Investigating the reaction of N-carbobenzoxy-O-carbobenzoxy-hydroxylamine with dimethyl sulfoxide: Formation of S,S-dimethyl-N-[(phenylmethoxy)-carbonyl] sulfoximine

Article abstract of DOI:10.1021/jo051158a

N-Carbobenzoxy-O-carbobenzoxyhydroxylamine (1a) underwent a thermally induced reaction in DMSO in which there is net N-α-eliminative oxidation with tandem oxidative incorporation of DMSO to yield S,S-dimethyl-N-[(phenyl- methoxy)carbonyl]sulfoximine. Mech

Full text of DOI:10.1021/jo051158a

Investigating the Reaction of
N-Carbobenzoxy-O-carbobenzoxy-
hydroxylamine with Dimethyl Sulfoxide:
Formation of
S,S-Dimethyl-N-[(phenylmethoxy)-
carbonyl]sulfoximine
Ron W. Darbeau,*,§ Noe T. Alvarez,§
Gregory A. Trahan,† and Frank R. Fronczek‡
Department of Chemistry, McNeese State University, Lake
Charles, Louisiana 70609, Department of Chemistry,
Louisiana State University, Baton Rouge, Louisiana 70803,
and Research & Development Division, Sasol North America
Inc., 2201 Old Spanish Trail, Westlake, Louisiana 70669
FIGURE 1. (a) N-Carbobenzoxy-O-carbobenzoxyhydroxyl-
amine (1a). (b) N-Carbobenzoxyhydroxylamine (1b).
rdarbeau@mail.mcneese.edu
Received June 8, 2005
FIGURE 2. 5-(Benzyloxy)-2,2-dimethyl-1,3,2λ4,4-dioxathiaz-
ole (2).
N-Carbobenzoxy-O-carbobenzoxyhydroxylamine (1a) under-
went a thermally induced reaction in DMSO in which there
is net N-R-eliminative oxidation with tandem oxidative
incorporation of DMSO to yield S,S-dimethyl-N-[(phenyl-
methoxy)carbonyl]sulfoximine. Mechanisms for the forma-
tion of the sulfoximine are presented as well as the product
characterizations, including the X-ray crystal structure.
FIGURE 3. Crystal structure of S,S-dimethyl-N-[(phenyl-
methoxy)carbonyl]sulfoximine.
cations react nearly quantitatively with arenes,2 and
disobey the Brown Selectivity Relationship.3
In the present work, 1a underwent an unknown
reaction in DMSO-d6 at elevated temperatures. Hydroly-
sis of 1a is not the operative route because, although
benzyl alcohol was formed, only traces of N-carbobenzoxy-
hydroxylamine4a (1b) were detected. Other unknown
trace minor products were formed; however, the uniden-
tified product (δ 5.01 in DMSO-d6) and benzyl alcohol
were the only major products. Interestingly, when 1a was
heated in methylene chloride (at 60 °C) and toluene (at
100 °C), no corresponding reaction occurred. The reaction
was originally thought to be a thermal isomerization in
which N,N-dicarbobenzoxyhydroxylamine (1c) was pro-
duced from 1a (Scheme 1) via the charge-separated
We recently reported1 the synthesis and characteriza-
tion of N-carbobenzoxy-O-carbobenzoxyhydroxylamine
(1a), whose nitrosation and subsequent decomposition
yields carbocations in multiple spacer-molecule separated
ion pairs,1,2 with reactivities and lifetimes significantly
larger than those of their standard deaminative coun-
terparts.1,2 Such hyperdeaminatively generated benzyl
* To whom correspondence should be addressed.
§ McNeese State University.
† Research & Development Division, Sasol North America Inc.
‡ Louisiana State University.
(1) (a) Darbeau, R. W.; Trahan, G. A.; Siso, L. M. Org. Biomol. Chem.
2004, 2, 695. (b) Siso, L. M. M.S. Thesis, McNeese State University,
Lake Charles, LA, 2003. (c) Trahan, G. A. M.S. Thesis, McNeese State
University, Lake Charles, LA, 2004. (d) Darbeau, R. W.; Trahan, G.
A.; Siso, L. M. Presented at the 228th National Meeting of the
American Chemical Society, Philadelphia, PA, Aug 2004.
(2) Darbeau, R. W.; Siso, L. M.; Trahan, G. A.; Nolan, R. S. Lett.
Org. Chem. 2005, 2, 54.
(3) Brown, H. C.; Nelson, K. L. J. Am. Chem. Soc. 1953, 75, 6292.
(b) Smoot, C. R.; Brown, H. C. J. Am. Chem. Soc. 1956, 78, 6249. (c)
Stock, L. M.; Brown, H. C. J. Am. Chem. Soc. 1959, 81, 3323. (d) Stock,
L. M.; Brown, H. C. Adv. Phys. Org. Chem. 1963, 1, 35.
(4) Benzyl N-hydroxycarbamate or N-(benzyloxycarbonyl)hydroxyl-
amine.
10.1021/jo051158a CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/18/2005
J. Org. Chem. 2005, 70, 9599-9602
9599

SCHEME 1. Proposed Mechanism for the Thermal Isomerization of 1a to 1c
transition state 1d. This putative reaction would proceed
in DMSO (ꢀ ) 46.7),5a,b but not in the low-polarity/
polarizability solvents CH2Cl2 (ꢀ ) 8.9)5a,b and toluene (ꢀ
) 2.4).5a,b We investigated this reaction and its kinetics
to test our hypothesis.
Cambridge Crystallographic Database contains the Nd
S(Me2)dO moiety.6c
FT-IR Spectrum of 3. A key feature of the spectrum
of 3 is the intense peak at 1657 cm-1 attributable to the
carbamate moiety. Bands at 1264, 1195, and 1019 cm-1
are characteristic of the -NdSdO moiety;7 those at 1456
cm-1 and at 1380 and 1329 cm-1 likely correspond to CH2
scissoring and C-H bending of the gem dimethyl groups
of 3.
A 20 mg sample of 1a was heated in 750 µL of DMSO-
d6 at 70 °C for 4 days. Removal of the solvent in vacuo
followed by washing the residue with 3 × 1 mL aliquots
of ether afforded a pure white solid (mp 99.5-100.5 °C)
that gave a single peak in the aliphatic region of the 1H
NMR spectrum at δ 5.01. Elemental analysis (C, H, N,
and O) of the unknown revealed that 17.35% of its mass
was unaccounted for, indicating the incorporation of
sulfur and perhaps deuterium from DMSO-d6. The reac-
tion was rerun in dry DMSO-h6, and the product was
1H NMR Spectrum of 3. Compound 3 gives rise to
an aromatic multiplet at δ 7.34 (5H), a benzylic singlet
at δ 5.12 (2H), and the methyl singlet at δ 3.25 (6H).
13C NMR Spectrum of 3. There are four types of
aromatic carbons assigned by their proximity to the
oxygen of the benzyloxy group. The ipso carbon appears
at δ 136.2, the para carbon (also the least intense) at δ
128.1. The meta carbons arise at δ 128.3, and the ortho
carbons at δ 128.5. The three types of nonaromatic
carbons are the carbamate C at δ 159.0, the methyl
carbons at δ 41.9, and the benzyl carbon at δ 67.9.
Potential Mechanisms for the Formation of 3.
Interestingly, there are three reasonable mechanisms
(Scheme 2) for the formation of 3 that differ in the roles
of DMSO and the intermediacy of a labile dioxathiazolone
(5) along the reaction pathway. There are several note-
worthy points regarding Scheme 2. In mechanism A,
DMSO acts as an electrophile,6d and suffers attack by
the N of 1a so that S-N bond formation is the first step.
Deprotonation of the first-formed intermediate yields a
species (4a) capable of eliminating the elements of the
benzyl carbonate ion as CO2 and PhCH2O- to form 3
directly (path A1). Alternatively, in path A2, 4a may
undergo a conformational change and then cyclization
to 5, from which decarboxylative fragmentation yields 3.
In mechanism B (Scheme 2), DMSO functions as a
nucleophile,6e attacking the carbonate group of 1a with
the expulsion of PhCH2O- to form the cationic species
4b. The latter cyclizes by formation of the S-N bond to
form 5, which decomposes, as in the previous case.
Regardless of the operative mechanism(s), however, the
overall reaction may be viewed as a net N-R-eliminative
oxidation with tandem oxidative incorporation of DMSO
or as a decarboxylative sulfoximation to yield 3.
1
isolated and characterized via H and 13C NMR, FT-IR,
UV-vis, and elemental analysis (C, H, N, O, and S).
A new singlet was observed at δ 3.25, which cor-
responded to H atoms previously replaced by D atoms in
the preceding synthesis. This latter peak bore a 6:5
integral ratio with the aromatic signal, and a 3:1 ratio
with the benzylic proton signal (δ 5.01), indicating that
it corresponded to 6 equivalent methyl protons via the
incorporation of DMSO into the new compound. No
heteroatom (broad, D2O-exchangeable) peak was ob-
served. The FT-IR indicated the lack of an imidodicar-
bonyl doublet in the approximate region 1840-1740 cm-1
.
Percentage composition from elemental analysis of the
product yielded an empirical formula of C10H13NO3S,
confirming that the product is not 1c. From the molecular
formula and spectral data, we proposed1c,d structure 2,
and reported the reaction as a decarboxylative [3+2]
cycloaddition of DMSO across the amido unit of 1a.1c,d
X-ray Diffraction Study on the Unknown. Struc-
ture determination of a crystal grown by evaporation of
an ethyl acetate-based solution of the unknown com-
pound showed that it is not heterocyclic,1c,d but possesses
a SdN bond6a-c and is, in fact, S,S-dimethyl-N-[(phenyl-
methoxy)carbonyl]sulfoximine (3). We have found only
one reference to this compound6b (but no crystal structure
(Figure 3)). Interestingly, only one other compound in the
(5) Buncel, E.; Rajagopal, S. Acc. Chem. Res. 1990, 23, 226 and
references therein. (b) Isaacs, N. S. Physical Organic Chemistry, 2nd
ed.; John Wiley and Sons: New York, 1995. (c) Pretsch, E.; Buhlmann,
P.; Affolter, C. Structure Determination of Organic Compounds: Tables
of Spectral Data, 3rd ed.; Springer: Berlin, 2000.
Determination of Activation Parameters for Sul-
foximine Formation. We investigated the kinetics of
the formation of 3 by heating a solution of 1a in DMSO-
d6 at 80, 90, and 95 °C in the NMR probe and taking
spectra at regular intervals. From the integrals as a
function of time,6f we made plots of the ln(% 3) versus
(6) (a) SciFinder identified one reference6b to this compound and one
other reference to a species with a NdS(Me2)dO moiety.6c (b) Kirby,
G. W.; Mackinnon, J. W. M.; Sharma, R. P. Tetrahedron Lett. 1977,
215. (c) Heine, H. W.; Empfield, J. R.; Golobish, T. D. J. Org. Chem.
1986, 51, 829. (d) Examples of DMSO as an electrophile include: Choi,
J. H.; Yoon, M. Y.; Yun, J. H.; Chung, D. W. Bull. Korean Chem. Soc.
1995, 16 (5), 416. Johnson, C. R.; Kirchhoff, R. A.; Reischer, R. J.;
Katekar, G. F. J. Am. Chem. Soc. 1973, 95 (13), 4287. (e) Examples of
DMSO as a nucleophile include: Dalton, D. R.; Dutta, V. P.; Jones, D.
C. J. Am. Chem. Soc. 1968, 90, 5498. Pfitzner, K. E.; Moffatt, J. G. J.
Am. Chem. Soc. 1965, 87, 5661. (f) See Supporting Information
(7) (a) Banks, R. E.; Prakash, A. Tetrahedron Lett. 1973, 99. (b)
Bentley, H. R.; Whitehead, J. K. J. Chem. Soc. 1950, 2081. (c)
Heintzelman, R. W.; Bailey, R. B.; Swern, D. J. Org. Chem. 1976, 41
(12), 2207.
9600 J. Org. Chem., Vol. 70, No. 23, 2005

SCHEME 2. Potential Routes for the Reaction of
N-Carbobenzoxy-O-carbobenzoxyhydroxylamine
with DMSO To Form S,S-Dimethyl-N-[(phenyl-
methoxy)carbonyl]sulfoximine
SCHEME 3. Proposed Mechanism for the
Reaction of 1b with DMSO To Form 3, and
Subsequent Hydrolysis of the Latter to Benzyl
Alcohol
circumvent the formation of 5 and proceed as in path A1,
for example.
Relative Stability of 1b in DMSO. When 1b is
incubated with DMSO, trace formation of 3 is detected
under only the more extreme conditions of 110 °C for 2
days. In contrast, with 1a, the reaction is complete after
11 h at 110 °C. This observation is interesting for several
reasons. (1) 1b being devoid of the benzyl carbonate
moiety would make it incapable of proceeding through
either the B or A2 pathway (Scheme 2). Consequently, if
1a and 1b follow analogous pathways to 3 (probable,
given the unusual nature of both the reagents and
product), then pathway A1 would appear to be more
likely than pathways B and A2. (2) The N in 1b should
be more nucleophilic than that in 1a because the lone
pairs on the O of the adjacent O-H are not in resonance
with a carbonate carbonyl as in 1a. Thus, a full R-effect
is felt by the lone pair on the nitrogen of 1b. Conse-
quently, nucleophilic attack by 1b on DMSO should be
more facile than the parallel attack by 1a. (3) However,
decarboxylative elimination from 1a (a relatively facile
process because of the positive ∆S* associated with one
particle forming three, including a gas) is replaced, in
1b, by the loss of OH-, a relatively poor nucleofuge,
without the benefit of a tandem decarboxylative event.
We postulate that the mechanisms for the thermal
reactions of 1a and 1b in DMSO both involve nucleophilic
attack by N on DMSO. This forms a zwitterionic entity
whose rapid deprotonation yields an anion that elimi-
nates either OH- (1b) or CO2 and PhCH2O- (1a) in the
RDS. This step is accompanied by increases in entropy
on going from the intermediate 4a (or 4a′ for 1b; Scheme
3) to the activated complex. The large ∆S* in the case of
1a is apparently critical for keeping the ∆G* low enough
for observable chemistry under less strenuous conditions.
As observed in the case of 1b, in which no decarboxyla-
tion occurs, the ∆S* is only slightly positive, and the ∆G*
rises, making the reaction more difficult,.
TABLE 1.
Kinetic Dataa for Thermolyses of 1a in DMSO
temperature
(°C)
85
90
95
rate constant
2.3
3.9
6.2
(×10-2 s-1
)
Activation Parametersb for Thermolyses of 1a in DMSO
Ea
∆H*
∆S*
parameter
value
A
(kJ mol-1
)
(kJ mol-1
)
(J mol-1 K-1
)
1.21 × 1014
107.7
104.7
14.7
a Data are averages of at least duplicate runs; max. std. dev. )
2 × 10-4
.
b Data are averages of at least duplicate runs; max. std.
dev. ) 2 × 10-2
.
time (R2 values ranged from 0.93 to 0.97). From these
plots, we determined the rate constants (equal to the
slopes) for each temperature (Table 1a), and Arrhenius
(R ) 0.9997)6f and Eyring (R ) 0.9996)6f plots were made.
The data (Table 1b) show a very large preexponential
factor consistent with a first-order reaction or, as postu-
lated here, a pseudo-first-order one in which DMSO
functions as both the solvent and reagent. Additionally,
the positive ∆S* value indicates that decarboxylation is
an event that occurs up to or including the rate-
determining step (RDS). Hence one of the following takes
place: (1) decarboxylation (in which ∆S* > 0) precedes
cyclization (in which ∆S* < 0), and the latter occurs after
the RDS; (2) decarboxylation but not cyclization occurs
along the reaction profile (path A1; Scheme 2); or (3)
decarboxylation and cyclization are concurrent processes
(paths A2 and B; Scheme 2). Option (1) appears unlikely,
given that the formation of 5 would be expected to be
somewhat difficult because of the stereochemical de-
mands of an sp2-hybridized N, a C, and a T-shaped S.
Thus, the Ea for cyclization would probably be too high
for the cyclization to not be involved in the RDS. Also, if
the cyclization is indeed difficult, the reaction path may
Kinetics of 1b + DMSO f 3. We attempted to
determine the kinetics of the formation of 3 from 1b and
DMSO. However, the reaction occurs only at observable
rates at ∼135 °C. Unfortunately, although some 3 is
observed, benzyl alcohol dominates the reaction product,
even when the reactions are carried out in sealed tubes
in extensively dried DMSO. We have accounted for this
event by the mechanism shown in Scheme 3. Thus, the
generation of 3 from 1b requires the tandem formation
of 1 equiv of water. At the elevated temperature, this (in
J. Org. Chem, Vol. 70, No. 23, 2005 9601

situ) water evidently hydrolyzes 3 into benzyl alcohol and
a sulfonyl derivative(s). This conclusion is supported by
the fact that the sum of the % composition of 1a, 3, and
benzyl alcohol approximates 100% throughout the reac-
tion (see Table 2, Supporting Information). The hydroly-
sis of 3 makes determining the kinetics of its formation
difficult. Nonetheless, the requirement of more strenuous
conditions for the initiation of the reaction from 1b as
compared to 1a lends strong support to our proposal.
A Brief Look at Sulfoximines. Sulfoximines are an
interesting class of compounds that have been generated
via reactions of azidopentafluorobenzene with DMSO,7a
oxidation of sulfilimines,7b treatment of amines with a
DMSO-tert-butyl hypochlorite complex,7c reaction of
o-quinone monoimides with sulfoxides,6c and reaction of
alkoxycarbonyl azides with DMSO (vide infra).6b They are
biologically active compounds with roles such as inhibi-
tion of GSH synthesis, which has implications in the
background resistance of cancer cells to chemotherapeutic
agents,8a protection of the mitochondrial electron trans-
port chain from damage by oxidative stress in astrocytes
and neurons,8b and survival of U937 cells via inhibition
of the apoptotic program downstream of the release of
cytochrome c.8c They also play roles in medicine by
decreasing blood-brain, large-neutral amino acid trans-
port in adults,8d and in agriculture by inhibiting phy-
tochelatin synthesis.8e Reviews of sulfoximine chemistry
are available.9
General Mechanism for Sulfoximine Formation
from Alkoxycarbonyls. Nitrogen atoms of alkoxycar-
bonyl compounds are evidently sufficiently nucleophilic
to attack even the relatively weakly electrophilic DMSO.
The reaction presented in this work appears to be a
variant of a side reaction observed by Kirby et al.,6d in
which an alkoxycarbonyl azide undergoes deaminative
sulfoximination to yield 3.6d It is noteworthy that the
carbamyl nitrogen in all three cases (1a, 1b, and 6) is
activated by either the R-effect (1a, 1b) or resonance (6)
to offset the drain of electron density from the adjacent
carbonyl (although that drain is somewhat diminished
by the carbamyl O).
The reaction of 6 with DMSO, and its mechanism, is
consistent with our observations and postulates (vide
supra).
Experimental Section
N-Carbobenzoxy-O-carbobenzoxyhydroxyl-
amine (1a). 1b (2.8 g) and dry NaHCO3 (9.8 g) were
placed in 200 mL of methylene chloride, and 3.1 g (1.1
equiv) of benzyl chloroformate was added dropwise with
stirring. The flask was charged with Ar and sealed with
a rubber septum, and a syringe needle attached to a
balloon of Ar was placed. The reaction was run for 24 h
at 20 °C with vigorous stirring. The suspension was
filtered, and the filtrate was evaporated in vacuo. The
residue was recrystallized from 1:3 ether:hexane in 52%
yield:10 mp 72-74 °C; IR (KBr) 3221, 1805, 1716, 1506,
1244, 1146 cm-1; 1H NMR (CDCl3) δ 5.23 (s, 2H), 5.26 (s,
2H), 7.35 (s, 5H), 7.37 (s, 5H), 7.83 (s, 1H); 13C NMR
(CDCl3) δ 68.5, 71.6, 128.4, 128.6, 128.7, 128.8, 128.9,
129.1, 134.3, 135.2, 155.5, 156.8; UV (CD3CN) λmax 252
(ꢀ ) 368), 257 (ꢀ ) 487), 262 (ꢀ ) 445), 268 nm (ꢀ ) 292).
Anal. Calcd for C16H15NO5: C, 63.78; H, 5.02; N, 4.65;
O, 26.55; Found: C, 63.69; H, 4.88; N, 4.63; O, 26.43.
S, S-Dimethyl-N-[(phenylmethoxy)carbonyl]sul-
foximine (3). 1a (180 mg) was dissolved in 5 mL of dry
DMSO in a pressure vial, and the vial was then sealed
with a Teflon-lined cap. The solution was placed at 80
°C for 2 days, after which the DMSO was removed in
vacuo. The remaining viscous liquid was washed twice
with 3 mL of boiling ether, after which precipitation of a
white solid occurred. The precipitate was then pumped
for 1 h to yield pure 3 in 16% yield:10 mp 99.5-100.5 °C;
IR (KBr) 3052, 3024, 2939, 1657, 1264, 1195, 1019, 994,
896, 790, 739, 696 cm-1; 1H NMR (CDCl3) δ 3.25 (s, 6H),
5.12 (s, 2H), 7.34 (m, 5H); 13C NMR (CDCl3) δ 41.9, 67.9,
128.1, 128.3, 128.5, 136.2, 159.0; UV (Acetonitrile) λmax
252 (ꢀ ) 286), 258 (ꢀ ) 354), 268 nm (ꢀ ) 196). Anal.
Calcd for C10H13NO3S: C, 52.85; H, 5.77; N, 6.16; O,
21.12; S, 14.11. Found: C, 52.75; H, 5.64; N, 6.03; O,
21.13; S, 13.83.
Acknowledgment. Acknowledgment is given to the
Louisiana Education Quality Support Fund, Grants
LEQSF(1999-2000)-ENH-TR-13, LEQSF-(2001-04)-RD-
A-23, and LEQSF-(2003-05)-RD-B-06, the Calcasieu
Parish Industrial and Development Board Endowed
Professorship, the Pinnacle Corporation, and the Mc-
Neese State University Chemistry Department. The
authors also thank Prof. Kenton H. Whitmire, Prof.
Peter Loos, and Teyeb O. Ely of the Rice University
Chemistry Department for their kind assistance in this
project.
(8) (a) Akan, I.; Akan, S.; Akca, H.; Savas, B.; Ozben, T. Eur. J.
Clin. Invest. 2004, 34 (10), 683. (b) Gegg, M. E.; Clark, J. B.; Heales,
S. J. R. Brain Res. 2005, 1036 (1-2), 1. (c) Filomeni, G.; Aquilano, K.;
Rotilio, G.; Ciriolo, M. R. Antioxid. Redox Signaling 2005, 7 (3-4), 446.
(d) De Marco, A.; Owczarek, M.; Raglione, M.; Lanza, B. Mutat. Res.
2005, 581 (1-2), 133.
(9) (a) Johnson, C. R. In Comprehensive Organic Chemistry; Jones,
N., Ed.; Pergamon Press: Oxford, 1979; Vol. 3, Chapter 11. (b)
Kennewell, P. D.; Taylor, J. B. Chem. Soc. Rev. 1975, 4, 180. (c)
Kennewell, P. D. Chem. Soc. Rev. 1980, 4, 477. (d) Truce, W. E.;
Klinger, T. C.; Brand, W. B. In Organic Chemistry of Sulfur; One, S.,
Ed.; Plenum Press: New York, 1977; Chapter 10.
Supporting Information Available: Crystallographic
information file (CIF) for compound 3. Kinetics tables and
Arrhenius and Eyring plots for thermolysis of 1a in DMSO.
This material is available free of charge via the Internet at
http://pubs.acs.org.
JO051158A
(10) Nonoptimized yield.
9602 J. Org. Chem., Vol. 70, No. 23, 2005