Louisiana State UniversityLSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1965

An Investigation of the Electronic Spectra of aSeries of Organic Sulfur Compounds.Samuel Donald ThompsonLouisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion inLSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contactgradetd@lsu.edu.

Recommended CitationThompson, Samuel Donald, "An Investigation of the Electronic Spectra of a Series of Organic Sulfur Compounds." (1965). LSUHistorical Dissertations and Theses. 1095.https://digitalcommons.lsu.edu/gradschool_disstheses/1095

This dissertation has been microfilmed exactly as received 6 6 -7 5 1

TH O M PSO N , S a m u e l D on a ld , 1 9 3 5 - A N IN V ESTIG A TIO N O F THE E L E C T R O N IC S P E C T R A O F A SE R IE S O F ORGANIC SU L F U R C O M PO U N D S.

L o u is ia n a S ta te U n iv e r s ity , P h .D ., 1965 C h e m is tr y , p h y s ic a l

U niversity M icrofilm s, Inc., A nn Arbor, M ichigan

AN INVESTIGATION OF THE ELECTRONIC SPECTRA OF A SERIES OF ORGANIC SULFUR COMPOUNDS

A Thesis

Submitted to the Graduate Faculty of the Louisiana State University and

Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in

The Department of Chemistry

bySamuel Donald Thompson

B.S., Little Rock University, 1960 August, 1965

ACKNOWLEDGMENT

The author would like to acknowledge the aid and encouragement

of his major professor Dr, S. P. McGlynn, the many fruitful discussions

with Dr. D. G. Carroll and the assistance of A. T. Armstrong on computer

work. The author would especially like to thank Jacqueline Thompson

for her many years of patience and understanding.

Financial support is gratefully acknowledged from The United

States Atomic Energy Commission - Biology Branch and the Dr. Charles

E. Coates Memorial Fund of the L. S. U. Foundation donated by George

H. Coates.

ii

I

TABLE OF CONTENTSPAGE

ACKNOWLEDGMENT .......................................... ii

LIST OF F I G U R E S ......................................... v

LIST OF TABLES ........................................ viii

ABSTRACT .............................................. ix

CHAPTER

I. INTRODUCTION .................................... 1

II. EXPERIMENTAL .................................... 4

A. Instrumentation and Technique . . . . . . 4

B. Chemicals . . . . . . . ............... 6

III. THEORETICAL .................................... 8

IV. RESULTS AND DISCUSSION.......................... 18

A. General.................................. 18

B. Hydrogen Sulfide ....................... 19

C. Methyl S u l f i d e ......................... 26

D. Methyl Mercaptan ....................... 32

E. Methyl Disulfide ....................... 35

F. Methyl Trisulfide ..................... 40

G. Effect of Variation of EffectiveCharges ( Z ) ........................... 40

V. FREE ELECTRON TREATMENT OF POLYSULFIDE SERIES 51

VI. SOLUTION STUDIES ON THE SERIES (CH3 )2Sn . . . . 54

VII. INVESTIGATION OF THE MECHANISM OF PHOTOLYSISOF CYCLIC DISULFIDES ......................... 67

iii

iv

CHAPTER PAGE

A. Introduction .......................... 67

B. Results ................................ 68

1 . Tetramethylene disulfide ........ 68

2, Pentamethylene disulfide ........ 71

C. Discussion ..................... 73

VIII. CONCLUSIONS......................................... 81

SELECTED BIBLIOGRAPHY........................................ 83

VITA ........................................................ 85

LIST OF FIGURES

FIGURE PAGE

1. Definition of Valence State Ionisation Potential . . 9

2. Coordinates of Hydrogen Sulfide, MethylMercaptan and Methyl Sulfide .......... . . . . . 14

3. Coordinates of Methyl Disulfide . IS

4. Coordinates of Methyl Trisulflde ................... 1635. Vapor Spectrum of Hydrogen Sulfide ................ 20

6 . Structure of the Hydrogen Sulfide 1965A BandObtained at - 8 0 ° C ............................... 21

7. Structure of the Deuterium Sulfide 1980A BandObtained at - 8 0 ° C ............................... 24

8 . Vapor Spectrum of Methyl Sulfide .......... . 27

9. Solution Spectrum of the-— 2350A Inflectionof Methyl Sulfide ......................... 28

310. Vapor Spectrum of Methyl Mercaptan ............... 34

11. Vapor Spectrum of Methyl Disulfide ............... 37

12. Splitting of 3d (Z»0) levels in(CH3 )2S2 ® -104° (e.v.) . . . . 39

13. Vapor Spectrum of Methyl Trisulfide ............... 41

14. Vapor Spectrum of Methyl Tetrasulfide ............. 44

15. Vapor Spectra of (CH^^S, (CH2 )2 S2 » (CH3 )2S3and (CH3 )2S4 45

16. Energy Level Diagram for Methyl Disulfide(9«104°) with Varying 3dg and 4sg EffectiveC h a r g e s .......................................... 46

v

vl

GE

48

49

55

57

60

61

62

69

69

69

69

69

69

72

72

7272

72

Splitting of the 4s Levels in (CH-)

<CH3>2S3 ....................................Energy Level Diagram for H^S, CH^SH,

(CH3)2S, <CH3)2S2 and ( C H ^ ^ withZ_, = 1.0 and Z. = 0.9 .......................s s

Solution Spectra of (CH3>2S .....................Solution Spectra of (CH3)2S2 .....................Solution Spectra of (CH3 )2S3 .....................Solution Spectra of ( C H ^ ^ S ^ .....................Long Wavelength Absorption of (CH3 >2 S^ ..........A: Absorption Spectrum of the Vapors Above Liquid

1,4-butanedithiol which had been Photolysed for 2 hours with an Unfiltered AH-6 Source . .

B: Absorption Spectrum of the Vapors above LiquidTetramethylene Disulfide Doped with NaOH which had been Photolysed for 45 Minutes with Filtered AH- 6 Source..........................

C: Absorption Spectrum of the Vapors above LiquidTetramethylene Disulfide which had been Photolysed for 1 Hour with Filtered AH- 6 Source ........................................

D: Vapor Spectrum of 1,4-butanedithiol ........E: Vapor Spectrum of Tetramethylene Sulfide . . .F: Vapor Spectrum of Tetramethylene Disulfide . .A: Absorption Spectrum of Vapors above Liquid

Pentamethylene Disulfide which had been Photolysed for 30 Minutes with Filtered AH- 6 Source....................................

B: Absorption Spectrum of Vapors above LiquidPentamethylene Disulfide Doped with NaOH which had been Photolysed for 30 Minutes withFiltered AH- 6 Source .........................

C: Vapor Spectrum of Pentamethylene Sulfide . . .D: Vapor Spectrum of Pentamethylene Disulfide . .

FIGURE

26. A

B

C

vii

PAGE

: Schematic One-Electron MO Energy Diagram ofa “C^-S^-Sg-Cg- System. 3pg Indicates Energy of the 3p Atomic Orbital of Sulfur. The Right Subscripts on the Carbons and Sul furs are Position-Ordering Indices ................... , 75

: Schematic One-Electron MO Energy Diagram ofA D i t h i o l ...................................... 75

; Schematic One-Electron MO Energy Diagram of an Anionic Species in which a Negative Charge is Supposed to Reside on One Sulfur. The 3pz orbital is Perpendicular to a Given C-SH Plane, while the 3p Orbital is in that Plane.As the Exact Location of the 3p Level isSomewhat Questionable, It is Shown in BothPossible Positions ............................. 75

LIST OF TABLES

TABLE

I. Coulomb Integrals (e.v.) ...........................

II, H2 S Transition Energies (e.v,) .....................

III. H2S and D2S Vibrational Frequencies (cm-^) ........

IV, HoS and D9S Fundamental Vibrational Frequencies2( c m - l ) .................................................................................................

V. Results of H2S Computations (Z3J - l'°S Z4sc-°*9 )- • •s sVI. (CH3 )2S Transition Energies (e.v.) .................

VII. Results of (CH3 >2 S Computations (Z jj "1.0; Z^s *0.9).S S

VIII. Vibrational Assignments in the (CH^^S 2150-2300and 1850-1960A bands . .......... ..............

IX. Comparison of the Vibrational Frequencies inthe (CH^^S 5.39 and 6.33 e.v. Bands . . . . . . .

X. CH3SH Transition Energies (e.v.) ...................

XI. Results of CH-SH Computations (Z3J -1.0; Z^ 8 -0.9)

XII. (CH^)^ Transition Energies (e.v.) . . . . . . . . .

XIII. Results of (0113)282 0 ■ 104° Computations

^Z3ds“1,0; Z4ss“0*9 .............................

XIV. (CH-j)2S3 Transition Energies (e.v.)... ..............

XV. Results of cis (CH^^S^ Computations (Z^g -0.9) . . .S

XVI. Results of Free Electron Treatment ofPolysulfide Series (cm”l) ..................... . .

XVII. Solution Spectra of the Series (CH3 )2Sn ............

PAGE

11

2222

22

23

26

29

30

31

32

33

35

36

42

43

53

63

viii

ABSTRACT

Although many excellent investigations have been performed

on n-electron molecules, relatively little has been done with satu­

rated systems particularly in regard to theoretical treatments of the

low energy electronic transitions. It was the purpose of this study

to experimentally determine the energy of these first few excitations

in a series of saturated sulfur compounds, gather empirical data

that would aid in their assignments and obtain a theoretical treat­

ment that would predict the energies and nature of the transitions.

To aid in the correlation of data the series of compounds

selected for investigation were chosen to fall into two groups: in

the first group the attached substituents were varied; in the second

group the number of sulfur atoms in the chain was varied. The empiri­

cal data collected included ultraviolet vapor spectra, solvent effects

on spectra and vibrational data.

Computations were performed on all compounds following an

approach closely related to that of Wolfsberg and Helmholz. Addition­

ally, a free electron treatment was applied to predict the energies of

the long wavelength transition in polysulfides.

The Wolfsberg-Helmholz type treatment closely approximates

the experimentally observed energies in the compounds containing

a single sulfur atom but encounters difficulties when two or more

sulfurs are present due to large predicted 3dg-3dg interactions.

ix

However, treatment of the long wavelength absorption in the

polysulfides as an n -• 4sg transition by the free electron method

yields good results. The calculations Indicate that the first several

transitions in all of the compounds investigated are to low lying

perturbed Rydberg levels.

A study was conducted into the mode of photolysis of cyclic

disulfides. The results indicate that the mechanism of photolysis

may vary depending upon conditions. Under one set of conditions

the photolysis proceeds by -S-S- scission while under a different

set of conditions the initial step is C-S scission.

To my FatherAn iota of the recognition due in life but which seldom comes until after death.

xi

CHAPTER I

INTRODUCTION

In the last twenty years much theoretical and empirical spectro­

scopic investigation has been conducted into the nature of the electronic

transitions in rr-electron systems. These studies were primarily concerned

with unsaturated hydrocarbons and to a lesser extent with unsaturated

heteroatomlc systems. They have provided most of our present knowledge

of excited electronic states. Very little systematic investigation has

been conducted on saturated systems. The few investigations that have

been performed are not generally of good theoretical caliber; most

authors usually seem satisfied with simple qualitative conclusions.

Several excellent investigations have been conducted on the

Rydberg transitions of systems similar to those investigated here,

but the study of the lower energy excitations of these molecules seems1 2 3to have been particularly neglected. ’ * These latter transitions,

which shall be termed non-Rydberg, Include excitations to sigma anti-

bonding levels and to the first few low energy members of the Rydberg

series. The low energy Rydbergs are placed in this category because

they undergo a sufficient interaction with their molecular environment

to lose a considerable amount of their "atomic" character.

Investigations of systems such as saturated sulfides have been

retarded by the experimental difficulties Inherent in working in the

2

far ultraviolet. Oxygen absorption produces minor problems. Solvents

useful in the visible and near ultraviolet are of no value. Molecular

transitions overlap and merge with Rydbergs and are generally broad

featureless bands. A second factor that has tended to discourage research

on such systems is an acute shortage of basic theoretical tools. Once

a reasonable foundation for calculations has been established, the

number of researchers in a given area tends to Increase rapidly.

In this thesis a systematic spectroscopic study of a number of

simple saturated sulfur compounds is reported. The compounds investi­

gated were selected in such a manner as to exhibit the effects of

varying two parameters. First, hydrogen sulfide, methyl mercaptan and

methyl sulfide demonstrate the effects of varying the substituent attached

to the sulfur. Second, methyl mono-, di-, tri-, and tetrasulfides show

the changes produced as the length of the sulfur chain is increased

and as the effects of the methyl end groups are thereby diminished.

Quantum chemical calculations followed an approach akin to4that of Wolfsberg and Helmholz. In addition to the orbitals which are

normally considered, the 3ds and 4sg atomic orbitals of sulfur have

also been included in an attempt to gain some insight into their role

in bonding and in the low energy electronic transitions.

An alternate approach was taken to the long wavelength tran­

sition in the polysulfides. Several authors have indicated that a

number of similarities exist between sulfur-sulfur chains and conjugated

h y d r o c a r b o n s . ^ In keeping with this thought, free electron-type

calculations were performed on the long wavelength electronic transitions

of dimethyl mono-, di-, tri- and tetrasulfides.

3

A number of experimental parameters are generally used as aids

in assigning electronic transitions. These include emission studies,

polarization studies and so on. The fact that the compounds investigated

here absorb at such high energies in the ultraviolet makes it difficult

to excite them with available sources. They fail to give any emission

at the sensitivities available to us. Thus, the tools at one's

disposal are rather limited. One device utilized here was to study

the spectral shifts Induced by solvents of different polarities. This

approach is of limited value particularly with regard to high energy

transitions due to the transmission limitations of available solvents.

Finally, one of the most relevant questions, at least insofar

as the biochemist is concerned, relates to the photoinduced reactions

of sulfur-sulfur linkages. Consequently an investigation has been

conducted into the mode of photolysis of cyclic disulfides.

CHAPTER II

EXPERIMENTAL

A. Inatrumentation and Technique

All ultraviolet absorption spectra were obtained with a Cary

model 14R Recording Spectrophotometer. For use below 2000A, the

instrument was purged with nitrogen to prevent absorption by oxygen.

Generally, there were large variations of intensity throughout the

spectral regions of interest and since it was less convenient to alter

vapor pressure of the samples, quartz cells of various path lengths

(0.1 to 100 cm.) were used. The 100 cm, gas cell was supplied as a

standard accessory item for the Cary 14. When it was used the normal

procedure followed was to evacuate it until a reasonably straight

baseline was obtainable in the region of oxygen absorption. The sample

was then introduced, a suitable period of time allowed for the vapor

to equilibrate and the spectrum recorded. Subsequently, either the

valve to the sample tube or the vacuum pump was again opened depending

upon whether the next region to be scanned was of higher or lower

intensity. This process was repeated until all areas of interest had

been scanned at reasonably reliable intensities. Unless otherwise

specified, all spectra were determined at ambient temperatures.

The molar extinction coefficients (e) of (CH3 )2 S were calculated

from the equation:"*€ - 760 x 22.4 x DT/273bp (1)

5

where D ■ optical density, T ■ temperature (°K), b ■ path length (cm.),

and p ■ pressure (mm.). The vapor pressure was obtained from the

expression:^

log p - 16.51798 - 1876.37/T - 3.0427 log T (2)oFor I - 293 K, p is found to be 395 mm. No information was available

regarding the vapor pressure of most of the compounds; consequently the

extinction coefficients for their vapor spectra are not presented.

However, a comparison of absorption Intensities may be made using the

e's calculated from the solution studies.

The photolysis studies were conducted by selectively irradiating

approximately ten mg. of sample in the region of its long wavelength

absorption band with a high pressure mercury lamp (G.E. AH-6 ). After

each stage of irradiation, product identification was made by Investi­

gating the short wavelength spectrum of the vapor in equilibrium with

the photolyzed materials. The photolyses were conducted in nitrogen

filled cells to prevent oxygen from interfering by either absorption

or by influencing the course of decomposition. An absorption cell

containing carbon tetrachloride and a Corning glass 7-54 were used

individually or in combination as appropriate filters of incident light.

Possible emission spectra of H2S, (CH^^S, ( ^ 3 ) 2 8 2 and ( C ^ ^ S ^

were sought using an Aminco-Kiers Spectrophosphorimeter. Measurements

were made in E.P.A. solvent (ether:isopentane:ethanol - 5:5:2) at

ambient and liquid nitrogen (77°K) temperatures. The results were

inconclusive. It is possible that weak emissions are present, but more

sensitive apparatus is needed to prove their existence conclusively.

The gas chromatograph used for purification of f<ome compounds

was a Beckman Model GC-2 equipped with the Beckman fraction collector

6

designed for attachment to the GC-2 and commercially available. The

column was packed with forty percent tritolyl phosphate on 42-60 mesh

firebrick; both materials were of commercial quality and were furnished

by Micro-Tek Instruments, Inc., Baton Rouge, La. Repeated Injections

of up to 0.5 ml. of the substance to be purified were made and the

desired component was collected at -76°C utilizing the fraction collector,

B. Chemicals

The methyl sulfide and disulfide utilized were the commercially

available white label grade reagents produced by Eastman Organic Chemicals,

Rochester 3, New York. Both were purified by triple distillation followed

by preparative gas chromatography. H2S and D2S were used as obtained

from the Matheson Company, East Rutherford, New Jersey and from Volk

Radiochemical Company, Skokie, Illinois. The methyl trisulfide was9prepared by the method of Strecker. Sodium polysulfide was heated

with dimethyl sulfate. The resulting mixture was poured over ice and

the heavy yellow oil which separated out was dried over CaCl2 and

distilled at reduced pressure yielding methyl trisulfide. Purification

was accomplished by vacuum sublimation and gas chromatography. Physical

constants were:b 15 - 74-76.8°C, np * 1.6044, d24»1.197.

The procedure of Feher, Krause, and Vogelbruch was utilized

to obtain methyl tetrasulfide.Methyl mercaptan was treated at -60°C

with sulfur monochloride in anhydrous ether. After completion of the

reaction and evaporation of solvent, the residue was repeatedly vacuum

distilled until the final distillation showed a change of only 0 . 0 0 0 2

units in the refractive index. The properties of the methyl tetrasulfide

Analysis: C H SCalcd. 15„19 3.80 81.01Found 15.53 3 81 80.85

Tetramethylene disulphide was prepared by the method of Calvin11et al. and purified by distillation followed by repeated sublimation

ountil a solid of m p. 30-32 was obtained. Pentamethylene disulfide

was obtained as a pale yellow liquid by the oxidation of 1 ,5-pentanedithiol

with an ethereal solution of anhydrous ferric chloride. Repeated

fractionation yielded a practically colorless liquid boiling at 70°/10mm.

All materials were refrigerated until investigated.

The solvents used were Hartman-Leddon Company fluorimetric grade

ethyl alcohol, methyl alcohol, cyclohexane and ultra-pure water. Hexane

was purified by drying over sodium wire, distilling and passing through

a six foot column of activated silica

CHAPTER III

THEORETICAL

The approach followed in the calculation of energy levels embodies

many of the attitudes advanced by Wolfsberg and Helmholz when they first

adapted the methods of molecular orbital calculations to computations4on inorganic complex ions. Basically, the nonvalence shell electrons are

assumed to be unaffected by bonding and, with the nuclei, they are taken

to constitute an effective core field in which the valence or optical

electrons are supposed to move.

Consequently for sulfur it was necessary to consider only the

3s§ and 3pg atomic orbitals (AO's); but the 3dg and 4sg atomic orbitals

were also considered in some computations in order to study the effects

of their inclusion. For hydrogen and carbon the orbitals of interest were

the Is -AO, and the 2s -AO and 2p„ -AO’s, respectively.H L CCoulomb integrals, , were evaluated in terms of neutral atom

valence state ionization potentials (V.S.I.E.), illustrated in Figure 112and defined by Moffittas:

Iv - Ig -f F+ - p0 (3)

where Iv ■ valence state ionization potential; I * ground state ioni-Ozation potential; P+ and PQ = promotional energies associated with the

electron in the ionic and ground states being in a molecular electronic

environment rather than in an atomic environment. V.S.I.E.'s are

calculated in such a manner as to attempt to reproduce the experimentally

8

FIGURE 1

DEFINITION OF VALENCE STATE IONIZATION POTENTIAL

10

observed variation of ionization potential (i.e. coulomb term) with the

charge on the atom in a molecule.

The promotional energies were computed by the method of Viste and

Gray as average energies of the appropriate configurations using spectral13 1Aterm values taken from Moore's tables. * The average was taken over

all of the terms weighted according to the degeneracy of each term.

As an example, the V.S.I.E. of a 3sg electron was calculated as follows:

The V.S.I.E. desired corresponds to the transition 3s^3p4" 3s3p^.

The ground state promotional energy, PQ , is obtained by averaging over2 Aall the possible term values associated with a 3s 3p configuration which

are *S, 3P and 1 D. Since I corresponds to the excitation 38^3?^- 3s2 3p3, P+4is computed as the average of all terms associated with 3s3p . The

latter configuration has the possible term values ^P and 2P.

Averaging 3s2 3p4 term values: 4Averaging 3s3p term values:

Term (cm 1) J x J Term (cm”1) J x J

3P2 0 0 0 . 0 5 0 0 0 . 0 4P X 2h 79394.8 6 476368.80

3Pl 396.8 3 1190.4 79757.9 4 319031.60

3Po 573.6 1 573.6 79968.0 2 159936.00

S 9239.0 5 46195.0 V 105599.02 4 422396.08

lSo 22181.4 __1 22181.4 2P^ 106044.16 2 212088.32

TOTAL 15 70140.4 TOTAL 18 1589820.80

P = E o av - 70140.4/15 - 4676.0 cm"1; P+ - Eav - 1589820.80/18 - 88323.37 cm" 1

From Moore's tables :14 I - 83539. 56cm”1 „ • -Iv - 83539.56 + 88323 .37 - 4676.0 - 167186.93 cm” 1 - 20.72 e. V.

11

The 3pxg orbitals were assumed to be perpendicular to the plane

of the molecule in H2S , CH3SH and (CH^^S thereby inhibiting their

interaction with the remaining orbitals. To correct for this obviously

erroneous situation, the coulomb terms for these highest energy non­

bonding electrons were assumed to be equal to their respective molecular

ionization potentials. The final values obtained for the various

coulomb integrals are presented in Table I.

TABLE I

Coulomb Integrals (e.v.)

ia -Hii ia -Hii3ss 20. 72 lsH 13.53'

3ps 11.61b°iSH

19.04

3pxS (I12S) 10.47b,c *8 . 0 2

3pxS(CH3SH) 9.44b »cspc 9.95

3pxS((CH3 )2S) 8 .6 8 b,d a 3spc * ^

12.40

3ds

4ss

3.67

3.76

a 3spc

7.50

a. Subscripts S,H,C denote respectively sulfur, carbon, hydrogen.

b. Coulomb terms for 3pxg orbitals were assumed equal to molecular ionization potentials in H2S , CH3SH and (CHj)2S otherwise their values would also have been 11.61.

c. Price, W.C., Teegan, J.P., and Walsh, A.D., Proc. Roy. Soc.. 201A.600 (1950).

d. Clark, L.B., Doctoral Thesis, University of Washington, 1963.

e. Handbook of Chemistry and Physics■ Cleveland, Ohio: TheChemical Rubber Publishing Co., 1956-1957 (Thirty-eighth Edition) pp. 2347.

f. Lossing, F. P., Ingold, K. U., and Henderson, I.H.S., J. Chem. Phys.. 22, 621 (1954).

12

Since Mulliken's original suggestion that resonance integrals,

, be taken as proportional to overlaps, S^j, various formulas have

been proposed to relate the two quantities.^ Most commonly equation (4)

is used in which k is simply a parameter selected to obtain a best fitp . 16of data.

Hij “ kSij (4)

It was found somewhat more satisfying to use equation (5), derived by

Cusachs, for these calculations.^ Cusachs found that coefficients

quite similar to those obtained empirically were derivable by use of

equation (5).

- (2 SiJ ) ( HH + ) St* (5)“ij 2Several authors have suggested using geometric rather than arithmetic

18 19means of coulomb terms. ’ However, the former yields imaginary

terms when one of the coulomb integrals is negative.

Overlap Integrals involving other than 3dg or 4sg AO's were20obtained from the tables of Mulliken, Rieke, Orloff, and Orloff.

The 3dg and 4 sg overlaps were calculated with an IBM 7040 utilizing

a program furnished by Cusachs.^ In both cases the orbitals utilized

were Slater type which predicated certain problems involving the use

of 3ds and 4sg orbitals. The effective charges, Z, of these orbitals21are predicted to be zero by Slater's rules; this would not appear to

be too realistic. Consequently, calculations were performed

utilizing Z values of 1.0, 0.5, and 0.0 for the 3dg's and 0.9 and 0.0

for the 4sg. Calculations were also begun using a value of 0.5 for the

4sg but were discontinued when it was found that the overlaps were

essentially the same as those obtained with Z = 0.0. Although there

13

Is little doubt that the charges used do not represent the true

situation, certainly they do represent the limits within which it lies,

since a charge of 1.0 would indicate an electron in the 3dg or 4sg AO

and a zero effective charge would by definition be the lower limit.

In the computations, no Initial hybridization of AO's was

assumed other than considering carbon hybrid orbitals to be sp^'s.

The coordinates chosen for the various molecules are Illustrated in

Figures 2, 3 and 4. Due to their symmetry (C2V ) and the relatively

shorter distances involved, several manipulations were possible with

H 2S and (CHgJjS that could not be performed with the remaining molecules.

First, group orbitals of the forms shown in equations 6 and 7 were

obtained from linear combinations of the hydrogen ls^'s and the

carbon sp^'s.

aA - ~ T <C1 + a2> (6 )

°B - ~ Y (°l - ct2) (7)

Subsequent applications of group theory resulted in the classification

of the orbitals according to irreducible representations as follows:

A].: 3sg, 3pzS, 3d^z2^s, 3d 2 _ y2^g, 4sg , ^

A2: 3d (xy)S

Bl: 3PXS* 3d (xz)S

B2: 3PyS* 3d (yz)S * Cfg

thus reducing the size of the secular determinants. In solving the

matrices, the approximations were made that:

Sii - 1 ; Stj - 0

FIGURE 2

COORDINATES OF HYDROGEN SULFIDE,

METHYL MERCAPTAN AND METHYL SULFIDE

R R B

R = H o r C

FIGURE 3

COORDINATES OF METHYL DISULFIDE

16

FIGURE 4

COORDINATES OF METHYL TRISULFIDE

c

3 ^

CIS4

TRANS

17

Consequently, the general form of the matrices was:

H 11 'E

H 21

H31

L

H12

H22 - E

H13

H23

H32

«12

H33 - E.

H13

H11

,H21

, H31

Hit - E

As a specific example the following secular determinant was

obtained from the symmetry orbitals of H2S with the effective charges

of the 3dg and 4s§ AO's being taken as 1.0 and 0.9 respectively:

CT1sH 3ss 3PzS ' 4sS 3d(z2)S 3d(x2

ClsH 19.04 - E 16.02 10.84 0.56 2.28 2.07

3ss 16.02 20.72 - E 0 0 0 0

3PzS 10.84 0 11.61 - E 0 0 0

4ss 0.56 0 0 3.76 - E 0 0

3d(z2)S 2.28 0 0 0 3.67 - E 0

3^(x2- y2)S 2.07 0 0 0 0 3.67

CHAPTER IV

RESULTS AND DISCUSSION

A. General

The vapor spectra and the results of the computations are

presented, compared and correlated for each compound individually.

Attention is focused only on those bands which appear to be non-Rydberg

in character (as defined in the introduction) since they are the ones

pertinent to this particular study and since Price and Clark have1 2 3extensively investigated the Rydbergs in most of these compounds. ’ ’

Similarly, several transitions were calculated to be in the region

of Rydberg absorption. Due to the multitude of sharp maxima observed

in these regions and the anticipated errors in computations, any

attempt to assign specific bands to these transitions in the absence

of additional information such as vibrational structure, polarization

data, etc. would of necessity be arbitrary and capricious. Consequently

no such attempt is undertaken here.

As some of the bands of interest were beyond the accessible

wavelength range of the instrumentation available in these laboratories,3several of Clark s spectra have been reproduced for completeness.

Most of the spectra presented for the dimethyl compounds have not

previously been reported.

The long wavelength bands in H2S and D2S were investigated

over the temperature range -80° to 25°C in an attempt to sharpen

18

19

the vibrational structure and to resolve the observed long-wavelength

tail. Moderate success was achieved in sharpening the structure but

the investigations failed to reveal any transitions in the absorption

tall. It was felt that the extremely broad, structureless band at

2500A in the disulfide might possibly harbor two transitions and,

therefore, attempts were made to Investigate it at liquid nitrogen

temperatures. However, efforts were thwarted by an absorption band at

2500A arising from the six layers of quartz necessary for the low

temperature study.

The results of the solution studies were such as to contribute

little to assignments or to correlation with calculations; therefore,

these results are presented separately, as are the effects of variation

of various parameters including inclusion and omission of 3dg and 4sg

orbitals and variations in effective charges.

In the subsequent discussion, the computed values taken for

comparison with experimental data are, in all cases except the tri­

sulfide, those values obtained when both 3dg and 4sg orbitals were

considered in the calculations with effective charges of 1 . 0 and

0.9 respectively. Trisulfide computations were not performed including

3dg's. In each case where the computations are discussed, the eigenvectors

and eigenvalues are also tabulated.

B. Hydrogen Sulfide

The vapor spectrum of is shown in Figs. 5 and 6 . Experi­

mentally observed transition energies are compared with the calculated

values in Table II.

FIGURE 5

3VAPOR SPECTRUM OF HYDROGEN SULFIDE

Moto

r Ex

tincti

on

Coef

ficie

nt

X 10

4 8

Hydrogen Sulfide

2*4

60 70 60 90Frequency kk

— vv i i— r ■ i - ■ *

_ Di F 1

. d

T ¥

T ” i i F •

i m

Hydrogen SoMo

r\,a a A v j

H

- M

-I*

Frequency kk

FIGURE 6

STRUCTURE OF THE HYDROGEN SULFIDE 1965 A BAND OBTAINED AT

2 0

O

x

2000 210019001800

W A V EL EN G T H (X)

22

TABLE II

H2S Transition Energies (e.v.)

Calculated 6.21 6.71 6.80 9.79 12.733Observed 6.51 7.85 8.03

The long wavelength transition (6.21 e.v.) is predicted by the

calculations to possess 75-80Z 3dg character. To aid in the comparison

of the observed and predicted values, the spectrum of D2S was also

run (Fig. 7) and vibrational analysis of the first band performed.

The latter operation indicated all of the vibrations to be members

of a single progression. The frequency differences between successive

vibrations are given in Tablelll and it will be noted that the average

differences are approximately 1118^20 cm and 822i"20 cm in H 2S and

D2S, respectively.

TABLE III

H2S and D2S Vibrational Frequencies (cm-^)

H2S 1145 1144 1112 1101 1086

D2S 804 817 804 871 828 794 834

22The three ground state fundamental vibrations are listed in TablelV.

22TABLE IV

H 2S and D2S Fundamental Vibrational Frequencies (cm *)

V1 V2 V3H2S 2611 1290 2684

D2S 1892 934 1999

TABLE V

Results of H„S Computations (Z_, = 1.0; Z, = 0.9)2 s 4ssSymmetry Eigenvectors Eigenvalues

1SH 3ss 3pzS 4ss 2az 2 2a x -y *

° - H 3V ayz 3pxS

6488 -.4523 -.5072 -.0605 -.2496 -.2266 0 0 0 0 2.260 0 0 0 0 0 -.7189 .5439 .4327 0 - 0 . 6 8

0 0 0 0 . 0 0 0 0 . 0 0 0 0 .0 0 0 0 .6722 -.7404 0 0 0 0 - 3.670058 -.0055 -.0080 -.9785 .1524 .1383 0 0 0 0 - 3.760 0 0 0 0 0 -.2921 .3285 -.8982 0 - 4.261743 -.1696 -.2570 .1964 .6770 .6147 0 0 0 0 - 4.260 0 0 0 0 0 0 0 0 1 . 0 0 0 -10.47b2219 -.5936 .7710 .0113 .0458 .0415 0 0 0 0 -14.730 0 0 0 0 0 .6307 . 7722 .0773 0 -18.377067 -.6436 -.2869 -.0114 -.0465 -.0422 0 0 0 0 -38.31

B,

B.

B,

a) 3dg AO'sb) Experimental ionization potential of H2S.Interactions of the 3d(xy)S and 3d(xz)S would be symmetry forbidden. Consequently they have not been included in the table and are assumed to have eigenvalues of -3.67 e.v.

NJu>

FIGURE 7

STRUCTURE OF THE DEUTERIUM SULFIDE 1980 % BAND OBTAINED AT

OPTI

CAL

DE

NSI

TY 1.0

0.8

0.6

0.4

0.2

1950 200019001850 2050

WAVELENGTH (A)

25

The vibration corresponds to the symmetrical stretch; V2 to t'ie

symmetric bend; the antisymmetric stretch. The ratio of the observed

D2S/H2S excited state vibrational frequencies is .735 comparing

favorably with the .718 calculated for the antisymmetric stretch from

ground state data. However, the frequency change necessitated by

considering the observed vibrations as members of a progression would

appear to be much too large. Moreover, since the molecule is considered

to undergo a symmetrical shape change and there is no clear evidence of

vibronic interactions, it is most reasonable to suppose that the active

vibrations belong to totally symmetric species.

This then leaves a choice between the symmetrical stretch and

the bending mode, both of which are of the A^ species. Once again

comparing the ground state vibrational frequencies,V2 would appear

to be the more reasonable assignment since it occurs at 1290 and 934 cm

in H 2S and D2S respectively. As further evidence for assignment of the

progression to a bending mode, it is observed in Table III that the

changes in vibrational frequencies in going to successive members of

the progressions are relatively minor. Consequently, any anharmonicity

of the excited state must be very small indicating that it is indeed a

tightly bound state such as might arise from excitation to a 3dg or 4sg AO

in agreement with predictions. The fact that the progression which

appears is a bending mode further tends to support such a hypothesis

since, were the excited level more of a sigma antibonding character,23 24as has been proposed by a number of authors, ’ one would expect a

preference for a stretching mode.

The two remaining diffuse systems occur at 7.85 and 8.03 e.v.

The calculated values are 6.71 and 6.80 e.v., somewhat low in energy, but,

26

relatively close together as are the observed bands. The 6.71 level

has about 90% 4s character while the 6.80 level is about 70% 3d_. The S Spredicted excitation energies to the antibonding levels, 9.79 (symmetry

forbidden) and 12.73 e.v, appear to be somewhat too high when compared

with the experimental data.

C . Methyl Sulfide

The vapor spectrum of dimethyl sulfide is presented in Fig. 8 .

Fig. 9 shows the long wavelength absorption in solution. The inflection

at approximately 235QA is so weak that it is detectable only in solution.

The observed and calculated transition energies are listed in Table VI,

TABLE VI

CH3^2^ Transition Energies (e.v.)

5.01 (A.)Calculated 4.92(A ) 7.13 (B2) 8.96 (A.)

1 5.01 (At)

5.01 (B2)

Observed 5.20 5.39 6.15 6.33

(Quantities in parentheses indicate symmetry of excited state.)

The systems at 5.39 and 6.33 e.v. exhibit considerable structure and

bear a remarkable resemblance to one another. In fact, it was possible

to completely analyze both bands in terms of progressions and combinations

of the symmetrical CSC stretch and the parallel methyl rock with one

exception: the weak second vibrational-electronic band which is a

torsional vibration.

FIGURE 8

VAPOR SPECTRUM OF METHYL SULFIDE

Log

€

56

34

30

2 2

1900W A V E L E N G T H A)

28

FIGURE 9

SOLUTION SPECTRUM OF THE ~2350 X INFLECTION OF METHYL SULFIDE

(CHg)z S

1 I2500

i

2 3001____ L -

2100W A V E L E N G T H (A)

TABLE VII

Symmetrya 3spc 3ss

Results of (CH^J^S Computations

Eigenvectors- „2a 2 2a 3pzS Z x -v s

(Zvi =3ds

*a 3spc

i*0 ; z a c -4sS

3fX cys

0.9)

YZ3 3pxS

Eigenvalues

A 1-.7458 .4418 . 4981 .0151 .0151 .0055 0 0 0 0 0.28

B 20 0 0 0 0 0 .7925 -.6092 -.0300 0 - 1.55

A 1 .0 0 0 0 .0000 .0000 -.7071 . 7071 . 0000 0 0 0 0 - 3.67

A 1 -.0136 .0099 .0136 -.7069 -.7069 .0046 0 0 0 0 - 3.67

B 20 0 0 0 0 0 -.0216 .0211 -.9995 0 - 3.76

A 1.0036 -.0026 -c0036 .0032 .0032 1 .0 0 0 0 0 0 0 0 - 3.67

B 10 0 0 0 0 0 0 0 0 1 . 0 0 0 - 8 .6 8b

A 1.2619 -.4933 .8295 . 0 0 2 0 .0020 .0008 0 0 0 0 -14.12

B 20 0 0 0 0 0 -.6095 -.7927 -.0035 0 -17.55

A 1.6124 .7492 . 2522 .0018 .0018 .0007 0 0 0 0 -30.89

a) 3d AO'sl)b) Experimental ionization potential ofInteractions of the 3d(xy)S and 3d(xz)S would be symmetry forbidden. Consequently they have not been included in the table and are assumed to have eigenvalues of -3.67 e.v.

30

The assignments, presented in Table VIII, would appear to be

quite reasonable but are not unambiguous.

TABLE VIII

Vibrational Assignments in the (c h 3)2s2150 - 2300 and 1850 - 196QA Bands

A v (cm I.R.a(cm”l) Symmetry Assignment1954 0

1943 237 284 (R) a 2 torsional vibr1928 637 692.5 A 1 I1916 962 1028 A 1 II1905. 5 1250 1385 A 1 2 x 1

1892 1624 1720.5 A 1 I + II1881 1933 2056 A 1 2 x II

or 2077.5 A 1 3 x 11869 2274 2413 A 1 (2 x I) + II1856 2649 2748.5 A 1 I + (2 x II)

or 2770 A 1 4 x 12276 0

2262 272 284 (R) A 2 torsional vibr2238. 3 742 692.5 A 1 I2 2 2 2 1067 1028 A 1 II2207 1373 1385 \ 2 x 1

2187 1787 1720.5 A 1 I + II2176 2018 2056 A 1 2 x II

or 2077.5A 1

3 x 12142 2748 2748.5 A 1 I + (2 x II)

or 2770A 1 4 x 1

I = symmetric CSC stretch; II « parallel CH^ rock.

a. Perchard, J. P., Forel, M. F. , and Josien, M. L. , J. Chem. Phys. , 61, 645 (1964).

(R). Raman active.

31

The similarity of the two systems is further illustrated by

comparing the vibrational frequencies in Table IX, The shift of

approximately 1 0 0 cm"* observed in each of the transitions would appear

quite reasonable since the latter corresponds to a higher excited electronic

level.

TABLE IX

Comparison of the Vibrational Frequencies in the

(CHj^S 5.39 and 6,33 e.v. Bands

Band (e.v.) v(cm~*)

5.39 272 742 1067 1373 1787

6.33 237 637 962 1250 1624

5.39 2018 2748

6.33 1933 2274 2649

The calculations predict that the lowest energy transition

should be predominantly 4s in character while the three degenerateOtransitions should be primarily 3dg. It does not appear unlikely in

view of the very small energy differences that the interaction between

the low lying levels might produce shifting in such a fashion as to

result in the 5.01 level becoming the lowest energy excited state

and thereby giving rise to the very weak inflection at 5.20 e.v.

Such a transition would be symmetry forbidden and thus account for its

low intensity. The extinction coefficient for this transition has3

been estimated to be about 15-20. The assignment of this band as

symmetry forbidden is substantiated by its strong appearance in t-butyl

32

3methyl sulfide. The extreme assymetry of the latter compound makes

the transition allowed. The system at 5.39 e.v. has an oscillator

strength of 0.016 and is assigned as arising from an excitation to

the second of the essentially 3dc levels while the third 3d level iss sassumed to be related to the 6.33 band. This leaves the transition

to the predominantly 4sg level to correspond to the diffuse system at

6.15 e.v. The similarity of the 5.39 and 6.33 e.v. bands offers some

justification for the assignment of both as involving excited states

that are essentially 3dg in character. Furthermore the increase in

intensity of the 0 - 0 relative to the other vibrations in the short

wavelength packet as compared to the intensity of the 0 - 0 to the other

members in the long wavelength packet is what would be expected as the

atomic nature of the excited state is increased.

D. Methyl Mercaptan

The predicted and observed transitions for CH^SH are compared3in Table X and its spectrum is presented in Fig. 10. As in l^S, the

lowest lying level is predicted to be predominantly 3dg followed in

order by essentially 4s„ and degenerate 3d levels. The agreemento Sbetween the calculated and the observed values appears to be surprisingly

good.

TABLE X

CH^SH Transition Energies (e.v.)

Calculated: 5.35 5.68 ^.77 9 02

3Observed: 5.39 6.08 6,75

TABLE XI

Results of CH_SH Computations (Z_, : 3 3ds

Eigenvectors

*H3spc 3ss 3pzS 3pyS 3pxS 4ss 2az

6554 -.3638 -.1922 -.5034 .2707 . 0 0 0 0 -.0393 .02682453 .6833 -.4575 -.1602 -.4599 . 0000 -.0283 -.00390 0 0 0 . 0 0 0 0 . 0 0 0 0 .0 0 0 0 . 0 0 0 0 . 0 0 0 0 .0000 -.66170001 .0154 -.0086 .0015 -.0154 . 0 0 0 0 -.0039 .74570077 -.0023 -.0040 -.0097 .0034 . 0 0 0 0 -.9819 -.01721814 -.0168 .1375 .2309 -.0077 .0 0 0 0 -.1829 .70430 0 0 0 . 0 0 0 0 .0 0 0 0 . 0 0 0 0 . 0 0 0 0 1 . 0 0 0 0 .0000 .00001384 -.1885 .4857 -.5541 -.6337 . 0 0 0 0 -.0050 .00073869 -.4577 -.1398 .5622 -.5497 .0 0 0 0 .0078 -.00675554 .3939 .6923 .2107 .1061 . 0 0 0 0 .0060 -.00200 0 0 0 . 0 0 0 0 .0 0 0 0 .0 0 0 0 . 0 0 0 0 . 0 0 0 0 .0000 .00000 0 0 0 .0 0 0 0 . 0 0 0 0 . 0 0 0 0 . 0 0 0 0 . 0 0 0 0 .0000 .0000

a) 3dg AO'sb) Experimental ionization potential of CH^SH,

1.0 ; Z = 0.9) 48S

Eigenvalues! 2a-y YZa XYa axz2152 -.1608 . 0 0 0 0 . 0 0 0 0 1.561169 -.1092 . 0 0 0 0 . 0 0 0 0 - 0.424998 .5589 . 0 0 0 0 . 0 0 0 0 - 3.673635 .5578 . 0 0 0 0 . 0 0 0 0 - 3.671501 .1138 . 0 0 0 0 . 0 0 0 0 - 3.767296 -.5690 . 0 0 0 0 . 0 0 0 0 - 4.090 0 0 0 . 0 0 0 0 . 0 0 0 0 . 0 0 0 0 - 9.4410 2 2 2 -.0191 . 0 0 0 0 . 0 0 0 0 -13.730443 .0317 . 0 0 0 0 . 0 0 0 0 -18.940289 .0235 . 0 0 0 0 . 0 0 0 0 -35.480 0 0 0 . 0 0 0 0 1 . 0 0 0 0 . 0 0 0 0 - 3.670 0 0 0 . 0 0 0 0 . 0 0 0 0 1 . 0 0 0 0 - 3.67

FIGURE 10

3VAPOR SPECTRUM OF METHYL MERCAPTAN

Motor

Citin

etiM

Co

offic

ioot

X

10*

I I I I

SS 3P S p

N9

Sitoi r

toto

v / \

\ / V

T

2ST

3r

r

M*5

toSI

Mtthyl Morcoptan

H

20

1 0

Fftortwy lik

35

E. Methyl Disulfide

The vapor spectrum of methyl disulfide Is shown in Fig. 11. The

computed and observed transition energies are presented in Table XII.

TABLE XII

(CH3 >2S2 Transition Energies (e.v.)

Predicted

i nl ^ V t nl ~ 9 i n 2 Vi n2 <Pi CT 1 “* Vi ai - Vi

e - 90° 0 - 104° 0 - 90° 6 - 104° 0 - 90° 0 = 104°

1 2.43 2.09 2 . 72 3.12 7 . U3 6 . W2 3.30 2.98 3.59 4.01 7.92 7.883 3.45 3.15 3.74 4.18 8.07 8.05

4 5.21 4.89 5. 50 5.92 9.83 9.79

5 5.29 5.07 5.58 6 . 1 0 9.91 9.97

6 5.57 5.19 5.86 6 . 2 2 10.19 10.09

7 9.70 9.39 9.99 10.42 14.32 14.29

8 9.71 9.53 1 0 . 0 0 10. 56 14.33 14.43

9 10.63 10.36 10.92 11.39 15.25 15.26

10 10. 70 10.39 10.99 11.42 15. 32 15.29

11 10.71 10.43 1 1 . 0 0 11.46 15.33 15.33

12 1 1 . 0 0 10.69 11.29 11.72 15.62 15.5913 11.30 11.15 11. 59 12.18 15.92 16.0514 11.84 11.26 12.13 12.29 16.46 16.1615 13.70 13.44 13.99 14.47 18.32 18.34

ni “ *i “

1th filled ith excited

nonbonding level, cpj

level, nj being the highest energy filled leve being the lowest energy vacant level.

Observed

E 4.96 5.89 6.32 6.65f 0.0312 0.0283 0.303 _ - - _

sC

19*St 8992'- 0CS2'- S5BS*- C290* 6960*- 9E12 * -196S * -C1C0* 2280*- 1992*- 1E10" 5600* 9020* £920*- 1£10*- 9210* 1100*-91£0* £910* 2000*-9010*-£010**80 '12 1019*- 6609* 05 59*- Eiso'­2S22 * -61S2' 1559* 2090* 6922* 9252*- 9910"- 06£0* osto*- 5£90* 69£0" 1E10* 9290*- 02£0* 9990’- 15E0*- 9110'-9110*-E6'02 099£*- 099f'-8690* ll* 628f*- 6209* 8690" 9152* B2BE*-9209* 6520* 56EO*- 1SE0* 16f 0 * -9E20*- 2520* 56£0*-1S£0* O6E0*-££20*- 8900*-8900'-er *91 OS 60* S901* -£062*- USE* 1689* 1291* 9962* itit*- 1119*- 0591*- 1200*- 6850*- 1910* 6£10* 1100*- 5120* 1620* 1E00* 8000*-0820' 0600* 0600 * -5t 0150*- 1590*- 2952*- 699E * -£682 * 2699* 6ES2*- 199E * -05 62* E299* 8210* 1910* 1010*- 6680*- 9910*- 6500*- 8100* 8210*- 9£ 60 * - 1110' 0620*- 0620'-SE*01 6602' 6602*-£161*-2195*- 1102* 8611' C161 * £195* 1102*- 8611*- 1610* 1E1T- 1681* 9520*-8£90*- 1610*-9£ 11 * 0061*- 5520" 1£90' S100* 5100'-£5*6 9850*- 9850*- 9800*- 6191 ‘ -8990*- 2 ICO* 9800*- 2891* 5990*- 21E0* 5920* 12£0*'£910* 9£15 * 251C * -9920* 91£0*-0110*-9£1S" 2S1f*-9900**9900*-91 -8 910:* 5101 * S6CI*- 190S * Eltf * siri* 9621**OS OS* S9EE * 9IE1 * 2910*- 1£81* 9691* 29£T* 1£90*- £910** 9181*' 91S1 * *29t1 *' 9E90*-0050*- 0050'-05*1 6610’ 661 O' 9C10 0010*-£ 9t 0' -9910*- SC10* 1010*- 29E0*-9910"- 2121*- £190*- 2910*- OEOO*- 0110* 2121*- 0190*- 5910*- 0£00* -0110' 1189 * -1189*-6C*1*99E0*-99E0'-S090* 0900'-9C90* 51E0'-5090* 6t 00 * -9E90* S1E0*- £159* £9£1* 1191* 1110*- 0200* £159* £9t I * 8191* 1110*- 0200* 9991*-9991*-01*1-£690'-£690"-BS90* 1861*- E9£l*- 8090'-8590* 1861 * -E9ET- 8090*- 1102*- 1t£9 * 8259* 2610*- 1200*- 1102*- 1t£9* 8259* 2610*- 1200*-£590*-£590'-86 "9 *9'10‘ 9110'- 0210'- 1022*- £f 02' 1800* 0210* 1022* EE02 * 1800*- £200* 6259* 9159*-otio*- 1220*- £200*- 6259*- 9159* OtlO* 1220' 9000* 9000*-90*9*9860' 9890'-9250'-9£10' SSIT- 8f £0' 9250* 9E10"- 5511* BEEO*- 5100* 6£20" 2020"- 185£ * -8850*-9i00*"6E20*-2020* 185 £ * 9185' £900* £900*-.'2*2-1912'- 1912'-9691* 2800* 99C2 * 1611*-9691* 2800* 99E2* 1611*- 9E50"-2161*- £ 191" 1812*- 9229*- 9E50'- 2161 * -E191 * 1812** 9929'- £600'-£600'-tro- 9991*-9991*- 1210* 9E10* £621' 9090*- 1210* 9E10 * - £621 * 9090*- 6920*- E09t *- 269E * £1£2 * 111E * 6920*-£09£ *-269£* E1E 2 * 111£ * 1200** 1200'-R-0- 6192* 6192 * -£220*- 96S0* 6181*- 2102* E221 * 9650*- 6181* 2102*- 1915" S5£1* 19£0* 5150* 2550* 1915*- 55£1*- 19C0*-6150*- 2550*- 6900* 6900*-EO'O- 2200' 2200'-£801*- 6000* 2100*- 1022*- C801 * 6000*-2100* 1022* £990* 1E10* 2210* 6500*- 6900* £990*- ItlO*- 2210*- 6500* 6900*- 8659* 8659'-10*0-£100* £100'-9000*- 6110'-2610*- COOO* 9000* 6110* 2610* £000*- 66E1* 0069* 0069*- 2000* £000* 96£1*- 0069* 0069* 2000*- £000'- 0500* 05 00'-12' 0 6190' 6190'-9082* 9210* 6690*-8E9S* 9082** C210*- 6690* BE95 *- 10£2‘- 1B10*- 9E90*-2£10* 6910* 10£2* 0810* 9E90* ino*- 6910'- 1052*' 1052*-8f *1 5059' -S059* 02 12' 6110"- 2E62 * 1190* 0212*- 6110* 2E62*- 1190*- 0£0£‘-£1£0*‘ 6E1I * 8£E1*- i£90*-0£0f- tito* 6£ 11 * -8EE1* 1£90* 11 £ 0' 11E0*-5 2*2 6629*- 6629*- 1281* 8800*- £612* 2660*- 2281* 8800*- E6I2* 2660*- 9990*- 6062' tsot*- 1901* otei* 9990"- 6062* 250t*- 1901* 0181* 0£00'- 0£00'-62*2 6090*- 6090* 6950* 9800*- 99C0* 6910** 6950*-9800* 99E0‘- 6910* 1920" 9900"-9210* 1685* 06it *- 1920*- 9900* 9210*- 1685 * -061E* 0110* 0110*-

«>3d.E (1>3d.C(2)S,f

(2)S*dt(2)Sidf<Z>SIdC (1>S1dt(1>SV CZ)*JE“>SV (2)«* ,(2)3*V •ttV #(1)i* (1) * * 2

••njtiwln no7?iAo*t?2s s

(6*0 - **2 :0‘1 * ^2) ■T»T3*in^oO 901 - 0 *Sf(£H3) 1° ■ll«W«ton n n m

FIGURE 11

VAPOR SPECTRUM OF METHYL DISULFIDE

LOG

0.5

0.0

-0 .5

1900 2100 2 3 0 0 2 5 0 0 2 7 0 0 2 9 0 0WAVELENGTH (i)

38

There appears to be a considerable disparity of results in view of

the transitions predicted to occur between 2.0 - 3.5 e.v. The error

apparently steins from the very large overlaps obtained for S, . (0.9)»4Sgand S7j -I. (0.6 - 0.9) and the consequent splitting produced. The

S’ Smagnitude of this effect may be gauged from Fig. 12 in which the 3dg

splitting for the case where the dihedral angle is 104° and the effective

charge is zero has been graphed. Under these conditions the 3dg orbitals

do not undergo any Interaction with the other available orbitals but the

3dg-3dg interactions are maximized. A number of levels are generated

sufficiently close to a predominantly 4Sg level at -7.49 e.v. and two

essentially nonbonding levels at -8.54 and -9.80 e.v. that a large

amount of mixing would be anticipated. Increasing the effective charge

on the 3dg's to 1.0 has relatively little effect on the 8 3 ^ 3^ values

and therefore should exert very little influence on the splitting. How­

ever, the additional charge does result in considerable mixing of the

lower lying 3dg's with the 4sg and nonbonding levels. Consequently the

two highest filled levels are formed from a combination of 3dg, 3pg 3and Sp orbitals but more important the first four excited levels are,c

in order, combinations of 4sc, 3d_, 3pe and <J 3 (predominantly 3d ); 4s0 0 0 8PC s

and 3dg (very predominantly 4sg); 4sg and 3dg (primarily 3d^); 3dg only.

The contributions of the 3dg' s to the next several excited levels are

much smaller and the transition energies correspond quite closely to■m

experimentally observed values. Thus the questionable transitions are

traceable to the large overlaps used. Although the use of smaller over­

laps cannot be theoretically justified, they almost undoubtedly would

produce better correlation with experiment.

39

FIGURE 12

SPLITTING OF 3ds (Z-O) LEVELS IN (CH3 )2S2 9- 104°; UNITS IN e.v.

+2.26

0.00

-1.40

-5 .93

-7.34

-9.61

AO

F . Methyl TrisulfIde

The spectrum of methyl trisulfide is presented in Fig. 13 and

the computational and experimental results are compiled in Table XIV.

The non-agreement of the results is not at all surprising in view of the

fact that 3dg AO's were not included in these computations. In the

calculations on all of the previous molecules it was found that the

excited levels corresponding to the observed transitions were generated

only when 3dg and 4sg AO's were considered. The excitation energies

to these levels in each case fell within the range of 5 - 7 e.v,, that

is, they fell right in the region of the observed methyl trisulfide band

maxima. There is no reason to expect that the inclusion of 3dg's in the

computations on this molecule would not have the same effect. Of course,

one would also expect, as was found in the disulfide, a number of

transitions to be predicted at lower energies assuming overlaps were

calculated by the same procedure used for the other molecules. Similarly,

the computed value of 1.90 e.v. undoubtedly arises from using too large

(0.9) a value for S4s .S ’ S

G, Effect of Variation of Effective Charges (Z)

A general picture of the effect of varying the effective charges

of the 3dg and Asg AO's is shown in Fig. 16 where the energy levels

obtained by using various Z values have been plotted for the case of

dimethyl disulfide with a dihedral angle of 10A°. These results are

more or less typical of those found in all of the molecules considered.

Inclusion of 4Sg AO's had essentially no effect upon the energy

of the levels generated in their absence. The maximum displacements

produced when they were added were of the order of 0.02 e.v. Consequently

FIGURE 13

VAPOR SPECTRUM OF METHYL TRISULFIDE

OPT

ICA

L D

EN

SIT

Y

0.7

0.5

0.3

1800 2200 2600 3000 3400

WAVELENGTH (A)

42

TABLE XIV

Transition Energies (e.v.)

Predicted

i

Cis: nf ^ n2 ^ i n3-*<Pi ar * i

Trans:

nr i n2~* <Pi n3~’ Vi a T Vi

1 1.90 2.69 3.91 6.45 1.03 1.78 2 . 8 8 5.22

2 7.13 7.92 9.14 1 1 . 6 8 7.21 7.96 9.06 11.40

3 8 . 2 0 8.99 1 0 . 2 1 12. 75 8.23 8.98 10.08 12.42

4 9.10 9.89 1 1 . 1 1 13.65 9.20 9.95 11.05 13.39

5 9.22 1 0 . 0 1 11.23 13.77 9.31 10.06 11.16 13.50

6 9.49 10.28 11.50 14.04 9. 59 10.34 11.44 13. 78

7 10.03 10.82 12,04 14.58 10.32 11.07 12.17 14.51

n^ ■ ith filled nonbonding level, n^ being the highest energy filled level. ■ ith excited level, cp being the lowest energy vacant level.

Observed

4.42 5.00 5.81 6.17 6.36 6.65

TABLE IV

**sCD 4*S(2) *“S(3) 3#S(l) 3p«s(i) **<!)

gesule* of CL* (CH^)^ Eigenvectors

^rSU) 3,S<2) 3p*S{2)

Computation* (2

3pyS{2) 3prS<2)

:4e * 0 S

3,S (3)

.9)

3PkS(3) 3pyS(3) 3P*S(3) ,PC(l) "PC( 2)

Eigenvalue*

-.2041 -.0319 .2092 -.0156 .1975 -.1545 .0481 -.1348 -.2244 -.0602 -.2015 .4161 -.0272 .3739 .0208 .2028 -.6166 .33-.2520 .0484 .2424 -.3335 .0968 -.4351 .0112 .0889 .0388 -.0431 .1279 -.1733 -.0554 -.2610 -.0106 .6481 .2420 .29-.4234 .8114 -.3876 .0101 -.0111 .0282 -.0076 .0299 -.0033 .0066 .0178 .0044 -.0293 .0315 -.0059 -.0401 -.0449 .02-.6137 -.0349 .6426 .1001 -.1116 .2337 -.0095 .0154 .0768 .0358 .0080 -.0405 .0281 -.0698 -.0149 -.3241 .1100 - .10-.0097 .0442 -.0315 -.0813 -.1731 -.0764 .1076 -.1728 .2072 -.1047 -.6239 -2219 . 5961 -.1679 .0146 .0635 .1836 -1.00-.0317 .0252 -.0004 .2687 . 5437 .0165 .0030 -.2740 -.5900 -.0625 -.0936 .0007 .0348 -.2790 .1334 -.0368 .3100 -2.07-.0396 -.0446 -.0441 .1133 -.1813 -.2780 -.0661 .0532 .0279 -.7690 .0889 .1698 -.1987 -.3767 .0047 -.1868 -.1380 - 7.30-.0131 -.0058 .0055 -.0616 -.2323 -.0349 -.6164 .0426 -.1485 .0437 .0753 .0317 .1700 .0721 .7025 .0269 .0005 -9.20-.0752 -.0905 -.0866 -.0926 -.3359 .0873 .5746 .3287 -.2163 .0705 -.2820 .0091 -.3340 -.0749 .4718 .0197 .0431 -9.99-.5659 -.5651 -.5634 .0497 .0032 -.0268 -.1092 -.0180 .0130 .0885 .0153 .0192 .0322 .0200 -.1183 .0416 .0423 -11.21-.0167 .0088 .0236 .3886 -.1801 -.6018 .2438 -.0920 -.0670 -.0061 .0843 -.2590 .1577 .4572 .0749 -.2004 .1650 -13.75-.0686 -.0642 -.0682 -.2433 .3671 .2729 .2870 -.0970 .2405 -.4310 .2416 -.2491 .2972 .2346 .3382 -.0305 -.0601 -17.04-.0133 .0081 -.0131 .1935 -.1469 -.0204 .2618 -.5690 .2614 .2834 .3617 .2061 -.0440 -.3381 .2609 .1340 -.1521 -18.25-.0006 .0007 -.0053 .0023 .4083 -.2226 -.1763 -.0880 .5320 .0838 -.3390 .0615 -.4579 .0928 .2248 -.2064 .1570 -19.96-.0096 -.0173 -.0060 -.0791 .2862 -.3331 .0537 .3604 .0341 .3020 .0374 -.1697 .3140 -.4089 .0660 -.307 5 -.4247 -23.00.0126 -.0043 -.0240 -.5894 -.0005 -.1924 .0999 -.0849 -.1671 .0924 .2762 .4780 .0200 .0992 -.0604 -.4005 .2867 -29.21.0245 .0183 .0224 .4191 .1916 .0480 .0778 .5146 .1959 .0111 .2491 . 5259 .1818 .0768 .0748 . L794 .2352 -36.86

FIGURE 14

VAPOR SPECTRUM OF METHYL TETRASULFIDE

2 200

2600 S000

3400 W

AV

EL

EN

GT

H

(X )

OPTICAL D E N S IT Y

(A

FIGURE 15

VAPOR SPECTRA OF (CH3 )2S, (CH3 )2S2, (CH3 )2 S2 AND (CH3)2S

OPT

ICA

L

DE

NSI

TY

ooV A P O R S P E C T R A

10 cm cellM e .S

M e , S . 0 5 c m cell

• • • • • • • • Me, S 9 0.5 cm cell

Me, S4 0 5 cm cellOOOOOOOO

0.8

0.6 -

0.4

0 «

0.2

‘•v.v.v.v.v.V,

2000 2 6 0 02200 2 4 0 0

WA V E L E N GT H, A

FIGURE 16

ENERGY LEVEL DIAGRAM FOR METHYL DISULFIDE (0

WITH VARYING 3dg AND 4ag EFFECTIVE CHARGES

(ch3»2s2 e» i04#

43(0.9) 3d(I.O) 48(0.3) 43 (0 .0 ) 43(0 .3) 43(0.3)3d{|.0) 3d(l.0) 3d <0.0> 3 d(O.O)

0.00

- 6.00

-x-x- -tt-K-

-9 .0 0

— 15.00

—27.00 -

47

it is not surprising that reduction of their effective charge from 0.9

to 0.0 leaves the levels virtually unchanged. Nonetheless, it was

observed that the 4s 's undergo quite large splittings in both the di-band trisulfides as exhibited in Fig. 17. Those levels generated by

Inclusion of the 4ag's are rather insensitive to variations of parameters,

exhibiting a maximum displacement of approximately 0 . 1 e.v. when

the 4s<j Z is decreased from 0.9 to 0.0.

As previously stated, Z values of 0.0, 0.5 and 1.0 were used for

3dg AO's, The 3dg AO's when inserted with effective charges of 0.0

or 0.5 remain degenerate and fail to perturb the levels already present

in the molecules containing a single sulfur atom. However, in the

disulfide, as previously discussed, the 3dg's exhibit considerable

splitting even with a Z value of zero. New levels appear at 2.26,

0.00 (triply degenerate), -1.40, -5.93, -7.34 (triply degenerate) and

-9,61 e.v. The triply degenerate levels correspond to bonding and anti­

bonding combinations of the xy, xz and yz AO's while the remaining four2 2 2levels consist of various combinations of the z 's and x -y 's. A

rather interesting point here is that the lowest of these new levels

(-9.61 e.v.) is approximately one e.v. lower in energy than one of the

previously filled levels (-8.54) which presumably was primarily nonbonding

in character. This situation continues to exist as the 3dg Z value is

increased. This then could offer some insight as to the role of the

3dg1s in octet expansion in polysulfide compounds. These results

would imply that octet expansion involves filling the vacated nonbonding

3pg level. Due to the relatively small energy difference, filling the

3ds level first would have little effect on the over-all stability of

the compound.

FIGURE 17

SPLITTING OF THE 4ss LEVELS IN

( C H , L S

-S.7«

//

//

/:}^3.76

4-0.02 - 0.10

FIGURE 18

ENERGY LEVEL DIAGRAM FOR H2 S, CH3 SH, (CH3)2S

(CH3 )2S2 AND (CH3 )2S3 WITH Z3d-1.0 AND Z^g-O

Ener

gy

(e.

v.)

M2S CHjSM (CH^ (ch 2 (ch3>2 ( C H ^ (CH^Sj• ■90 e -1 0 4 CIS TRANS

- 0.00

- 12.00

■="18.00

E 27 .00

— 30.00

—36.00

50

Increasing the charge to 0.5 removes the remaining degeneracy

and results in considerable mixing of the 3dg's with nearby levels. The

levels are consequently shifted by amounts ranging from 0.1 to 1.5 e.v.

with smaller shifts being more common.

An effective charge of 1.0 produces shifts and at least some

splitting in all compounds except (CHj^S. Again the changes Induced

ranged from a few tenths of an e.v. to 1.0 e.v. The 3dg splitting

increased in going from (CH^^S to CH^SH to I^S.

CHAPTER V

FREE ELECTRON TREATMENT OF POLYSULFIDE SERIES

In all of the foregoing it was observed that the transition

to the excited level most nearly 4Sg in nature always corresponded

to one of the longest wavelength bands. A number of authors have

suggested that sulfur-sulfur chains demonstrate properties quite

similar to conjugated carbon-carbon systems.^ ^ Consequently, it

was decided to attempt to treat the long wavelength transition in the

(CH3 >2Sn series and the($CH2 )2Sn aeries of Minoura from a free electron

approach assuming the excited state to be the 4sg level and the highest

energy filled level to be a nonbonding ^Pg*^ Neglecting hybridization

is an oversimplification but should not affect the qualitative results.

Application of simple L.C.A.O. theory leads to wave functions

of the form:3ps n 3p

Y - > C f S (8 )n Y «1 Y »n Y

and4s 4s

Y S -,> d ft S (9)m Y*1 Y »ra Y

where is the nonbonding 3pg AO of the yth sulfur atom. Solution

of the appropriate secular equations for the coefficients yields:

C » d mf 2 sin __1 ’,1 v,l v N + 1■y, 1 y , W H + 1 N + l (10)

51

52

and for the energy levels:E3pS . 0 3pS + 2 e 3pSn cos n

N+l(ID

E4sS . q 4sS + 2 S 4ssm K cos m

N+l(12)

where and are coulomb and resonance Integrals respectively.

The lowest energy excitation would correspond to a transition

from the highest 3pg level (n«N)tothe first 4sg level (n-1) for which

the excitation energy would be given by:

The coulomb and resonance terms were evaluated empirically

by considering the cases where N=1 and N=2. Minoura did not investigate

sufficiently far down in the ultraviolet to locate the n~* 4Sg transition,

consequently, 47,620 cm * is assumed as a rather reasonable value. The

results for both the dimethyl and Minoura’s dibenzyl series are given

in Table XVI. Agreement is good for both series although the divergence

tends to increase as N becomes larger.

If the n "*4sg band is indeed moving as indicated then it would

seem reasonable to anticipate that the excitation energy would approach

the corresponding transition in the linear Sq molecule in the limit as

the length of the sulfur chain increases and the end effects of the

methyl groups become negligible. However, this can not be substantiated

since the n-*4Sg transition has not yet been assigned in the Sg molecule.

E Q.N+l (13)

53

TABLE XVI

Results of^Free Electron Treatment

of Polysulfide Series (cm *■)1,251 Expmt. Theory Expmt. Theory

(ch3 )2 sn (CH2 )2Sn

1 50,000 50,000 47,620

2 39,700 39,700 38,760 39,700

3 34,480 34,400 36,760 36,420

4 33,610 33,300 34,960 34,800

5 33,560 33,880

6 32,680 33,350

7 31,640 32,980

8 30,770 32,730

1 . Minoura, Y., J. Chem. Soc (1952).

., Japan, Pure Chem. Sect 73, 131, 244

2 . Ibid., Trans. Far. Soc., 59, 1019 (1963).

CHAPTER VI

SOLUTION STUDIES ON THE SERIES (CH3 )2 SN

The absorption spectra of dimethyl mono-, di-, tri- and

tetrasulfides were investigated in a series of solvents of different

polarities in an attempt to gain added insight into the nature of

the various electronic transitions. The spectral range covered in the

Investigations was from 3500A to the particular lower wavelength

transmission limit of each solvent. As a consequence of the solvent

limitations and/or blue shifts of several bands, the higher energy

transitions which were observable in the vapor could not be detected in

solution.

The spectra of the monosulfide (Fig. 19) exhibit the greatest

overall change in appearance in going from the vapor to solution.

First, the weak inflection at 2330A (Fig. 9), unobservable in the

vapor, appears in solution as previously mentioned. Apparently, due

to its very low intensity, this band is masked in the vapor spectrum

by the much stronger nearby 2125-2290A system. However, in solution

the vibrational structure of the latter system is suppressed and the

entire band is blue shifted to the extent that it appears as a shoulder

on the side of the strong 2000A excitation rather than as a separate,

well defined band. The elimination of structure and the increased

separation due to the blue shift combine to increase the resolution

54

FIGURE 19

SOLUTION SPECTRA OF (CHj^S

OPT

ICA

L D

EN

SIT

Y(ChLLS

Sotvtnt: | . H^O

2 . EtOH3. M#OH4 . Hexan*

*•

0.8

0.6

0 . 4 -

0.2

2 3 0 02200210020001900

WAVELENGTH (A)

56

of the weaker band in solution. However, the ~2300A band was still

too ill-defined to allow any definite statements regarding solvent

effects.

The previously noted blue shift of the 2125-2290A band continues

to increase as the polarity of the solvent increases until in water its

appearance has become that of an inflection. The increase in intensity

observed at the same time is rather strange in that all of the other

bands in this as well as in the remaining compounds exhibit decreases

in intensity in similar circumstances Consequently, it is more likely

that, contrary to the experimental data, the excitation is actually

decreasing in intensity as the solvent polarity is increasing. The

observed increases in absorption may be reasonably accounted for if it

is assumed that gains from increased overlapping with the much stronger

2000A band more than compensate for solvent induced losses. Solvent

effects on the 2000A transition are similar except that this transition

does exhibit a decrease in intensity with increasing solvent polarity.

Solvent transmission limitations plus a probable blue shift prevented

a complete investigation of the band observed in the monosulfide vapor

spectrum between 1860-1950A. The large slit widths required for

solvents other than hexane below ~1950 renders data collected below

this point as somewhat questionable However, in hexane it was observed

that the structure present in the 1860-1950A system in the vapor had

vanished and a definite blue shift had occurred just as in the 2125-2290A

band. The similar behavior of both bands tends to lend further support

to the assignment of both as resulting from n" 3dg transitions.

The general features of the disulfide solution spectra

(Fig. 20) are essentially the same as those of the vapor--a broad

FIGURE 20

SOLUTION SPECTRA OF (CH3>2S2

OPTI

CAL

DENSITY

Solvent: I. EtOH 2.MeOH

0.84. H exane

Not*: 0.1 OtO. unit oddod to Honoito

0.6

0.4

0.2

30002 6 0 02 4 0 0 2 6 0 02200W A V E L E N G T H (A)

58

unresolved band with a maximum in the region of 250QA, a shouldero

between 2000-2075A and a comparatively sharp band at ~1950A. However,

there are certain qualitative differences. A particularly interesting

one is in the 2500A band. Earlier, in the free electron treatment,

this band was assigned as corresponding to excitation to an essentially

4sg level. The computations further predicted, in apparent agreement

with experiment, that the transition would be shifted to longer

wavelengths as the number of sulfur atoms in the chain increased. Yet,

in addition to the long wavelength excitation in the tri- and tetra-

sulfides, there is still a band at 2500A. It is felt that the solution

to this dilemma lies in there being two transitions beneath the 2500A

envelope in the disulfide. The logical choice for the second excited★

level would appear to be oQ c. Certainly the band is sufficientlyJ “ wbroad to accommodate both transitions. Consideration of an angular

molecular model predicts that the n-*j should be independent of the

number of sulfur atoms in the chain. Such a conclusion is also

justifiable on the basis of available experimental data. Thus the

energy of the bonding and antibonding levels would be a function of the

sulfur-sulfur distances. The separation is found to be essentially

constant in the di- and tri-sulfides. Data on the geometry of the tetra-

sulfide is not available but should also be virtually the same.

If the transition is indeed independent of the number of sulfur

atoms then we should anticipate its occurrence in the same region in

the linear Sg molecule. Bass has in fact reported the presence of ao 25broad, unresolved band at 2550A in sulfur vapor spectra at 250 C. As

final evidence there is the efficiency of 250QA radiation in effecting

59

S-S scission in photolytic experiments. The rather peculiar flattening

of this band observed in these solution studies could then be attributed

to the achievement of a very slight amount of resolution of the two

transitions.

The shoulder at 2095A in the vapor spectrum has become sub­

merged in the stronger 1950A band to such an extent as to make an

accurate determination of its position impossible. The 1950A transition

is found to blue shift and decrease in intensity as the solvent polarity

increases.

As in the disulfide, the vapor and solution spectra of both

the tri- and tetrasulfides (Figs. 21, 22) are virtually analagous. In

fact the spectra of the two compounds are virtually identical, the only

major differences being in intensities and positions of the long wave­

length transitions. Thus in the trisulfide the long wavelength band

occurs at ~2850A and is merely an inflection while the 2500A absorption

is very obvious. However in the tetrasulfide the situation is reversed.

The first transition occurs at ~3000A as a well defined though relatively

diffuse system while the second transition has degenerated considerably.

Examination of the data in Table XVII reveals that the high

frequency electronic transitions which occur at approximately 2170 and

2000A shift to the blue in solvents of large hydrogen bonding ability.

This tendency is further confirmed by examining the shifts which are

observed in 107, V:V water: ethanol containing different amounts of

sodium hydroxide. The presence of alkali induces a shift to the red.

Now to explain these shifts in terms of structure is not an

easy task, because there are no simple spectroscopic criteria for

60

FIGURE 21

SOLUTION SPECTRA OF (CH3 )2S3

OPTI

CAL

DEN

SITY

1.0(C H ,L S

Solvent t I. H txont 2. EtOH9. MtOHr •

0.6

0.6

0.4

0.2

3 0 0 02200 2400 2 6 0 0 2 8 0 02000WAVELENGTH (A)

FIGURE 22

SOLUTION SPECTRA OF (CH3>2S4

OPTI

CAL

DEN

SITY

1.0

0 . 0 “

0.2

Solvent: I. MtOH 2.EI0H3.CycloM xon«4 . H 2 0

<CH* S4

All.JL2100 2 5 0 0 2 9 0 0 5 9 0 0 9 7 0 0

WAVELENGTH <1*

FIGURE 23

LONG WAVELENGTH ABSORPTION OF (CH3)2S

OP

TIC

AL

D

EN

SIT

Y

Solvent: I. MtOH

2400 2500 2600 2700 28 00 2900 3000 3100 3200

WAVELENGTH ( A)

TABLE XVII

n X max log e

Solution Spectra of the

X max log e X max

Series (CHJ„S 3 2 n

log e X max log e X max log eVapor 1 - -__ ___ 2125-2290B 2025B ____ 1850-1955BHexane 1 2330i — 1 .9 2140 2.80 2 0 0 0 3.38 1905 3.40EtOH 1 2330i - 2.0 2 1 2 0 2.95 1990 3.28Me OH 1 2330i — 1 • 9 2135 2.94 1990 3.24h20 1 2340i - 2.0 2 1 0 0 3.08 1935 3.17

Vapor 2 2515 209 5S 1958 ,*__Hexane 2 2530 2.62 (*) 1945 3.84EtOH 2 2530 2.59 1945 3.73Me OH 2 2525 2.52 1930 3.72H2° 2 2510 2.52 1890 3.73

Vapor 3 ~2900i --- 2480 --- 2130 - _ • • 2010S -- - 1950 - *Hexane 3 ~2850i -2.5 2530S 2.61 2 1 2 0 4.08 -2000S ---EtOH 3 -2900i -2.5 2530S 2.59 2105 3.98 --- ---Me OH 3 -2900i — 2. 5 2525S 2.58 2105 3.97 — 1980 ---h 2° 3 -2900i -2.5 2510S 2.53 2025S 3.83 --- ---

Vapor 4 2975 — — -2475S --- 2120S ---- 2 0 1 2 --- 1948 ---Cyclohexane 4 3015 3.27 -2575i 3.30 2070 4.27 20 0 0 ---EtOH 4 3000 3.28 -2575i 3.32 2070 4.11 --- ---Me OH 4 3000 3.30 — 2520i 3.34 2065 4.17 --- ---

H2° 4 2980 --- -2540i --- 2045 --- 1980

B = Bandi = inflection(*) Brandt, G.R.A., Emeleus, H.J., and Hazeldine, R.N., J. Chem. Soc., 1952, 2549, have reported absorption

at 2030-2070 which presumably would correspond to absorption in this region in view of their reported extinction coefficients of —3.2.

64

differentiating between changes In the ground and excited states.

The theory of solvent effects as developed by Bayllss and McRae

envisages the shifts in the absorption spectra as the consequences of

different kinds of forces such as dispersion, dipole-polarization,26and dipole-dipole forces. Their treatment, however, does not give

adequate attention to hydrogen bonding or to other electronic effects,

such as hyperconjugative or inductive effects, which when present

dominate dipole, polarization or dispersion effects. In addition,

the fact that dimethyl sulfide, disulfide, trisulfide and tetrasulfide

are substances of low polarity Indicates that there are no very

significant solute-solvent orientation forces. It remains, therefore,

to consider the solvent effects on the short wavelength absorption

of sulfides etc. as electronic in nature. In other words, if it is

assumed that the absorption of light leads to the formation of Ionic

forms which are further stabilized by resonance involving 3d<, orbital

expansion of the sulfur atom, as Is shown below;

[CH3-S-CH3]— H- [CH2-S-CH3]and

[CH2-S-CH3]— CCH2-S-CH3]

then, the energy of the transitions will vary with the polarity and

pH of the medium in which the molecule exists. In solvents of low

concentration in hydrogen ions there is a tendency for the formation

of sulfide anion and subsequent mesomerization to the structure which

involves expansion of 3dg orbital of sulfur.The above mechanism is supported by the fact that alkyl or

aryl sulfides are readily isomerized by base to the corresponding

65

propenyl sulfides through the carbanlons as the following equations 27Indicate:

r-s-ch2ch-ch2

(-)R-S-CH-CH - CH2

C2H5ONa (-) ' C2H^OH

C 2H5OHR-S -CH-CH - CH, R-S-CH - CH-CH,

(-)R-S-CH - CH-CH'

At this point it Is perhaps worthwhile to make the following

observation: by examining the frequency shifts of disulfide, tri­

sulfide and tetrasulfide in n-hexane and cyclohexane, we can isolate26some of the influential factors considered by Bayliss and McRae,

and on that basis obtain an estimate of the relative magnitudes of

dipole moments in the excited and ground states. Indeed, if in the

general expression for the solvent shifts derived by Ooshika and28,29 30McRae, the approximations made by Ito ^ al. are applied, the

following simple expression is obtained for the solvent induced frequency

shift Av :

Av

where:B =

nD ” 1 B ------ + C2nD + 1

(M0o ) 2

D-l

D+2

<” ll > 2

nD " 1

nD + 2(14)

hC

2

hC

A3

Motf ^ l0o - Mil)

Mqq and Mt¥ are the dipole moment vectors of the solute in

the ground and excited states respectively, and A is the Onsager

reaction radius. D is the dielectric constant of the solvent and nDthe refractive index for the sodium line. Now, for nonpolar solvents,

D is nearly equal to nD and the second term in (14) vanishes. Consequently,

the sign o£ AT! will be determined by the sign of B or the relative magnitude

of and If the ground state has a greater dipole moment than the

excited state, B is positive and vice versa. The sign of B may be2determined from a plot of jv vs. n^ - 1. By comparing the values of

2 ^ 7 7

2 n2(n^ - 1 )/(2d + 1 ) for the solvents n-hexane and cyciohexane:

n-hexane: np(25°) • 1.372; (n^ - 1)/ 2n2 + 1) ■ 0.185

cyciohexane: np(25°) - 1.423; (n^ - l)/(2n^ + 1) - 0.203

with the frequencies of the transitions in the two solvents we see

that B is positive and consequently Mii.

CHAPTER VII

INVESTIGATION OF THE MECHANISM OF PHOTOLYSIS OF CYCLIC DISULFIDES

A. Introduction

Photochemical observations on cyclic disulfides are of some

Importance because one of the photosynthetic events Is supposed to

Involve a rupture of the disulfide bond of thioctlc acid. The

photolysis of dlthlolane (trimethylene disulfide) solutions has been11studied by Calvin £t jal. and the results have been used to establish

31the modes of photochemical transformations. Forbes at j l. have

studied the photolysis of cystine solutions and the products have been

identified by means of chromatography. Most of the above it) vitro

observations have been made on solutions. In the course of the

present studies on cyclic disulfides a hitherto unexplored but un­

doubtedly useful method of following the course of photochemical

changes was recognized: measurement of the short wavelength gaseous

absorption spectrum of the vapor which is in equilibrium with the

photolysed liquid. These observations also differ in that liquid

photolysis is carried out more or less in the absence of solvents.

The advantages of the methods used here are that irradiation

times are reduced considerably and photoproduct identification is

immediate. This time decrease reduced the probability of the formation

of secondary degradation products. Yet another advantage is that

67

68

photolysis of the pure substrate is readily effected. The technique

does have several disadvantages, however; the results are difficult

to make quantitative and the spectra of less volatile photoproducts

may be completely masked by strong absorption due to the more volatile

components.

B. Results

1. Tetramethylene Disulfide: The results are illustrated in

Fig. 24. Tetramethylene disulfide (Fig. 24,F), in addition to the

2860A absorption band, exhibits two broad unresolved bands with

centers at 2110A and 2020A and a few others at higher energy.

Irradiation of the cell containing pure disulfide with unfiltered

AH- 6 output produced marked changes in the spectra. The band at

2110A decreased in intensity and the band at 2020A grew in intensity;

the intensity of the bands in the shorter wavelength region remained

soim?what higher than that of the 2020A band (Fig. 24,C). It may

be noted that 1,4-butanedithiol (Fig. 24,D) exhibits a rather

pronounced band around 2020A and does not exhibit shorter wavelength

bands. All these facts suggest that the photochemical transformation

of tetramethylene disulfide results in the production of dithiol.

Similar results were obtained when carbon tetrachloride and Corning

7-54 filters were placed between the source of light and the

irradiated material.

Identical studies were made on tetramethylene disulfide which

had been doped with sodium hydroxide solution. When irradiated with

total AH- 6 output, chemical transofrmation occurred approximately

four times faster than did reaction of the undoped disulfide and the

69

FIGURE 24

A: Absorption Spectrum of the Vapors above Liquid 1,4-butanedithiolwhich had been Photolysed for 2 hours with an Unfiltered AH- 6 Source

B: Absorption Spectrum of the Vapors above Liquid TetramethyleneDisulfide Doped with NaOH which had been Photolysed for 45 Minutes with Filtered AH- 6 Source

C: Absorption Spectrum of the Vapors above Liquid TetramethyleneDisulfide which had been Photolysed for 1 hour with Filtered AH- 6 Source

D: Vapor Spectrum of 1,4-butanedithiol

E: Vapor Spectrum of Tetramethylene Sulfide

F: Vapor Spectrum of Tetramethylene Disulfide

O.D0.6

0.3

O.D.0.5

O.D

0.7

0.5

0.2

1200

600

400

1 1 0 0

eo o

19501900

W A V E L E N G T H . A

70

vapor spectrum showed different characteristics (Fig, 24,B). Tooth­

shaped bands at 2093, 2064, 2045, 2028, 2011, 1995, 1980, 1965 and

1956A appeared; the cell extract was shown to contain hydrogen sulfide

even though the gas phase did not darken lead acetate paper; there

was also some turbidity in the cell, which turbidity was taken to

indicate polymerization of the disulfide. The sharp tooth-shaped

bands at 2093, 2064 and 2045A clearly reveal the presence of tetra­

methylene sulfide (Fig. 24,E). Virtually identical results were

obtained when carbon tetrachloride and Corning 7-54 filters were

interposed in the light beam.

Photolysis of 1,4-butanedithiol (Fig. 24,D) with unfiltered

AH- 6 output was approximately two times slower than that of tetra­

methylene disulfide. When the photolysis had occurred the spectrum

exhibited bands at 2093, 2063, 2035, 2028, 2011, 1995, 1980, 1965,

1950, 1935 and 1923A (Fig. 24,A). The cell contained a large amount

of hydrogen sulfide. When the dithiol was doped with alkali, as

previously, photodecomposition proceeded approximately twelve times

faster and the spectrum showed very sharp bands at 2093, 2064 and

2036A and diffuse bands at other positions (virtually identical to

those of Fig. 24,A). The nearly identical spectra of the vapors

above irradiated alkali-treated tetramethylene disulfide and those

above the irradiated (doped and undoped) 1 ,4-butanedithiol strongly

suggest that the volatile products are identical in both cases.

It appears relatively well substantiated then that pure

liquid tetramethylene disulfide, when photolysed at 2860A, produces

1,4-butanedithiol. It seems equally well proven that alkali-doped

71

tetramethylene disulfide, when photolized at 286QA, produces

tetramethylene sulfide. Photolysis of 1,4-butanedithiol, in the

presence or absence of alkali dopant, with light of 2020A, produces

tetramethylene sulfide, the rate of such production being signifi­

cantly faster in the presence of alkali.

2. Pentamethylene Disulfide: Results are shown in Fig. 25 and

are similar to, but not as clear as those obtained in the case of tetra­

methylene disulfide. Pentamethylene disulfide, in addition to the

2650A absorption band which is not shown, exhibits two broad bands

(Fig. 25,D) at 2060 and 2000A and a diffuse band at 1915A. When

irradiated with the AH- 6 lamp (with or without the Corning 7-54 filter)

the 2060A band disappeared and there was an abnormal increase in the

intensity of the shorter wavelength absorption region (Fig. 25,A).

The 1,5-pentanedithiol spectrum exhibits a band maximum at 2020A. All

these facts suggest that 1 ,5-pentanedithiol is produced in this case.

When the disulfide was doped with alkali and Irradiated

(with or without the Corning 7-54 filter) the decomposition proceeded

at least three times faster than when the alkali was absent. The

spectrum obtained after photo-decomposition of alkali-doped disulfide

showed a distinct band at 2082A and diffuse bands at 2040, 2030,

1960 and 1920A (Fig. 25,B). There was also extensive polymerization.

The extract of the contents of the cell imparted a yellow color to

lead acetate paper. The appearance of bands at 2082, 2040, 2030,

1960 and 1920A is certainly suggestive of the formation of pentamethylene

sulfide (Fig. 25,C).

72

FIGURE 25

A: Absorption Spectrum of Vapors above Liquid PentamethyleneDisulfide which had been Photolysed for 30 Minutes withFiltered AH- 6 Source

B: Absorption Spectrum of Vapors above Liquid PentamethyleneDisulfide Doped with NaOH which had been Photolysed for 30 Minutes with Filtered AH- 6 Source

C: Vapor Spectrum of Pentamethylene Sulfide

D: Vapor Spectrum of Pentamethylene Disulfide

0. 0.

o.»O.D.

0.6

900

600

300

2200

1900

1600

•300

210020 00•900

WAVELENGTH, A

73

C. Discussion

It Is proper to seek some understanding as to why the cyclic

disulfides studied should produce the corresponding cyclic monosulfides

in the presence of alkali at greatly enhanced rates, while yielding

only the dithlols in its absence. It would also appear proper to

assume that a somewhat general photochemical rule for solution or any

condensed medium might read somewhat as follows: Photodecomposition

will proceed primarily from the lowest excited state unless the

irradiation energy exceeds the ionization limit; furthermore, dependent

on the magnitude of intersystem crossing rate constants, this lowest

excited state may be singlet, triplet, or perhaps both. Unfortunately,

nothing is known of the triplet states of smaller sulfur molecules

since such molecules neither phosphoresce nor show any susceptibility

to triplet enhancement techniques. Consequently, any discussion must

be couched in such knowledge of the singlet states of these molecules

as exists and even this information is very scanty and certainly

incomplete. Furthermore conclusions with regard to singlet states need

not necessarily apply even to triplet states of the same orbital electronic

configuration.

A molecular orbital scheme of one-electron energies for a

-C^-S^-Sg-Cg- system, is shown in Fig. 26,A; the notation used is more

or less self-evident and it is understood that the -S-S dihedral angleois not 90 . The tt_ orbital is the out-of-phase combination of the two

3p orbitals of the sulfur atoms. The ground state configurations is:TT

<VsA>2 %-s/ <Vsb)2 < 2 ("-)2 (a)

32 33and the lowest excited configuration is: ’

Sa -Sb 5' < b >

and is primarily descriptive of the states achieved by 2860A absorption

in tetramethylene disulfide and 2560A absorption in pentamethylene

disulfide. Since this excitation involves removal of a sulfur non­

bonding electron and its placement in an S/^-Sg antibonding orbital, and

since it is the lowest energy excitation leading to a singlet state, such

excitation should result in scission of the SA-SB bond, and indeed this

is what is observed. Mechanistically, the reaction probably proceeds

through initial formation of a diradical, subsequent hydrolysis by

moisture contamination to the mercapto-sulfenic acid and finally the

dithiol in the fashion suggested by Barltrop, Hayes, and Calvin for the

photolysis of trimethylene disulfide in solution.

The configurations:

(T1.)1 (C’C a -Sa )1 <c >

<'*cB-sB>1are degenerate and will exhibit a first order configuration interaction.

Consequently an additional absorption is expected in the cyclic disulfides*and will lie at higher energy than the tt —» o Q « absorption band. In

A B

the tetra- and pentamethylene dithiols where SA and are separated by*

large distances,tt _ and tt+ will be degenerate, <?sA -sB and ° Sa-Sb wifi no

longer exist and, as shown in the M.O. energy level diagram of the dithiol

given in Fig. 26,B, the lowest energy configurational excitation will be:

75

FIGURE 26

A: Schematic One-Electron MO Energy Diagram of a "C^-S^-Sg-Cg- System.3pg Indicates Energy of the 3p Atomic Orbital of Sulfur.The Right Subscripts on the Carbons and Sulfurs are Position- Ordering Indices.

B: Schematic One-Electron MO Energy Diagram of a Dithiol

C: Schematic One-Electron MO Energy Diagram of an Anionic Species in which a Negative Charge is Supposed to Reside on One Sulfur.The 3pz orbital is Perpendicular to a Given C-SH Plane, while the 3py Orbital is in that Plane. As the Exact Location of the 3py Level is Somewhat questionable, It is Shown in both Possible Positions.

A

3A

P A f

<7C-S^ B =} ^c-s

■ < £ s‘TT B

■ * *7T-> + '

B C

HSCH ) SH2 4

* S-Hcrv «

% h

% h

8 * * ----- X * 7Tj

yyygn}-7r-------------- *x— 7r°z XX— 1A

'B

5 * 3 0 = 5 - 0 ^ - ---------------- XX------ o r _ g

XXX X ■}~(ji------- XX---S.-H

76

Since such excitation involves population of a C-S antibonding orbital

it should result in weakening of the C-S bond. Consequently, excitation

of the dithiol should result in preferential C-S scission either in the

presence or absence of alkali. Excitation of the dithiol is primarily

due to the absorption at 2020A. However, in mercaptans a weaker

absorption is generally detectable at ~2300A; its extinction coefficient

is so low ( e <50) that it may be due to impurity, but if so it is a

remarkably tenacious and ubiquitous contaminant. If this 2300A absorption

be real, it is tempting to assign it as transition (e) above, in which

case the 2020A band might be assigned as:

<V4 ••••(V3 <£> (TTZ )4 ....(TT2 ) 3 (43s) 1 (g)

where the 4sg orbital in question is located on the sulfur atom. It is

not possible to decide between these two alternatives. If the 2300A

absorption be real, excitation would be effected by transitions (f) or

(g) at 2020A; the resulting state would internally convert to the: 3 * 1. • • • (tt ) state, from whence both photochemical C-S scission,

and considerable internal conversion to the ground state would occur. If

the 2300A absorption of the dithiol be spurious then the 2020A absorption

must be assigned to transition (e), in which case, this being the lowest

singlet state, both C-S scission and considerable internal conversion

to the ground state should occur. In other words, the conclusions are

independent of whether the 2300A absorption is real or is due to impurity;

the primary neglect is composed of triplet states and the intersystem

crossing processes which populate them.

77

Let us now ask why the rate of dithiol photolysis is increased

in the presence of alkali dopant. Let us presume that in the presence of

alkali, the initial step in the reaction is the removal of a proton:

- (ch2 V + OH — <ch 2yr

S .H ! !;ah

Where the subscripts are used only to differentiate similar atomic centers.

A schematic MO energy level diagram descriptive of the negative ion is

shown in Fig. 26,C; the orbitals involving will remain unchanged in

energy, while those involving Sg will probably change as shown because

of the extra electron density now resident on this sulfur atom. The lowest

energy excitation should now correspond to the transition:

• • • • ^ B - S B * 00and if the contention that the Cg-Sg bond is weakened relative to that

in the dithiol be correct, scission of this bond should occur more

readily. It is also predicted that the energy of this transition,

relative to that in dithiol, will decrease, and indeed large red shifts

of the 2020A band are observed when dithiol is dissolved in alkaline

solutions. Since photolysis was effected with an AH- 6 lamp, the output

of which is very low at 2020A, it is felt that the increased energy

intake due to the redshift of the 2020A band in the alkali-doped

medium coupled with the proposed weakening of the Cg-Sg sigma bond are

sufficient to account for the observed photolytic rate increase.

The next question concerns the reaction which takes place when

doped tetra- and penta-methylene disulfides are photolized in the region

of their long wavelength absorption bands (2860 and 2560A respectively):

78

(CH~ )2 'IT_h^OH

i-S-

In the absence of alkali such photolysis produces the dithiol; photolysis

of the dithiol to the cyclic monosulfide occurs rapidly in alkaline media

when irradiated with the unfiltered output of the AH- 6 source. One might

then suggest that the overall reaction in the doped medium is a two

photon process: cyclic disulfide dithiol cyclic sulfide. Unfortu­

nately, dlthiols in alkaline or non-alkaline media do not absorb at

2860 or 2560A as far as is known, and such a process seems impossible

under the conditions of irradiation used by us. One could of course

still contend that the dithiol dark-reacts to the cyclic sulfide, but

since this has been observed not to occur at any appreciable speed in

the dark, one is forced to conclude that the dithiol produced in the

initial photoiytic step must contain considerable vibrational energy.

It is not possible to positively refute this suggestion, but it does seem

better to seek an alternative interpretation. We accordingly postulate

a radical mechanism, the initial step in which implies a C-S scission.

■(CH- >2 nrCH'

S-S-

CH, _hvJ °H.. E ■<c h2)2— | i-<CH2>r|+

} yT 2 — s— 1s-sI H2°

I ^ **2 ^ + H2S +

The removed sulfur may be very reactive at the time of removal and

may initiate polymerization of the material. It may also react with the

79

photo-products to give polysulfides. Calvin's observation that irradiated

alkaline solutions of dithiolane in ethanol exhibit a band at about11 31250QA may be due to such polysulfides. Forbes et al. have similarly

suggested that irradiation of cystine solutions at pH values above S

causes a preferential C-S scission. It appears that we are forced to

concur. Let us now ask if we can provide any MO rationale for the

occurrence of such C-S scission. Consider the two excited configurations

of the cyclic disulfide (c) and (d) above, in which a nonbonding sulfur

electron has been promoted into a C-S antibonding orbital. Some first

order mixing of these two configurations occurs; we will designate the

lower of the two components as (c+d) without regard to phase. It seems

not unreasonable to assign one or the other of the 2110 or 2020A

absorptions of tetramethylene disulfide (2080 or 2000A of pentamethylene2 1 1disulfide) to the transition (n_) (tt_) (c+d) . In the presence

of alkali, according to the usual conceptions, the -C^- group is

polarized and the hyperconjugated structure -CH ■ S-S- is supposed tok+

contribute significantly to the ground state. The resulting distortion,

if it is also present in the excited states, may increase considerably1 1 1 * 1 the mixing of the two configurations C7. ) (c+d) and..... (*rr_) (Og_g) ,

and thereby introduce a considerable component of antibonding C-S character

into the lowest excited singlet state of the cyclic disulfides. The

corresponding C-S scission is not then totally unexpected. The above

suggestions are of some interest in the case of the photolysis of bls-

trifluoromethyl disulfide in non-alkaline media. The initial step in

this photolysis has been suggested to be S-S scission;^ it could with

equal validity be C-S scission. Electron attracting groups such as

80

CF^ attached to the S-S bond considerably increase the strength of this

bond and produce a considerable blue-shift of the lowest energy★

absorption band. The consequent approach energetically of o and

excitations increases the previously specified configuration

interaction and may even cause the lowest excited state of this molecule

to have predominant C-S antibonding character. Such scission could

account satisfactorily for the observed photolytic products.

CHAPTER VIII

CONCLUSIONS

The modified Wolfsberg-Helmholz procedure used here appears

to be capable of adequately predicting the eigenvalues for the lower

lying levels in compounds having a single sulfur atom but runs into

difficulties when additional sulfurs are added. Although the bases for

most correlations were of necessity somewhat tenuous in nature, the

agreement between prediction and experiment was found to be good in

those few cases where vibrational structure was present to serve as an

additional check on assignments. Intensity computations would be an

invaluable tool in further verifying the predictions.

The difficulties attendant to applying the method to polysulfide

systems and obtaining improved monosulfide excitation energies might

be surmounted by making the calculations self-consistent with respect

to charge. It was pointed out earlier that coulomb terms (and therefore

resonance integrals) were obtained from neutral atom valence state

ionization potentials and that the large and values

produced the extremely low lying excited states in the polysulfides.

Calculations performed recently on water indicate that making the

computations self-consistent would tend to transpose the entire set of

3dg levels to higher energies through alteration of the coulomb terms

rather than affecting the overlaps and thereby the splitting.

82

As a first approximation, treatment of the long wavelength

transition in polysulfides by the free electron approach yields quite

reasonable results. This would infer that the 4Sg levels are indeed

delocalized to a considerable extent in compounds containing two or

more sulfurs.

The computations indicate that although improved calculations

might result in the interchanging of some levels the first few members

of the Rydbergs (3dg and 4sg) will continue to be extremely important

in any considerations of excited states particularly in accounting for

the lower energy excitations. It is rather surprising and somewhat

open to question that transitions to the levels that are primarily anti­

bonding in nature were predicted to be at such high energies (7-12 e.v.).

Another rather debatable conclusion suggested by the calculations

is in regard to the r^le of 3d_ orbitals in bonding in the polysulfides.OThe usual explanation is that sulfur expands its octet by placing

additional electrons in a vacant 3dg orbital. These results indicate

that as a result of large splittings, the highest filled level in

bivalent polysulfides is primarily 3d_ in character while a close lying

excited state contains a significant 3pg contribution.

SELECTED BIBLIOGRAPHY

1. Price, W. C., J. Chem. Phys., 4, 147 (1936),

2. Price, W. C., Teegan, J. P., and Walsh, A, D., Proc. Roy. Soc.,A201. 600 (1950).

3. Clark, L. B., Doctoral Thesis, University of Washington, 1963.

4. Wolfsberg, M., and Helmholz, L., J. Chem. Phys., 20, 837 (1952).

5. Brandt, G.R.A., Emel/us, H. J., and Haszeldlne, R. N., J. Chem. Soc.,1952, 2549.

6 . Minoura, Y., J. Chem. Soc., Japan. Pure Chem. Sect., 73, 131, 244(1952).

7. Feher, T. and Munzner, H. , Bet., j)6 , 1131 (1963).

8 . Jordan, T. E., "Vapor Pressure of Organic Compounds," IntersciencePubl., New York, 1954, Pp. 224.

9. Strecker, W., Ber.. 41. 1105 (1908).

10. Feher, T., Krause, G. and Vogelbruch, K., Chem. Ber., 90, 1570(1957).

11. Barltrop, J. A., Hayes, P. M., and Calvin, M., J. Am. Chem. Soc.,76, 4348 (1954).

12. Moffitt, W,, Phys. Soc. Repts.. Prog. Phys., 17. 173 (1954).

13. Vlste, A. and Gray, H. B., J. Inorg. Chem.. _3, 1113 (1964).

14. Moore, C. E., "Atomic Energy Levels," N.B.S, Circular 467, 1949 and1952.

15. Mulliken, R. S., J. Chim. Phys.. 46, 497, 675 (1949).

16. Streitwieser, A., "Molecular Orbital Theory for Organic Chemists,"Wiley, New York, 1964.

17. Cusachs, C., Private Communication.

83

84

18. Ballhausen, C. J, and Gray, H. B., "Molecular Orbital Theory,"Benjamin, New York, 1964, Pp . 118.

19. Ballhausen, C. J. and Gray, H. B. , Inorg. Chem. . _1, 111 (1962).

20. Mulliken, R. S., Rieke, C. A., Orloff, D. and Orloff, H., J. Chem.Phys., 17, 1248 (1949).

21. Slater, J. C,, Phys. Rev.. 36, 57 (1930).

22. Herzberg, G., "Molecular Spectra and Molecular Structure.II. Infrared and Raman Spectra of Polyatomic Molecules," D. Van Nostrand Co., Princeton, 1945, Pp. 161.

23. Bergson, G., Doctoral Thesis, University of Uppsala, 1962.

24. Koch, H. P., J. Chem. Soc., 1952, 387.

25. Bass, A. M., J. Chem. Phys.. 21, 80 (1953).

26. Bayliss, N. and McRae, E. G., J. Phys. Chem., 58, 1002 (1954).

27. Cilento, G., Chem. Revs., 60, 147 (1960).

28. Ooshika, Y., J. Phys. Soc. Japan. 594 (1954).

29. McRae, E. G., J. Phys. Chem., 61. 862 (1957).

30. Ito, M., Iunzuka, K. and Imaniski, S., J. Am. Chem. Soc., 82,1317 (1960).

31. Forbes, W. F., and Savige, W. E., Photochem. Photobiol., 1 (1962);Bogle, G. S., Burgess, V. R., Forbes, W. F. and Savige, W. E., ibid. I, 277 (1962).

32. McGlynn, S. P., Nag-Chaudhurl, J. and Good, M., J. Am. Chem. Soc.,84, 9 (1962).

33. Schoberl, A. and Grafge, H., Angew. Chem., 69, 713 (1957).

VITA

Samuel Donald Thompson was born in Mobile, Alabama on July 31,

1935. He attended public schools in Mobile; Franklinton, Louisiana; and

North Little Rock, Arkansas. After one and one-half years of college,

he entered the service and was sent to England where he met and wed

the charming Jacqueline Haycock. For the last seven and one-half

years, he has been occupied with attending college and begetting

children. He has met with more success in the latter area--he now

has at least three children: Shaun, Vanessa, and Richard: but only

one degree--B.S., Little Rock University, 1960. He is presently a

candidate for the Doctor of Philosophy degree in chemistry at Louisiana

State University.

85

EXAMINATION AND THESIS REPORT

Candidate: Samuel Donald Thompson

Major Field: Chemistry

Title of Thesis: An Investigation of the Electronic Spectra of a Series of

Organic Sulfur CompoundsApproved:

__ ____Major Professor and Chairman

Dean of the Graduate School

EXAMINING COMMITTEE:

Date of Examination:

July 27, 1965