Synthetic routes to, transformations of, and rather surprising stabilities of (N-Methyl-N-phenylcarbamoyl)sulfenyl Chloride, ((N-Methyl-N-phenylcarbamoyl) dithio)carbonyl chloride, and related compounds

Article abstract of DOI:10.1021/jo201329n

(Figure presented) The title compound classes, (carbamoyl)sulfenyl chlorides and ((carbamoyl)dithio)carbonyl chlorides, have been implicated previously as unstable, albeit trappable, intermediates in organosulfur chemistry. The presentwork reports for each of these functional groups: (i) several routes to prepare it in the N-methylaniline family; (ii) its direct structural characterization by several spectroscopic techniques; (iii) its rather unexpected stability and its ultimate fatewhen it decomposes; (iv) a series of further chemical transformations that give highly stable derivatives, each in turn subject to thorough characterization. Relevant kinetic and mechanistic experiments were carried out, including some with p-methyl- and 2,6-dimethyl-substituted N-methylanilines. Given that the title compounds can be isolated and are relatively stable, they may find applications in the preparation of thiolyzable and/or photolabile protecting groups for the sulfhydryl function of cysteine and for the development of new protein synthesis and modification reagents

Full text of DOI:10.1021/jo201329n

ARTICLE
pubs.acs.org/joc
Synthetic Routes to, Transformations of, and Rather Surprising
Stabilities of (N-Methyl-N-phenylcarbamoyl)sulfenyl Chloride,
((N-Methyl-N-phenylcarbamoyl)dithio)carbonyl Chloride, and
Related Compounds†
Alex M. Schrader,‡ Alayne L. Schroll,§ and George Barany*,‡
‡Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455, United States
§Saint Michael's College, Colchester, Vermont 05439, United States
S Supporting Information
b
ABSTRACT:
The title compound classes, (carbamoyl)sulfenyl chlorides and ((carbamoyl)dithio)carbonyl chlorides, have been implicated previously as
unstable, albeit trappable, intermediates in organosulfur chemistry. The present work reports for each of these functional groups: (i) several
routes to prepare it in the N-methylaniline family; (ii) its direct structural characterization by several spectroscopic techniques; (iii) its
rather unexpected stability and its ultimate fate when it decomposes; (iv) a series of further chemical transformations that give highly stable
derivatives, each in turn subject to thorough characterization. Relevant kinetic and mechanistic experiments were carried out, including
some with p-methyl- and 2,6-dimethyl-substituted N-methylanilines. Given that the title compounds can be isolated and are relatively
stable, they may find applications in the preparation of thiolyzable and/or photolabile protecting groups for the sulfhydryl function of
cysteine and for the development of new protein synthesis and modification reagents.
’ INTRODUCTION
’ RESULTS AND DISCUSSION
A seminal 1970 review by Zumach and K€uhle1 made pass-
ing reference to both (carbamoyl)sulfenyl chlorides (1) and
((carbamoyl)dithio)carbonyl chlorides (2), each of which was
implied to be an unstable, transient species that underwent facile
loss of elemental sulfur2 (and in the second case, carbonyl sulfide
as well) to furnish carbamoyl chlorides (3) (Scheme 1). Never-
theless, each of the putative intermediates 1 and 2 was reported
to form a characterizable adduct (7 and 8, respectively; Scheme 1)
upon trapping with cyclohexene (for 1) or methanol (for 2).
Primary references cited by Zumach and K€uhle1 were exclusively
to unpublished studies from industrial laboratories, and a com-
prehensive, current literature search revealed a paucity of addi-
tional information on these species.3ꢀ7 In contrast, the present
paper reports not only the capability to generate the title
functional groups by several complementary routes ꢀ as sup-
ported by thorough spectroscopic characterization as well as
studies on several further transformations ꢀ it also documents
their extraordinary (and rather surprising) stabilities. This work
may have applications in the preparation of thiolyzable and/or
photolabile protecting groups for the sulfhydryl of cysteine8ꢀ10
and for the development of protein modification reagents.11
Preparation of (N-Methyl-N-phenylcarbamoyl)sulfenyl
Chloride (1). While the ZumachꢀK€uhle review1 alluded to a
range of primary and secondary, aliphatic and aromatic carba-
moyl moieties (legend to Scheme 1), we elected to study princi-
pally compounds based on N-methylaniline, especially because
an extensive set of derivatives13,14 were available in our laboratory
that could be used as reference compounds in elucidating the
subsequent chemistries. To facilitate and standardize spectro-
scopic characterization and comparisons, the described reactions
(Scheme 2) were generally carried out in CDCl3 solution,
although for completeness, selected experiments were repeated
in other aprotic solvents, specifically CD3CN, dioxane-d8, and
benzene-d6. Outcomes of such experiments were similar, although
kinetics varied, consistent with the range of polarities for the sol-
vents tested.
Several in situ routes produced 1 in the N-methylaniline family
(Scheme 2). First, treatment at 25 °C of (N,N0-dimethyl-N,
N0-diphenylcarbamoyl)sulfenamide (5)13 with anhydrous hydrogen
Received: June 24, 2011
Published: August 26, 2011
r
2011 American Chemical Society
7882
dx.doi.org/10.1021/jo201329n J. Org. Chem. 2011, 76, 7882–7892
|

The Journal of Organic Chemistry
ARTICLE
Scheme 1. Summary of ZumachꢀK€uhle Observations and Conclusions a
a In ref 1 for chemistry on the left side of this scheme, R2N = Me2N, nPr2N, morpholino, iPrNH, tBuNH, and PhNH; for chemistry on the right side, R2N
= Me2N and R0 = nBu. Intermediates 1 and 2 were both stated to be rather unstable [hence shown here within square brackets], a conclusion not
supported by the presently reported studies for which R2N = Ph(Me)N and related. Compound 7 is drawn with trans stereochemistry not due to specific
spectroscopic evidence but rather based on general literature precedents.12
Scheme 2. Syntheses and Transformations Related to (Carbamoyl)sulfenyl Chloride 1 aꢀc
a The majority of these syntheses and transformations are new; previously known compounds are referenced in the text. Structure 9 is presumed based
on NMR detection of a reaction intermediate (mixture of two configurations that have identical reactivities); it is therefore shown in square brackets
to indicate its transient nature. Much of this chemistry was demonstrated not only for derivatives of unsubstituted N-methylaniline but also
for appropriately substituted N-methylanilines. More specifically, phenyl (no further punctuation) was sometimes replaced by p-methylphenyl
(prime = 0 series) or 2,6-dimethylphenyl (double prime = 00 series). b When this manuscript was under review, an anonymous referee suggested that
cleavage of 13 with chlorine might generate 1 (see also ref 4). In fact (details in Experimental Section), reaction of 13 with sulfuryl chloride (a safer
chlorine equivalent) led to direct formation of carbamoyl chloride 3 plus heterocycle 14 (in a 1:1 ratio), presumably because diacyl disulfanes are
asymmetrically cleaved as has been noted for other examples summarized by K€uhle (ref 15). c The decomposition of 1 to 3 occurred only when 1
had been generated by the chemistry in the upper right corner of this scheme. For all other methods to generate 1, this compound was eventually
converted to heterocycle 14.
chloride [delivered either as gas bubbled through a CDCl3
solution or from a 1 M stock solution in dioxaneꢀCDCl3
(3:7)] rapidly generated16 the sulfenyl chloride 1, along with a
stoichiometric amount of N-methylaniline hydrochloride. Second,
7883
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
reactions7 of O-alkyl N-methyl-N-phenylthiocarbamates (6)13,17
with sulfuryl chloride (1 equiv) gave desired 1, together with the
appropriate alkyl chloride, within 1 h at 25 °C. While starting thio-
carbamates 6 disappeared within moments of addition of sulfuryl
chloride, short-lived intermediates did appear; these have been tenta-
tively assigned structure 9 (Scheme 2). Interestingly, 9 existed in two
more or less equally populated configurations for the methyl (a) and
ethyl (b) series, albeit in a single configuration for isopropyl (c).
Further conversion of 9 (both configurations, reacting at the same
rates) to 1, and concomitant alkyl chloride formation, occurred at
25 °C with t1/2 ∼ 8 min for 6a (R0 = methyl), t1/2 ∼ 10 min for 6b
(R0 = ethyl), and t1/2 ∼ 3 min for 6c (R0 = isopropyl). These relative
reactivities of methyl vs ethyl vs isopropyl were confirmed when all of
the appropriate precursors (6a, 6b, and 6c) were combined into a
single tube and reacted with SO2Cl2.
Further approaches to generate (carbamoyl)sulfenyl chloride 1,
while not recommended for preparative purposes, served to
elucidate additional aspects of the chemistry, as well as to confirm
spectroscopic assignments. Given that reaction of chlorocarbo-
nylsulfenyl chloride (4)1,13 with excess N-methylaniline leads
rapidly and quantitatively to bis(substituted) compound 5, it was
of interest to note that when 4 was in excess (i.e., N-methylaniline
was limiting), compound 1 was among the species formed. As
might be expected, this mode of reaction was accompanied by
formation of some 5 from substitution at the less reactive sulfenyl
chloride function (above and beyond the preferred substitution
at the carbonyl). Also, because N-methylaniline serves not only
as the nucleophile but also as a base to absorb the hydrogen
chloride coproduct, the corresponding hydrochloride salt was
observed. To simplify stoichiometric calculations and provide
more straightforward results, variant acylations were carried out
when the substrate was the N-trimethylsilylated derivative18ꢀ21
10 of the secondary amine. As predicted from precedents in other
systems,22 these reactions resulted in quantitative formation of
chlorotrimethylsilane (unreactive under these conditions) as the
coproduct. Most relevant to the goals of this work, a 1:1 ratio of 4
plus 10 was transformed, essentially instantaneously at 25 °C, to
1 plus TMSCl (quantitative; does not react further), while
reaction of 4 (1 equiv) plus 10 (2 equiv) gave the bis(sub-
stituted) adduct 5, plus TMSCl (2 equiv). The fact that 5 was not
detected from the 1:1 reaction suggests that a silylated amine is
considerably more reactive toward the acyl chloride moiety of 4,
with respect to a significantly more sluggish reaction at the
sulfenyl chloride moiety.
Trapping of Compound 1. (Carbamoyl)sulfenyl chloride 1,
generated by any of these methods, was trapped through
quenches with N-methylaniline or simple alkanethiols, providing
respectively the previously described (carbamoyl)sulfenamide
513 or the appropriate alkyl (N-methyl-N-phenylcarbamoyl)disulfane
(11).23 Interestingly, when 1 was generated from SO2Cl2 plus 6,
N-methylaniline quenches conducted even before intermediate 9
had been fully converted (with simultaneous formation of alkyl
chloride) gave only (carbamoyl)sulfenamide 5 and no alkyl
group-containing products that might have corresponded to
direct trapping of 9. Of further interest, 1 did not react whatso-
ever with neat methanol at 25 °C, as evidenced by direct spec-
troscopic examination and confirmed by experiments in which
the N-methylaniline quenches were repeated after an extended
delay period. Because results on the delayed quenches were the
same as for experiments with freshly made 1, it was reasonable to
conclude that methanol and 1 do not react readily. Finally, we
reproduced the observation of ZumachꢀK€uhle1 that 1 can be
trapped with cyclohexene to provide racemic 7 with presumed
trans stereochemistry.12 Compound 7 from the experiments in
this work was evaluated by HPLC and gave a single peak
consistent with a racemic pair of defined stereochemistry; two
diastereomeric peaks would have been expected if both cis and
trans addition had occurred.
Compelling further evidence consistent with structure 1 was
provided by reductive dimerizations carried out on solutions of 1
in CDCl3. To achieve the required transformation, the organic
phase was washed with aqueous potassium iodide24,25 to quickly
give primarily bis(N-methyl-N-phenylcarbamoyl)disulfane (13),13,26
along with much smaller amounts of the trisulfane and tetra-
sulfane counterparts.14 Disulfane 13 was also formed, in excellent
yield and purity, when thiocarbamate 6c (2 equiv) was reacted
with SO2Cl2 (1 equiv) in a process that undoubtedly goes through
1 as an intermediate. Thus, compound 1 formed from reaction of
the first equivalent of 6c is assumed to react with the second
equivalent of 6c to produce 13 (lower middle of Scheme 2; note that
similar results, albeit with somewhat lower yields and purities,
were also observed when starting with substrates 6a and 6b).
These routes to 13 scale up readily and provide an original
and efficient alternative pathway to a compound that was made
previously by using N-methylaniline to quench bis(chloro-
carbonyl)disulfane (structure 20, introduced later in this text
and in Scheme 3; a multistep route to 20 is described in ref 13).
When (carbamoyl)sulfenyl chloride 1 was simply maintained
under ambient conditions in the absence of a trapping agent, it
underwent a clean intramolecular aromatic substitution/cycliza-
tion to provide 3-methyl-2(3H)-benzothiazolone (14), a known
heterocycle that was made previously by careful reaction1,13 of N-
methylaniline with chlorocarbonylsulfenyl chloride (4) or by
unrelated chemistry.28,29 At 25 °C, formation of 14 occurred with
t1/2 ∼ 24 h when 1 was generated from 4 plus 10 (final con-
centrations 0.1 M), with t1/2 ∼ 1 h when the same method was
used but at higher concentration (i.e., 1 M), with t1/2 ∼ 4 h when
1 was made by reaction of SO2Cl2 plus 6, and with t1/2 ∼ 30 min
when hydrogen chloride (acting as a catalyst) was present, as in
those studies where (carbamoyl)sulfenamide 5 was the precur-
sor. Within experimental error, such rates vary linearly with
concentration. For example, when 1 was prepared from 6c plus
SO2Cl2, the rate of conversion of 1 to 14 decreased approxi-
mately 10-fold upon 10-fold dilution of the reaction mixture.
For further clarity, 1 was trapped by N-methylaniline deriva-
tives that have additional substituents in the aromatic ring, i.e.,
p-methyl-N-methylaniline and N,2,6-trimethylaniline.30ꢀ32 Such
reactions provided the novel mixed sulfenamides 50 and 500, in
which the sulfenyl component was contributed regioselectively
by the modified N-methylaniline (Scheme 2).
Stability of Compound 1. The stability of 1 was readily
assessed both by direct spectroscopic examination over an
extended time frame and by substantially delayed quenches.
After correction for the already described relatively slow forma-
tion of heterocycle 14, the title compound was remarkably stable,
as evidenced by observation of the same quench patterns as many
as 3 days later, at 25 °C. Attempts to isolate 1 by simple eva-
poration under reduced pressure, immediately after it had been
generated from 6 plus SO2Cl2, resulted in its direct and quan-
titative decomposition to carbamoyl chloride 3,13 along with
elemental sulfur. Compound 3 was recognized spectroscopically,
and also by its very sluggish trappings33 by either methanol to
provide urethane 15,13 or by N-methylaniline to provide sym-
metrical urea 16.13 As a technical point, the 1H chemical shifts of
7884
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
Scheme 3. Syntheses and Transformations Related to ((Carbamoyl)dithio)carbonyl Chloride 2a
a Conventions used are the same as in Scheme 2. The further decomposition of 2 to 1 occurred when it was generated from 6 plus 4 or from heating the
reaction mixture that started from 21, but not when 20 was the starting material, even upon heating. The further decomposition of 1 to 3, a rather minor
pathway to begin with, was seen exclusively in experiments that started with 6 plus 4.
1 and 3 were the same, although 3 gave a broader singlet; the
further trappings were necessary to establish unambiguously that
1 had been transformed to 3. In contrast, when 1 was generated
from (i) 4 (excess) plus N-methylaniline (limiting) (Scheme 2,
upper left), (ii) equimolar amounts of 4 and silylated amine 10
(Scheme 2, upper left), or (iii) HCl cleavage of 5 (Scheme 2,
upper middle), subsequent evaporation did not result in the
formation of 3. Instead, these latter experiments demonstrated
exclusively the rapid and complete conversion of (carbamoyl)-
sulfenyl chloride 1 to heterocycle 14.
With the goal to create a system in which the (carbamoyl)-
sulfenyl chloride functionality might have a longer lifetime, some
of the aforementioned chemistry was repeated with the usual
N-methylaniline group replaced by an N,2,6-trimethylaniline
group. The presence of methyl substituents at both ortho positions
in putative (N,2,6-trimethyl-N-phenylcarbamoyl)sulfenyl chloride
(100) was expected to preclude this intermediate from cyclization,
a contrast to the already described transformation of unblocked 1
to 14. In practice, synthesis of 100 was achieved through treatment
of chlorocarbonylsulfenyl chloride (4) with limiting N,2,6-tri-
methylaniline, as well as through SꢀN bond cleavage of the novel
(carbamoyl)sulfenamide 1700 by HCl gas (Scheme 2). Whereas 1
prepared by the HCl acidolysis method cyclized to 14 completely
within 2 h at 25 °C, 100 prepared in an analogous fashion
remained intact for at least 1 week. Confirmation of the proposed
structure of 100 was provided by trapping with alkanethiols,
N-methylaniline, and N,2,6-trimethylaniline, yielding respectively
disulfane 1100, mixed (carbamoyl)sulfenamide 17, and (carbamoyl)-
sulfenamide 1700. Furthermore, reductive dimerization of 100
by aqueous potassium iodide treatment gave symmetrical bis-
(carbamoyl)disulfane 18. Upon drying and evaporation, 100 made
by any of the aforementioned procedures did not lose elemental
sulfur, as proven through identical spectroscopic and quench
data, as well as by direct elemental analysis.
SO2Cl2 (1 equiv), it decomposed rapidly (t1/2 ∼ 3 min) to
carbamoyl chloride 300 plus elemental sulfur, even as the instantly
formed intermediate 9b00 (single configuration by 1H NMR) was
still losing ethyl chloride to generate more 100. The fact that the
stability of 100 depends so markedly on the method by which it
was formed is quite remarkable, but we cannot point to any
component in the reaction mixtures that might respectively
inhibit and/or catalyze its decomposition.
Preparation of ((N-Methyl-N-phenylcarbamoyl)dithio)-
carbonyl Chloride (2). The title species 2 was accessed by
several in situ routes (Scheme 3). The most straightforward
(ease, clean, and unambiguous formation of desired product) of
these was treatment of O-alkyl N-methyl-N-phenylthiocarba-
mates (6)13,17 with chlorocarbonylsulfenyl chloride (4)1,13 (1equiv);
this follows the ZumachꢀK€uhle precedent1 (Scheme 1) which in
turn builds on work of Harris.34 Desired 2 formed relatively
rapidly and quantitatively, along with the appropriate alkyl
chloride (complete within 40 min at 25 °C); an initial adduct
formed within moments of addition and was assigned structure
19 (again, two configurations). Intermediate 19 was transformed
to 2 plus alkyl chloride with t1/2 ∼ 1.5 min for R0 = Me, t1/2
∼
40 min for R0 = Et, and t1/2 ∼ 1 min for R0 = iPr. This
nonmonotonic trend in conversion rates as a function of alkyl
group was confirmed in another experiment in which 6a, 6b, and
6c were mixed and collectively treated with 4; intermediate 19c
converted the fastest, followed closely by 19a, and intermediate
19b converted substantially slower. Most likely, for R0 = Me or
Et, a limiting SN2 displacement is followed, while for R0 = iPr, the
mechanism switches to SN1. Alternatively, (dithio)carbonyl
chloride 2 was made by reaction of bis(chlorocarbonyl)disulfane
(20)13,26 with limiting N-methylaniline to provide a mixture
comprising 2, unreacted 20, and the bis(substituted) adduct 13, a
readily recognizable compound that had been made previously13,14,26
by reaction of 20 with excess N-methylaniline, and in the present
work (vida infra) by reductive dimerization of (carbamoyl)sul-
fenyl chloride 1. In addition, compounds 2 and 13 were formed
In contrast to the results reported in the preceding paragraph,
when 100 was prepared by treatment of thiocarbamate 6b00 with
7885
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
in similar amounts (3:2 molar ratio), along with TMSCl
formed quantitatively, from reaction of equimolar amounts of N-
trimethylsilyl-N-methylaniline (10)18ꢀ21 plus bis(chlorocarbonyl)-
disulfane (20).13,26 Last, a stirred emulsion of (trichloromethyl)-
(N-methyl-N-phenylcarbamoyl)disulfane (21)13 (dissolved in
CDCl3) in the presence of aqueous sulfuric acid, emulating
several precedents1,13,26,27 for hydrolyzing a trichloromethylsul-
fenyl group to a chlorocarbonylsulfenyl group, gave a complex
mixture with 2 as a significant component (∼45% conversion of
21 after 24 h at 55 °C). Lengthy additional heating resulted in a
reaction mixture that comprised approximately equal amounts of
unreacted starting 21 and heterocycle 14 (t1/2 ∼ 7 days; ∼47%
based on starting 21). A likely explanation for this result is that all
of intermediate 2 that had formed during the course of the
reaction had decomposed to carbonyl sulfide plus (carbamoyl)-
sulfenyl chloride 1, which in turn converted almost entirely to
heterocycle 14 according to a previously elucidated transformation.
Trapping of Compound 2. Irrespective of how generated,
((carbamoyl)dithio)carbonyl chloride 2 was readily trapped by
quenches with methanol or N-methylaniline, providing respec-
tively the previously described disulfanes 813 and 13.13,14,26 For
further clarity, 2 was trapped by substituted N-methylaniline
derivatives to provide unsymmetrical bis(carbamoyl)disulfanes
130 and 1300. Just as the quenching of intermediate 9 gave the
same results as the quenching of compound 1, so did inter-
mediate 19 and compound 2 display the same quenching patterns,
i.e., no alkyl-containing quench products.
Stability of Compound 2. The stability of 2 was monitored
primarily by repeated recording, at a variety of time points, of
spectra of mixtures in which this species was a component. In
addition, quenches with methanol or with N-methylaniline were
carried out on a delayed basis. Results depended on the mode by
which 2 had been generated (footnote to Scheme 3). Remark-
ably, even after 4 weeks under ambient conditions, 2 that had
been prepared by monosubstitution of 20 remained essentially
unchanged. Moreover, moderately pure 2 (∼95%, with the
remainder 13) could be isolated by flash chromatography,
followed by simple evaporation of solvent. However, when 2
had been created from reaction of 4 plus 6, decomposition to
sulfur to form carbamoyl chloride 3. Analogues based on N-2,6-
trimethylaniline, where heterocyclization to a product akin to 14
is impossible, provided additional insights. Reasons for the selec-
tive stabilities of 1 or 2 as a function of their precursors, and/or
the conditions of the reactions that convert those precursors to 1
or 2, remain mysteries.
’ EXPERIMENTAL SECTION
General. Most methods and materials used were described in our
previous publications,13,14 although many of the spectroscopic and
analytical methods are substantially more effective. Labile acid and
sulfenyl chlorides were converted to the corresponding amides and
sulfenamides by adding the compounds to a sufficient excess of
N-methylaniline or a related amine, 2 M in CHCl3 at 4 °C, unless otherwise
specified; after 15 min reaction, extractive workups were carried out as
described previously.13 1H NMR spectra were recorded in CDCl3 (or
any indicated solvent) at 300 MHz (some) or 500 MHz (mostly, and
assumed if not specified otherwise). 13C NMR spectra were recorded in
CDCl3 on the same instruments at 75 or 125 MHz, with CDCl3
normalized to δ 77.0 ppm. IR spectra were recorded in CH2Cl2 or
CHCl3 solutions. Electrospray ionization was used in most cases to
acquire high resolution mass spectra. A spectrometer coupled with a gas
chromatograph or solids probe was used to acquire chemical ionization
(with methane and ammonia) high resolution mass spectra for chlorine-
containing compounds. Elemental analyses were obtained by combus-
tion analysis (for C,H,N,S) and by ion chromatography (for Cl). HPLC
data was acquired using a dual solvent system with UV/vis detection,
equipped with a 250 ꢁ 4.6 mm reversed phase (C18) column (5 μm
particle size). Solvent was methanolꢀwater (4:1, unless otherwise
specified), with 1 mL/min flow rate. All solvent evaporation procedures
were conducted under aspirator vacuum.
(N-Methyl-N-phenylcarbamoyl)sulfenyl Chloride (1)
(Scheme 2). Method A. Neat SO2Cl2 (180 μL, 2.2 mmol) was added
to a solution of O-methyl thiocarbamate 6a (372 mg, 2.0 mmol) in
CDCl3 (2.0 mL). Reaction progress at 25 °C was tracked by repeat 1H
NMR scans of the mixture, as well as N-methylaniline quenches at
appropriate time points. Within moments of addition, peaks attributed
1
to intermediate 9a were noted: H NMR (500 MHz) δ 7.5ꢀ7.1 (m,
5 H), 4.70 and 4.48 (two singlets of equal heights, representing two
configurations, 3 H total), 4.04 and 3.59 (two singlets of equal height,
3 H total); 13C NMR (125 MHz) diagnostic peaks at δ 63.1, 62.8, 45.5,
(carbamoyl)sulfenyl chloride 1 plus COS occurred (Me: t1/2
∼
10 min; Et: t1/2 ∼ 30 min; iPr: t1/2 ∼ 2 min), up to a saturation
point, i.e., ratio of 2:1 = 2:1 for Me, 1:3 for Et, and 4:1 for iPr, with
no further 1 produced. Subsequently, for this set of experiments,
1 lost HCl to form heterocycle 14 (t1/2 ∼ 10 h for all R). In
addition, a modest level of bis(carbamoyl)disulfane 13 was
noted, presumably from reaction of 1 with 6, as hypothesized
to explain other results discussed earlier.
1
42.2. Then, with t1/2 ∼ 8 min, title compound 1 appeared: H NMR
δ 7.65ꢀ7.35 (m, 5 H), 3.38 (s, 3 H); 13C NMR δ 164.0, 138.6, 130.4,
130.1, 128.8, 38.9; along with MeCl formed at the same rate: 1H NMR
δ 3.01 (s, 3 H); 13C NMR δ 25.8. An N-methylaniline quench after 5 min
showed (carbamoyl)sulfenamide 5 as the major product [90% yield;
1
diagnostic H NMR signals at δ 3.38 (s, 3 H) and 3.27 (s, 3 H), as
reported previously13]. After 3 days, heterocycle 14 [diagnostic signal at
δ 3.45 (s, 3 H)] and carbamoyl chloride 3 [δ 3.38 (broad s, 3 H)] both
formed, in a 4:1 ratio. A similar experiment, with CH2Cl2 as solvent and
with the reaction being followed by IR, led to assignment of the main
’ SUMMARY AND CONCLUSIONS
This research has documented a number of ways (Schemes 2
and 3, respectively) to generate two types of species, i.e. (car-
bamoyl)sulfenyl chlorides 1 and ((carbamoyl)dithio)carbonyl
chlorides 2, that had hitherto been thought to be quite unstable.
Instead, compelling spectroscopic and chemical evidence in
support of these structures was obtained. Numerous derivatives,
some new and others matching materials known by alternative
routes, were prepared and characterized. While both 1 and 2
proved to be unusually stable as a result of most, but not all, of the
methods used to obtain them, the pathways of their decomposi-
tions were also elucidated: 2 loses COS to form 1, which either
loses hydrogen chloride to form heterocycle 14 or loses elemental
peaks due to 1: 3054 (m), 1732 (vs) cmꢀ1
.
Similar experiments were carried out starting with O-ethyl thiocarba-
mate 6b (390 mg, 2.0 mmol) or O-isopropyl thiocarbamate 6c (418 mg,
2.0 mmol). Results were similar, except for the kinetics (summarized in
text). The intermediates were 9b [1H NMR (500 MHz) δ 7.5ꢀ7.1 (m,
5 H), 5.15 and 4.98 (two q of equal heights, representing two config-
urations, J = 7.0 Hz, 2 H total), 3.98 and 3.55 (two singlets, 3 H total),
1.62 and 1.32 (t, J = 7.0 Hz, 3 H total); 13C NMR (125 MHz) diagnostic
peaks at δ 176.8, 176.6, 74.7, 74.3, 45.1, 42.1, 14.1, 13.8] and 9c [1H
NMR (500 MHz) δ 7.5ꢀ7.1 (m, 5 H), 6.02 (septet, J = 6.5 Hz, 1 H),
3.98 (s, 3 H), 1.32 (d, J = 6.5 Hz, 6 H); 13C NMR (125 MHz) diagnostic
peaks at δ 86.1, 44.9, 21.9]. Coproduct peaks were noted as follows: EtCl
7886
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
[1H NMR δ 3.57 (q, J = 7.2 Hz, 2 H), 1.48 (t, J = 7.2 Hz, 3 H); 13C NMR
δ 39.8, 18.7] and iPrCl [1H NMR δ 4.19 (septet, J = 6.5 Hz, 1 H), 1.50
(d, J = 6.5 Hz, 6 H); 13C NMR δ 53.8, 27.3]. After 3 days, the ratio of 14
to 3 was 3:1 in the case of series b, and 4:1 in the case of series c.
To confirm the surprising trend involving rates by which intermediate
9 converts to (carbamoyl)sulfenyl chloride 1 as a function of the R group
(i.e., iPr > Me > Et), a mixture of 6a (90 mg, 0.5 mmol), 6b (97 mg,
0.5 mmol), and 6c (104 mg, 0.5 mmol) in CDCl3 (1.5 mL) was treated
with neat SO2Cl2 (138 μL, 1.7 mmol), and reaction progress was tracked
by 1H NMR. At 2 min, 9c was in a 1:1 ratio with iPrCl; at 8 min, 9a was in
a 1:1 ratio with MeCl; and at 12 min, 9b was in a 1:1 ratio with EtCl.
Note that the kinetics of reaction with SO2Cl2 when 6a, 6b, and 6c were
admixed are close to, but not identical to, those determined when each of
these pure compounds was reacted separately with SO2Cl2.
The effect of concentration on the rate of decomposition of 1 to 14
was studied by diluting the reaction mixture of 6b plus SO2Cl2. To create
0.50 or 0.25 M mixtures, neat SO2Cl2 (45 or 22 μL, 0.55 or 0.27 mmol)
was added to a solution of 6b (98 mg or 49 mg, 0.50 or 0.25 mmol) in
CDCl3 (1.0 mL for both experiments). A 0.1 M mixture was created by
adding SO2Cl2 (19 μL, 0.23 mmol) to 6b (39 mg, 0.20 mmol) in CDCl3
(2 mL). Reactions were tracked by 1H NMR, and 1:1 ratios of 1 to 14
were seen at ∼8 h for the 0.50 M mixture, ∼12 h for the 0.25 M mixture
and ∼32 h for the 0.1 M mixture.
and then allowed to warm to 25 °C. Complete conversion of 10 was
1
observed by H NMR within 3 min after addition was finished, to
provide 1 (∼95%) and 5 (∼5%) along with TMSCl (quantitative).
Further time points revealed that 1 decomposed to 14 with a t1/2 ∼ 1 h,
at 25 °C. When a similar experiment was carried out at a final
concentration of 0.1 M, the initial reaction was again very fast, and the
decomposition of 1 to 14 was substantially slower (t1/2 ∼ 24 h).
Further Reactions To Quench or Trap (Carbamoyl)sulfenyl
Chloride 1 (Scheme 2). Once 1 had been generated by any of the
described methods (Scheme 2), the followup transformations described
herein were carried out to support the structural conclusions of this
work. While results were consistent across-the-board, somewhat purer
products were obtained when the starting material for generation of 1
was the O-isopropyl thiocarbamate 6c and when quenching reactions
were conducted at 4 °C within 5 min of the initial reaction. In a few cases,
i.e., for species 50, 500, and 7, the indicated novel products were isolated
and characterized further. For the vast remainder of these studies,
products from the quenches matched reference compounds made
previously and/or in this work by alternative routes. For such latter
cases, mixtures obtained after quenching reactions were spiked with
known amounts of the relevant reference compounds to confirm
identifications of reaction products.
Method A. An aliquot of a reaction mixture containing 1 (typically
∼0.5 mmol scale; concentration ∼0.5 M) in CDCl3 was added quickly
to the appropriate volume of a 2 M solution of an alkanethiol RSH
(R = Me, Et, iPr, tBu) in CHCl3, so that the thiol was in slight excess
(1.2 equiv). After 1 h at 25 °C, these quenched reactions were
evaporated, providing the appropriate carbamoyl disulfide 11 (see later
for reference compounds and their properties). In a typical experiment,
one-quarter portions of the reaction of 6c with SO2Cl2 (previously
described quantities) were quenched after 1 min with each of the four
alkanethiols (0.6 mmol per experiment) to provide the appropriate
11aꢀd in ∼95% purity upon concentration. More specifically, 11a was a
brown oil (101 mg, 95%), 11b was a brown oil (108 mg, 95%), 11c was a
white powder (113 mg, 93%) with mp 45ꢀ49 °C, and 11d was a white
powder (119 mg, 93%) with mp 38ꢀ41 °C. Note that the melting points
of 11c and 11d were essentially identical to authentic materials made as
described later. Quenching conducted after 20 min gave the same
results, with a slight reduction in the purities of 11 to ∼90%, the rest
being mostly 14 and 3.
Method B. Mixtures containing 1 (0.5ꢀ2 mmol; concentration
0.5ꢀ1.0 M) in CDCl3 were added to methanol (equal volume), with
the thought that 1 might react with the solvent, but in fact no
substitution was noted. In one experiment, 1 was made from 6c and
SO2Cl2 (previously described quantities) and “quenched” into CD3OD
(equal volume) after 5 min, and the mixture was followed by 1H NMR.
At the 2-h point, only 1, 3 (not converted to 15), and 14 were observed,
in a ratio of 12:1:4. At the 4-h point, an aliquot of the mixture was reacted
with N-methylaniline to produce 5 (from trapping unreacted 1), 14
(from the continuing intramolecular cyclization of 1), and 3 (from
decomposition of 1; note that it did not react further with methanol to
give 15) in a ratio of 6:4:1. Extractive workup and concentration at this
point gave a tan powder (384 mg, 86%), showing the same ratio of
compounds.
In another experiment, the same reaction of 6c and SO2Cl2 was added
to CD3OD after 5 min and left for 3 days. Concentration at that point
gave only heterocycle 14 and carbamoyl chloride 3 in the usual 4:1 ratio
(305 mg, 92%).
Method C. Mixtures containing 1 (∼2 mmol; concentration ∼1 M)
in CHCl3 were added to the appropriate volume of a 2 M solution of
cyclohexene (1.2 equiv) in CHCl3, for 1 h reaction at 25 °C. Then
concentration and purification by flash chromatography [SiO2, eluted
with hexanes:EtOAc (6:1)] were performed. This process, applied to the
reaction of 6c and SO2Cl2 (previously described quantities), gave 7
The reaction of O-ethyl thiocarbamate 6b with SO2Cl2 (previously
described quantities) was also conducted in the same manner, varying
only the solvent: CD3CN, benzene-d6, and dioxane-d8. Relevant chemical
shifts are reported in the Supporting Information (Supporting Table 1).
Results were the same as in CDCl3, in terms of product distributions,
stabilities, decompositions, and trappings, but there were differences in
terms of kinetics. More precisely, initial adduct 9b formed essentially
instantly in all cases, but the rates of concomitant formation of 1 and
ethyl chloride were increased ∼5-fold in CD3CN (t1/2 ∼ 2 min), slowed
∼5-fold in benzene-d6 (t1/2 ∼ 45 min), and slowed ∼20-fold in dioxane-
d8 (t1/2 ∼ 3.5 h).
Method B. A commercially available solution of hydrogen chloride
(4.0 M) in dioxane was diluted into CDCl3 to create a 1.0 M solution
[HCl (4.7 mmol) in dioxane (1.4 mL) and CDCl3 (3.3 mL)] which was
added to a solution of (carbamoyl)sulfenamide 5 (605 mg, 2.2 mmol) in
CDCl3 (23 mL), under N2 at 25 °C. 1H NMR examination after 15 min
revealed a 1:1 mixture of starting 5 and product 1, with no further
decomposition products noted. At 1 h, the ratio of 5 to 1 was 1:7, and
heterocycle 14 comprised ∼5% of the mixture. At 6 h, starting 5 was no
longer apparent and the ratio of 1 to 14 was 1:1. After 24 h, 1H NMR
showed only heterocycle 14, along with N-methylaniline hydrochloride
[diagnostic signal at δ 3.03 (s, 3 H)].
Method C. Hydrogen chloride gas was bubbled at a moderate rate
through a solution of (carbamoyl)sulfenamide 5 (136 mg, 0.5 mmol) in
CDCl3 (5 mL), under N2, at 25 °C for 1 min. A 1H NMR (500 MHz)
spectrum taken 20 min later revealed that starting 5 was entirely
consumed, while 1 and 14 were present in a 3:1 ratio and N-methylani-
line hydrochloride had formed quantitatively. A time point at 1 h showed
a 1:3 ratio of 1 to 14, and after 2 h, only 14 and the amine salt were noted.
Method D. A 2 M solution of N-methylaniline (0.5 mL, 2.0 equiv) in
CDCl3 (2.0 mL) was added slowly to a solution of chlorocarbonylsulfe-
nyl chloride (4) (42 μL, 0.5 mmol) in CDCl3 (0.5 mL) at 4 °C and then
allowed to warm to 25 °C. 1H NMR examination after 30 min revealed
the presence of 1 and heterocycle 14 in a 2:1 ratio, while after 1 h, these
species were in equal quantities, and after 8 h, 1 had disappeared entirely
and only 14 plus a trace amount of disubstituted compound 5 were
noted. At all time points, a quantitative amount of N-methylaniline
hydrochloride (1 equiv) was present.
Method E. A solution of silylated amine 10 (180 mg, 1.0 mmol) in
CDCl3 (0.5 mL) was added over 1 min to a solution of chlorocarbo-
nylsulfenyl chloride (4) (83 μL, 1.0 mmol) in CDCl3 (0.5 mL) at 4 °C
7887
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
(a new compound) as a clear, colorless oil (325 mg, 65%), with dia-
stereomeric purity indicated by HPLC (MeOH/H2O = 9:1, tr = 7.51 min.);
1H NMR (500 MHz) δ 7.5ꢀ7.2 (m, 5 H), 4.1ꢀ4.0 (m, 1 H), 3.8ꢀ3.7
(m, 1 H), 3.32 (s, 3 H), 2.3ꢀ2.1 (m, 2 H), 1.80ꢀ1.69 (m, 2 H),
1.58ꢀ1.20 (m, 4 H); 13C NMR (75 MHz) δ 167.4, 141.8, 129.4, 128.3,
128.2, 61.9, 50.0, 38.3, 33.7, 30.3, 24.0, 22.7; HRMS (ESI): m/z [M +
Na+] calcd for C14H18ClNOS: 306.0690; found: 306.0695; IR 3053 (s),
of bis(carbamoyl)disulfane 13 (166 mg, 0.5 mmol) and SO2Cl2 (45 μL,
0.55 mmol) in CDCl3 (2.5 mL) was followed by H NMR at 25 °C.
1
Within minutes of addition, carbamoyl chloride 3 and heterocycle 14
were observed in a 1:1 ratio; this ratio was unchanged throughout the
duration of the reaction and was also found once the end point was
reached (no starting 13 remaining). The transformation of 13 to the
aforementioned products proceeded with t1/2 ∼ 45 min. As a practical
matter, intermediate time points required careful scrutiny because of the
similar chemical shifts of starting 13 (sharp singlet) and product 3
(broad singlet).
(N,2,6-Trimethyl-N-phenylcarbamoyl)sulfenyl Chloride
(100). Method A (preferred). Hydrogen chloride gas was bubbled at
a moderate rate through a solution of sulfenamide 1700 (164 mg,
0.5 mmol) in CDCl3 (5 mL), under N2, at 25 °C for 1 min. 1H NMR
revealed that formation of 100 and the amine hydrochloride was complete
within 10 min, and extractive workup and concentration gave 100 as a
brown oil (106 mg, 93%); 1H NMR (500 MHz) δ 7.5ꢀ7.1 (m, 3 H),
3.26 (s, 3 H), 2.28 (s, 6 H); 13C NMR (75 MHz) δ 164.2, 139.1, 130.7,
129.3, 129.0, 36.0, 17.4 (2 C); HRMS (CI) m/z [M + H+] calcd for
C10H12ClNOS: 230.0401; found: 230.0419; IR 3025 (w), 2926 (s),
2696 (s), 1696 (s) cmꢀ1; Anal. Calcd for C10H12ClNOS: C, 52.28; H,
5.27; N, 6.10; S, 13.96; Cl, 15.43. Found: C, 52.70; H, 5.46; N, 5.94; S,
14.12; Cl, 15.48.
Method B. A 2 M solution of N,2,6-trimethylaniline (0.5 mL,
2.0 equiv) in CDCl3 was added slowly to a solution of chlorocarbonyl-
sulfenyl chloride (4) (42 μL, 0.5 mmol) in CDCl3 (0.5 mL) at 4 °C
and then allowed to warm to 25 °C. Tracking by 1H NMR revealed
reaction completion within 10 min and formation of a trace amount
of disubstituted compound 1700. The usual workup gave 100 as a brown
oil (196 mg, 84%).
Method C. Neat SO2Cl2 (180 μL, 2.2 mmol) was added to a solution
of O-ethyl thiocarbamate 600 (223 mg, 2.0 mmol) in CDCl3 (2.0 mL).
Reaction progress at 25 °C was tracked by repeat 1H NMR scans of the
mixture, as well as N,2,6-trimethylaniline quenches at appropriate time
points. Within moments of addition, peaks attributed to intermediate 900
were noted: 1H NMR (500 MHz) δ 5.07 (q, J = 7.0 Hz, 2 H), 3.97 (s, 3
H), 2.31 (s, 6 H), 1.39 (t, J = 7.0 Hz, 3 H); 13C NMR (125 MHz)
diagnostic peaks at δ 74.4, 42.8, 17.3 (2 C), 14.3. Then, with t1/2 ∼ 3 min,
title compound 100 appeared, along with EtCl formed at the same rate. An
N,2,6-trimethylaniline quench after 30 min showed (carbamoyl)sulfen-
amide 500 as only a minor product (5%), with carbamoyl chloride 300
comprising the rest [diagnostic signal at δ 3.25 (s, 3 H)]. After 2 h, the
mixture was concentrated, giving 300 as a tan powder (347 mg, 88%); mp
73ꢀ76 °C (lit.36 mp 78ꢀ79 °C).
2986 (m), 1653 (s) cmꢀ1
.
Method D. Mixtures containing 1 (∼2 mmol; concentration ∼1 M)
in CHCl3 were added, with stirring, to the appropriate volume of a 0.5 M
aqueous solution of potassium iodide (5 equiv). After being stirred for
10 min, the resultant mixture was washed with 1 M aqueous sodium
thiosulfate (3 ꢁ 1 volume) until it was colorless. Extractive workup and
concentration gave a white powder that was predominantly disulfane
1313,26 but included traces of its tri- and tetrasulfane counterparts.14 For
example, when 6c was reacted with SO2Cl2 (previously described
quantities) to produce 1, followed 5 min later by the aforementioned
iodide treatment, 13 represented 90% of the product mixture (265 mg,
79%); mp 225ꢀ230 °C (lit.13 mp 240ꢀ243 °C).
Method E. Mixtures containing 1 (0.5ꢀ2 mmol; concentration 0.5ꢀ
1.0 M) in CHCl3 were added to a 2 M solution of N,2,6-trimethylaniline
(2.2 equiv) in CHCl3 at 4 °C. The derivatization reaction was complete
after 15 min. The case when 1 was generated by reaction of 6c and
SO2Cl2 (previously described quantities), followed by normal workup
procedures, purification by flash chromatography [SiO2, eluted with
hexanes:EtOAc (6:1)], and crystallization under hexanes (6 mL, ꢀ20 °C)
for 1 week, resulted in formation of 500 as white needles (408 mg, 68%);
mp 93ꢀ95 °C; 1H NMR (500 MHz) δ 7.5ꢀ6.9 (m, 8 H), 3.29 (s, 3 H),
3.17 (s, 3 H), 2.41 (broad s, 6 H); 13C NMR (75 MHz) δ 171.9, 148.7,
140.6, 136.6, 129.4, 128.6, 128.4, 128.0, 126.0, 47.2, 38.2, 19.0 (2 C);
HRMS (ESI): m/z [M + Na+] calcd for C17H20N2OS: 323.1189; found:
323.1194; IR 3053 (s), 2986 (w), 1668 (m) cmꢀ1
.
Method F. In a manner similar to method E, starting from 6c and
SO2Cl2 (previously described quantities), followed by an N-methyl-
p-toluidine quench, gave after workup and purification by flash chro-
matography [SiO2, eluted with hexanes:EtOAc (6:1)] compound 50 as a
viscous, brown oil (418 mg, 73%): 1H NMR (300 MHz) δ 7.5ꢀ7.1 (m,
5 H), 3.35 (s, 3 H), 3.26 (s, 3 H), 2.24 (s, 3 H); 13C NMR (125 MHz)
δ 170.0, 147.7, 130.4, 129.5, 129.1, 128.7, 127.8, 122.2, 115.6, 44.6, 37.7,
20.2; HRMS (ESI): m/z [M + Na+] calcd for C16H18N2OS: 309.1032;
found: 309.1028; IR 3027 (w), 2937 (m), 1679 (vs) cmꢀ1
.
Method G. To test the quenching profile of intermediate 9, the
mixture after 1 min of the reaction between 6b and SO2Cl2 (previously
described quantities; a time when the original reaction mixture com-
prised primarily 9b, with little of either starting material 6b or product 1)
was treated with N-methylaniline (2.2 equiv, from a 2 M stock solution
in CDCl3). A 1H NMR spectrum recorded 3 min after the quenching
(no workup) showed the presence of (carbamoyl)sulfenamide 5 in 95%
purity (the rest being 13), along with the stoichiometric amount of EtCl.
No change was seen thereafter. Extractive workup and concentration at
this point gave the mixture as an off-white powder (445 mg, 85%); mp
66ꢀ70 °C (lit.13 mp 71ꢀ72 °C).
Quenches and Trappings of 100. This section refers to quenches
of experiments where (carbamoyl)sulfenyl chloride 100 was generated
in situ and followed closely the procedures already outlined for
unsubstituted 1 (including confirmation of products by doping). As
indicated above, it was also possible to isolate and characterize 100
(methods A and B) and that material could be reacted in similar fashion.
Method A. Quenching mixtures containing 100, 5 min after they were
made from 6b00 plus SO2Cl2 (previously described quantities) with 2 M
solutions of the alkanethiols (1.2 equiv) in CDCl3, 1 h reaction at 25 °C,
gave (after concentratrion) mixtures with 11a00ꢀd00 as ∼35%, with the
rest being 300, which forms rapidly under the reaction conditions. The
MeSH quench gave, after concentration, a brown oil (102 mg), EtSH
quench also gave a brown oil (102 mg), iPrSH quench gave a white powder
(106 mg), and tBuSH quench also gave a white powder (109 mg).
Method B. Reaction of 100 with aqueous KI using the previously
described procedure gave 18 quantitatively based on 100; however, due to
the partial decomposition of 100 to 300 by the time the KI reaction was run,
the same amount of 300 (which does not react with KI) was carried over.
For reaction of 6b00 and SO2Cl2 (previously described quantities),
trapping with KI after 5 min gave a mixture that comprised 18 and 300
Method H. Following a quenching procedure that B€ohme et al.35 used
successfully on chlorocarbonyldisulfanyl chloride, a portion (0.5 mmol)
of the reaction mixture of 6c and SO2Cl2 (previously described
quantities) was added, 3 min after it had reacted to form 1 plus iPrCl,
to a solution of 1,3,5-trimethoxybenzene (117 mg, 0.7 mmol) in CDCl3
1
(0.5 mL) at 25 °C. H NMR after 24 h indicated that there was no
reaction, showing only peaks corresponding to iPrCl, as well as 14 and 3
(ratio 4:1; these are decomposition products of 1, which was no longer
present), and the unreacted starting material, 1,3,5-trimethoxybenzene
(diagnostic peak at δ 3.75; confirmed by doping).
Reaction of Bis(N-methyl-N-phenylcarbamoyl)disulfane
(13) with Sulfuryl Chloride (Scheme 2, footnote b). A solution
7888
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
in a 3:7 ratio. When the KI procedure was applied to the reaction mixture
of N,2,6-trimethylaniline and excess 4 (previously described quantities),
filtration, extractive workup, and concentration gave a white powder (70
mg, 72%), mp 187ꢀ193 °C, of which 18 comprised ∼80% by 1H NMR,
with two further species (∼10% each) [1H NMR respectively δ 3.20 (s),
2.32 (s) and δ 3.13 (s), 2.22 (s)], also present.
To confirm the surprising trend involving the rates by which
intermediate 19 is converted to (dithio)carbonyl chloride 2 as a function
of R group (i.e., iPr > Me > Et), a mixture of 6a (90 mg, 0.5 mmol), 6b
(97 mg, 0.5 mmol), and 6c (104 mg, 0.5 mmol), together in CDCl3
(1.5 mL) was treated with neat 4 (142 μL, 1.7 mmol), and reaction
progress was tracked by 1H NMR. At 3 min, 19c was not seen (complete
conversion to 2 plus iPrCl), 19a was in a 1:1 ratio with MeCl, and 19b
was mostly unconverted. At 16 min, 19a and 19c were gone, the
amounts of iPrCl and MeCl were quantitative, and 19b was in a 1:1
ratio with EtCl.
Method C. Quenching mixtures containing 100 (0.5ꢀ2.0 mmol) with
the appropriate amount of N-methylaniline also produced expected
results, giving 17. The reaction of 6b00 and SO2Cl2 (previously described
quantities) was quenched after 3 min, giving a mixture that comprised 17
and 300 in a 1:1 ratio, along with N-methylaniline hydrochloride commen-
surate with the amount of 17 formed. Extractive workup and concentration
gave the mixture as a tan powder (448 mg, 90%).
Method B. A mixture of concentrated H2SO4 (1.0 mL) and water
(0.2 mL) was added to a solution of carbamoyl disulfide 21 (150 mg,
0.48 mmol) in CDCl3 (2.4 mL), and the heterogeneous mixture was
stirred vigorously at 55 °C for 24 h while HCl gas was released
continuously. At this point, the CDCl3 phase was removed and
Method D. A solution of pure 100 (115 mg, 0.5 mmol) in CHCl3
(0.5 mL) was added to CD3OD (equal volume), and the mixture was
maintained for 24 h at 25 °C, with no changes in its 1H NMR spectrum.
At this point, the mixture was added to a 2 M solution of N-methylaniline
(118 mL, 2.2 equiv) in CDCl3, and 17 was observed as the sole product.
Extractive workup and concentration gave mixed (carbamoyl)sulfenamide
17 as the expected tan powder (137 mg, 92%); mp 144ꢀ148 °C.
((N-Methyl-N-phenylcarbamoyl)dithio)carbonyl Chloride
(2). Method A. Neat chlorocarbonylsulfenyl chloride (4) (184 μL, 2.2
mmol) was added to a solution of O-methyl thiocarbamate 6a (373 mg,
2.0 mmol) in CDCl3 (2.0 mL). Reaction progress at 25 °C was tracked
by repeat 1H NMR scans as well as N-methylaniline quenches at
appropriate time points. Within moments of addition, peaks attributed
to intermediate 19a were noted: 1H NMR (500 MHz) δ 7.5ꢀ7.1 (m,
5 H), 4.48 (more prevalent) and 4.70 (two singlets, ratio 6:5, 3 H total),
4.04 and 3.59 (two singlets, ratio 6:5, 3 H total); 13C NMR (125 MHz)
diagnostic peaks at δ 62.8, 61.6, 46.5, 45.5. Then, with t1/2 ∼ 1.5 min,
title compound 2 appeared: 1H NMR δ 7.5ꢀ7.3 (m, 5 H), 3.40 (s, 3 H);
13C NMR δ 171.4, 164.6, 130.1, 129.9, 128.7, 113.7, 39.6, along with
MeCl formed at the same rate. An N-methylaniline quench after 10 min
showed disulfane 13 and (carbamoyl)sulfenamide 5 in equimolar
1
examined by H NMR, revealing a mixture of product 2 and starting
material 21 in a 9:11 ratio, with no other species present. However,
after 7 days of additional heating, unreacted 21 and heterocycle 14
(from decomposition of 2 to 1 plus COS, and subsequent cyclization
of 1) were present in a 1:1 ratio, with no discernible 2. Concentra-
tion at this point gave a tan powder (95 mg, 82%) which was a 1:1
mixture of 21 and 14.
Method C (preparative). A solution of N-methylaniline (1.29 g,
12 mmol, 2.0 equiv) in CHCl3 (6.0 mL) was added slowly to a solution
of bis(chlorocarbonyl)disulfane (20) (1.15 g, 6.0 mmol) in CHCl3
(6.0 mL) at 4 °C. After 15-min reaction, the mixture was concentrated,
and flash chromatography [SiO2, eluted with hexanes:EtOAc (1:1)]
provided 2 as a clear viscous colorless oil [672 mg, 43%; mostly
separated from starting 20 and bis(carbamoyl)disulfane 13]; HRMS
(CI): m/z [M + H+] calcd for C9H8ClNO2S2: 261.9758; found:
261.9754; IR 3055 (s), 2983 (m), 1779 (w), 1687 (s) cmꢀ1; Anal.
Calcd for C9H8ClNO2S2: C, 41.30; H, 3.08; N, 5.35; S, 24.50; Cl, 13.54.
Found: C, 41.69; H, 3.22; N, 5.27; S, 26.27; Cl, 12.05. The aforemen-
tioned data are admittedly different from ideal, and this is attributed to
the presence of 13 (∼5%) that had not been entirely removed during the
chromatography step.
Method D. A solution of silylated amine 10 (181 mg, 1.0 mmol) in
CDCl3 (0.5 mL) was added slowly to a solution of bis(chlorocar-
bonyl)disulfane (20) (191 mg, 1.0 mmol) in CDCl3 (0.5 mL) at 4 °C.
1H NMR examination within 3 min revealed that the reaction was
complete (starting 10 consumed; stoichiometric amount of TMSCl
formed), with products 2 and 13 present in a 3:2 molar ratio. The
relative amount of 2 remained unchanged after even a day at 25 °C.
Concentration of the reaction mixture after 15 min revealed the
presence of 2 and 13 in the same ratio and accounting for essentially
all of the material (287 mg, 95%).
Quenches and Trappings of Compound 2. Quenches of all
mixtures containing 2 were carried out in the same way as that
performed for mixtures containing 1 or 100. In situ reactions described
next were used only for diagnostic purposes and not to isolate products.
All quenches were confirmed by doping with authentic products,
prepared as described later. Note that 2, once isolated, reacts with the
various trapping reagents to provide the appropriate products in high
yields and purities.
1
quantities [93% yield; diagnostic H NMR signals of 13 at δ 3.36 (s,
3 H) as reported previously13]; this result reflects relatively rapid formation
of compound 2 and decomposition of approximately half of that compound
2 to (carbamoyl)sulfenyl chloride 1 plus COS over the 10-min time span.
An N-methylaniline quench was carried out 2 days after the original reaction
and lasted for an additional 24 h to allow any carbamoyl chloride 3 to react
to form urea 16. This experiment showed 13 and 14 in a 2:1 ratio, with urea
16 comprising a further ∼3% of the mixture. This result was interpreted to
mean that decomposition of compound 2had stopped and that compound 1,
formed by decomposition of 2 earlier in the process, had by now entirely
cyclized to 14. To confirm this, the 2-day reaction mixture was
evaporated directly to provide a brown oil (422 mg, 92%) which
comprised stable 2 and heterocycle 14 in a 2:1 ratio.
Similar experiments were carried out starting with O-ethyl thiocarba-
mate 6b (390 mg, 2.0 mmol) or O-isopropyl thiocarbamate 6c (418 mg,
2.0 mmol). Results were similar, except for the kinetics (summarized in
text). The intermediates were 19b [1H NMR (500 MHz) δ 7.5ꢀ7.1 (m,
5 H), 5.21 and 5.01 (two quartets of equal heights, representing two
configurations, J = 6.0 Hz, 2 H total), 4.00 and 3.55 (two singlets, 3 H
total), 1.63 and 1.33 (t, J = 6.0 Hz, 3 H total); 13C NMR (125 MHz)
diagnostic peaks at δ 176.8, 176.7, 74.5, 74.2, 45.0, 42.0, 14.1, 13.8] and
19c [1H NMR (500 MHz) δ 7.5ꢀ7.1 (m, 5 H), 6.17 and 6.02 (two
septets, J = 6.0 Hz, 1 H total), 3.99 and 3.54 (two singlets, 3 H total), 1.68
and 1.42 (two doublets, J = 6.0 Hz, 6 H total); 13C NMR (125 MHz)
diagnostic peaks at δ 86.1, 80.7, 44.9, 43.6, 22.3, 21.9]. EtCl and iPrCl
were the respective coproducts of these transformations. The 3-day
reaction mixtures were evaporated to give brown oils [starting with 6b,
344 mg (91%) comprising 2 and 14 in a 1:3 ratio, with a small amount
(2%) of 3; starting with 6c, 455 mg (95%) comprising 2 and 14 in a 4:1
ratio, with 3 not observed].
Method A. Mixtures containing 2 (0.5ꢀ2.0 mmol; concentration
0.5ꢀ1.0 M) in CDCl3 solution were added to a 2 M solution (2.2 equiv)
of N,2,6-trimethylaniline at 4 °C, to ultimately provide unsymmetrical
disulfane 1300. When the reaction mixture of 6c and 4 (previously
described quantities) was quenched after 30 min, (carbamoyl)disulfane
1300 was observed alongside (carbamoyl)sulfenamide 500 in a 4:1 ratio;
this result was interpreted to indicate that about 20% of ((carbamoyl)-
dithio)carbonyl chloride 2 had decomposed to sulfenyl chloride 1, which
would react further to give 500. Also, N,2,6-trimethylaniline hydrochlor-
ide [diagnostic 1H NMR (500 MHz) signals at δ 3.07 (t, J = 5.5 Hz, 3 H),
7889
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
2.67 (s, 6 H)] was observed in the immediate reaction mixture.
Extractive workup and concentration gave the 4:1 mixture of 1300 and
500 (but without the amine hydrochloride) as a white powder (626 mg,
90%); mp 126ꢀ131 °C.
(28.5 mL) was added to a solution of N,2,6-trimethylaniline (8.1 g,
60 mmol, 2.1 equiv) in CHCl3 (30 mL). Extractive workup gave 6b00 as a
brown oil (4.58 g, 69%). The compound showed two conformations,17
1
in a 2:1 ratio (major listed first, then minor); H NMR (500 MHz)
Method B. Trapping in the same fashion as method A was conducted
but using N-methyl-p-toluidine. When the reaction mixture of 6c and 4
(previously described quantities) was quenched after 2 min, 130
comprised nearly all of the mixture, and the corresponding amount of
(major) δ 7.2ꢀ7.0 (m, 3 H), 4.47 (q, J = 7.0 Hz, 2 H), 3.48 (s, 3 H), 2.13
(s, 6 H), 1.33 (t, J = 7.0 Hz, 3 H); (minor) δ 7.2ꢀ7.0 (m, 3 H), 4.59 (q,
J = 7.0 Hz, 2 H), 3.24 (s, 3 H), 2.21 (s, 6 H), 1.43 (t, J = 7.0 Hz, 3 H); 13C
NMR (125 MHz) (major) δ 188.3, 141.4, 134.0, 128.1, 127.5, 67.1, 41.0,
17.3, 14.1; (minor) δ 187.9, 142.9, 134.5, 128.5, 127.7, 67.4, 36.7, 17.4,
14.2; HRMS (ESI): m/z [M + Na+] calcd for C12H17NOS: 246.0923;
1
N-methyl-p-toluidine hydrochloride [diagnostic H NMR signals at δ
2.86 (s, 3 H), 2.27 (s, 3 H)] was present. Extractive workup and
concentration gave 130 as a white powder (588 mg, 85%); mp 126ꢀ130 °C.
Method C. When mixtures containing 2 (0.2ꢀ2.0 mmol; concentra-
tion 0.2ꢀ2.0 M) in CDCl3 solution were added quickly to excess
methanol (equal volume) at 25 °C, ((methoxycarbonyl)dithio)carbamoyl
derivative 8 formed, as observed 1 h later upon evaporation. For example,
when compound 21 (0.2 mmol) was partially hydrolyzed with H2O/H2SO4
at 55 °C for 24 h (method B to form 2), a methanol quench showed a 9:11
ratio of derivative 8 to unreacted starting 21. Concentration gave a tan
powder (46 mg, 79%) with the same ratio of 8 to 21.
found: 246.0926; IR 3017 (m), 2980 (s), 2927 (m), 1675 (m) cmꢀ1
.
This procedure was carried out both in 1983 and in 2010; the older
material that had been stored under ambient conditions was still usable
for reactions carried out in 2010.
N-Trimethylsilyl-N-methylaniline (10). Modifying a procedure
outlined briefly by Cullis et al.,21 a solution of N-methylaniline (2.0 g,
19.0 mmol) and N,O-bis(trimethylsilyl)trifluoroacetamide (5.0 g, 19.5
mmol) in dry acetonitrile (26 mL) was refluxed at 105 °C for 24 h.
Subsequent solvent removal and vacuum distillation provided the title
product as a pale yellow oil (2.04 g, 60%), bp 32ꢀ33 °C (0.7 mm) [lit.21
59ꢀ61 °C (1.9 mm); lit.18 77 °C (5.5 mm); lit.19 91 °C (14 mm)].
When the limiting reagent was the silylating agent, as in the original
publication, it was difficult to separate fully unreacted N-methylaniline
from the title product.
Method D. To test the quenching profile of intermediate 19, the
mixture after 1 min of the reaction between 6b and chlorocarbonylsul-
fenyl chloride (4) (previously described quantities; a time when the
original reaction mixture comprised primarily 19b, with little of either
starting material 6b or product 2) was treated with N-methylaniline (2.2
1
equiv, from a 2 M stock solution in CDCl3). A H NMR spectrum
Alkyl (N-Methyl-N-phenylcarbamoyl)disulfanes (11). In a
typical procedure, a solution of ((isopropyl)dithio)carbonyl chloride
(12c) (339 mg, 2.0 mmol) in CHCl3 (1.0 mL) was added to a solution of
N-methylaniline (470 mg, 2.2 equiv) in CHCl3 (2.2 mL). Extractive
workup, followed by flash chromatography [SiO2, eluted with hexanes:
EtOAc (6:1)] and crystallization under hexanes (5 mL) at ꢀ20 °C for 24
h, gave 11c as a white powder (344 mg, 62%); mp 49ꢀ51 °C (lit.13 mp
54ꢀ55 °C); 1H NMR (300 MHz) δ 7.5ꢀ7.3 (m, 5 H), 3.37 (s, 3 H),
3.05 (septet, J = 6.6 Hz, 1 H), 1.27 (d, J = 7.0 Hz, 6 H); 13C NMR
(75 MHz) δ 166.8, 141.3, 129.6, 128.8, 128.4, 41.1, 39.3, 22.1 (2 C);
HRMS (ESI): m/z [M + Na+] calcd for C11H15NOS2: 264.0487; found:
264.0480; IR 3018 (s), 2965 (m), 1673 (vs) cmꢀ1. Applying the same
procedure with tBu reagent 12d (368 mg, 2.0 mmol) provided 11d as
recorded 3 min after the quenching (no workup) showed the presence of
disulfane 13 in 90% purity (the rest being 5), along with the stoichio-
metric amount of EtCl. No change was seen thereafter. Extractive
workup and concentration at this point gave the mixture as an off-white
powder (541 mg, 86%); mp 221ꢀ225 °C (lit.13 mp 240ꢀ243 °C).
(N,N0-Dimethyl-N,N0-diphenylcarbamoyl)sulfenamide (5).
Neat chlorocarbonylsulfenyl chloride (4) (42 μL, 0.5 mmol) was added
to a solution of silylated amine 10 (180 mg, 1.0 mmol) in CDCl3
(1.0 mL) at 4 °C. 1H NMR of the reaction mixture, recorded within 3
min, showed that 10 was no longer present, and that 5 (0.5 equiv) plus
TMSCl (1 equiv) had formed nearly quantitatively. Concentration gave
title product 5 as an off-white powder (256 mg, 94%); mp 68ꢀ70 °C
(lit.13 mp 71ꢀ72 °C).
1
white needles (357 mg, 70%); mp 40ꢀ42 °C; H NMR (300 MHz)
δ 7.5ꢀ7.2 (m, 5 H), 3.33 (s, 3 H), 1.26 (s, 9 H); 13C NMR (75 MHz)
165.5, 140.6, 128.9, 128.1, 127.7, 47.0, 38.6, 28.8 (3 C); HRMS (ESI):
m/z [M + Na+] calcd for C12H17NOS2: 278.0644; found: 278.0636; IR
O-Methyl N-Methyl-N-phenylthiocarbamate (6a). Material
made by quenching methoxythiocarbonyl chloride with N-methylani-
line, followed by the standard quantitative aqueous extractive workup,13
was still good and usable 30 years later, having been stored under
ambient conditions; 1H NMR17 (300 MHz, 50 °C) δ 7.5ꢀ7.2 (m, 5 H),
3.95 (broad s, 3 H), 3.64 (broad s, 3 H); with negligible MeS(CdO)
isomer,13 diagnostic δ 3.33 (s, 3 H), 2.26 (s, 3 H). Further characteriza-
tion of this compound was reported previously.13
O-Ethyl N-Methyl-N-phenylthiocarbamate (6b). N-Methyl-
aniline (14.3 mL, 0.13 mol) was added to a solution of bis(ethoxy-
(thiocarbonyl)) sulfide36 (28.0 g, 0.13 mol) in CHCl3 (200 mL). After 2
days at 25 °C, extractive workup provided 6b as a yellow oil (21.2 g,
82%); 1H NMR17 (300 MHz, 50 °C) δ 7.5ꢀ7.1 (m, 5 H), 4.50 (q, J = 7.2
Hz, 2 H), 3.58 (s, 3 H), 1.22 (theoretical triplet appearing as broad
singlet, 3 H). Further characterization of this compound was reported
previously.13
3061 (w), 2960 (s), 1684 (vs) cmꢀ1
.
Alkyl (N,2,6-Trimethyl-N-phenylcarbamoyl)disulfanes (1100).
A solution of ((methyl)dithio)carbonyl chloride (12a) (283 mg, 2.0
mmol) in CHCl3 (1.0 mL) was added to a solution of N,2,6-trimethy-
laniline (593 mg, 2.2 equiv) in CHCl3 (2.2 mL). Extractive workup,
followed by flash chromatography [SiO2, eluted with hexanes:EtOAc
(6:1)], provided 11a00 as a white powder (386 mg, 80%); mp 68ꢀ70 °C;
1H NMR (500 MHz) δ 7.3ꢀ7.1 (m, 3 H), 3.25 (s, 3 H), 2.38 (s, 3 H),
2.24 (s, 6 H); 13C NMR (75 MHz) 166.4, 137.8, 137.5, 129.4, 128.9,
36.1, 23.3, 17.6 (2 C); HRMS (ESI): m/z [M + Na+] calcd for C11H15-
NOS2: 264.0487; found: 264.0494; IR 3018 (s), 1671 (s) cmꢀ1. The
same procedure, including purification, using Et derivative 12b (312 mg,
1
2.0 mmol) provided 11b00 as a brown oil (418 mg, 82%); H NMR
O-Isopropyl N-Methyl-N-phenylthiocarbamate (6c). N-Methyl-
aniline (20 mL, 185 mmol) was added to a solution of bis(isopropoxy-
(thiocarbonyl)) sulfide14 (10.8 g, 45 mmol) in CHCl3 (185 mL) at 25 °C.
After 2 days, extractive workup gave the crude product, which was then
placed under petroleum ether for 24 h at ꢀ20 °C, giving 6c as white
crystals (7.4 g, 78%); mp 72ꢀ74 °C (lit.14 mp 67ꢀ70 °C); 1H NMR13
(300 MHz, 50 °C) δ 7.5ꢀ7.1 (m, 5 H), 5.56 (septet, J = 6.0 Hz, 1 H), 3.60
(s, 3 H), 1.20 (poorly resolved doublet, J = 6.0 Hz, 6 H).
(300 MHz) δ 7.3ꢀ7.1 (m, 3 H), 3.24 (s, 3 H), 2.70 (q, J = 7.5 Hz, 2 H),
2.25 (s, 6 H), 1.26 (t, J = 7.5 Hz, 3 H); 13C NMR (75 MHz) δ 166.5,
137.8, 137.4, 129.3, 128.8, 36.1, 32.7, 17.5, 13.6; HRMS (ESI): m/z [M +
Na+] calcd for C12H17NOS2: 278.0644; found: 278.0640; IR 3029 (w),
2984 (w), 1674 (s) cmꢀ1. Use of iPr derivative 12c (339 mg, 2.0 mmol)
provided 11c00 as a viscous, brown oil after purification (446 mg, 83%);
1H NMR (500 MHz) δ 7.3ꢀ7.1 (m, 3 H), 3.20 (s, 3 H), 2.99 (septet, J =
6.5 Hz, 1 H), 2.23 (s, 6 H), 1.23 (d, J = 6.5 Hz, 6 H); 13C NMR (75
MHz) δ 166.6, 137.8, 137.4, 129.2, 128.8, 41.0, 36.1, 22.0 (2 C), 17.5
(2 C); HRMS (ESI): m/z [M + Na+] calcd for C13H19NOS2: 292.0800;
O-Ethyl N,2,6-Trimethyl-N-phenylthiocarbamate (6b00). A
solution of ethoxythiocarbonyl chloride13 (3.55 g, 28.5 mmol) in CHCl3
7890
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
found: 292.0804; IR 3022 (w), 2959 (s), 2922 (s), 2862 (m), 1687
(vs) cmꢀ1. Finally, using tBuderivative 12d (369 mg, 2.0 mmol), followed
bypurification and then recrystallization under hexanes atꢀ20 °C over 24
h, gave 11d00 as white needles (418 mg, 74%); mp 92ꢀ94 °C; 1H NMR
(300 MHz) δ 7.3ꢀ7.1 (m, 3 H), 3.24 (s, 3 H), 2.28 (s, 6 H), 1.28 (s, 9 H);
13C NMR (75 MHz) δ 166.6, 138.1, 137.5, 129.2, 128.9, 47.7, 36.4, 29.5
(3 C), 17.6 (2 C); HRMS (ESI): m/z [M + Na+] calcd for C14H21NOS2:
306.0957; found: 306.0962; IR 3023 (w), 2959 (m), 1685 (s) cmꢀ1; Anal.
Calcd for C14H21NOS2: C, 59.32; H, 7.47; N, 4.94; S, 22.62. Found: C,
59.22; H, 7.40; N, 4.84; S, 22.47.
138.1, 137.5, 129.7, 129.5, 128.9, 128.6, 128.4, 39.3, 36.2, 17.6 (2 C);
HRMS (ESI): m/z [M + Na+] calcd for C18H20N2O2S2: 383.0858;
found: 383.0860; IR 3017 (m), 2939 (w), 1684 (s), 1676 (s) cmꢀ1
.
(N,N0,2,6-Tetramethyl-N,N0-diphenylcarbamoyl)sulfenamide
(17). A solution of pure (carbamoyl)sulfenyl chloride 100 (115 mg, 0.5
mmol) in CHCl3 (0.5 mL) was added to a solution of N-methylaniline
(118 mL, 2.2 equiv) in CHCl3 (0.5 mL). Extractive workup and purifi-
cation by flash chromatography [SiO2, eluted with hexanes:EtOAc
1
(6:1)] gave 17 as a tan powder (130 mg, 86%); mp 76ꢀ78 °C; H
NMR (500 MHz) δ 7.3ꢀ6.8 (m, 8 H), 3.38 (s, 3 H), 3.17 (s, 3 H), 2.35
(broad d, J = 13.5 Hz, 6 H); 13C NMR (75 MHz) δ 170.0, 150.2, 138.2,
136.5, 129.5, 129.0, 128.7, 119.5, 115.5, 44.7, 35.1, 17.7 (2 C); HRMS
(ESI): m/z [M + Na+] calcd for C17H20N2OS: 323.1189; found:
((Alkyl)dithio)carbonyl Chlorides (12). While these com-
pounds (R = Me, Et, iPr,39 tBu40) have all been described previously,
including purifications by distillation,1,38 the small-scale procedures for
12c and 12d that follow were newly developed to suit the needs of this
research. A solution of chlorocarbonylsulfenyl chloride (4) (200 μL,
2.4 mmol) in CHCl3 (2.4 mL) was added to a solution of iPrSH
(198 mg, 2.6 mmol) in CHCl3 (2.4 mL). After 1 h at 25 °C, the mixture
was concentrated, providing 12c as a clear, colorless oil (375 mg, 92%);
1H NMR (500 MHz) δ 3.24 (septet, J = 7.0 Hz, 1 H), 1.36 (d, J = 7.0 Hz,
6 H); 13C NMR (75 MHz) 167.0, 42.6, 22.3 (2 C); GC-MS: (60 m ꢁ
0.25 mm ID, DB-5, 50 °C for 1 min, then increase by 15 °C per min until
323.1195; IR 3019 (s), 1668 (m) cmꢀ1
.
(N,N0,2,20,6,60-Hexamethyl-N,N0-diphenylcarbamoyl)sulfen-
amide (1700). A solution of chlorocarbonylsulfenyl chloride (4) (84 μL,
1.0 mmol) in CHCl3 (2.0 mL) was added to a solution of N,2,6-
trimethylaniline (594 mg, 4.4 equiv) in CHCl3 (2.2 mL), followed by
extractive workup and purification by flash chromatography [SiO2,
eluted with hexanes:EtOAc (6:1)] to give 1700 as short, white needles
(233 mg, 71%); mp 146ꢀ149 °C; 1H NMR (500 MHz) δ 7.5ꢀ7.2 (m, 6
H), 3.18 (s, 3 H), 3.16 (s, 3 H), 2.38 (broad s, 6 H), 2.21 (s, 6 H); 13C
NMR (75 MHz) δ 171.8, 148.6, 137.5, 137.0, 136.5, 129.0, 128.7, 128.5,
125.7, 47.2, 35.3, 19.0 (2 C), 17.6 (2 C); HRMS (ESI): m/z [M + Na+]
calcd for C19H24N2OS: 351.1502; found: 351.1510; IR 3054 (s), 2986
+
320 °C; tR = 6.50 min); HRMS (CI): m/z [M + NH4 ] calcd for
C4H7ClOS2: 187.9965; found: 187.9973; IR 2973 (s), 2927 (m), 2866
(m), 1782 (vs) cmꢀ1. The same procedure was used with tBuSH (234
mg, 2.6 mmol), providing 12d as a translucent, brown oil (411 mg,
93%); 1H NMR (300 MHz) δ 1.38 (s, 9 H); 13C NMR (75 MHz) 167.2,
34.4, 29.6 (3 C); GC-MS: (same conditions as for 12c; tR = 6.98 min);
(m), 1671 (vs) cmꢀ1
.
Bis(N,2,6-trimethyl-N-phenylcarbamoyl)disulfane (18). A
solution of bis(chlorocarbonyl)disulfane (20) (761 mg, 4.0 mmol) in
CHCl3 (4.0 mL) was added to a solution of N,2,6-trimethylaniline (2.32
g, 4.3 equiv) in CHCl3 (8.0 mL). Extractive workup gave 18 as a white
powder (1.36 g, 88%); mp 195ꢀ197 °C; 1H NMR (300 MHz)
δ 7.3ꢀ7.1 (m, 6 H), 3.24 (s, 6 H), 2.35 (s, 12 H); 13C NMR (75 MHz)
δ 165.0 (2 C), 138.2 (2 C), 137.6 (4 C), 129.6 (4 C), 129.0 (4 C), 36.2
(2 C), 17.7 (4 C); HRMS (ESI): m/z [M + Na+] calcd for
C20H24N2O2S2: 411.1171; found: 411.1181; IR 3025 (w), 2965
+
HRMS (CI): m/z [M + NH4 ] calcd for C5H9ClOS2: 202.0122; found:
202.0126; IR 2965 (s), 1785 (vs) cmꢀ1
.
Bis(N-methyl-N-phenylcarbamoyl)disulfane (13). Method A.
A solution of SO2Cl2 (42 μL, 0.5 mmol) in CDCl3 (0.5 mL) was added
over 1 min to a solution of O-isopropyl thiocarbamate 6c (209 mg,
1
1.0 mmol) in CDCl3 (0.5 mL) at 25 °C. H NMR data at 10 min
showed completion of the reaction to produce 13 (quantitative, with
all starting material consumed and converted to iPrCl). Direct concen-
tration gave 13 as a white powder, ∼98% pure by 1H NMR (156 mg, 94%);
(w), 2926 (w), 1674 (s) cmꢀ1
.
1
mp 227ꢀ231 °C (lit.13 mp 240ꢀ243 °C); H NMR (500 MHz) 7.5ꢀ
(Trichloromethyl)(N-methyl-N-phenylcarbamoyl)disulfane
(21). A solution of perchloromethyl mercaptan (0.6 g, 3.2 mmol) in
CH2Cl2 (3 mL) was added to a solution of O-isopropyl N-methyl-
N-phenylthiocarbamate (0.6 g, 2.7 mmol) in CH2Cl2 (2.7 mL), and the
reaction proceeded for 15 min at 25 °C. Concentration gave the crude
product (0.9 g, 2.8 mmol, quantitative), a portion (0.54 g, 1.7 mmol) of
which was dissolved in hot hexanes (5.4 mL). White needles formed
upon cooling (0.17 g total for two crops, 32%); mp 50ꢀ52 °C (lit.13 mp
54ꢀ55 °C).
7.2 (m, 10 H), 3.36 (s, 6 H).
Method B. Neat bis(chlorocarbonyl)disulfane (20)13 (95 mg, 0.5
mmol) was added to a solution of silylated amine 10 (181 mg, 1.0 mmol)
1
in CDCl3 (1.0 mL). A H NMR spectrum of the reaction mixture
recorded after 3 min showed that 10 had reacted completely, while 13
(0.5 equiv) and TMSCl (1 equiv) had formed. Concentration at this
stage gave 13 as an off-white powder (305 mg, 92%); mp 230ꢀ233 °C
(lit.13 mp 240ꢀ243 °C).
(p-Methyl-N-phenylcarbamoyl)(N-methyl-N-phenyl)disulfane
(130). A solution of ((carbamoyl)dithio)carbonyl chloride 2 (270 mg,
1.0 mmol) in CHCl3 (1.0 mmol) was added to a solution of N-methyl-p-
toluidine (264 mg, 2.2 mmol) in CHCl3 (1.0 mL). The usual workup,
followed by flash chromatography [SiO2, eluted with hexanes:EtOAc
(6:1)], provided 130 as a white powder (311 mg, 90%); mp 127ꢀ
130 °C; 1H NMR (500 MHz) 7.5ꢀ7.2 (m, 9 H), 3.36 (s, 3 H), 3.33 (s,
3 H), 2.38 (s, 3 H); 13C NMR (125 MHz) δ 166.7, 141.3, 129.6, 128.8,
128.3, 41.1, 39.2, 22.0; HRMS (ESI): m/z [M + Na+] calcd for
C17H18N2O2S2: 369.0702; found: 369.0703; IR 3019 (s), 2977 (w),
’ ASSOCIATED CONTENT
1
S
Supporting Information. A table of H NMR shifts in
b
various solvents, a description of chemical exchange in com-
pound 6, and copies of key NMR spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
1683 (m), 1676 (m) cmꢀ1
.
Corresponding Author
*E-mail: barany@umn.edu.
(N,2,6-Trimethyl-N-phenylcarbamoyl)(N-methyl-N-phenyl)-
disulfane (1300). A solution of ((carbamoyl)dithio)carbonyl chloride 2
(270 mg, 1.0 mmol) in CHCl3 (1.0 mL) was added to a solution of
N,2,6-trimethylaniline (297 mg, 2.2 mmol) in CHCl3 (1.0 mL). Ex-
tractive workup and purification by flash chromatography [SiO2, eluted
with hexanes:EtOAc (6:1)] gave 1300 as a white powder (317 mg, 88%);
mp 131ꢀ133 °C; 1H NMR (500 MHz) 7.5ꢀ7.1 (m, 8 H), 3.35 (s, 3 H),
3.24 (s, 3 H), 2.34 (s, 6 H); 13C NMR (75 MHz) δ 164.6, 163.2, 140.9,
’ ACKNOWLEDGMENT
We thank Steven J. Eastep for preliminary studies in the N,2,
6-trimethylaniline family, Al Mutassim Abu-Khdeir, Matthew J.
Bongers, Kristin M. Coari, David K. Ford, David A. Halsrud,
Lesia Tchobaniouk, Chelsea R. Vandegrift, and Xiaolu Zheng for
7891
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892

The Journal of Organic Chemistry
ARTICLE
assistance with preparation of some starting materials, Letitia Yao
and Kent Mann for helpful discussions about the temperature
dependence of NMR spectra of thiocarbamates, Jed Fisher and
Robert P. Hammer for critical readings of the manuscript, and
anonymous referees for useful suggestions that improved the
final manuscript.
magnetic field strength (i.e., 80 MHz). These new results are attributed
to conformational averaging (medium exchange) on the NMR time
scale; the fact that thiocarbamates can populate discrete cis and trans
conformational states was first discussed by Bauman, R. A. J. Org. Chem.
1
1967, 32, 4129–4132. Consistent with this hypothesis, high-field H
NMR spectra of 6 run at higher temperature gave sharper peaks with the
expected multiplicities (details in Supporting Information, pages
S2ꢀS5, with representative spectra on subsequent pages). Also, the
1H NMR spectrum of the corresponding 2,6-dimethyl compounds, 600,
showed the normal expected sharp peaks at 25 °C, indicating that cis and
trans isomers of 600 are likely to be in slow exchange.
(18) From N-methylaniline plus chlorotrimethylsilane in the pre-
sence of triethylamine, but without experimental details: Klebe, J. F.;
Bush, J. B., Jr.; Lyons, J. E. J. Am. Chem. Soc. 1964, 86, 4400–4406.
(19) Prepared by using NaH to create anion of N-methylaniline and
then alkylating with chlorotrimethylsilane: Rauchschwalbe, G.; Ahlbrecht,
H. Synthesis 1974, 663–665.
’ DEDICATION
†Dedicated to Kate Bꢀarꢀany (April 29, 1929 to June 13, 2011),
gifted teacher, renowned biophysical chemist, and champion for
women in academia.
’ REFERENCES
(1) Zumach, G.; K€uhle, E. Angew. Chem., Int. Ed. Engl. 1970, 9, 54–63.
(2) Reference 1 also draws the analogy to the well-known industrial-
scale process for production of N,N-dialkylthiocarbamoyl chlorides, i.e.,
(20) Procedure similar to ref 18 with 65% yield: Smith, C. J.; Early,
T. R.; Holmes, A. B.; Shute, R. E. Chem. Commun. (Cambridge, U. K.)
2004, 17, 1976–1977.
1
/[R2N(CdS)S]2 + 1/Cl2 f [R2N(CdS)SCl] f R2N(CdS)Cl + S, to
2
2
(21) From N-methylaniline plus N,O-bis(trimethylsilyl)trifluoro-
acetamide, and reproduced herein (other procedures proved difficult
to carry out to completion), as reported in: Cullis, C. F.; Herron, D.;
Hirschler, M. M. Combust. Flame 1985, 59, 151–165.
(22) Barany, M. J.; Hammer, R. P.; Merrifield, R. B.; Barany, G. J. Am.
Chem. Soc. 2005, 127, 508–509 and references cited therein.
(23) Authentic 11, several of which were new compounds, were
prepared by acylation of N-methylaniline by the corresponding
(alkyldithio)carbonyl chlorides 12, which in turn were derived from
regiospecific reactions of thiols with chlorocarbonylsulfenyl chloride 4,
as previously described.1,13,37
(24) Precedent: Kharasch, N.; Wald, M. M. Anal. Chem. 1955, 27,
996–998.
(25) Precedent: Barany, G.; Mott, A. W. J. Org. Chem. 1984, 49,
1043–1051 and references cited therein.
(26) Kobayashi, N.; Osawa, A.; Fujisawa, T. Chem. Lett. 1973,
12, 1315–1318.
support their view that R2N(CdO)SCl should be unstable.
(3) (Carbamoyl)sulfenyl chloride R2N(CdO)SCl for R2N = iPr2N
was reported as an unstable intermediate in a complicated rearrange-
ment/elimination that occurred upon oxidation of certain 3-chloroallyl
thiocarbamates. See: Schuphan, I.; Casida, J. E. Tetrahedron Lett. 1979,
10, 841–844.
(4) Notwithstanding the caveat of ref 2, a full paper following up on
ref 3 sketched how with R2N = iPr2N that 1/[R2N(CdO)S]2 + 1/Cl2 f
2
2
[R2N(CdO)SCl] (1), sufficiently stable to allow trapping with
cyclohexene (compare to Scheme 1) prior to decomposing to R2N-
(CdO)Cl (3). Experimental details and spectral data were not
reported. See: Schuphan, I.; Casida, J. E. J. Agric. Food Chem. 1979, 27,
1060–1066.
(5) A recently described one-pot synthesis of 3 by reaction of
secondary amines with 4 undoubtedly goes through 1 as an intermediate
(compare to Scheme 1). See: Adeppa, K.; Rupainwar, D. C.; Misra, K.
Can. J. Chem. 2010, 88, 1277–1280.
(27) Moltzen, E. K.; Senning, A. Sulfur Lett. 1986, 4, 93–96.
(28) Besthorn, E. Ber. Dtsch. Chem. Ges. 1910, 43, 1519–1526.
(29) Fife, T. H.; Hutchins, J. E.; Wang, M. S. J. Am. Chem. Soc. 1975,
97, 5878–5882.
(30) Shapiro, S. L.; Weinberg, K.; Bazga, T.; Freeman., L. J. Am. Chem.
Soc. 1958, 80, 3734–3738.
(31) Gassman, P.; Campbell, G.; Frederick, R. J. Am. Chem. Soc.
1972, 94, 3884–3891.
(32) Mariappan, P.; Jayakumar, K.; Bharathi, P. J. Org. Chem. 2000,
65, 3548–3550.
(33) While the overwhelming majority of mono- and bifunctional
acid and sulfenyl chlorides react quickly with methanol or N-methylani-
line (reactions under conditions described in Experimental Section are
complete within 15 min), carbamoyl chloride 3 reacts at a very slow rate
for reasons on which we can only speculate. To achieve complete
conversion respectively to urethane 15 or symmetrical urea 16, reaction
times of 48 h at 50 °C were necessary.
(34) Harris, J. F., Jr. J. Am. Chem. Soc. 1960, 82, 155–158.
(35) Bo€hme, H.; Brinkmann, M.; Steudel, H. P. Liebigs Ann. Chem.
1981, 1244–1251.
(36) Sievertsson, H.; Nilsson, J.; Lars, G. Acta Chem. Scand. 1970,
24, 939–945.
(37) Barany, G.; Fulpius, B. W.; King, T. P. J. Org. Chem. 1978, 43,
2930–2932.
(38) Barany, G. Int. J. Pept. Protein Res. 1982, 19, 321–324.
(39) Zalipsky, S.; Albericio, F.; Slomczynska, U.; Barany, G. Int. J.
Pept. Protein Res. 1987, 30, 740–783 for R = iPr.
(40) W€unsch, E.; Moroder, L.; Nyfeler, R.; Jaeger, E. Hoppe Seyler's
Z. Physiol. Chem. 1982, 363, 197–202 for R = tBu.
(6) A handful of additional “hits” for R2N(CdO)SCl (1) in the
patent literature turn out, upon close reading, to actually refer to
R2N(CdS)Cl that had been erroneously indexed as 1.
(7) From our own laboratories, tantalizing hints at the existence,
trapping, and stabilities of 1 and 2 are found in footnote 14 and the text
accompanying Scheme III of ref 14 (with brief accompanying experi-
mentals for compounds 20 and 22 as numbered in that paper) and on
pages 39 and 59 of the Ph.D. Thesis of Schroll, A. L. (1986), University
of Minnesota.
(8) Schroll, A. L.; Barany, G. J. Org. Chem. 1989, 54, 244–247.
(9) Chen, L.; Golser, R.; Machova, A.; Slaninova, J.; Barany, G.
J. Med. Chem. 1999, 42, 5002–5009.
(10) Kolano, C.; Sander, W. Eur. J. Org. Chem. 2003, 6, 1074–1079.
(11) Morin, T. J.; Kobertz, W. R. Proc. Natl. Acad. Sci. U.S.A. 2008,
105, 1478–1482 including Supporting Information which has relevant
experimental procedures.
(12) The stereochemistry of addition of a variety of sulfenyl chlor-
ides to alkenes is reviewed by: K€uhle, E. Synthesis 1971, 563–586.
(13) Barany, G.; Schroll, A. L.; Mott, A. W.; Halsrud, D. A. J. Org.
Chem. 1983, 48, 4750–4761 and references cited therein.
(14) Schroll, A.; Barany, G. J. Org. Chem. 1986, 51, 1866–1881 and
references cited therein.
(15) Review: K€uhle, E. Synthesis 1970, 561–580.
(16) Precedents for this chemoselective acidolytic cleavage of a
sulfenamide bond are found in ref 1 and in: Schroll, A; Eastep, S. J.;
Barany, G. J. Org. Chem. 1990, 55, 1475–1479 and references cited
therein.
(17) Attention is called to the unusually broad 1H NMR peaks of 6a,
6b, and 6c as observed at 25 °C and 500 MHz for the present study,
which contrast to the classical expected multiplicities when these same
compounds were examined years earlier (refs 13 and 14) at lower
7892
dx.doi.org/10.1021/jo201329n |J. Org. Chem. 2011, 76, 7882–7892