Sulfur-controlled and rhodium-catalyzed formal (3 + 3) transannulation of thioacyl carbenes with alk-2-enals and mechanistic insights

Article abstract of DOI:10.1039/d1ob00116g

A rhodium-catalyzed denitrogenative formal (3 + 3) transannulation of 1,2,3-thiadiazoles with alk-2-enals is achieved, producing 2,3-dihydrothiopyran-4-ones in moderate to excellent yields. An inverse KIE of 0.49 is obtained, suggesting the reversibility of the oxidative addition of thioacyl Rh(i) carbenes to alk-2-enals. The late-stage structural modifications of steroid compounds are realized. Moreover, our studies show that thioacyl carbenes have different reactivities to those of α-oxo and α-imino carbenes, and highlight the importance of heteroatoms in deciding the reactivities of heterovinyl carbenes.

Full text of DOI:10.1039/d1ob00116g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Sulfur-controlled and rhodium-catalyzed formal
(3 + 3) transannulation of thioacyl carbenes with
alk-2-enals and mechanistic insights†
Cite this: Org. Biomol. Chem., 2021,
19, 3173
Qiuyue Wu, Ziyang Dong,
Jiaxi Xu * and Zhanhui Yang
*
A rhodium-catalyzed denitrogenative formal (3 + 3) transannulation of 1,2,3-thiadiazoles with alk-2-enals
is achieved, producing 2,3-dihydrothiopyran-4-ones in moderate to excellent yields. An inverse KIE of
0.49 is obtained, suggesting the reversibility of the oxidative addition of thioacyl Rh(I) carbenes to alk-2-
enals. The late-stage structural modifications of steroid compounds are realized. Moreover, our studies
show that thioacyl carbenes have different reactivities to those of α-oxo and α-imino carbenes, and high-
light the importance of heteroatoms in deciding the reactivities of heterovinyl carbenes.
Received 22nd January 2021,
Accepted 15th March 2021
DOI: 10.1039/d1ob00116g
rsc.li/obc
and coworkers in 1985,8 with a Cu(II) catalyst, the α-oxo car-
benes generated from α-diazoketones underwent (3 + 2) trans-
Introduction
The transannulations of heterovinyl transition-metal-carbenes annulation with the CvO bonds of alk-2-enals to produce 1,3-
with unsaturated compounds represent some of the most dioxoles (Scheme 1a). In 2013, a novel reaction of alk-2-enals
important approaches to heterocycles. Transition metal cata- and α-imino carbenes derived from N-sulfonyl-1,2,3-triazoles
lyzed transannulations of α-oxo and α-imino carbenes have was developed by Miura and Murakami’s group to stereoselec-
provided various elegant constructions of structurally diverse tively synthesize trans-2,3-disubstituted 2,3-dihydropyrroles
oxa- and azacycles, respectively.1,2 Since 2016, Gevorgyan’s,3 (Scheme 1b).9 Mechanistically, the (3 + 2) transannulation of
Lee’s,4 and Nakajima and Nishibayashi’s5 groups demon- α-imino carbenes with the formyl group of alk-2-enals and a
strated that thioacyl carbenes (also referred to as thiavinyl car- subsequent ionic rearrangement of the resultant 4-oxazoline
benes) were able to transannulate with alkynes, alkenes, intermediates were involved. In these two cases, these two
nitriles, and phosphaalkynes to give five-membered thiacycles types of products were virtually decided by the electrophilic
in the presence of a rhodium catalyst. It must be mentioned nature of acyl carbenes and α-imino carbenes. In other words,
that it is Gevorgyan’s group that made seminal breakthroughs the different electronegativity and nucleophilicity of hetero-
in generating α-imino and thioacyl carbenes and disclosing atoms in the heterovinyl metal–carbenes led to different
their reactivities.3a,6 Nowadays, one of the central and impor- reactivities.
tant research areas in the thioacyl carbene field is to exploit
On the basis of the reports mentioned above and from our
new transannulative partners and explore new and unique background of sulfur chemistry10 and carbene chemistry,11 we
reactivities of thioacyl carbenes.
envisioned that replacement of the respective oxygen atom or
Alk-2-enals are versatile building blocks in the synthesis of
heterocycles and natural products.7 Initially we envisioned that
they might be potential transannulative partners of thioacyl
carbenes. A literature survey revealed that the transition metal-
catalyzed transannulations of alk-2-enals with α-oxo and
α-imino carbenes had been exploited. As reported by Alonso
Department of Organic Chemistry, College of Chemistry, Beijing University of Chemical
Technology, Beijing 100029, P. R. China. E-mail: zhyang@mail.buct.edu.cn,
jxxu@mail.buct.edu.cn; Fax: +86 10 64435565
†Electronic supplementary information (ESI) available: Detailed experimental
procedures, analytical data and copies of NMR (1H, 19F, and 13C) spectra of pro-
ducts 1–6, 12 and 13; copies of 1H NMR spectra of crude reaction mixtures.
CCDC 1919688. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/d1ob00116g
Scheme 1 Reactions of α-oxo, α-imino, and α-thioxo metal carbenes
with alk-2-enals.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 3173–3180 | 3173

View Article Online
Paper
Organic & Biomolecular Chemistry
NSO2R group of the aforementioned two classes of carbenes (entries 12 and 13), only gave a trace amount of product 3aa.
with the sulfur atom might enable a new transannulation Using toluene as a solvent gave a moderate 54% yield (entry
mode in the reaction of alk-2-enals. By harnessing the stronger 14). Reactions at lower temperature gave poor results (entries
nucleophilicity of the sulfur atom in the thioacyl group, we, 15 and 16). Even under the optimal conditions (entry 1), 1a
herein, realize a formal (3 + 3) transannulation of thioacyl Rh– was not completely consumed, and some unidentified mix-
carbenes with alk-2-enals, giving 2,3-dihydrothiopyran-4-ones tures, as shown by 1H NMR, were also isolated.
as products (Scheme 1c). Our studies highlight the importance
of heteroatoms in heterovinyl carbenes in deciding their reac-
Substrate scope and substituent effect
tivity, and provide a novel method to construct 2,3-dihy-
drothiopyran-4-ones,12 which are important intermediates in
In our previous work on the (3 + 3) transannulation of thioacyl
carbenes with alk-2-ynals,14 some examples of the (3 + 3) trans-
organic synthesis and the preparation of high electrical con-
ductivity materials, and also show some biological activities.13
annulation of thioacyl carbenes with alk-2-enals were divulged.
However, alk-2-enals 2 possess more substituents (R1, R2, R3 in
Table 2) at the α- and β-positions. Thus, it is necessary to
further disclose the details of the transannulations of alken-2-
als with thioacyl carbenes. The substrate scope and substituent
effect of alk-2-enals on the (3 + 3) transannulation have been
studied and the results are presented in Table 2.
Results and discussion
Optimization of conditions
The optimization of reaction conditions was performed by
using 1,2,3-thiadiazole 1a as the α-thioxo carbene precursor
and 3-methylcrotonaldehyde (2a) as a transannulation partner
(Table 1, for more details see Table S1 in the ESI†). Under stan-
dard conditions, the desired (3 + 3) product 3aa was isolated
in 72% yield (entry 1). No reaction occurred without the ligand
DPPF (entry 2). The use of [RhCp*Cl2]2 led to a slightly lower
63% yield (entry 3). Changing the selected diphosphine
ligands did not afford as good results as using DPPF (entries
4–8). Monodentate ligands, whether electron-deficient triaryl-
phosphines (entries 9–11) or electron-rich phosphinates
Table 2 Scope of alk-2-enals
Table 1 Optimization of reaction conditionsa
Entry
Cat.
Ligand (mol%)
Yieldb (%)
1
2
3
4
5
6
7
8
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[RhCp*Cl2]2
DPPF (12)
—
72
Trace
63
Trace
6
8
DPPF (12)
DPPM (12)
DPPE (12)
DPPP (12)
DPPB (12)
DPPPenta (12)
PPh3 (24)
P(4-CF3C6H4)3 (24)
P(C6F5)3 (24)
P(OEt)3 (24)
P(OPh)3 (24)
DPPF (12)
DPPF (12)
DPPF (12)
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
[Rh(COD)Cl]2
15
3
9
Trace
Trace
Trace
Trace
Trace
54
10
11
12
13
14c
15d
16e
3
Trace
a Reactions were performed using 1a (0.25 mmol) and 2a (0.50 mmol)
under the indicated conditions. b Yield obtained by column chromato- a Together with 21% yield of 4-nitrostyrene and 24% recovery of
graphy. c Toluene as the solvent. d Reaction temperature 100 °C. 4-nitrocinnamaldehyde (2k). b The dr could not be determined by 1H
e Reaction temperature 60 °C.
NMR. c 10 mol% [Rh(COD)Cl]2, 24 mol% DPPF, 12 h.
3174 | Org. Biomol. Chem., 2021, 19, 3173–3180
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 3 Scope of 1,2,3-thiadiazoles
We first investigated the β-substituent effect of (E)-alk-2-
enals 2c–2q by reacting them with thiadiazole 1a under the
optimal conditions (Table 2a). Compared with the reaction of
acrylaldehyde (R1 = R2 = R3 = H, 2b), the presence of a
β-substituent (R2) greatly facilitated the transannulations, and
the product yields were improved to 39–80% (3ac–3ap). The
electronic properties of β-aryls were relevant to the yields.
Strong electron acceptors such as the nitro group on the aryl
group gave 3ak in a lower yield of 39%. The reaction of 3-
(naphthalen-2-yl)acrylaldehyde (2l) and thiadiazole 1a furn-
ished 3al in 69% yield, showing their excellent reactivity
toward (3 + 3) transannulation. The (3 + 3) transannulation
was promoted by the β-alk-1-enyl groups of alk-2-enals 2m and
2n, regardless of their steric hindrance, and the desired pro-
ducts 3am and 3an were isolated in 74% and 80% yields,
respectively. Also investigated was the β-alkyl effect of alk-2-
enals 2o and 2p. The branched secondary alkyls (2o and 2p)
and long-chain branched primary alkyl (2q) did not
suppress the transannulations, with the desired products
obtained in 59–72% yields. Notably, when a steric center was
present in alk-2-enals 2, two diastereomers were generated
(3ap and 3aq) because a new stereocenter (C6 position) was
forged in 2,3-dihydrothiopyran-4-ones when the transannula-
tion occurred.
a 0.25 mmol of 1, 0.5 mmol of 2a, and 1 mL PhCl were used.
b 0.1 mmol of 1, 0.3 mmol of 2a, 10 mol% [Rh(COD)Cl]2, 24 mol%
DPPF, 20 mol% AgBF4, and 0.5 mL PhCl were used. c Reaction for 12 h.
d Reaction for 10 h.
Alk-2-enals 2r–2ab with two β-substituents (R2 and R3) were
also tested (Table 2b). The reactions of β,β-dimethyl (2a,
Table 1), β-methyl-β-(4-methyl)pent-3-enyl (2r), β-methyl-
β-phenyl (2s), and β-cyclopropyl-β-methyl enals (2t) with 1a
took place readily, and the desired products 3aa, 3ar, 3as, and
3at were isolated in 70–86% yields. In Lee’s report, alkenes
readily reacted with 1,2,3-thiadiazole 1a under similar con-
ditions. However, in our work, the trisubstituted (2q and 2r)
and disubstituted alkene (2m and 2y) moieties remained
intact. In addition, the ring-opening of cyclopropane moieties
was frequently involved in Rh(I)-catalyzed carbocycle synth-
eses.15 However, they were well tolerated under our conditions.
These cases indicated the priority of alk-2-enals over alkenes
and cyclopropanes when they reacted with thioacyl Rh–car-
benes. The reactions of β,β-(1,n-alkylidene)enals or their ana-
logues 2u–2ab, namely, R2 and R3 were connected together,
gave spiro products 3au–3aab in moderate to good yields. The
ring size of these cycloalkylideneacetaldehydes and their ana-
logues, together with other factors, if any, largely affected the
transannulations. The four-, five-, and six-membered enals 2u–
2y transannulated well to give the desired spiro products 3au–
3ay in 64–85% yields, while those with seven-, twelve-, and
fifteen-membered rings only gave the corresponding spiro pro-
ducts 3az–3aab in moderate 42–48% yields.
the optimal conditions (Table 3). A variety of ethyl 5-aryl-1,2,3-
thiadiazole-4-carboxylates 1 were able to undergo the trans-
annulation to give the desired 2,3-dihydrothiopyran-4-ones 3
in 46–78% yields of the isolated products. The well-tolerated
functional groups include but are not limited to methyl-
enedioxy, halogen atoms, cyano, fur-2-yl, and thiophen-2-yl.
However, the reaction of 5-(fur-2-yl) thiadiazole (1k) gave a
relatively low yield of 46%. Lee and coworkers have demon-
strated that aryl nitriles were viable substrates to undergo (3 +
2) transannulation with thioacyl Rh(I)–carbenes derived from
1,2,3-thiadiazoles. However, in the case of 1j, the cyano group
survived.
In the above reports, most of the β-monosubstituted alke-
nals 2b–2p possessed the E-configuration. We studied the con-
figuration effect of the alkenyl moiety on the (3 + 3) transannu-
lation. With 2 equivalents of (E)-2g or (Z)-2g, the same product
3ag was obtained in 85% or 86% yield, respectively (Scheme 2a
and b). With 1 equivalent of (E)-2g and 1 equivalent of (Z)-2g,
3ag was also obtained in 86% yield (Scheme 2c). These results
show that the (Z)- or (E)-configurations of alken-2-als do not
affect their reactivity toward the formal (3
+
3)
transannulations.
The α-monosubstituted and α,β-disubstituted alkenals
2ac–2ae showed low activity toward the (3 + 3) transannulation
with thioacyl Rh–carbenes, and the desired products 3aac–
3aae were isolated in only 28–50% yields (Table 2c). These
examples demonstrated that the presence of an α-substituent
of alkenals 2 disfavored the desired (3 + 3) transannulation.
On the other hand, the scope of 1,2,3-thiadiazoles 1 was
also screened by reacting with 3-methylbut-2-enal (2a) under Scheme 2 Effect of the configuration of alkenals 2g.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 3173–3180 | 3175

View Article Online
Paper
Organic & Biomolecular Chemistry
Mechanistic studies and proposal
hydride as the turnover-limiting step. In the mechanism, a
primary kinetic isotope effect (KIE) with kH/kD = 3.07 was
observed. Since alk-2-enals 2 are structurally different from
alk-2-ynals, does the present transannulation of alk-2-enals
follow the same mechanism as that of alk-2-ynals?
We sought to probe the key intermediates in the present
formal (3 + 3) transannulation of alk-2-enals 2. Although thio-
acyl Rh–carbenes were proposed as intermediates in
Gevorgyan’s, Lee’s, and Nakajima and Nishibayashi’s work to
react with alkynes, alkenes, nitriles, and phosphaalkynes,3–5
they were never probed, isolated or characterized. Initially, we
attempted to isolate the thioacyl Rh–carbenes by directly
mixing thiadiazole 1a with [Rh(COD)Cl]2 and different biden-
tate phosphine ligands. However, none of the trials prevailed.
For example, heating 1a, [Rh(COD)Cl]2, and DPPF at 130 °C for
6 h caused the decomposition of thiadiazole 1a, and the
carbene dimer 4 and bisphosphine sulfide 5 were isolated in
46% and 7% yields, respectively, together with the recovery of
1a in 21% yield (Scheme 3a). Apparently, compound 4 was gen-
erated from dimerization of the thioacyl Rh–carbene, with the
desulfurization occurring before or after the dimerization. The
lost sulfur was partially accepted by the phosphine ligand. In
other words, the presence of the carbene dimer 4 was indirect
evidence for the existence of thioacyl carbenes. Recently, Bao
and coworkers proposed that the denitrogenation of 1,2,3-thia-
diazoles promoted by the Rh(I) catalyst might not afford the
commonly proposed α-thiavinyl Rh-carbenoid intermediates,
based on their DFT calculation results on the Rh(I)-catalyzed
transannulation of 1,2,3-thiadiazoles with alkenes, alkynes,
and nitriles.16 However, our experimental result (Scheme 3a)
suggested the carbene intermediates. Additionally, alk-2-enal
2l was also subjected to the optimal conditions, and decarbo-
nylation product 6 was isolated in 91% yield, together with 5%
recovery of 2l (Scheme 3b). It was evidenced that the acyl
rhodium hydride was involved as the key intermediate in the
decarbonylation of alkenals.17 The above two intermediate
probing experiments suggested that thioacyl Rh–carbenes or
acyl rhodium hydrides were probably involved in the dehydro-
genative transannulation of alk-2-enals 2 and thiadiazoles 1.
In our previous work on the Rh-catalyzed transannulation
between alk-2-ynals and 1,2,3-thiadiazoles 1,14 we proposed a
1,1-hydroacylation and subsequent 6-endo-dig cyclization
mechanism involving the Rh-catalyzed 1,1-hydroacylation of
thioacyl carbenes with the formation of the acyl rhodium
To answer this question, we first performed an isotope
tracing experiment by reacting thiadiazole 1a with cinnamalde-
hyde-d (2g-d) (Scheme 4a). The desired product 3ag-dh was iso-
lated in 55% yield, and the deuterium atom of 2g-d was par-
tially transferred to the 5-position of 3ag-dh, and a 4 : 1 H/D
1
ratio was determined by the H NMR analysis of 3ag-dh. The
low deuterium incorporation was attributed to the highly
acidic C–D bonds of intermediate 3ag-d, and probably resulted
from D–H exchange of 3ag-d with environmental water during
purification or the NMR test. This result also implied that it
was inappropriate to use the intermolecular competition
experiment with equimolar 2g-d and 2g in one pot to deter-
mine the KIE of the transannulation. Thus, we performed the
KIE studies using the initial rate method with two parallel
experiments of 2g-d and 2g (Scheme 4b). To our surprise, an
inverse primary KIE of kH/kD = 0.49 was obtained, in sharp con-
trast with our previous KIE studies (kH/kD = 3.07) on the trans-
annulation of alk-2-ynals and 1,2,3-thiadiazoles 1. Therefore,
the current transannulation mechanism appeared to be
different from that for the alk-2-ynal transannulation.
Based on the intermediate probing experiments, two plaus-
ible mechanistic pathways are proposed (Scheme 5). The
ligand exchange between DPPF and [Rh(COD)Cl]2 affords an
Scheme 4 (a) Isotope tracing experiment; (b) KIE studies using the
Scheme 3 Key intermediate probing experiments.
3176 | Org. Biomol. Chem., 2021, 19, 3173–3180
initial rate method.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 5 Proposed mechanism involving thioacyl Rh–carbenes.
active catalyst [Rh] (eqn (1)). In pathway i, the catalyst [Rh] observed by numerous groups. In our previous work associated
undergoes an oxidative addition to form acyl rhodium with hydroacylation of carbenes with aldehydes,14 a primary
hydrides A (step a). Subsequently, denitrogenative carbene for- KIE of 3.07 was observed. In these reports, the irreversible oxi-
mation between thiadiazoles 1 (possibly via their diazo form dative addition of Rh(I) to the formyl C–H bond (step a), which
1′) and A occurs to give carbene intermediates B (step b). leads to the formation of acyl rhodium hydrides A, is proposed
Alternatively, in pathway ii, the first decomposition of thiadia- as the rate-determining step. In addition, KIEs of 1.4–1.7 were
zoles 1 gives thioacyl carbenes F (step e). Because of the rela- also reported for other hydroacylation reactions, indicating
tively high nucleophilicity of the sulfur atom of the thioacyl that migratory insertion (similar to step c) or reductive elimin-
group, we suggest that the sulfur coordinates with the center ation (similar to step d) might act as the rate-determining
Rh(I) metal to render stabilized carbenes G (step g). Oxidative step.22 In light of the above reports, if the denitrogenative
addition of F to alk-2-enals 2 also furnishes B (step f). The for- transannulation of 1,2,3-thiodiazoles 1 and alk-2-enals 2
mation of oxonium ylides from F and aldehydes is not con- follows pathway i, a KIE of around 2.4–3.07 or 1.4–1.7 might
sidered, as these intermediates probably lead to direct (3 + 2) be obtained, depending on which step is the rate-determining
or (3 + 4) transannulations, not in accordance with our (3 + 3) step. However, in fact, we measured a KIE of 0.49, and this
transannulations.18 Intramolecular 1,2-hydride transfer of B could not be explained by the mechanism of pathway i.
delivers intermediates C (step c), whose reductive elimination
Faced with the contradiction, we turned to the mechanism
regenerates the catalyst and produces β-thioacyl enones D (step of pathway ii, and now the key problem is how to explain the
d). Tautomerization of D to E followed by intramolecular thia- inverse KIE of 0.49. Within this pathway, the first step involves
Michael addition delivers the desired products 3, which is in the irreversible denitrogenative carbene formation, and has no
high accordance with Christoffers’s synthesis of 2,3-dihy- relation to the KIE. The hydride transfer (step c) or reductive
drothiopyran-4-ones from pent-1-en-4-yn-3-ones and sodium elimination (step d), according to the aforementioned previous
sulfide or sodium hydrosulfide in the presence of water (eqn work,18–20 occurring either irreversibly fast,20,21 or relatively
(3)).19
slow to contribute to a KIE of 1.4–1.7,22 cannot account for a
We believe that the KIE (kH/kD) may play an important role KIE of 0.49. Recent advances on the equilibrium isotope effect
in distinguishing the two mechanistic pathways. The key (EIE) have revealed that a low inverse KIE value (for example,
differences between the two mechanistic pathways (i and ii) less than 0.6) in the C–H activation process may arise from the
rely on how intermediates B are formed (steps a and b, or equilibrium between oxidative addition and reductive elimin-
steps e and f). Pathway i features acyl rhodium hydrides A, and ation.23 Inspired by this, we suggest the oxidative addition of
is quite similar to that of the alk-2-ynal transannulation pro- carbene-ligated Rh(I) (F) to alk-2-enals 2 (step f) to form
posed in our previous work.14 Pathway ii is characteristic of carbene-ligated acyl rhodium hydrides B reversibly. The EIE
thioacyl carbenes F, and is consistent with various transannu- value is theoretically the ratio of the forward and reverse KIEs.
lations reported by Gevorgyan’s,3 Lee’s,4 and Nakajima and In step f, intermediates F are the reactants and B are the pro-
Nishibayashi’s5 groups. In the case of Rh(I)-catalyzed reactions ducts. Thus, oxidative addition from F to B is the forward reac-
involving acyl hyrides A, for example, hydroacylations of tion, and reductive elimination from B to F is the reverse reac-
ketones and alkenes with aldehydes20 as well as decarbonyla- tion. Obviously, the thioacyl carbene ligand is important to
tion of aldehydes,21 primary KIEs varying from 2.4 to 2.9 were render step f reversible. From a simple electronic perspective,
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 3173–3180 | 3177

View Article Online
Paper
Organic & Biomolecular Chemistry
the coordination of thioacyl carbenes with the catalyst [Rh]
makes the metal center more electron-rich (16e configuration
of F, and 18e configuration of G), and lowers their ability to
undergo oxidative addition. On the other hand, the electronic
richness of the metal center (18e configuration) of B grants
them a high tendency to undergo reductive elimination. As a
result, a reversibility between F and B emerges and leads to an
EIE of 0.49. Thus, the mechanism containing thioacyl car-
benes F (pathway ii) is plausibly favored.
Scheme 7 Modification of androst-4-en-3,17-dione.
Returning to Scheme 1, the transannuloselectivity can be
well explained. Both the α-oxo and α-imino carbenes are elec-
trophilic, and the carbenic centers directly react with the
nucleophilic oxygen atom of the alk-2-enals to form ylide inter-
mediates; subsequent cyclization (for Scheme 1a)8 or
rearrangement–cyclization (for Scheme 1b)9 gave the desired
five-membered cycles. It is the electrophilic nature of the two
kinds of carbenes that governs the transannuloselectivity
toward five-membered products. However, in the case of thio-
acyl carbenes, it is the reversible oxidative addition between
the carbene-ligated metal and alk-2-enals that decides the
transannuloselectivity toward six-membered cycles; the
carbene center does not directly participate in the formation
of the C(carbene)–C(formyl) bond. In other words, the α-oxo
and α-imino carbenes act as actor ligands, while the thioacyl
carbenes act as spectator ligands. Such a reactivity difference
is attributed to the heteroatoms of the α-heterovinyl groups of
the three types of carbenes.
Synthetic applications
Our current rhodium-catalyzed formal (3 + 3) transannulation
can be used in the late-stage structural modification of natu-
rally occurring products. As exemplified in Scheme 7, by using
our previously developed iridium-catalyzed reduction–elimin-
ation process, androst-4-en-3,17-dione (9) was readily con-
verted to steroidal 3,5-dien-17-one 11;24 transformation of
ketone 11 to enal 12 and subsequent transannulation with
thioacyl Rh–carbene derived from thiadiazole 1a gave penta-
cyclic product 13 in 91% yield, with the diene moiety excel-
lently tolerated.
Conclusions
We have realized a sulfur-controlled and rhodium-catalyzed
formal denitrogenative (3 + 3) transannulation of 1,2,3-thiadia-
zoles with alk-2-enals, affording structurally diverse 2,3-dihy-
drothiopyran-4-ones in moderate to excellent yields, with good
substrate scope and functional group tolerance. The substitu-
ent effects of alk-2-enals are investigated. The presence of an
α-substituent suppresses the transannulation, while the other
substituents promote the transannulation. Detailed mechanis-
tic studies, especially the inverse KIE of kH/kD = 0.49, suggest
an equilibrium of oxidative addition and reductive addition
between thioacyl rhodium carbenes and thioacyl carbene-
ligated acyl rhodium hydrides. The present formal (3 + 3)
transannulation has been demonstrated to be applicable in
the late-stage structural modification of naturally occurring
products. Moreover, the importance of the heteroatoms of het-
erovinyl carbenes in deciding their reactivities toward transan-
nulations is demonstrated.
Gram-scale synthesis and trials on asymmetric catalysis
The transannulation was scaled up to a gram-scale synthesis.
For example, the reaction of 1.17 g of thiodiazole 1a with
1.32 g of cinnamaldehyde (2g) gave 1.27 g of the desired
product 3ag in 75% yield (Scheme 6a). The possibility of realiz-
ing an asymmetric catalytic reaction by changing the bidentate
DPPF ligand to other chiral bisphosphine ligands was also
exploited. However, the use of the (R)- or (S)-BINAP ligand did
not render any optically active product (Scheme 6b). The
results are in high accordance with the proposed mechanism
in which the chiral center in the final product is formed by an
intramolecular thia-Michael addition of nonchiral intermedi-
ate E, having no concern with the bisphosphine ligand.
Conflicts of interest
There are no conflicts of interest to declare.
Acknowledgements
This work was financially supported by the National Natural
Science Foundation of China (No. 21602010 to Z. Y.) and the
Scheme 6 Gram-scale synthesis and trials on asymmetric catalysis.
3178 | Org. Biomol. Chem., 2021, 19, 3173–3180
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Fundamental Research Funds for the Central Universities
(XK1802-6 to Z. Y. and J. X.; No. 12060093063 to Z. Y.).
J. Chem. Soc., Chem. Commun., 1988, 454; (c) C. H. Chen,
G. A. Reynolds and B. C. Cossar, J. Org. Chem., 1981, 46,
2752; (d) M. S. J. Briggs, M. Helliwell, D. Moorcroft and
E. J. Thomas, J. Chem. Soc., Perkin Trans. 1, 1992, 2223;
(e) S. M. Jeffery, A. G. Sutherland, S. M. Pyke, A. K. Powell
and R. J. K. Taylor, J. Chem. Soc., Perkin Trans. 1, 1993,
2317; (f) V. K. Kansal and R. J. K. Taylor, J. Chem. Soc.,
Perkin Trans. 1, 1984, 703; (g) C. H. Chen, G. A. Reynolds
and J. A. Van Allan, J. Org. Chem., 1977, 42, 2777;
(h) C. H. Chen and G. A. Reynolds, J. Org. Chem., 1979, 44,
3144; (i) D. E. Ward, Y. Gai and Y. Lai, Synlett, 1996, 261.
Notes and references
1 (a) H.-S. Yeom and S. Shin, Acc. Chem. Res., 2014, 47, 966;
(b) M. P. Doyle, M. A. McKervey and T. Ye, Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds, Wiley,
New York, NY, 1998; (c) T. Ye and M. A. McKervey, Chem. Rev.,
1994, 94, 1091; (d) Z. Zhang and J. Wang, Tetrahedron, 2008, 64,
6577; (e) A. Ford, H. Miel, A. Ring, C. N. Slattery, A. R. Maguire 13 (a) G. A. Reynolds, C. H. Chen and J. A. Van Allan, J. Org.
and M. A. McKervey, Chem. Rev., 2015, 115, 9981.
Chem., 1979, 44, 4456; (b) S. Kumaresan, M. V. Krishna and
S. R. Ramadas, Phosphorus Sulfur Relat. Elem., 1987, 31, 43;
(c) D. E. Ward and T. E. Nixey, Tetrahedron Lett., 1993, 34,
947; (d) M. Pieroni, M. Dimovska, J. P. Brincat, S. Sabatini,
E. Carosati, S. Massari, G. W. Kaatz and A. Fravolini, J. Med.
Chem., 2010, 53, 4466; (e) K. Yamamoto, S. Yamazaki,
I. Murata and Y. Fukazawa, J. Org. Chem., 1987, 52, 5239.
2 (a) B. Chattopadhyay and V. Gevorgyan, Angew. Chem., Int.
Ed., 2012, 51, 862; (b) A. V. Gulevich and V. Gevorgyan,
Angew. Chem., Int. Ed., 2013, 52, 1371; (c) H. M. L. Davies
and J. S. Alford, Chem. Soc. Rev., 2014, 43, 5151;
(d) Y. Jiang, R. Sun, X.-Y. Tang and M. Shi, Chem. – Eur. J.,
2016, 22, 17910; (e) E. Aguilar and J. Santamaria, Org.
Chem. Front., 2019, 6, 1513; (f) X. H. Tian, L. N. Song, 14 B. Zhou, Q. Wu, Z. Dong, J. Xu and Z. Yang, Org. Lett.,
A. Hashmi and K. Stephen, Chem. – Eur. J., 2020, 26, 3197. 2019, 21, 3594.
3 (a) D. Kurandina and V. Gevorgyan, Org. Lett., 2016, 18, 15 (a) V. Pirenne, B. Muriel and J. Waser, Chem. Rev., 2021,
1804; (b) For a recent review, see: Y. Shafran, T. Glukhareva,
W. Dehaen and V. Bakulev, Adv. Heterocycl. Chem., 2018,
126, 109.
121(1), 227; (b) M. Meazza, H. Guo and R. Rios, Org.
Biomol. Chem., 2017, 15, 2479.
16 K. Lv and X. Bao, Org. Chem. Front., 2021, 8, 310–318.
4 (a) J.-Y. Son, J. Kim, S. H. Han, S. H. Kim and P. H. Lee, 17 J. Tsuji and K. Ohno, Tetrahedron Lett., 1965, 6, 3969.
Org. Lett., 2016, 18, 5408; (b) B. Seo, Y. G. Kim and 18 Three-membered intermediates formed by [2 + 1] cycload-
P. H. Lee, Org. Lett., 2016, 18, 5050; (c) B. Seo, H. Kim,
Y. G. Kim, Y. Baek, K. Um and P. H. Lee, J. Org. Chem.,
2017, 82, 10574; (d) J. E. Kim, J. Lee, H. Yun, Y. Baek and
P. H. Lee, J. Org. Chem., 2017, 82, 1437.
5 W. Liang, K. Nakajima and Y. Nishibayashi, Eur. J. Org.
Chem., 2020, 3879.
ditions of carbenes were proposed in the previous works by
Gevorgyan, Lee, and Nakajima and Nishibayashi (ref. 3a,
4a,b, and 5); thus one may also consider an epoxide inter-
mediate in this work. However, complex and obscure evol-
ution (especially the H migration) of an imagined epoxide
to compound 3 prevented us from considering such a poss-
ible mechanism. In addition, we also tried to verify the
mechanism experimentally, designing the synthesis of
such an intermediate by Rh-catalyzed epoxide formation
from PhC(O)C(N2)CO2Et and 2l and subsequent thionation
of benzoyl with the Lawson reagent. However, the first step
could not be realized, although various conditions were
tried.
6 S. Chuprakov, F. W. Hwang and V. Gevorgyan, Angew.
Chem., Int. Ed., 2007, 46, 4757.
7 (a) L. V. R. Reddy, V. Kumar, R. Sagar and A. K. Shaw,
Chem. Rev., 2013, 113, 3605; (b) V. Marcos and J. Alemán,
Chem. Soc. Rev., 2016, 45, 6812; (c) N. A. Keiko and
N. V. Vchislo, Asian J. Org. Chem., 2016, 5, 1169;
(d) Z. Wang, Molecules, 2019, 24, 3412.
8 M. E. Alonso, M. d. C. Garcia and A. W. Chitty, J. Org. 19 A. Rosiak, R. M. Müller and J. Christoffers, Monatsh. Chem.,
Chem., 1985, 50, 3445. 2007, 138, 13.
9 T. Miura, T. Tanaka, K. Hiraga, S. G. Stewart and 20 (a) K. G. M. Kou, D. N. Le and V. M. Dong, J. Am. Chem.
M. Murakami, J. Am. Chem. Soc., 2013, 135, 13652.
10 (a) Y. Chen, H. Qi, N. Chen, D. Ren, J. Xu and Z. Yang,
J. Org. Chem., 2019, 84, 9044; (b) D. Fu, J. Dong, H. Du and
Soc., 2014, 136, 9471; (b) S. K. Murphy, A. Bruch and
V. M. Dong, Chem. Sci., 2015, 6, 174; (c) R. Guo and
G. Zhang, J. Am. Chem. Soc., 2017, 139, 12891.
J. Xu, J. Org. Chem., 2020, 85, 2752; (c) J. Dong, H. Du and 21 For a mechanism for Rh-catalysed decarbonylation, see:
J. Xu, J. Org. Chem., 2019, 84, 10724; (d) Z. Yang, W. Xu,
Q. Wu and J. Xu, J. Org. Chem., 2016, 81, 3051.
P. Fristrup, M. Kreis, A. Palmelund, P.-O. Norrby and
R. Madsen, J. Am. Chem. Soc., 2008, 130, 5206.
11 (a) Z. Yang and J. Xu, Chem. Commun., 2014, 50, 3616; 22 (a) G. L. Moxham, H. Randell-Sly, S. K. Brayshaw,
(b) P. Huang, Z. Yang and J. Xu, Tetrahedron, 2017, 73,
3255; (c) W. He, J. Zhuang, Z. Yang and J. Xu, Org. Biomol.
Chem., 2017, 15, 5541; (d) W. He, J. Zhuang, H. Du, Z. Yang
and J. Xu, Org. Biomol. Chem., 2017, 15, 9424.
A. S. Weller and M. C. Willis, Chem. – Eur. J., 2008, 14,
8383; (b) Z. Shen, P. K. Dornan, H. A. Khan, T. K. Woo and
V. M. Dong, J. Am. Chem. Soc., 2009, 131, 1077;
(c) A. B. Chaplin, J. F. Hooper, A. S. Weller and M. C. Willis,
J. Am. Chem. Soc., 2012, 134, 4885; (d) I. Pernik,
J. F. Hooper, A. B. Chaplin, A. S. Weller and M. C. Willis,
12 (a) K. Nakasuji, M. Nakatsuka and I. Murata, J. Chem. Soc.,
Chem. Commun., 1981, 1143; (b) G. Casy and R. J. K. Taylor,
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 3173–3180 | 3179

View Article Online
Paper
ACS Catal., 2012, 2, 2779; (e) A. Prades, M. Fernández,
Organic & Biomolecular Chemistry
Res., 2003, 36, 140; (c) For a selected example, see:
D. G. Churchill, K. E. Janak, J. S. Wittenberg and G. Parkin,
J. Am. Chem. Soc., 2003, 125, 1403.
S. D. Pike, M. C. Willis and A. S. Weller, Angew. Chem., Int.
Ed., 2015, 54, 8520.
23 For reviews, see: (a) M. Gomez-Gallego and M. A. Sierra, 24 J. Li, W. Tang, D. Ren, J. Xu and Z. Yang, Green Chem.,
Chem. Rev., 2011, 111, 4857; (b) W. D. Jones, Acc. Chem. 2019, 21, 2088.
3180 | Org. Biomol. Chem., 2021, 19, 3173–3180
This journal is © The Royal Society of Chemistry 2021