Sulfur imidations by light-induced ruthenium-catalyzed nitrene transfer reactions

Article abstract of DOI:10.1002/ejoc.201500220

N-Acyl nitrenes have been generated from a range of heterocyclic precursors, and their applications in light-induced ruthenium-catalyzed sulfur imidations have been studied. Analyzing the reaction scope and determining the structural requirements of the in situ formed electrophilic nitrogen species for effective nitrene transfer allowed a mechanistic scheme to be proposed. The mechanistic conclusions were substantiated by the identification of potential intermediates.

Full text of DOI:10.1002/ejoc.201500220

FULL PAPER
DOI: 10.1002/ejoc.201500220
Sulfur Imidations by Light-Induced Ruthenium-Catalyzed Nitrene Transfer
Reactions
Vincent Bizet[a] and Carsten Bolm*[a]
Keywords: Synthetic methods / Reaction mechanisms / Ruthenium / Photochemistry / Decarboxylation
N-Acyl nitrenes have been generated from a range of hetero-
cyclic precursors, and their applications in light-induced
ruthenium-catalyzed sulfur imidations have been studied.
Analyzing the reaction scope and determining the structural
requirements of the in situ formed electrophilic nitrogen spe-
cies for effective nitrene transfer allowed a mechanistic
scheme to be proposed. The mechanistic conclusions were
substantiated by the identification of potential intermediates.
wide range of sulfur substituents to be accessed (dialkyl,
diaryl, and alkyl aryl). In addition, a “one-pot” sulfur imid-
ation/oxidation sequence was developed by taking advan-
tage of the same metal (ruthenium) in the reaction mixture
for both oxidative steps. Accordingly, N-acyl sulfoximines 5
could be rapidly accessed in yields of up to 99% under mild
reaction conditions starting from the corresponding sulfides
2 (Scheme 1, top). This protocol was subsequently applied
to the synthesis of methionine sulfoximine (MSO, 6a) and
buthionine sulfoximine (BSO, 6b),[10] which are two well-
established chemotherapeutic agents that are known to re-
duce the level of glutathione in cells (Scheme 1, bottom). A
comparison of the “one-pot” imidation/oxidation pro-
Introduction
The development of original synthetic methods for the
preparation of highly functionalized molecules is an ongo-
ing challenge. Sulfimides[1] and sulfoximines[2] are charac-
terized by core fragments with a high heteroatom content.[3]
Recently, newly discovered biological properties led to a
clear recognition of the value of such compounds in medici-
nal chemistry[4] and crop protection.[5] In the preparation
of sulfimides and sulfoximines, sulfur imidations play a
prominent role.[6–12] Surprisingly, only a few procedures al-
low the direct conversion of a sulfide or sulfoxide into the
corresponding N-acyl sulfimide or sulfoximine. A common
feature of such reactions is the involvement of an N-acyl
nitrene or a metallo-nitrenoid.[7] In this context, an early
finding by Sauer and Mayer caught our attention.[8] While
investigating the reactivity of 3-substituted 1,4,2-dioxazol-
5-ones 1 under photochemical or thermal conditions they
found that heating 1 to 150 °C in dimethyl sulfoxide
(DMSO) resulted in decarboxylation and formation of the
corresponding N-acyl sulfoximine 5. Hypothesizing that im-
provements to such a decarboxylative imidation process
may be possible by applying a combination of metal cataly-
sis and photoinitiation, we recently discovered an advanced
procedure allowing the conversion of a wide range of sulf-
ides 2 and sulfoxides 3 by utilizing a combination of 1 as
N-acyl nitrene precursor and a catalytic amount of a ruth-
enium(II) porphyrin under irradiation with visible light.[9]
By using this strategy, the corresponding N-acyl sulfimides
4 and sulfoximines 5 could be obtained in good to excellent
yields (up to 99%) at ambient temperature. Noteworthy was
the broad substrate scope, which allowed products with a
[a] Institute of Organic Chemistry, RWTH Aachen University,
Landoltweg 1, 52056 Aachen, Germany
E-mail: Carsten.Bolm@oc.RWTH-Aachen.de
http://bolm.oc.rwth-aachen.de/
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500220.
Scheme 1. Ruthenium-catalyzed sulfur imidation/oxidation by
light-induced decarboxylation of 1 (top); application of this proto-
col to the synthesis of MSO and BSO (bottom).
2854
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 2854–2860

Sulfur Imidations by Nitrene Transfer Reactions
cedure with alternative approaches towards MSO/BSO re- ide (1ad; 86% yield), and 2,2-diethoxypropane in the pres-
vealed the superiority of the former method.
ence of 10-camphorsulfonic acid (CSA) providing 5,5-di-
Herein, we report the details of our investigation and de- methyl-3-phenyl-1,4,2-dioxazole (1af; 77% yield). 5-Phenyl-
scribe the results of mechanistic studies leading to an im- 1,3,4-oxathiazol-2-one (1ac) was prepared from benzamide
proved understanding of the light-induced catalytic sulfur (9a) in 68% yield by treatment with an excess of chloro-
imidation.
carbonylsulfenyl chloride in refluxing toluene. 5-Phenyl-
1,3,2,4-dioxathiazole 2,2-dioxide (1ae) was obtained from
1ad in 60% yield by using a mild ruthenium-catalyzed oxid-
ation under phase-transfer conditions. Finally, reaction of
benzohydrazide (10a) with sodium nitrite in the presence of
acetic acid led to benzoyl azide (1ag) in 75% yield
(Scheme 2).
Results and Discussion
Choice of the N-Acyl Nitrene Precursor
The study was initiated by the synthesis and evaluation
of phenyl-substituted compounds 1aa–ag (Scheme 2),
which we considered to be potential sources for N-benzoyl
nitrene (8a). In this series, the only variation came from the
nature of the leaving group: CO2 for 1aa, COS for 1ab and
1ac, SO2 for 1ad, SO3 for 1ae, acetone for 1af, and N2 for
1ag. Most of these heterocycles were obtained from N-
hydroxybenzamide (7a), which was treated with a specific
reagent (Scheme 2). Those were carbonyldiimidazole (CDI)
for the synthesis of 3-phenyl-1,4,2-dioxazol-5-one (1aa;
91% yield), thiophosgene in the presence of triethylamine
for 3-phenyl-1,4,2-dioxazole-5-thione (1ab; 97% yield),
thionyl chloride to give 5-phenyl-1,3,2,4-dioxathiazole 2-ox-
The potential of compounds 1aa–ag to serve as a source
for N-benzoyl nitrene (8a) in photochemically-induced
ruthenium-catalyzed amidations was tested in reactions
with thioanisole (2a) and methylphenylsulfoxide (3a). The
expected products were N-benzoyl sulfimide 4aa and N-
benzoyl sulfoximine 5aa, respectively. As catalyst, 1 mol-%
Ru(TPP)CO (12a) was applied. As reported previously,[9]
1aa was a highly efficient precursor of 8a, leading to 4aa
and 5aa in essentially quantitative yields after 4 h of irradia-
tion at room temperature (Table 1, entry–1). Its C=S ana-
logue 1ab showed moderate reactivity, affording 4aa and
5aa in yields of 87 and 56%, respectively (entry–2). In con-
trast, 5-phenyl-1,3,4-oxathiazol-2-one (1ac), which is an O/
S isomer of 1ab, did not react under these conditions. The
S=O analogue of 1aa, compound 1ad, showed a low reac-
tivity, giving 4aa and 5aa in only 36 and 5% yield, respec-
tively (entry–4). Almost no reaction occurred by applying
heterocycles 1ae and 1af (entries 5 and 6). Finally, the use
of benzoyl azide (1ag), which is a common source of N-
benzoyl nitrene (8a), was tested; again, the yields of 4aa
(5%) and 5aa (10%) were low (entry–7).
Table 1. Light-induced ruthenium-catalyzed imidations of 4aa and
5aa with heterocycles 1aa–ag.[a]
Entry
1a
X = Y
Yield of 4aa [%][b] Yield of 5aa [%][b]
1
2
3
4
5
6
7
1aa
1ab
1ac
1ad
1ae
1af
1ag
CO2
COS
COS
SO2
SO3
acetone
N2
99
87
0
36
5
0
10
99
56
0
5
0
0
5
[a] Reaction conditions: 1a (0.25 mmol), 2a or 3a (0.25 mmol), tol-
uene (1 mL), irradiation with a 125 W high-pressured mercury
lamp, room temperature, 4 h. [b] After column chromatography.
Scheme 2. Syntheses of heterocycles 1aa–ag to be used as precur-
sors of N-benzoyl nitrene (8a).
Eur. J. Org. Chem. 2015, 2854–2860
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2855

V. Bizet, C. Bolm
FULL PAPER
The data presented in Table 1 show that 1,4,2-dioxazol- Table 2. Applications of 1aa–oa in the imidations of sulfide 2a and
sulfoxide 3a.
5-one 1aa was superior to the other heterocycles in generat-
ing an intermediate with relevance for the applied light-in-
duced ruthenium catalysis. Generally, sulfide 2a appeared
to be more reactive than sulfoxide 3a.
Scope and Reactivity of 1,4,2-Dioxazol-5-ones
The application of 1,4,2-dioxazol-5-ones 1ba–oa, having
substituents other than phenyl (as in 1aa) at the 3-position
of the heterocycle, were then studied (Table 2). Based on
our earlier findings in imidations of variously substituted
sulfur reagents,[9] sulfide 2a and sulfoxide 3a were selected
as representative model substrates. High efficiencies of the
photochemically-induced ruthenium-catalyzed imidations
were observed, with heterocycles 1aa and 1ba providing the
corresponding N-benzoyl and N-acetyl sulfimides (4aa and
4ba) and sulfoximines (5aa and 5ba) in yields up to 99%
(entries 1 and 2). Applying 1,4,2-dioxazol-5-one 1ca, having
a strong electron-withdrawing CF3 group at C3, gave less
satisfactory results (entry–3). Only sulfide 2a reacted in the
expected manner, providing 4ca in 57% yield. Mixing of
1ca with sulfoxide 3a led to deoxygenation of the sulfur
reagent and sulfide 2a was formed with strong efferves-
cence. Apparently, 1,4,2-dioxazol-5-one 1ca was highly re-
active towards traces of water or other nucleophiles, render-
ing further studies with this heterocycle unattractive. The
use of 3-ethyl- and 3-benzyl-substituted 1,4,2-dioxazol-5-
one derivatives 1da and 1ea in the imidations of sulfide 2a
led to sulfimides 4da and 4ea in 92 and 86% yield, respec-
tively, indicating that a moderate increase in size of the sub-
stituent at C3 of the 1,4,2-dioxazol-5-one had only a minor
effect on the reactivity towards the sulfide. This behavior
was in sharp contrast to that observed in imidations of sulf-
oxide 3a, which showed low conversions in reactions with
1da and 1ea, providing the corresponding products 5da and
5ea in only 30 and 7% yield, respectively.
The same trend was observed when 1,4,2-dioxazol-5-ones
1fa (with a 3-isopropyl) and 1ga (with a 3-tert-butyl group)
were treated with 2a and 3a (Table 2, entries 6 and 7). From
these four experiments, only a single product (4fa) could be
isolated, and the yield of this sulfimide remained moderate
(57%). No products were obtained from the other three re-
actions. Imidations of sulfide 2a with 1,4,2-dioxazol-5-ones
bearing 2-furyl (1ha) and 2-thiofuranyl substituents (1ia)
worked well, providing sulfimides 4ha and 4ia in 96 and
[a] Yield after column chromatography. [b] The 1,4,2-dioxazol-5-
one is poorly soluble in toluene.
64% yields, respectively (entries 8 and 9). The analogous had a positive effect on the imidation of sulfide 2a but ren-
reactions with sulfoxide 3a were less effective, resulting in dered the conversion of sulfoxide 3a more difficult. As a
yields of only 14% for 5ha and 12% for 5ia. Applying consequence, sulfimide 4la was obtained in 99% yield,
pentafluorophenyl-substituted 1,4,2-dioxazol-5-one 1ja in whereas sulfoximine 5la could only be isolated in 35% yield
the imidations of 2a and 3a led, to our delight, to the pre- (entry–12). This difference in reactivity between the sulfide
viously unreported N-pentafluorobenzoylated sulfimide 4ja and the sulfoxide was also revealed in attempts to imidate
and sulfoximine 5ja in excellent yields (entry–10). The for- 2a and 3a with ortho-methoxy-substituted dioxazolone
mation of products 4ka and 5ka starting from 1ka with a 1ma. In this case, only sulfide 2a reacted, and sulfoxide 3a
p-nitro-substituted aryl group also proceeded well (entry– remained untouched. Presumably, the steric effect of the or-
11). Changing the electron-withdrawing para-nitro group to tho substituent of 1ma was responsible for the low yield
a para-methoxy substituent and using 1la as nitrene source (40%) of sulfimide 4ma (entry–13). No reaction took place
2856
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 2854–2860

Sulfur Imidations by Nitrene Transfer Reactions
with either 2-hydroxyphenyl- or 4-pyridinyl-substituted taking 1aa as a representative example (Scheme 4). As de-
heterocycles 1na and 1oa with either sulfide 2a or sulfoxide scribed before (Table 1, entry–1), reactions with equimolar
3a (entries 14 and 15). Assuming that the lack of reactivity amounts of 1aa and sulfide 2a afforded sulfimide 4aa in
was due to the poor solubility of these two 1,4,2-dioxazol- 99% yield. Given that 4aa could be further oxidized to sulf-
5-ones in toluene, N,N-dimethylformamide (DMF) was oximine 5aa, it was hypothesized that sulfondiimide 11aa
added to the reaction mixture. As expected, the heterocycles might be formed in reactions of 2a with an excess of the
dissolved, but, unfortunately, no reactions occurred. We as- imidating agent (Scheme 4, a). However, this was not the
sume that the heteroatoms of the pyridinyl and the phenol case; even with 2 equiv. 1aa, sulfondiimide 11aa remained
substituents of 1na and 1oa interacted with the ruthenium undetected and the yield of sulfimide 4aa was unchanged
complex, resulting in catalyst inhibition.
(99%). Apparently, compared with the sulfide sulfur in 2a,
Reviewing and evaluating the data presented in Table 2 the mono-imidated sulfur in 4aa was not sufficiently
led to the conclusion that the light-induced ruthenium ca- nucleophilic to be imidated for a second time, rendering the
talysis for sulfur imidations was highly effective in providing overall process highly selective even in the presence of an
a wide range of N-acylated sulfimides and sulfoximines in excess of the imidating agent.
moderate to excellent yields. Particularly noteworthy are the
sulfide imidations because they lead to products that are
commonly difficult to prepare by alternative routes because
they require N-acylation reactions of rather unstable and
difficult to handle NH-sulfimides.[11]
The initially described one-pot sulfur imidation/oxid-
ation sequence (Scheme 1) allows the direct transformation
of sulfides into N-acyl sulfoximines under essentially neu-
tral conditions.[9] Both reaction steps are catalyzed by the
same ruthenium source. We then investigated how much the
catalyst loading could be reduced in up-scaled reactions;
the results are shown in Scheme 3. Under the original reac-
tion conditions involving the use of 1 mol-% Ru(TPP)CO
(12a) as catalyst on a 0.25 mmol scale, product 5ba was ob-
tained in 92% yield. To our delight, increasing the reaction
scale to 2.5 mmol allowed the amount of catalyst to be re-
Scheme 4. Comparative sulfur imidations.
duced to 0.2 and 0.1 mol-%, providing 5ba in even higher
The same trend was observed when the catalysis was per-
yields of 97 and 95%, respectively. Furthermore, an im- formed with a mixture of equimolar amounts of 1aa, sulfide
proved reactivity in the imidation step was observed, lead- 2a, and sulfoxide 3a (Scheme 4, b). In this case, sulfimide
ing to complete sulfide conversion after only 4–5 h. Al- 4aa was also the main product (81% yield), and the yield
though the subsequent oxidation was comparably slow of sulfoximine 5aa was only 5%.
(overnight), it is clear that this newly developed one-pot
The importance of using a well chosen N-acyl nitrene
procedure is superior to the existing methods towards N- source was demonstrated by a crossover experiment involv-
acyl sulfoximines. This is particularly true for sensitive com- ing 3-methyl-1,4,2-dioxazol-5-one (1ba), benzamide (9a),
pounds, which benefit from the neutral conditions of both and sulfide 2a. Applying these compounds in an equimolar
of the aforementioned oxidative transformations.
ratio led to the exclusive formation of sulfimide 4ba by reac-
tion of 1ba with 2a (Scheme 4, c). The lack of reactivity of
9a was confirmed in a separate experiment with 2a as sub-
strate. Again, none of the expected product 4aa was ob-
served, showing that for the newly developed sulfur imid-
ation, the use of compounds such as 1,4,2-dioxazol-5-one 1
was essential.
In our initial work we demonstrated that the combina-
tion of ruthenium and light was essential for achieving
sulfur imidations.[9] We now wondered whether the applied
ruthenium porphyrin [Ru(TPP)CO, 12a] could be replaced
by another photocatalyst (Figure 1). The first answers were
given by performing imidations with ruthenium complexes
having ligands with substituted aryl groups.[9,13] As for 12a
with phenyl substituents, the use of porphyrin 12b, having
tolyl groups, revealed an exceptionally high catalytic ac-
Scheme 3. Scale and required catalyst loadings.
Mechanistic Investigations
With the goal of gaining deeper mechanistic insights, tivity in the N-benzoyl nitrene transfer from 1aa onto 2a,
competition experiments were performed. First, the imidat- leading to 4aa in 99% yield. In contrast, mesityl-bearing
ing power of the 1,4,2-dioxazol-5-ones was addressed by ruthenium porphyrin 12c showed low reactivity, providing
Eur. J. Org. Chem. 2015, 2854–2860
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2857

V. Bizet, C. Bolm
FULL PAPER
sulfimide 4aa in only approximately 10%. Most likely the a significant but not necessarily decisive role in the newly
eight ortho-methyl substituents on the four arenes led to discovered imidation reaction.
steric hindrance at the metal core, hampering the formation
With the goal of identifying possible ruthenium-contain-
of relevant intermediates. Ruthenium porphyrin 12d, with ing intermediates such as rutheno N-acyl nitrenes,[16] high-
pentafluorophenyl substituents, also showed a low catalytic resolution mass spectrometry (MS) experiments were per-
activity, which we interpreted as a hint for the requirement formed. After mixing of 12a and 1ba in a 1:20 ratio under
of having a metal center with reasonable electron density. visible light irradiation, aliquots of the reaction mixture
When the latter was reduced by the presence of strongly were taken after 10, 30, and 60 min. The samples were then
electron-withdrawing substituents on the ligand, the cata- dissolved in acetonitrile and analyzed by MS spectrometry
lytic activity was diminished. When the ruthenium catalyst using the electrospray ionization mode.[17] After 10 min,
was replaced by the non-metallic photosensitizers Rose MS analysis (Figure 2) showed a predominance of the start-
Bengal (13a) and Rose Bengal lactone (13b), no reactions ing ruthenium complex B with a cluster peak at m/z
occurred.[14] This observation strengthened our hypothesis 744.13312 (m/z calcd. 744.14576). Another weaker signal at
that N-bound nitrene/ruthenium species were relevant for m/z 714.13564 (m/z calcd. 714.13520) indicated the forma-
the success of the imidation process.[15,16]
tion of a ruthenium complex A formed from B by loss of
CO. Of main significance was the cluster peak at m/z
796.18869 (m/z calcd. 796.16208), which we attributed to a
rutheno N-acyl nitrene complex (with an associated Na+)
C, formed from complex A and 1ba upon loss of CO2. After
30 and 60 min, similar cluster peaks were obtained along
with other signals above m/z 1400, probably stemming from
fragmentations and associations of the aforementioned spe-
cies in the spectrometer. Although no rutheno bis(N-acyl
nitrene) complex was detected in this ESI-MS experiment,
the formation of such species cannot be ruled out.[17] An
attempt to apply phenyl-substituted 1,4,2-dioxazol-5-one
1aa (instead of 1ba) in an analogous MS study was unsuc-
cessful. Presumably due to a rapid Curtius-type rearrange-
ment of the intermediately formed phenylacyl nitrene, no
related rutheno species could be detected.
Figure 1. Ruthenium prophyrins and non-metallic photocatalysts
applied in the imidation process.
Commonly, a high-pressure mercury lamp (200–600 nm)
was used for the photochemical activation of the catalyst.
Given that the maximum absorption (λmax) of 12a was de-
termined to be at 410 nm, it was envisaged that the effi-
ciency of the catalysis could be increased by performing the
reaction in a photochemical reactor with a fixed wavelength
at 400 nm. However, this attempt proved unfruitful. First,
the reactivity was essentially unchanged, confirming our as-
sumption that the reaction was induced by visible light. Sec-
ond, the temperature increase in the photochemical reactor
led to unproductive side reactions (such as Curtius re-
arrangements), rendering this approach synthetically less
attractive.
Hypothesizing that the light-induced decarboxylation of
1 involved electron transfer processes, which could be affec-
ted by good electron acceptors, the imidation of 3a with
1aa was performed in the presence of 20 mol-% p-dinitro-
benzene. As a result, the yield of 5aa decreased to only
23%. Using 20 mol-% terephthalonitrile as electron ac-
Figure 2. Structures of proposed ruthenium intermediates observed
in MS experiments.
Based on the detection of N-acyl nitrene-bound ruth-
ceptor led to the formation of 5aa in 29% yield. Almost the enium complex C, the following reaction mechanism for the
same result (22% yield of 5aa) was observed when the sulfur imidation can be proposed (Scheme 5). Upon light
radical
scavenger
2,2,6,6-tetramethylpiperidine-1-oxyl irradiation, ruthenium porphyrin 12a loses CO, leading to
(TEMPO; 1 equiv.) was added to the reaction mixture. For- a new complex I (which was observed as complex A by MS;
mation of a TEMPO adduct was not observed. Taking the Figure 2) having an open coordination site. Through one
results together, we regard the decreased yields in these ex- of the heteroatoms, 1,4,2-dioxazol-5-one 1 interacts with I,
periments as evidence for electron transfer events playing forming ruthenium complex II.
2858
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 2854–2860

Sulfur Imidations by Nitrene Transfer Reactions
Experimental Section
General Procedure for the Sulfur Imidation: To a solution of sulfide
2
or sulfoxide 3 (0.25 mmol) and 1,4,2-dioxazol-5-one 1
(0.25 mmol) in anhydrous toluene (1 mL) under argon was added
Ru(TPP)CO (1.9 mg, 0.0025 mmol). The reaction mixture was irra-
diated with a 125 W high-pressure mercury lamp at room tempera-
ture until full conversion was reached (as analyzed by TLC). The
mixture was concentrated under reduced pressure and purified by
column chromatography over silica gel (pentane/ethyl acetate) to
give the corresponding N-acyl sulfimide 4 or sulfoximine 5.
General Procedure for the “One-Pot” Sulfur Imidation/Oxidation:
Upon completion of the imidation step, the mixture was concen-
trated under reduced pressure to remove toluene and was then dis-
solved in dichloromethane (2.5 mL). A solution of sodium per-
iodate (107 mg, 0.5 mmol) in water (1.25 mL) was added and vigor-
ous stirring of the reaction mixture was performed by using a mag-
netic cross-shaped stir bar. After full conversion (one night at room
temperature), the mixture was extracted with dichloromethane,
dried with sodium sulfate, and concentrated under reduced pres-
sure. The product was then purified by column chromatography
over silica gel (pentane/EtOAc) to give the corresponding N-acyl
sulfoximine 5.
Acknowledgments
V. B. acknowledges the Alexander von Humboldt foundation for a
postdoctoral fellowship. The authors thank Dr. Wolfgang Bettray
(RWTH Aachen University) for the ESI-MS measurements, techni-
cal advice, and helpful discussions.
Scheme 5. Proposed mechanism for the light-induced ruthenium-
catalyzed sulfur imidation/oxidation reactions.
[1] For reviews on sulfimides, see: a) T. L. Gilchrist, C. J. Moody,
Chem. Rev. 1977, 77, 409–435; b) N. Furukawa, S. Oae, Ind.
Eng. Chem. Prod. Res. Dev. 1981, 20, 260–270; c) I. V. Koval,
Sulfur Rep. 1993, 14, 149; d) P. C. Taylor, Sulfur Rep. 1999, 21,
241–280.
[2] For reviews related to sulfoximine chemistry, see: a) C. R. John-
son, Acc. Chem. Res. 1973, 6, 341–347; b) M. Reggelin, C. Zur,
Synthesis 2000, 1–64; c) M. Harmata, Chemtracts 2003, 16,
660–666; d) H. Okamura, C. Bolm, Chem. Lett. 2004, 33, 482–
487; e) C. Worch, A. C. Mayer, C. Bolm, in: Organosulfur
Chemistry in Asymmetric Synthesis (Eds.: T. Toru, C. Bolm),
Wiley-VCH, Weinheim, Germany, 2008, pp 209–232; f) R.
Bentley, Chem. Soc. Rev. 2005, 34, 609–624; g) H.-J. Gais,
Heteroat. Chem. 2007, 18, 472–481; h) V. Bizet, R. Kowalczyk,
C. Bolm, Chem. Soc. Rev. 2014, 43, 2426–2438; i) X. Shen, J.
Hu, Eur. J. Org. Chem. 2014, 4437–4451.
Loss of CO2 provides rutheno N-acyl nitrene complex III
(which was observed as Na+ adduct C by MS; Figure 2).
Interaction of III with the sulfur atom of sulfide 2 or sulfox-
imine 3 allows a N-acyl nitrene transfer reaction (via a com-
plex such as IV), providing the imidated products 4 or 5,
respectively. Concomitantly, ruthenium complex I is regen-
erated, starting a new catalytic cycle. The subsequent
sulfimide oxidation can then either involve an oxo-ruth-
enium complex such as V[18] and proceed via VI, or it may
be catalyzed by a new ruthenium species generated by de-
gradation of 12a upon reaction with the oxidant (NaIO4).
[3] For a recent example demonstrating the use of sulfoximines in
asymmetric catalysis, see: M. Frings, I. Thomé, I. Schiffers, C.
Bolm, Chem. Eur. J. 2014, 20, 1691–1700.
Conclusions
[4] For selected recent examples, see: a) D. P. Walker, M. P. Zawis-
toski, M. A. McGlynn, J. C. Li, D. W. Kung, P. C. Bonnette,
A. Baumann, L. Buckbinder, J. A. Houser, J. Boer, A. Mistry,
S. Han, L. Xing, A. Guzman-Perez, Bioorg. Med. Chem. Lett.
2009, 19, 3253–3258; b) S. J. Park, H. Baars, S. Mersmann, H.
Buschmann, J. M. Baron, P. M. Amann, K. Czaja, H. Hollert,
K. Bluhm, R. Redelstein, C. Bolm, ChemMedChem 2013, 8,
217–220; c) U. Lücking, Angew. Chem. Int. Ed. 2013, 52, 9399–
9408; Angew. Chem. 2013, 125, 9570–9580; d) F. W. Goldberg,
J. G. Kettle, J. Xiong, D. Lin, Tetrahedron 2014, 70, 6613–6622.
[5] For selected recent examples, see: a) H. Yu, Z. Qin, H. Dai, X.
Zhang, X. Qin, T. Wang, J. Fang, J. Agric. Food Chem. 2008,
56, 11356–11360; b) Y. Zhu, M. R. Loso, G. B. Watson, T. C.
Sparks, R. B. Rogers, J. X. Huang, B. C. Gerwick, J. M. Bab-
cock, D. Kelley, V. B. Hegde, B. M. Nugent, J. M. Renga, I.
Denholm, K. Gorman, G. J. DeBoer, J. Hasler, T. Meade, J. D.
We have studied various N-acyl nitrene precursors in
light-induced ruthenium-catalyzed imidations of sulfides
and sulfoxides and found that 3-substituted 1,4,2-dioxazol-
5-ones 1 were the most reactive species. In this manner, a
wide range of N-substituted sulfimides and sulfoximines
has been prepared. Crossover experiments revealed an ac-
tivity grading, and potentially catalytic intermediates were
identified by ESI-MS. The evaluation of all reaction details
allowed a mechanistic scheme summarizing the observed
proceedings to be proposed. Further studies regarding the
use of 1 as N-acyl nitrene precursors in reactions with other
nucleophiles are ongoing in our laboratories, and the results
will be published in due course.
Eur. J. Org. Chem. 2015, 2854–2860
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2859

V. Bizet, C. Bolm
FULL PAPER
Thomas, J. Agric. Food Chem. 2011, 59, 2950–2957; c) C.
Gnamm, A. Jeanguenat, A. C. Dutton, C. Grimm, D. P. Kloer,
A. J. Crossthwaite, Bioorg. Med. Chem. Lett. 2012, 22, 3800–
3806; d) T. C. Sparks, G. B. Watson, M. R. Loso, C. Geng,
J. M. Babcock, J. D. Thomas, Pest. Biochem. Physiol. 2013, 107,
1–7.
[12] For a fundamentally different approach towards N-substituted
sulfimides starting from sulfoxides, see: a) A. K. Sharma, T.
Ku, A. D. Dawson, D. Swern, J. Org. Chem. 1975, 40, 2758–
2764; b) S. L. Huang, D. Swern, J. Org. Chem. 1978, 43, 4537–
4538; c) C. Urban, Y. Macé, F. Cadoret, J.-C. Blazejewski, E.
Magnier, Adv. Synth. Catal. 2010, 352, 2805–2814; d) S.
Raghavan, S. Mustafa, K. Rathore, Tetrahedron Lett. 2008, 49,
4256–4259; e) C. M. M. Hendriks, P. Lamers, J. Engel, C.
Bolm, Adv. Synth. Catal. 2013, 355, 3363–3368.
[6]
For selected examples, see: a) T. Bach, C. Korber, Tetrahedron
Lett. 1998, 39, 5015–5016; b) T. Bach, C. Körber, Eur. J. Org.
Chem. 1999, 1033–1039; c) W. Ou, Z.-C. Chen, Synth. Com-
mun. 1999, 29, 4443–4449; d) A. L. Marzinzik, K. B. Sharpless,
J. Org. Chem. 2001, 66, 594–596; e) G. Y. Cho, C. Bolm, Org.
Lett. 2005, 7, 4983–4985; f) G. Y. Cho, C. Bolm, Tetrahedron
Lett. 2005, 46, 8007–8008; g) O. García Mancheño, C. Bolm,
Org. Lett. 2006, 8, 2349–2352; h) O. García Mancheño, C.
Bolm, Chem. Eur. J. 2007, 13, 6674–6681; i) O. García Man-
cheño, O. Bistri, C. Bolm, Org. Lett. 2007, 9, 3809–3811; j) O.
García Mancheño, J. Dallimore, A. Plant, C. Bolm, Org. Lett.
2009, 11, 2429–2432; k) O. García Mancheño, J. Dallimore, A.
Plant, C. Bolm, Adv. Synth. Catal. 2010, 352, 309–316; l) A.
Pandey, C. Bolm, Synthesis 2010, 2922–2925; m) J. Miao,
N. G. J. Richards, H. Ge, Chem. Commun. 2014, 50, 9687–
9689.
a) H. Kise, G. F. Whitfield, D. Swern, J. Org. Chem. 1972, 37,
1121–1125; b) G. F. Koser, P. B. Kokil, M. Shah, Tetrahedron
Lett. 1987, 28, 5431–5434; c) C. S. Tomooka, D. D. LeCloux,
H. Sasaki, E. M. Carreira, Org. Lett. 1999, 1, 149–152; d) C. S.
Tomooka, E. M. Carreira, Helv. Chim. Acta 2002, 85, 3773–
3784; e) H. Okamura, C. Bolm, Org. Lett. 2004, 6, 1305–1307;
f) J. Gries, J. Krüger, Synlett 2014, 25, 1831–1834.
[13] Abbreviations used in the ligand nomenclature: TPP = tetra-
phenylporphyrinato, TTP = tetra-(p-tolyl)porphyrinato, TMP
=
tetramesitylporphyrinato, F20TPP
= tetra(pentafluoro-
phenyl)porphyrinato.
[14] For a recent example of a Rose Bengal-catalyzed decarboxyl-
ation, see: L. Chen, C. S. Chao, Y. Pan, S. Dong, Y. C. Teo, J.
Wang, C.-H. Tan, Org. Biomol. Chem. 2013, 11, 5922–5925.
[15] For mechanistically related transformations starting from 3-
aryl-1,4,2-dioxazol-5-ones providing oxazoles and oxazolines
from alkynes and alkenes, respectively, by ruthenium catalysis,
see: C. L. Zhong, B. Y. Tang, P. Yin, Y. Chen, L. He, J. Org.
Chem. 2012, 77, 4271–4277.
[16] For recent examples describing the formation of ruthenium
imido intermediates starting from ruthenium porphyrin com-
plexes, see: a) S. Fantauzzi, E. Gallo, A. Caselli, F. Ragaini, N.
Casati, P. Macchi, S. Cenini, Chem. Commun. 2009, 3952–3954;
b) D. Intrieri, A. Caselli, F. Ragaini, P. Macchi, N. Casati, E.
Gallo, Eur. J. Inorg. Chem. 2012, 569–580; c) W. Xiao, C.-Y.
Zhou, C.-M. Che, Chem. Commun. 2012, 48, 5871–5873; d) G.
Manca, E. Gallo, D. Intrieri, C. Mealli, ACS Catal. 2014, 4,
823–832; e) J. Wei, W. Xiao, C.-Y. Zhou, C.-M. Che, Chem.
Commun. 2014, 50, 3373–3376; f) K.-H. Chan, X. Guan, V. K.-
Y. Lo, C.-M. Che, Angew. Chem. Int. Ed. 2014, 53, 2982–2987.
[17] For MS spectra and related assignments, see the Supporting
Information.
[18] For selected examples describing the use of oxo-ruthenium por-
phyrin complexes in oxygen transfer reactions, see: a) C.-M.
Che, J.-L. Zhang, R. Zhang, J.-S. Huang, T.-S. Lai, W.-M. Tsui,
X.-G. Zhou, Z.-Y. Zhou, N. Zhu, C. K. Chang, Chem. Eur.
J. 2005, 11, 7040–7053; b) E. Vanover, Y. Huang, L. Xu, M.
Newcomb, R. Zhang, Org. Lett. 2010, 12, 2246–2249; c) S. Fu-
nyu, M. Kinai, D. Masui, S. Takagi, T. Shimada, H. Ta-
chibana, H. Inoue, Photochem. Photobiol. Sci. 2010, 9, 931–
936; d) A. Ishikawa, S. Sakaki, J. Phys. Chem. A 2011, 115,
4774–4785.
[7]
[8]
J. Sauer, K. K. Mayer, Tetrahedron Lett. 1968, 9, 319–324.
[9] V. Bizet, L. Buglioni, C. Bolm, Angew. Chem. Int. Ed. 2014,
53, 5639–5642.
[10] L. Buglioni, V. Bizet, C. Bolm, Adv. Synth. Catal. 2014, 356,
2209–2213.
[11] For additional synthetic methods towards sulfimides with vari-
ous N-substituents, see: a) H. Fujita, T. Uchida, R. Irie, T.
Katsuki, Chem. Lett. 2007, 36, 1092–1093; b) M. Ochiai, M.
Naito, K. Miyamoto, S. Hayashi, W. Nakanishi, Chem. Eur. J.
2010, 16, 8713–8718; c) Y. Liu, C.-M. Che, Chem. Eur. J. 2010,
16, 10494–10501; d) C. C. Farwell, J. A. McIntosh, T. K. Hys-
ter, Z. J. Wang, F. H. Arnold, J. Am. Chem. Soc. 2014, 136,
8766–8771; e) H. Lebel, H. Piras, J. Bartholoméüs, Angew.
Chem. Int. Ed. 2014, 53, 7300–7304; Angew. Chem. 2014, 126,
7428–7432; f) T. Uchida, T. Katsuki, Chem. Rec. 2014, 14, 117–
129.
Received: February 12, 2015
Published Online: March 17, 2015
2860
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 2854–2860