an investigation of the electronic spectra of a series of
Post on 01-Jan-2022
5 Views
Preview:
TRANSCRIPT
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
top related