Products and Mechanisms in the Oxidation of Phosphorus by Sulfur at Low Temperature

Article abstract of DOI:10.1021/ic9614879

The oxidation of phosphorus by sulfur at low temperatures (< 100 °C) has been shown to produce complex mixtures that include 12 of the 17 known binary phosphorus sulfides. Two of these 17 sulfldes have been observed in these mixtures for the first time. The rate-determining and first step in the reaction appears to be the formation of the S₈ diradical. This proposal is supported by the observed rate at which the sulfldes are formed and the distribution of the sulfide product stoichiometries. Photoinitiation of this oxidation at 0 °C produces a similar array of sulfldes. The differences in the product distributions between the thermal and photochemical processes facilitate the understanding of the mother-daughter relationships between the products. The effects of oxidation by sulfur and reduction by phosphorus have been determined for several of the known phosphorus sulfides. The characterizations of phosphorus compounds in molten mixtures of phosphorus and sulfur were performed by ₃₁P NMR and Raman spectroscopy directly on the reaction mixtures.

Full text of DOI:10.1021/ic9614879

Inorg. Chem. 1997, 36, 2633-2640
2633
Products and Mechanisms in the Oxidation of Phosphorus by Sulfur at Low Temperature
Mark E. Jason,* Toan Ngo, and Shahidur Rahman
Monsanto Company, Performance Materials, 800 N. Lindbergh Blvd., T3W, St. Louis, Missouri 63167
ReceiVed December 13, 1996X
The oxidation of phosphorus by sulfur at low temperatures (<100 °C) has been shown to produce complex mixtures
that include 12 of the 17 known binary phosphorus sulfides. Two of these 17 sulfides have been observed in
these mixtures for the first time. The rate-determining and first step in the reaction appears to be the formation
of the S8 diradical. This proposal is supported by the observed rate at which the sulfides are formed and the
distribution of the sulfide product stoichiometries. Photoinitiation of this oxidation at 0 °C produces a similar
array of sulfides. The differences in the product distributions between the thermal and photochemical processes
facilitate the understanding of the mother-daughter relationships between the products. The effects of oxidation
by sulfur and reduction by phosphorus have been determined for several of the known phosphorus sulfides. The
characterizations of phosphorus compounds in molten mixtures of phosphorus and sulfur were performed by 31
NMR and Raman spectroscopy directly on the reaction mixtures.
P
Introduction
The phosphorus sulfides are a class of compounds whose
roots go back to the earliest experimentation with these two
elements.1,2 The richness and complexity of these compounds
and their chemistry recently returned as an area of increased
activity. The work of Bjorholm and Jakobsen3 is noteworthy
for both the identification of several new phosphorus sulfides
and the use of chemical exchange NMR spectroscopy to
establish the structural relationships between pairs of intercon-
verting compounds. Their work has become invaluable for its
assignments of both structures and NMR chemical shifts.
Recently, Blachnik4 has extended the family of known phos-
phorus sulfides by two members through the analysis of
quenched samples from high-temperature equilibrations of
phosphorus-sulfur mixtures.
A common analytical theme in these recent developments
has been the use of 31P NMR spectroscopy to analyze mixtures
of phosphorus sulfides that contain from trace to high concen-
trations of the various structures. Prior to 1991 there were 10
sulfides known that had been characterized either by X-ray
diffraction or by convincing application of NMR spectroscopy.
Since that time, five new sulfides have been added to the
collection by Bjorholm and Blachnik. Two more will be added
with this report. All of the phosphorus sulfides that have been
structurally characterized are shown in Figure 1.
The nomenclature of the phosphorus sulfides is muddled by
a combination of Greek letter prefixes and Roman numeral
suffixes. This communication uses the older practice of
attaching a lower case Greek letter to those sulfides that display
more than one isomer, advancing through the alphabet in the
order in which they were reported. The names accompany the
structures in Figure 1. The structures are uniquely named, and
Figure 1. Known and proposed phosphorus sulfides.
X Abstract published in AdVance ACS Abstracts, May 15, 1997.
(1) Hoffman, H.; Becke-Goering, M. In Topics in Phosphorus Chemistry;
Griffith, E. J., Grayson, M., Eds.; Interscience Publishers: New York,
1976; Vol. 8, pp 193-271.
all of the old prefixed names are retained. Every attempt to
design names that were structurally informative made them too
long to be useful as nicknames.
Given the age of this class of compounds, it is remarkable
how little of its reaction chemistry has been studied. Until
recently the focus of most of the work with phosphorus sulfides
has been the phase chemistry of phosphorus and sulfur and the
structural and spectroscopic characterization of stable species.
Mechanistic chemistry has been nonexistent. In the case of the
(2) Mellor, J. W. A ComprehensiVe Treatise on Inorganic and Theoretical
Chemistry; Longmans, Green and Co.: London, 1931; Vol. VIII, pp
1047-1080.
(3) Bjorholm, T.; Jakobsen, H. J. J. Am. Chem. Soc. 1991, 113, 27-32.
(4) Blachnik, R.; Peukert, U.; Czediwoda, A.; Engelen, B.; Boldt, K. Z.
Anorg. Allg. Chem. 1995, 621, 1637-1643.
(5) Van Wazer, J. R. Phosphorus and Its Compounds; Interscience
Publishers, Inc.: New York, 1958; Vol. 1.
S0020-1669(96)01487-5 CCC: $14.00 © 1997 American Chemical Society

2634 Inorganic Chemistry, Vol. 36, No. 12, 1997
Jason et al.
The program gNMR (Cherwell, Ltd., available from SoftShell) was
used in the same manner as the other pair. All of the chemical shifts
and coupling constants for the species reported in this study were
obtained by fitting the spectroscopic data with either of these two
programs. A table of this data can be found in the Supporting
Information.
Reaction of 1:1/8 P4/S8 at 80 °C in an NMR Tube. A mixture of
3.54 g (0.114 mol P) of P4 that had been filtered through charcoal and
0.916 g (0.0286 mol of S) of S8 was prepared in a Schlenk tube under
argon. The material was heated until it was molten. In a glovebag,
0.7 mL samples were removed with a syringe and filtered directly into
NMR tubes containing a DMSO-d6 capillary insert. The tubes were
sealed at 12 Torr.
In one experiment, one of the NMR tubes containing P4/S8 was
heated in a mechanically stirred oil bath 80 ( 0.2 °C. The tube was
removed from the bath and cooled to room temperature to record a
spectrum; the spectrometer was operated at 25 °C. Spectra were taken
before the heating began and after 30, 120, 240, 360, 1320, 2400, and
3720 cumulative minutes at 80 °C. The tube was then placed in a 150
°C oil bath for 50 min and 3960 cumulative minutes. The composition
of this mixture is described in the Results section.
Another tube was used to perform an 80 °C experiment directly in
an NMR probe, so that the tube did not have to be removed during the
experiment. Each spectrum was the result of 30 min of accumulation.
After 7 h of data collection, the tube was removed from the spectrometer
and placed in an 80 °C oil bath. The tube was monitored continuously
to determine if equilibration to P4S3 had occurred. See the Results
and Discussion sections for details. Representative NMR spectra from
these studies can be found in the Supporting Information.
NMR Study of the Reactions of 1:n P4/S8 Mixtures Where n >
5/8. Various amounts of crushed sulfur were added to each of eight
NMR tubes. A mixture of 1:1/8 P4/S8 composition was prepared from
6.80 g of white phosphorus and 1.80 g of sulfur. The NMR tubes
were filled as described above. One tube was also prepared with only
the 1:1/8 P4/S8 material. The tubes were sealed under vacuum and heated
in a water bath set at 80 and later 90 °C. Because of the slurry content
and high viscosity, the NMR spectra of all samples except the 1:1/8
P4/S8 had to be obtained at 80 °C. Results of these experiments are
presented in the Results and Discussion sections.
Reactions of Phosphorus Sulfides in Liquid Phosphorus (P4). The
following procedure was typical: In a glovebag, 5 mg of R-P4S5 was
placed in an NMR tube followed by a piece of solid white phosphorus
weighing approximately 1.0 g. Phosphorus was melted into the lower
portion of the tube and the tube sealed under vacuum. The 31P NMR
spectrum of the mixture was recorded at 60 °C. The NMR tube was
heated in an oil bath at 90 °C for 8 h and the NMR spectrum again
recorded.
phosphorus sulfides, novel and intriguing paths must exist that
interconvert the members of the class, and these paths have
never been explored. Sulfur exists in these structures in two
forms: as part of the cage (endocyclic) and as a thiono group
(exocyclic). Phosphorus appears in the formal oxidation states
from zero to five, inclusive.
This work is the beginning of a search for the transformations
and mechanisms involved in the bimolecular reactions of
phosphorus and sulfur. During the course of this effort, two
new phosphorus sulfides have been identified. Several mecha-
nistic explanations for the initially formed product distribution
observed in the low-temperature (80 °C) reaction of P4 and S8
are also outlined. The most likely of these, on the basis of
present observations, can be tied to the accepted mechanism
for the polymerization chemistry of S8.
Experimental Section
General Information. The raw materials for these studies included
white phosphorus, sulfur, and commercial samples of P4S3 (2) and P4S10
(18). The phosphorus was obtained from Monsanto Co. Some of the
described reactions used phosphorus that had been passed over a column
of activated carbon. No difference in chemistry was observed between
the standard and the purified versions of white phosphorus. Sulfur
was taken from three sources: Monsanto Co., Fisher Scientific, and
99.998% purity grade from Aldrich Chemical Co. There has been no
evidence of a difference among any of these sulfur samples. P4S3 was
obtained from Fluka Chemical Co. It was recrystallized from CS2 prior
to use. P4S10 was purified by Soxhlet extraction; the process is
described below. Carbon disulfide used for spectroscopy was used as
received (Fisher); the carbon disulfide used as solvent for chemical
reactions was dried by refluxing over P4S10. R-P4S5 (5),6 â-P4S5 (6),7
R-P4S7 (11),8 and R-P4S9 (16)9 were prepared according to literature
procedures. All work was performed using either Schlenk or glovebox
techniques. Because of the fact that P4 is sensitive to light,5 precautions
were taken to make certain the long exposure reactions were protected
from room light. Red phosphorus was not produced in the thermal
experiments or by exposure of the samples to room light after they
had undergone extensive reaction.
The NMR spectroscopy was performed on both 300 and 400 MHz
(proton) spectrometers, all from Varian. Lock capability was obtained
by using sealed melting point capillaries containing DMSO-d6. The
melting point capillaries were then used in standard 5 mm NMR tubes,
without concern for centering the capillary. The spectra obtained under
these conditions usually had adequate, and sometimes outstanding,
resolution. If the spectrum of both P4 and the phosphorus sulfides were
to be observed in the same experiment, only the 300 MHz instruments
could be used because the sweep width limitations on the 400 MHz
instruments would not allow establishing the necessary 800 ppm
window. Spectroscopy on samples in liquid P4 or P4/S8 alloys
performed on the 400 MHz instruments was commonly conducted with
presaturation of the phosphorus peak. A delay between pulses of 2 s
and standard homonuclear phosphorus decoupler power was employed.
The decoupler was kept on during the delay and turned off during the
pulse and acquisition. Referencing of the spectra was not performed.
Instead, the instrumental offset for DMSO-d6 was held constant. The
chemical shifts were reproducible within 0.5 ppm, considered acceptable
for this application.
The system P4S3/P4 was more extensively investigated. Dilute
solutions of P4S3 in P4 were prepared and tested as above. In addition,
mixtures with the composition P4S2 (2:1 P4S3/P4), which are liquid at
room temperature, were prepared. The latter mixtures were heated at
94 °C for 70 h and at 250 °C for 5 h.
Reactions of Phosphorus Sulfides in Liquid 1:1/8 P4/S8. A liquid
mixture of 13.55 g of phosphorus (0.438 mol of P) and 3.51 g of sulfur
(0.109 mol of S) was prepared. NMR tubes adapted for sealing under
vacuum were loaded with a DMSO-d6 capillary insert and 3-5 mg of
phosphorus sulfide (unweighed). The tubes were flushed with nitrogen.
The phosphorus-sulfur mixture was filtered into the NMR tubes in a
glovebag. The sealed tubes were heated at 80 °C for 30 min and quickly
cooled; the 31P NMR spectra were recorded at 25 °C. The tubes were
heated for an additional 90 min and the spectra retaken. Analysis of
the spectra is presented in the Results and in the Discussion sections.
Photochemical Initiation of the Oxidation of Phosphorus by
Sulfur. NMR tubes containing a 1:1/8 P4/S8 mixture and capillary insert
sealed under vacuum were prepared in the manner described above
and also by distilling P4 at 0.2 Torr onto sulfur and then filling NMR
tubes that had been sealed to the receiver. The tubes were kept in the
dark at 4 °C (i.e., in a refrigerator). Photolyses were performed by
placing a tube in a circulating 0 °C bath. Although this temperature is
below the liquidus curve for the phosphorus-sulfur system, the tube
could be maintained unsolidified for several hours if undisturbed. The
temperature during the photolyses was held in the range 0.5-3.0 °C.
Spin simulation and refinement calculations were performed using
two different methods: the Serena Software program PMR and the
PC version of LAOCN-510 were used in tandem for the early work.
(6) Brauer, G. Handbook of PreparatiVe Inorganic Chemistry; Ferdinand
Enke Verlak: Stuttgart, Germany, 1960; p 506.
(7) Bues, W.; Somer, M.; Brockner, W. Z. Anorg. Allg. Chem. 1981, 476,
153-158.
(8) Thamm, R.; Heckmann, G.; Fluck, E. Phosphorus Sulfur 1981, 11,
273-278.
(9) Brylewicz, Z.; Rudnicki, R. Phosphorus, Sulfur Silicon 1994, 89, 173-
179.
(10) Program QCMP-049, available from QCPE, University of Indiana,
Bloomington, IN.

Oxidation of Phosphorus by Sulfur
Inorganic Chemistry, Vol. 36, No. 12, 1997 2635
Irradiation came from a 150 W incandescent spotlight bulb positioned
5 cm from the tube. The 31P NMR spectrum was recorded as a function
of time. The length of time the sample spent at room temperature was
held to a minimum; however, the NMR spectra had to be recorded at
room temperature in order to prevent the precipitation of R-P4S7 and
the solidification of the reaction mixture. These “high temperature
excursions” were limited to 30 min and a temperature maximum of
25 °C.
In one such experiment, the tube was irradiated for 3 h at low
temperature, and the 31P NMR spectrum was recorded. The tube was
then heated at 80 °C for 1 h, the sample cooled, and the spectrum
retaken. This experiment was used to judge the thermal reactions of
the sulfides produced by photolysis and is discussed below.
Raman Spectroscopic Investigation of the Reaction of Phospho-
rus and Sulfur. Sulfur and phosphorus, in the desired stoichiometry,
were added as solids to a quartz vessel that consisted of a 49 mm tube
having three ground joints sealed to the top. A 2.5 mL cup with a 1
cm quartz flat as a bottom was sealed to the bottom of the tube. A
mechanical paddle stirrer was used to agitate the mixture. The cup
was wrapped with nichrome wire; a J-Kem temperature controller was
used to maintain temperature. The thermocouple, placed in the molten
mixture, was stainless steel coated with tantalum. Sufficient material
was used to fill the cup when the stirrer and thermocouple were in
place. The fiber optic probe from the Raman spectrometer was placed
beneath the tube, touching the quartz flat. Spectra were recorded at
regular time intervals during a run that lasted 8-24 h.
1
Figure 2. Product profile for the reaction of P4 + /8 S8 at 80 °C.
Only the six most abundant products are shown.
Results
Identification of Phosphorus Sulfide Structures. All
structural characterizations were performed by 31P NMR
spectroscopy. The chemical shifts and coupling constants of
the known sulfides were used as a starting point. The
differences between the chemical shifts measured in liquid
phosphorus/sulfur melts and those measured in carbon disulfide
fall in the range from +15.7 to -14.7 ppm, depending on the
structure. The chemical shifts observed in liquid phosphorus
more closely resemble those observed in CS2 as opposed to those
observed by Bjorholm and Jakobsen in molten “P4S4”.3 Though
the chemical shifts were sensitive to the “solvent”, there was
very little variation in the measured coupling constants. The
largest difference, approximately 10 Hz, was observed in R-P4S6
(8). It was the close correspondence of the coupling constants
that allowed for straightforward identification of previously
reported phosphorus sulfides. The assignment of the P4S10
singlet was based on the spectrum of a dilute solution of R-P4S9
and P4S10 dissolved in P4. The NMR spectra of the two new
sulfides were pieced together from unassigned coupling patterns
and refined. A discussion of the spectra and a proof of structure
based on reaction chemistry is presented in the accompanying
paper. All of the measured chemical shifts and coupling
constants (in CS2, P4, and, where reported, “P4S4”) are found
in the Supporting Information.
Figure 3. First 7 h of the reaction shown in Figure 2, showing the
“primary” products. See Discussion for details.
There are a few peaks in the 31P NMR spectra of the materials
Figure 4. First 7 h of the reaction shown in Figure 2, showing the
“secondary” products. See Discussion for details.
produced by reaction at 80 °C that remain unassigned.
A
number of small and often broad singlets were observed in the
region from 20 to -5 ppm, consistent with phosphate and
phosphite species. A singlet often appeared at approximately
19.5 ppm for which an assignment has not been made. The
chemical shift is reminiscent of a peak found in high sulfur
P-S melts.11
reaction mixtures. These are as follows: P4S3 (2), R-P4S5 (5),
R-P4S6 (8), â-P4S6 (9), γ-P4S6 (10), R-P4S7 (11), â-P4S7 (12),
R-P4S8 (13), â-P4S8 (14), γ-P4S8 (15), R-P4S9 (16), and P4S10
(18). R-P4S6 (8) is the most rapidly formed phosphorus sulfide
during the first several hours of the reaction. Formed at nearly
the same rate is R-P4S7. Finally, the third major product is
R-P4S9. Although the concentration of most of the sulfides
continue to grow during the reaction, that of R-P4S7 levels off
at approximately 6 h, a result of the low solubility of R-P4S7 in
any solvent including liquid phosphorus. The crystals of R-P4S7
that grew in the static 1:1/8 P4/S8 reaction mixtures were large
and very well formed. No other sulfide was observed to
precipitate from solution during the 80 °C portion of the reaction.
Reactions at Stoichiometry 1:1/8 P4/S8. Sulfur as a dilute
solution in liquid phosphorus reacts with phosphorus in a slow
and controlled fashion to give a complex mixture of phosphorus
sulfides. The product distribution of such a reaction at 80 °C
as a function of time is shown in Figures 2-4. In total, 12 of
the 17 known or suspected sulfides appear as products in these
(11) Demarcq, M. C. Phosphorus Sulfur 1987, 33, 127-134.

2636 Inorganic Chemistry, Vol. 36, No. 12, 1997
Jason et al.
The composition of the mixtures produced during the first
several hours was difficult to obtain with good precision. The
T1’s of the sulfides in these mixtures are difficult to measure
because of their low concentrations and the complexity of the
spectra; delay time was also sacrificed in order to collect
sufficient transients to obtain usable spectroscopy. The integrals
often had to span coupling patterns that were more than 6 ppm
wide (over 960 Hz at a proton frequency of 400 MHz). Thus,
baseline drift and phasing problems presented a continual source
of error. The last limitation was posed by the underdigitization
of spectra taken at very wide sweep widths. However, the
smoothness of many of the curves in Figure 2, and the flat
concentration profile for R-P4S7, indicate these errors do not
overwhelm the task of studying this reaction.
When an NMR tube containing a 1:1/8 P4/S8 mixture was held
at 80 °C for 55 days, the reaction mass remained a two-phase
system comprised of precipitated solid, identified as R-P4S7,
and a phosphorus-rich liquid. No change in the appearance of
the tube occurred after the first several days of heating. The
composition of the liquid portion, though measured from a low-
resolution spectrum, showed a continued increase in the
concentration of P4S3, the appearance of the two P4S4 isomers,
and a gradual loss of the higher sulfides. This NMR tube was
then heated in several steps in order to bracket the temperature
at which equilibration of the sulfur in the R-P4S7 was redis-
tributed into lower sulfides. No changes in the amount of
precipitated solid occurred during heating at 100 °C for 7 days
and 120 °C for 4 days. Upon heating of this mixture to 150 °C
for 3 days, the mixture was transformed into a clear solution of
P4S3 in P4, containing minor amounts of other phosphorus
sulfides. This collection included the following: R-P4S4 (3),
â-P4S4 (4), R-P4S5 (5), γ-P4S5 (7), â-P4S6 (9), R-P4S7 (11),
R-P4S8 (13), and several multiplets that could not be assigned
to known compounds. (Spectra for these unassigned multi-
plets can be found in the Supporting Information.) Three of
these sulfides, R-P4S4, â-P4S4, and γ-P4S5, were never found
in any of the low temperature (80-90 °C) reaction mixtures
during the first few days. The concentration of R-P4S7 was very
low, and no trace of the other P4S8 isomers or higher sulfides
could be found. The mixture was again heated to 150 °C for
an additional 4 days; disappearing from the spectrum were
R-P4S5, γ-P4S5, â-P4S6, R-P4S7, and R-P4S8, while the intensities
of the unknown phosphorus resonances increased. These two
changes are probably not coupled. Re-equilibration of this same
material at 80 °C did not change the composition of the melt
measurably.
Though integration of the 31P NMR spectra of these mixtures
is not very precise, one can obtain kinetic information from the
early portion of these reactions. An initial slope analysis (based
on loss of S8) of the sum of sulfide concentrations for the first
6 h of the reaction yields an apparent first-order rate at 80 °C
of 2.4 × 10-6 s-1. The assumptions made in this analysis
include the following: All of the products arise from a single,
first-formed intermediate. There is no effective change in the
S8 concentration. None of the intermediates formed on the way
to the stable products “activates” another 1 mol of S8.
Reactions at Stoichiometries between 1:1/8 P4/S8 and P4S10.
Another set of experiments involved the use of sulfur-
phosphorus mixtures, again prepared in NMR tubes, having
much higher S:P4 ratios: from 5 to 10. In all of these high-
sulfur experiments the predominant species were the same (in
order of decreasing relative concentration): R-P4S9, R-P4S7,
R-P4S6, and â-P4S8. Upon continued heating of samples at 90
°C, several other phosphorus sulfides joined the first-formed
four, including P4S3 and P4S10.
Reactions of Phosphorus Sulfides with Liquid Phosphorus.
Dilute solutions of four of the synthetically available phosphorus
sulfides in liquid phosphorus were prepared and heated for
extended periods at 80-90 °C, and the mixtures were analyzed
for the products of sulfur exchange. It is well-known that at
high (200-400 °C) temperatures sulfur is rapidly exchanged
between species forming an equilibrium distribution of sul-
fides.3,4
Only P4S10 is rapidly and completely reduced by phosphorus
at these temperatures. Approximately 75% of the P4S10 was
reduced to R-P4S9 during the time it took to tune the spectrom-
eter and collect a spectrum at 60 °C, about 30 min. After 8 h
at 90 °C no P4S10 remained in the reaction mixture; the principal
product was R-P4S9 (74%) along with five other minor
components: R-P4S8 (5%), R-P4S7 (9%), â-P4S6 (9%), R-P4S4
(1%), and P4S3 (3%). Only small changes in composition were
apparent on continued heating of the mixture at 90 °C.
Three lower sulfides, P4S3, R-P4S5, and R-P4S7, did not
undergo significant reaction with P4 at temperatures between
80 and 90 °C. Small (<1 mol %) quantities of various sulfides
were observed in some spectra but were not quantified.
Continued heating of dilute solutions gave rise to no changes
in the spectra of the compounds, no precipitation, and no loss
of integrated intensity relative to an external standard. P4S3
was studied extensively at both low and high concentration in
P4. A mixture having the composition 1:0.36 P4S3/P4 was heated
for 18 h at 90 °C. Less than 1% of the P4S3 was converted
into other sulfides. These included R-P4S4, â-P4S4, R-P4S5, and
â-P4S7. When a mixture having the composition 1:2.0 P4S3/P4
(equivalent to the composition P4S2) was heated to 250 °C, the
outcome was similar: More than 99% of the P4S3 was
unchanged. However, among the extra resonances in the
spectrum only R-P4S4 and â-P4S4 could be identified as members
of the known sulfides.
Reactions of Phosphorus Sulfides with a Liquid Phospho-
rus-Sulfur Mixture. The rate at which several of the easily
prepared sulfides react with sulfur in liquid phosphorus was
probed at 80 °C. The intent was to determine if any of these
sulfides react rapidly to form higher sulfides. A tube containing
only the 1:1/8 P4/S8 mixture was used as the control.
P4S3 is remarkably stable under these reaction conditions.
After 30 min no reaction that could be ascribed to P4S3 was
observed. At 2 h, the composition was similar to the control
with a few exceptions: R-P4S5 is one of the major sulfides in
the P4S3 reaction and missing from the control. The concentra-
tions of some of the higher sulfides, γ-P4S8 and R-P4S9, were
measurably smaller than the control.
R-P4S5 displays the most extensive chemistry of the studied
sulfides. The three P4S6 isomers are all seen in approximately
three times their respective concentration found in the control.
R- and â-P4S7 are seen at twice their concentrations in the
control.
R-P4S6 was shown to be oxidized to R-P4S7 though only at a
pace that roughly doubled the R-P4S7 concentration.
R-P4S7 appears to be nearly inert to this form of oxidation
by sulfur. As a result of exposing R-P4S7 to 1:1/8 P4/S8, the
only sulfide whose concentration is meaningfully different from
the control was R-P4S9, which was observed at half the
concentration of the control.
R-P4S9 showed only a small increase in the R-P4S7 concentra-
tion relative to the control. All other sulfides showed concen-
trations identical to the control.
The outcome of this series of experiments confirms that the
sulfides seen by 31P NMR spectroscopy in these molten reaction
mixtures are not in rapid equilibrium with each other or with

Oxidation of Phosphorus by Sulfur
Inorganic Chemistry, Vol. 36, No. 12, 1997 2637
Table 1. Comparison of the Product Distributions for the
mixtures. The first is an isomer of P4S6 that exhibits a plane
of symmetry. In liquid phosphorus, this sulfide has the
following phosphorus spectroscopic parameters: P1, 160.71
ppm; P2,3, 99.39 ppm; P4, 61.77 ppm; J(1,2), 43.2 Hz; J(1,4),
43.9 Hz; J(2,4), -303.8 Hz. Among the logical struc-
tures, only γ-P4S6 fit the 31P NMR spectroscopy well. A strong
case for the assignment of structure 10 to this compound is made
in the accompanying paper.15 The second compound is formed
as a very minor component of these mixtures and has the fol-
lowing spectroscopic parameters: P1,2, 90.2 ppm; P3, 51.4 ppm;
P4, 42.3 ppm; J(1,3), 19.4 Hz; J(1,4), 33.1 Hz; J(3,4), 157.9
Hz. The fact that there is no large (>200 Hz) coupling constant
observable implies that the compound has no PP(dS) fragment.3
The molecule must also have C2 or Cs symmetry, more likely
the latter. With these restrictions, the candidate that most
intelligently fits this NMR data is γ-P4S8, structure 15.
The kinetic profiles for many of the products observed can
be divided into two categories: those found in Figure 3 that
are visible in the earliest spectrum, recorded at 30 min, and
those that delay their appearance from 2 to 6 h, found in Figure
4. As discussed below, no observed sulfide can be a primary
product of the initial reaction. However, the group in Figure 3
(P4S3, R-P4S6, R-P4S7, R-P4S9) is considered “primary” because
the concentration profiles appear not to be dependent on the
concentrations of other obserVed products. The compounds in
Figure 4 (R-P4S5, γ-P4S6, R-P4S8, â-P4S8), called “secondary”,
delay their onset until the “primary” products are firmly
established in the reaction mixture. The mechanisms by which
two of the primary products are formed are considered below;
the secondary products are assumed to require sulfur transfer
to one of the primary products.
Photolysis and Thermolysis of a 1:1/8 P4/S8 Mixturea
3 h of photolysis
3 h of photolysis
and then 1 h of
4 h of heating
compd
at 0 °C
heating at 80 °C
at 80 °C
P4S3 (2)
7
0
16
0
3
31
5
0
21
3
13
0
4
1
12
1
1
63
2
1
2
1
13
0
8
1
30
3
0
39
8
0
3
0
9
R-P4S5 (5)
R-P4S6 (8)
â-P4S6 (9)
γ-P4S6 (10)
R-P4S7 (11)
â-P4S7 (12)
R-P4S8 (13)
â-P4S8 (14)
γ-P4S8 (15)
R-P4S9 (16)
P4S10 (18)
0
a The data are relative mol percent figures. See the Experimental
and Discussion sections for details.
the elements. Examples of these reactions in both the oxidizing
and the reducing directions have been demonstrated while the
rates of these reactions are observed to be slow (half-lives greater
than 2-3 h for all studied sulfides).
Photochemical Initiation of the Oxidation of Phosphorus
by Sulfur. Visible light irradiation of a 1:1/8 P4/S8 mixture at
0 °C led to the rapid appearance of several of the known sulfides.
The composition of this mixture is given in Table 1 and is not
corrected for a small amount of R-P4S7 that precipitated during
the photolysis and remained undissolved while taking the NMR
spectrum of the mixture.
Discussion
The identity of the products of this oxidation and the profile
data found in Figures 2-4 provide sufficient information to
propose (though not establish) a mechanism for the start of these
reactions. The complexity of the products and the induction
period in the kinetic profile both necessitate one or more steps
linking the starting materials with the observed products.
The two-component phase diagram of white phosphorus, P4,
and sulfur, S8, is indistinct above 100 °C because the reaction
between these two species becomes uncontrollably rapid at
temperatures between 120 and 130 °C.5 The temperature at
which this reaction becomes uncontrollable depends on the size
of the sample, the rate of heating, and any number of impurities.
However, if the phase diagram below 100 °C is strictly
interpreted, the composition of materials represented on the
diagram should not depend on the age of the sample. In other
words, the diagram should represent a mixture at equilibrium.
One of the conclusions from this work must be that mixtures
of P4 and S8 at 80 °C, and most certainly at even lower
temperatures, are neither thermodynamically nor kinetically
stable, being transformed, albeit slowly, to complex mixtures
of phosphorus sulfides.
Recent work by Griffith and Ngo12 has made use of the slow
reactions taking place in mixtures of phosphorus and sulfur at
temperatures in the range 100-120 °C. Using mixtures with
P4S10 stoichiometry, they succeeded in producing products
nearly identical to those made from P4S10 itself. Blachnik has
also reported seeing evidence of the same chemistry.4
Because only stable, characterized products are observed by 31
P
NMR spectroscopy, the first step in their formation must be
slow, followed by a cascade of more rapid steps that terminate
in the kinetic sink of one of the observed species. But what
intermediate is formed in that first, slow step?
A number of candidates can be found in the literature. These
include reactions centered on sulfur: the transfer of one sulfur
atom16 from S8 to P4; equilibration of S8 to give smaller and
more reactive sulfur rings such as S6;17 formation of S2;18-21
the formation of cycloheptasulfur thionsulfoxide;22 the formation
of S8 diradical by S-S homolysis.23 Two phosphorus candi-
dates are possible: P-P homolysis to give the P4 diradical;
cleavage to form two P2 molecules. Finally, several bimolecular
reactions are plausible, including ring-opening addition of P4
to S8 or S8 to P4 and a complex formed by electron transfer to
give a radical cation-radical anion pair.
Twelve unique phosphorus sulfides have been observed as
the products of the low-temperature oxidation of phosphorus
by sulfur. These include the well-known and structurally char-
acterized sulfides P4S3,1 R-P4S5,1,8 â-P4S6,3,4 R-P4S7,1,13 R-P4S9,1,9
and P4S10.1 Four other sulfides, which have been previously
characterized only spectroscopically, were also observed:
R-P4S6,3,4 â-P4S7,4 R-P4S8,14 and â-P4S8.4 Finally, two sulfides
not previously described in the literature were observed in these
(15) Jason, M. E. Inorg. Chem. 1997, 36, 2641-2646.
(16) Baechler, R. D.; Stack, M.; Stevenson, K.; Vanvalkenburgh, V.
Phosphorus, Sulfur Silicon 1990, 48, 49-52.
(17) Meyer, B. Chem. ReV. 1976, 76, 367-388.
(18) Hills, A. J.; Cicerone, R. J.; Calvert, J. G.; Birks, J. W. J. Phys. Chem.
1987, 91, 1199-1204.
(19) Steliou, K.; Salama, P.; Yu, X. J. Am. Chem. Soc. 1992, 114, 1456-
1462.
(20) Tardif, S. L.; Williams, C. R.; Harpp, D. N. J. Am. Chem. Soc. 1995,
117, 9067-9068.
(21) Bosser, G.; Paris, J. New J. Chem. 1995, 19, 391-399.
(22) Suontomo, R. J.; Laitinen, R. S. Main Group Chem. 1996, 1, 241-
251.
(23) Kende, I.; Pickering, T. L.; Tobolsky, A. V. J. Am. Chem. Soc. 1965,
87, 5582-5586.
(12) Griffith, E. J.; Ngo, T. M. Patent WO 9526357 A1, 1995.
(13) Dixon, D. T.; Einstein, F. W. B.; Penfold, B. R. Acta Crystallogr.
1965, 18, 221-225.
(14) Barieux, J.-J.; Demarcq, M. C. J. Chem. Soc., Chem. Commun. 1982,
176-177.

2638 Inorganic Chemistry, Vol. 36, No. 12, 1997
Jason et al.
It is difficult to ignore the fact that a mixture of 1:1/8 P4/S8
stoichiometry generates mostly products of P4S6 and higher
stoichiometries. One molecule of each starting material appears
to react to form a manifold of intermediates; these intermediates
then rapidly incorporate sulfur into the phosphorus cage and
only reluctantly lose sulfur back to the reaction mass. Though
there are other explanations for the observed products, the nearly
complete consumption of a molecule of S8 by a molecule of P4
is one of the most straightforward.
An initial-slope kinetic analysis provides an estimate of the
rate of this first slow step, about 2.4 × 10-6 s-1. The activation
•
parameter data for the reaction S8 w S-S6-S• can be taken
from the extensive rubber technology literature,23,27 with the
understanding that these rates are appropriate for the cross-
linking of olefinic hydrocarbons. Since the capture of a sulfur
radical by an olefin may be less efficient than capture by P4,
the preexponential term taken from the vulcanization studies
may be smaller than appropriate for the oxidation of phosphorus.
An activation energy of approximately 146 kJ/mol (35 kcal/
•
mol) has been determined for the first-order formation of S-
S6-S• from S8. The pre-exponential term has a large reported
variation but typically falls in the range 1014-1015. A rate at
80 °C of 2 × 10-7 s-1 is calculated from the activation energy
and the largest pre-exponential term. The calculated rate is
within a factor of 12 of the observed rate of oxidation.
Several oxidations were performed at low temperature (∼0
°C) and initiated by visible light. It is well-known and well-
documented that the polymerization of both phosphorus5 and
sulfur28,29 can be initiated photochemically. The S8 diradical
intermediate involved in the photochemical polymerization of
sulfur has been observed by ESR spectroscopy;28 the suspected
diradical involved in phosphorus polymerization has not been
observed. The consequences of the production of either
diradical are easily drawn and are found in Figure 6. Both
radicals can either homo-oligomerize, an outcome not observed
at least for phosphorus, or react with the other material in the
melt. Cross addition beginning with either of the two diradicals
will lead to the same intermediate, diradical 20.
Figure 5. Proposal for the first four steps in the reaction of sulfur and
phosphorus at temperatures below 100 °C.
25
Both the P4 diradical24 and P2 are inaccessible at the
temperatures of the reaction, making sulfur the far more likely
source of the first-formed intermediate. The transfer of one or
two sulfur atoms to a P4 molecule would appear to lead to
compounds having initially low S/P4 ratios, inconsistent with
the observed products. P4S3 is formed early in the reaction,
but R-P4S5 is not. In fact, R-P4S5 is best explained by the slow
oxidation of P4S3 (see below). Though the stability of the P4S4
isomers to the reaction conditions has not been studied, it is
clear that P4S3 and R-P4S5 are sufficiently stable to be observed
as the major products early in the reaction if the formation of
the sulfides were to occur by the stepwise addition of one or
two sulfur atoms at a time.
Bimolecular, 2-electron reactions between P4 and S8 are
expected to be very slow because neither is a good nucleo-
phile.17,26 One-electron transfer from sulfur to phosphorus
would generate, after radical pair collapse, the same intermediate
discussed below and therefore is mechanistically impossible to
distinguish. Initiation of this oxidation reaction by the formation
of a high-energy Sn species (S8 diradical, thionsulfoxide, or S6)
is therefore a logical starting point. The S8 diradical (19) is
the best known of the three and will be the focus of the following
discussion. Either of the other two species would give rise to
cage intermediates similar or identical to those produced by the
S8 diradical and can thus be considered alternate entries into
the same manifold. The S8 diradical is also consistent with
many of the observations made in this study: the stoichiometry
of the early products, the kinetics of the slow step in the reaction,
and the similarity of the thermally and photochemically initiated
reactions. If the S8 diradical is the species formed in the initial,
slow step, a proposal for the start of this reaction can be drawn
as shown in Figure 5.
Among the unique alternatives discussed above, the initial
•
formation of S-S6-S• from S8 is most consistent with both
the properties of the reactants and the experiments performed.
The fact that elemental sulfur easily forms diradicals and the
known sensitivity of phosphorus toward formation of bonds with
radicals26,30,31 support the choice of building a model of this
reaction based on an initial addition of 19 to P4. The next
question to be answered is whether the formation of 20 and 21
can lead to the observed products.
The chemistry of 21 is more difficult to specify. Several
choices exist, including homolytic S-S bond cleavages and
2-electron processes, the products of which are also shown in
Figure 5. Compound 21 must decompose more rapidly than
the “initiation” reaction because, if the disappearance (in a first-
order or pseudo-first-order fashion) of 21 were slow on this
time scale, it would be visible in the 31P NMR spectrum. It is
not. Assuming that one part in ten thousand of a new
(27) Tobolsky, A. V.; Eisenberg, A. J. Am. Chem. Soc. 1959, 81, 780-
782.
(28) Steudel, R.; Albersen, J.; Zink, K. Bunsenges. Phys. Chem. 1989, 93,
502-509.
(29) Geyer, J.; Stu¨lpnagel, H.; Rademann, K. Angew. Chem., Int. Ed. Engl.
1992, 31, 874-876 and references cited therein.
(24) DeWitt, T. W.; Skolink, S. J. Am. Chem. Soc. 1946, 68, 2305-2309.
(25) Gurvich, L. V.; Veyts, I. V.; Alcock, C. B. Thermodynamic Properties
of IndiVidual Substances, 4th ed.; Hemisphere Publishing Corp.: New
York, 1989.
(30) Bentrude, W. G. The Chemistry of Organophosphorus Compounds;
Hartley, F. R., Ed.; John Wiley and Sons, Ltd.: New York, 1990;
Vol. 1, pp 531-566.
(31) Prokof’ev, A. I.; Khodak, A. A.; Malysheva, N. A.; Bubnov, N. N.;
Solodovnikov, S. P.; Belostotskaya, I. S.; Ershov, V. V.; Kabachnik,
M. I. Dokl. Akad. Nauk SSSR, Int. Ed. 1978, 240, 222-226.
(26) Maier, L. Fortschr. Chem. Forsch. 1971, 19, 1-59 and references
cited therein.

Oxidation of Phosphorus by Sulfur
Inorganic Chemistry, Vol. 36, No. 12, 1997 2639
Figure 6. Production and fate of diradical species from P4 and S8 in
mixtures of the two elements.
phosphorus species would be just visible, the first-order rate of
disappearance of 21 or any other intermediate must be 3 orders
of magnitude faster than its appearance in order for it to remain
submerged from view. An estimate of the maximum half-life
for such an intermediate would then be about 10 min. A good
model for the formation of 22 from 21 is taken from the work
of Danen and Newkirk32 in which they find the S-S homolytic
bond strength of compounds of general formula R2NS-SNR2
is about 25 kJ/mol lower than the equivalent bond in RSS-
SSR. Were all of this difference applicable to the PS-SS bond
cleavages in these cage polysulfides, intermediate 21 would react
too rapidly to be visible.
Assuming the formation of 22, there are many ways of
rearranging sulfur and phosphorus atoms so as to give logical
entries into the observed products. A collection of mechanistic
steps that transforms 21 into R-P4S6 and R-P4S7 is shown in
Figure 7. This scheme is meant only to be illustrative; yet most
of the steps have a precedent in the chemistry of phosphorus
(or carbon) radicals. The reactions envisioned include S-S
bond cleavage, quaternization (a diradical recombination),
rearrangement with ring expansion, and loss of sulfur. The
intermediate 27 can continue on to give R-P4S7 by performing
one more ring expansion. The series of steps shown for the
transformation of 21 to 24 in Figure 7 constitutes a cleavage-
addition-migration sequence to accomplish an oxidation and
ring enlargement. Two of the reactions in this sequence have
precedents in the rearrangements of a phosphite ester observed
by Hudson.33 Their work also demonstrated both the radical
nature of the reaction of their phosphite ester and the fact that
the rate was rapid at temperatures below 0 °C. The path drawn
from 24 to 11 is but one of many. It relies on the same reactions
used to transform 21 into 24, adding only the loss of sulfur at
the end. The path was drawn to retain the necessary P-P bond,
to retain sulfur in the cage as long as possible, and to make
every nonradical daughter structure appear to be more stable
Figure 7. Proposal for the formation of R-P4S6 and R-P4S7 from the
P4S8 intermediate 21.
than its parent. Compound 27 cannot use this same collection
of reactions and give rise to γ-P4S6, â-P4S7, any of the P4S8 or
P4S9 isomers, or P4S10. At some point prior to the formation
of 27, phosphorus no. 1 (labeled in structure 26) must be
oxidized to a tetracoordinate phosphorus. This oxidation could
involve an intra- or intermolecular transfer of sulfur.
The mechanism illustrated in Figure 7 makes the assumption
that sulfur is not lost from an intermediate until the final step
in the formation of a stable (i.e., observed) sulfide. This choice
was made in order to make use of the observation that
compounds of low S:P4 ratio are not made most rapidly. There
is good precedent from the chemistry of elemental sulfur for
the intramolecular cyclization to form small rings and small
fragments from larger diradical chains.26,34,35 Intermediates such
as 22 could opt for S-S bond formation with a sulfur in the
chain other than the S-radical terminus, forming instead a species
like 21, only with fewer sulfur atoms in the chain, and a Sn
fragment. That this is not a major path is supported by the fact
that P4S3 and R-P4S5 are not among the most rapidly formed
products.
The low-temperature photochemical reactions have contrib-
uted far more than support for the S8 diradical initiation of this
reaction. The phosphorus sulfides in the photochemical experi-
ments are taken from the same collection as those found in the
thermal reactions but in remarkably different proportions (Table
1). R-P4S6 and R-P4S7 are two of the major products in both
reactions. But P4S3 and R-P4S9 are the only other products
observed in the early part of the thermal reaction, and they are
(34) Meyer, B.; Gouterman, M.; Jensen, D.; Oommen, T. V.; Spitzer, K.;
Stroyer-Hansen, T. Sulfur Research Trends; Advances in Chemistry
Series 110; Miller, D. J., Wiewiorowski, T. K., Eds.; American
Chemical Society: Washington, DC, 1972; pp 53-72.
(35) Barton, D. H. R.; Castagnino, E.; Jaszberenyi, J. C. Tetrahedron Lett.
1994, 35, 6057-6060.
(32) Danen, W. C.; Newkirk, D. D. J. Am. Chem. Soc. 1976, 98, 516-
520.
(33) Brown, C.; Hudson, R. F.; Maron, A.; Record, K. A. F. J. Chem. Soc.
D 1976, 663-664.

2640 Inorganic Chemistry, Vol. 36, No. 12, 1997
Jason et al.
absent from the photochemical reaction. Replacing them are a
manifold of products built from γ-P4S6: γ-P4S6, â-P4S7, â-P4S8,
and γ-P4S8. These four sulfides are not as thermally stable as
other sulfides. Their sensitivity to higher temperature was
confirmed by heating the irradiated mixture (total irradiation
time of three hours) to 80 °C for 45 min and observing a
significant decrease in the concentrations of these four sulfides.
The decreases in relative concentration for the four were 80%,
60%, 90%, and 80%, respectively. Since the appearance profile
for these sulfides in the thermal 80 °C reaction is relatively
smooth, their observed concentrations are the result of both
formation and either sulfur exchange or rearrangement to form
other sulfides. Thus, sulfides observed only at low concentration
appear to be formed and rapidly transformed, occupying part
of the network of reactive intermediates. They may also be
responsible for the attachment and removal of exocyclic sulfur
atoms, as is necessary for the formation of R-P4S9. Until the
syntheses of these unusual phosphorus sulfides have been
accomplished, this analysis cannot be tested independently.
cleavage of an S-S bond. The choice of the diradical is
supported by the kinetics of the reaction, the sulfur content of
the major products, and the similarity between the thermally
and photochemically initiated reactions. The initially formed
products are found with high S:P4 ratios, even from starting
mixtures having low S:P4 ratios, arguing that the products are
generated in a series of rapid steps that are not dominated by
the formation of one or more of the lower sulfides. These two
observations were used to construct a mechanism for this
oxidation, recognizing that the proposal is meant to stimulate
interest in answering the questions that underlie these mecha-
nistic steps. Finally, the family of phosphorus sulfides has been
increased by the addition of two new members: γ-P4S6, 10,
whose structure is reasonably secure, and γ-P4S8, 15, whose
identity is less secure and should be trusted only after more
supporting evidence is accumulated.
Acknowledgment. We thank Claude Jones for his assistance
with the NMR spectroscopy and Enriqueta Cortez and Greg
Gervasio for help with the Raman spectroscopy. Thanks are
due Ed Griffith and Frank Pacholec for encouraging us to pursue
this remarkable chemistry. The reviewers have offered correc-
tions and provided comments that have been incorporated into
this report.
Conclusions
A number of unique results have been demonstrated by this
exercise in phosphorus and sulfur chemistry. Foremost is the
observation of slow but significant oxidation of phosphorus by
sulfur at temperatures much lower than used to equilibrate these
elements. The reaction appears to be initiated by the formation
of a reactive sulfur species. At the moment the most likely
candidate for this species is the S8 diradical formed by the
Supporting Information Available: Listings of 31P NMR data and
representative NMR and Raman spectra (15 pages). Ordering informa-
tion is given on any current masthead page.
IC9614879