0*

INTRA

INTER

3Oa-30 0 60

(0c)

42

Fig. 9--Temperature dependence of relative scatteringintensities in DCE (solid circles) andchloroethanol (open circles) .

43

1-

0

0 0

3.0 3.5 4.0

1000/'T K

-2.0

z-i

-3.01-

-4.0 F

I1'. - I I I a

(K~ 1 )

44

compared to 2.9+0.3 kcal/mole in chloroethanol (10).

The overall enthalpy change should, in principle,

correspond to the difference between the heats of formation

of the inter- and intramolecular hydrogen bonds. If it is

assumed that the intermolecular hydrogen bonds in these

molecules are formed primarily through the oxygen atom, AHF

(inter) should remain approximately constant for halogenated

ethanol (19), with a value on the order of -5 kcal/mole

(reported for ethanol itself). Given this assumption, the

above results imply that the intramolecular hydrogen bond

strength is somewhat greater in DCE than in chloroethanol.

It is satisfying to note that this conclusion is consistent

with the results obtained in earlier investigations of the

two alcohols in dilute solution (3).

CHAPTER BIBLIOGRAPHY

1. Mizushima, S.,, Structure of Molecules and InternalRotation, New York, Academic Press, 1954.

2. Orville-Thomas, W. J. , Internal Rotation in Molecules,New York, John Wiley and Sons, 1974.

3. Krueger, P. J., and Mettee, H. D., Can. J. Chem. 42,(1964), 326.

4. Wyn-Jones, E., and Orville-Thomas,, W. J. , J. Mole'c.Struct. 1, (1967), 79.

5. Kastha, G. S., Roy, S. B., and Nandy, S. K., IndianJ.,Phys.6, (1972), 293.

6. Hirokawa, T., Sumida, Y., Hayashi, M., and Murata, H.,J.;Sci. Hiroshima Univ. '38, (1974), 281.

7. Davenport, D., and Schwartz, M., J. Molec. Struct. 50,(1978), 259.

8. Pertilla, M., Spectrochim. Acta 35A, (1979) , 37.

9. Maleknia, S., Friedman, B. Ri., Abedi, N., andSchwartz, M., Spectrosc. Letters 13, (1980), 777.

10. Gupta, A., Davenport, D., and Schwartz,, M., Spectrochim.Acta 36A, (1980) , 601.

11. The conformation termed G' in this paper has alterna-tively been labelled as the trans configurationin some earlier work (ref. 6) .

12. Determination of AH in the neat liquid utilizingvarious pairs of band intensities, as expected,yielded values that were the same to withinexperimental error.

13. Maddams, W. F., Appl. Spec'trosc. 34, (1980) , 245.

14. Savitzky, A., and Golay, M. J. E.,' Anal. Chem. _36,(1964) , 1627.

45

46

15. Schwartz, M., Spectrochin. Acta 33A, (1977), 1025.

16. Table of Dielectric Constants of Pure Liquids, NBSCircular 514, (1951).

17. Johnston, M. D., Jr., and Barfield, M. , J. Chem.Phys. 54, (1971), 3083.

18. Although not pictured, Raman spectra in the 0-H valenceregion of DCE dissolved in both solvents showthe virtually complete absence of the bandassigned to intramolecularly hydrogen bondedhydroxyl groups. This behavior was also observedfor 2-chloroethanol in the same solvents (Fig. 1of ref. 9).

19. This assumption is felt to be reasonable, sincevarious thermodynamic studies have shown thatintermolecular hydrogen bonds of alcohols withalkyl halides are significantly weaker than withethers; e.g. West, R., Powell, D. L., Whatley,L. S., Lee, M. K. T., and Schleyer, P. Von R.,J. Am. Chem. Soc. 84, (1962) , 3221.

APPENDIX

Characteristic to the analysis of high frequency Raman

spectral data is the problem of resolving overlapping bands

into individual components. In these investigations, this

difficulty was overcome by using the computer program T3BW

(1), written specifically to input digitized spectra and

output resolved peaks derived from a non-linear least-squares

algorithm.

The high frequency Raman spectra of halogenated ethanols

contains two bands due to inter- and intramolecular hydrogen

bonding. The former is a very broad band centered around

3400 cm and the latter is a narrow band at approximately

3600 cm . In addition, there exists a contribution from

the residual intensity arising from the C-H stretch, cen-

tered at 3000 cm~. This residual intensity was subtracted

from the intensity of hydrogen bonding peaks using a simple

exponential function of the form:

Residual = wA * EXP(-wE * (w-o ))

where Iresidual is the intensity at the frequency w, wow is

the center frequency of the C-H stretch, wA is the peak

intensity, and wE is an exponential parameter. The spectral

47

48

bands due to inter- and intramolecular hydrogen bonding

were assumed to be a combination of Gaussian and Lorentzian

lineshapes described by:

(3-Lk 2-LN2 ()

I = Ae DGaussian

-A

Lorentzian _+(' )1+ (W-WQ)

D

where wO is the center frequency, A is the height, and D is

the half width at half height of the bands. In these parti-

cular studies a pure Gaussian lineshape was found to be an

optimum model for the hydrogen bonding bands. The computer

program then adjusts the exponentail function for the

residual C-H stretch intensity and the Gaussian lineshape

for the unresolved bands until the sum of the three model

functions approximate the experimental data to the desired

degree of accuracy. The resulting functions are inspected

to obtain the height, width, and frequency for the hydrogen

bonding bands.

The input data set for this computer program consists

of the following:

1) 201 equally spaced intensity values representing the

experimental high frequency Raman spectrum.

2) The starting frequency and the frequency increment

of the spectrum.

49

3) Initial estimates of woo, wA, wE, and wo, A, and D

for each of the bands.

4) Mask array, specifying which parameters should be

initially held constant.

The estimates for wo, A, and D are obtained by inspec-

tion of the experimental spectrum. The center frequency

(wow) for these studies was estimated at 3700 cm . However,

the results were found to be insensitive to the precise

origin of the exponential function. Initial estimates for

wA and wE were determined from the first two points of the

recorded spectrum using the formulae:

A = yl e lE (wow-OW)

wE = LN (y2/yl)

where wo1 and wo2 are the frequency of the points; y1 and Y2

are the corresponding intensities. In the case where the

initial estimates deviate greatly from the final values,

the parameters for the narrow peak may be greatly in error

due to its small size relative to the neighboring band.

These errors are minimized by initially fixing the narrow

band parameters using the mask array.

The output of the program is the height, width, and

center frequency of the resolved bands, and the parameters

of the residual intensity due to the C-H stretch.

APPENDIX BIBLIOGRAPHY

1. Burrill, J. H., Jr., Progran ' 66, Quan tum Chemi'strProgram Exchange, University of Indiana,Bloomington, Indiana.

50

BIBLIOGRAPHY

Books

Barnes, A. J., Vibrational Spectroscopy Modern Trends,New York, Elsevier, 1977.

Horak, M., Interpretation and Processing of VibrationalSpectra, New York, John Wiley and Sons, 1978.

Mizushima, S., Structure of Molecules and Internal Rotation,New York, Academic Press, 1954.

Orville-Thomas, W. J., Internal Rotation in Molecules,New York, John Wiley and Sons, 1974.

Articles

Buckley, P., and Giguere, P. A., Can. J._ Chem. 45, (1967) ,397.

Buckley, P., Giguere, P. A., and Schneider, M., Can. J.Ch'em. 47, (1969) , 901.

Davenport, D., and Schwartz, M., J. Molec. Struct. 50,(1978) , 259.

Gupta, A., Davenport, D., and Schwartz, M., Spectrochim.Acta 36A, (1980) , 601.

Hirokawa, T., Sumida, Y., Hayashi, M., and Murata, H.,J. Sci. Hiroshima Univ. 38, (1974) , 281.

Johnston, M. D., and Barfield, M., J. Chem. Phys. 54,(1971) , 3083.

Kastha, G. S., Roy, S. B., and Nandy, S. K., Indian J.Phys. 46, (1972) , 293.

Krueger, P. J., and Mettee, H. D., Can. J. Chem. 42,(1964) , 326.

Maddams, W. F., Appl. Spectrosc. 34, (1980), 245.

51

52

Maleknia, S., Friedman, B. Ri., Abedi, N., and Schwartz, M.,Spectrosc. Letters 13, (1980) , 777.

Perttila, M., Spectrochim. Acta 35A, (1979), 37.

Pruettiangkura, P., Ho, S., and Schwartz, M., Spectrosc.Letters 12, (1979) , 679.

Savitzky, A., and Golay, M. J. E., Anal. Chem.' 36, (1964) ,1627.

Schwartz, M., Spectrochim. Acta 33A, (1977), 1025.

Wyn-Jones, E., and Orville-Thomas, W. J., J. Molec. Struct.

1, (1967), 79.

Reports

Burrill, J. H., Jr., Program 66, Quantum Chemistry ProgramExchange, University of Indiana, Bloomington, Indiana.