The energy-transfer-enabled biocompatible disulfide–ene reaction

Article abstract of DOI:10.1038/s41557-018-0102-z

Sulfur-containing molecules participate in many essential biological processes. Of utmost importance is the methylthioether moiety, present in the proteinogenic amino acid methionine and installed in tRNA by radical-S-adenosylmethionine methylthiotransferases. Although the thiol–ene reaction for carbon–sulfur bond formation has found widespread applications in materials or medicinal science, a biocompatible chemo- and regioselective hydrothiolation of unactivated alkenes and alkynes remains elusive. Here, we describe the design of a general chemoselective anti-Markovnikov hydroalkyl/aryl thiolation of alkenes and alkynes—also allowing the biologically important hydromethylthiolation—by triplet–triplet energy transfer activation of disulfides. This fast disulfide–ene reaction shows extraordinary functional group tolerance and biocompatibility. Transient absorption spectroscopy was used to study the sensitization process in detail. The hereby gained mechanistic insights were successfully employed for optimization of the catalytic system. This photosensitized transformation should stimulate bioimaging applications and carbon–sulfur bond-forming late-stage functionalization chemistry, especially in the context of metabolic labelling.

Full text of DOI:10.1038/s41557-018-0102-z

Articles
https://doi.org/10.1038/s41557-018-0102-z
The energy-transfer-enabled biocompatible
disulfide–ene reaction
Michael Teders1, Christian Henkel2, Lea Anhäuser3, Felix Strieth-Kalthoff1, Adrián Gómez-Suárezꢀ ꢀ1,
Roman Kleinmans1, Axel Kahntꢀ ꢀ2, Andrea Rentmeisterꢀ ꢀ3,4, Dirk Guldiꢀ ꢀ2* and Frank Gloriusꢀ ꢀ1*
Sulfur-containing molecules participate in many essential biological processes. Of utmost importance is the methylthioether
moiety, present in the proteinogenic amino acid methionine and installed in tRNA by radical-S-adenosylmethionine methylthio-
transferases. Although the thiol–ene reaction for carbon–sulfur bond formation has found widespread applications in materials
or medicinal science, a biocompatible chemo- and regioselective hydrothiolation of unactivated alkenes and alkynes remains
elusive. Here, we describe the design of a general chemoselective anti-Markovnikov hydroalkyl/aryl thiolation of alkenes and
alkynes—also allowing the biologically important hydromethylthiolation—by triplet–triplet energy transfer activation of disul-
fides. This fast disulfide–ene reaction shows extraordinary functional group tolerance and biocompatibility. Transient absorption
spectroscopy was used to study the sensitization process in detail. The hereby gained mechanistic insights were successfully
employed for optimization of the catalytic system. This photosensitized transformation should stimulate bioimaging applica-
tions and carbon–sulfur bond-forming late-stage functionalization chemistry, especially in the context of metabolic labelling.
eveloping new reaction methodologies that give access to thiol–ene reaction (Fig. 1b)18–20. Significantly milder reaction con-
unprecedented reactivity modes or provide more selective, ditions of photocatalytic systems enable application in bioconjuga-
D
efficient and environmentally sustainable alternatives to estab- tion reactions and polymer synthesis21,22. Despite such advantages,
lished approaches is the central goal of many research programmes the photoinitiated thiol–ene reaction still suffers from three major
throughout chemistry1. Inspiration for the design of new reactions limitations. First, the inherent reactivity of thiols (for example,
is rooted in the understanding of reaction mechanisms and the fun- thiol–Michael additions) goes hand-in-hand with a lack of che-
damental reactivity principles that underpin them2. Methylthioether moselectivity, especially in complex molecules bearing suitable
functionalities, which are ubiquitous and of utmost importance in Michael acceptor functionalities23. Second, the hydromethylthiola-
nature, are inserted into tRNA and ribosomal proteins by highly tion of unactivated alkenes and alkynes under mild and biocompat-
effective radical-S-adenosylmethionine (SAM) methylthiotransfer- ible conditions has never been reported, due to the need to handle
ases bearing [4Fe-4S] redox-active clusters (Fig. 1a)3–5. Surprisingly, highly toxic, gaseous methanethiol as reagent. Third, a chemoselec-
the way that organisms construct carbon–sulfur bonds during the tive generation of thiyl radicals from thiols in biological systems is
biosynthesis of sulfur-containing residues in living organisms is hardly possible due to the presence of cellular thiols, limiting the
barely understood5. Thiyl radicals are essential intermediates in overall value of this approach for labelling applications.
many biosynthetic pathways6,7, for example, in the deoxygenation of
Inspired by the S–S bond dissociation energies of aliphatic disul-
ribonucleotides8, a transformation occurring in living organisms as fides of ~65kcalmol−1 (ref. 24), reports on the direct sensitization of
part of the de novo generation of DNA precursors. The omnipres- aryl disulfides using UV light11, and the mild reaction conditions
ence of thiyl radicals in biological systems9 along with their vari- of a sensitization protocol, we hypothesized using suitable photo-
ous synthetic opportunities to access organosulfur moieties widely catalysts for the energy transfer activation of alkyl disulfides, such
represented within natural products, pharmaceuticals or material as dimethyl disulfide (2), to access alkylthiyl radicals (Fig. 1c)25.
products10, has been an inspiration to investigate the utilization of This energy-transfer-enabled disulfide–ene reaction would allow
these highly reactive species in diverse applications11. Thiol–ene the biologically relevant chemoselective hydromethylthiolation of
click reactions12, in which a thiyl radical generated from the cor- alkenes and alkynes under mild conditions, namely visible-light
responding thiol regioselectively adds in an anti-Markovnikov irradiation26,27. The use of visible light as an abundant natural
fashion to an alkene, are among the most prominent approaches to resource to promote chemical reactions offers many advantages28.
carbon–sulfur bond formation13. They have been applied in biocon- Considering the costs and environmental sustainability, photo-
jugates14,15, pharmaceutical chemistry16 and polymer science17. In the catalytic reactions have attracted widespread attention, with many
past, UV-light irradiation or radical initiators have mostly been used examples demonstrating unusual or exotic reactivity patterns that
for the generation of thiyl radicals11. However, the functional group do not occur in the absence of light29,30. Ground-breaking transfor-
tolerance, the need for costly set-ups for UV irradiation and the pro- mations utilizing photoexcited catalysts to engage in electron trans-
duction of huge amounts of waste have been limiting factors.
fer processes with organic molecules or other metal complexes have
The aforementioned drawbacks set up the development of been achieved recently29–31. The other fundamental pathway for the
visible-light photocatalytic systems as a means to initiate the bimolecular decay of photoexcited states, namely energy transfer,
1Organisch-Chemisches Institut, Westfälische Wilhelms-Universität Münster, Münster, Germany. 2Department für Chemie und Pharmazie,
Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany. 3Institut für Biochemie, Westfälische Wilhelms-Universität Münster, Münster,
Germany. 4Cells-in-Motion Cluster of Excellence (EXC1003-CiM), Westfälische Wilhelms-Universität, Münster, Germany. *e-mail: dirk.guldi@fau.de;
glorius@uni-muenster.de
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
has rarely been used in organic synthesis29. Triplet–triplet energy
transfer (TTEnT) from visible-light photocatalysts has only been
utilized to achieve a small number of organic transformations29.
These are roughly divided into three categories: E/Z isomerization
of (activated) alkenes32, [2+2] photocycloadditions33,34 and sensiti-
zation of transition metal complexes35.
a
HN
OH
HN
OH
SAM, sulfur source
O
O
O
O
H
RimO
H
H
S
methylthiotransferase
Asp89 of ribosomal
S12 protein in E. coli
Sulfur-containing drugs
Results and discussion
H
H
O
S
Luminescence screening and optimization. We began our stud-
ies with a hypothesis-driven approach by applying our recently
reported luminescence quenching screening36,37 to investigate the
ability of dimethyl disulfide (2) to interact with various excited-
state photocatalysts featuring different excited-state triplet ener-
gies. In particular, four different Ir-based photocatalysts gave rise to
significant luminescence quenching when dimethyl disulfide (2) is
present. The photocatalyst [Ir(dF(CF3)ppy)2(dtbbpy)](PF6) ([Ir–F],
dF(CF3)ppy = 2-(2,4-difluorophenyl)-3-trifluoromethylpyridine,
dtbbpy=4,4ʹ-di-tert-butyl-2,2ʹ-bipyridine) was the most efficient
among them, with a quenching fraction of 42% (Fig. 2a). The
luminescence quenching correlated with the triplet excited-state
energies of the photosensitizers. Notably, catalysts with exception-
ally high excited-state oxidation or reduction potentials showed
no or lower luminescence quenching. From this we assume that a
TTEnT reaction from the excited photocatalyst to the disulfide is
the modus operandi.
NH
S
OH
N
NH2
Methionine
Pergolide
R1
R2
b
c
S
R3
R3
R1 SH
Initiation by
R1
S
+
R2
H
anti-Markovnikov
hydrothiolation
Radical initiators (AIBN, (t-BuO)2)
UV irradiation (λ < 300 nm)
Visible light
S–S
BDE
CF3
tBu
N
F
(kcal mol–1
)
N
Ir
50
S
S
Ar
Ar
S
S
F
F
+
To investigate the chemo- and regioselectivity of the hydro-
methylthiolation reaction, carvone (1a), dimethyl disulfide (2) and
[Ir–F] were reacted in acetonitrile under visible-light irradiation
(λmax =455nm). Of great relevance is the fact that the yield of the
methylthiolated product 3a correlated perfectly with the screening
results. Due to the electrophilic nature of the methylthiyl radical,
the exocyclic, more electron-rich alkene functionality of carvone
(1a) was chemo- and regioselectively hydromethylthiolated in an
anti-Markovnikov fashion (Fig. 2a). Our TTEnT activation hypoth-
esis was independently supported by control reactions, in which the
need for the [Ir–F] photosensitizer and light were demonstrated.
To further support this hypothesis, benzophenone, a classical triplet
sensitizer, was used under 365nm irradiation, affording the desired
product 3a in 25% yield. Further optimization of parameters, such
as the irradiation wavelength or the solvent, allowed the yield of 3a
to be increased to 74% (Supplementary Section 3.1). The reaction
is characterized by a fast reaction progress, as documented in the
reaction profile (Fig. 2b).
N
N
tBu
Alkyl
65
Alkyl
F
CF3
[Ir-F]
ET = 60.8 kcal mol–1
(PF6)
R4
S
Ir
R2
S
R4
R2
R3
R1
+
R4
R1
S
R3
Triplet
sensitization
H
General method for the hydrothiolation of
oxidation- or reduction-sensitive substrates
O
O
NH
H
HN
O
S
S
H
Fig. 1 | Chemoselective anti-Markovnikov hydrothiolation of alkenes—
allowing the biologically important hydromethylthiolation—enabled
by triplet–triplet photosensitization of disulfides. a, Methylthioether
scaffolds are involved in diverse biosynthetic processes. Furthermore,
this functional unit can be identified in important drugs, such as in the
essential amino acid methionine or in pergolide, an ergoline-based
dopamine receptor agonist used for the treatment of Parkinson’s disease.
b, The photocatalytic thiol–ene click reaction allows carbon–sulfur bond
formation under extremely mild conditions and is therefore widely used in
bioconjugate chemistry, polymer science and pharmaceutical synthesis.
However, this methodology is still limited by chemoselectivity issues and
by oxidative or reductive reaction environments limiting the functional
group tolerance of the overall reaction. The biologically important
hydromethylthiolation is challenging, and biochemical applications
are hardly possible due to the presence of cellular thiols participating
in undesired side reactions. c, The design of a biocompatible energy-
transfer-enabled disulfide–ene reaction would allow the chemo- and
anti-Markovnikov selective construction of important methylthioethers.
Triplet–triplet sensitization of disulfides by photocatalysts with sufficient
triplet excited-state energy may allow the mild installation of alkyl- and
arylthioether scaffolds. Tuning of the photocatalysts redox potentials
would even allow the hydrothiolation of oxidation- or reduction-sensitive
compounds. BDE, bond dissociation energy.
Substrate scope and limitations. With the optimized reaction con-
ditions in hand, we investigated the scope and synthetic limitations
of this disulfide–ene reaction (Fig. 3). The hydromethylthiolation of
alkenes tolerates a wide range of functional groups, such as ketones
(3b), aldehydes (3c), epoxides (3f, 3v), amides (3g, 3w), alcohols
(3h, 3p), nitriles (3r), esters (3l, 3t), sulfonamides (3x) or hetero-
cycles (3i, 3k, 3w, 3x). In all cases, the corresponding products are
formed in moderate to good yields (42–85%). Substrates exhibiting
more than one unsaturated carbon–carbon bond are chemoselec-
tively methylthiolated in an anti-Markovnikov fashion at the more
electron-rich, sterically more easily accessible alkene functionality.
Aliphatic disulfides with longer alkyl chains (3y and 3z) or with a
hydroxy group (3aa) could also be employed using this protocol,
whereas in the case of sterically demanding disulfides, such as
tert-butyl disulfide, no reactivity is observed (3ab; Supplementary
Section 4.11). A slight modification of the reaction conditions
allows the use of diaryl disulfides as thiyl radical precursors for the
synthesis of arylthioethers (3ac–3ae, 84–89%). Furthermore, the
functionalization of an azide derivative (3af) and of amino acids
(3ag–3ai) as bioconjugates delivers the respective hydrothiolated
products in good yields, paving the road for potential in vitro and in
vivo applications. The vic-di-methylthiolation of different alkynes
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
a
Finally, the functional group tolerance and preservation of this reac-
tion was analysed by performing additive-based robustness screens,
indicating a high robustness and functional group preservation of
O
O
Photocatalyst
S
+
S
MeCN (0.1 M)
Blue LEDs (λmax = 455 nm)
r.t., 16 h
S
the overall transformation (Supplementary Table 3)39,40
.
1a
2
3a
(1.0 equiv.)
(2.0 equiv.)
Mechanistic investigations. To corroborate the reaction mecha-
nism and to support the postulated TTEnT, photosensitization of
dimethyl disulfide (2) and/or 1-octene by [Ir–F] was studied in
detail. Experiments showed that methyl(octyl)sulfane formation is
based on irradiation with light (400nm<λ≤455nm) and the pres-
ence of the photocatalyst. From steady-state UV–vis absorption
spectroscopy we conclude that visible light (400nm <λ≤455nm)
is exclusively absorbed by the [Ir–F] photocatalyst and that neither
2 (λmax =255nm) nor 1-octene (λmax <200nm) is appreciably photo-
activated (Fig. 4a). In other words, the photocatalyst is crucial in the
visible-light-driven disulfide–ene reaction.
II/III*
III*/IV
Photocatalyst
Yield 3a (%)
ET
E1/2
(V)
E1/2
(V)
Quenching
fraction F
(kcal mol–1
)
[Ir-F]
60.8
59.1
+1.21
+0.34
–0.89
–1.44
42
24
58
53
fac-[Ir(dF(ppy))3]
[Ir(ppy)2(NHC-F2)]
57.8
57.8
+0.38
+0.31
–1.49
–1.73
31
19
47
46
fac-[Ir(ppy)3]
[Ir(ppy)2(dtbbpy)](PF6)
[Ru(bpz)3](PF6)2
49.2
48.4
46.8
46.5
+0.66
+1.45
+0.82
+0.77
–0.96
–0.26
–0.87
–0.81
<5
<5
<5
<5
0
0
0
0
To shed light on the mechanistic aspects, phosphorescence
quenching of [Ir–F] on addition of 2 and/or 1-octene was investi-
gated by Stern–Volmer analyses: 2 is an effective phosphorescence
quencher, while 1-octene is not at all. Our findings are in agree-
ment with the screening results and suggest favourable interac-
tions between [Ir–F] and 2 (Fig. 4b), but no interactions between
[Ir–F] and 1-octene. Consequently, we characterized the nature of
the [Ir–F]/(2) interaction. The triplet excited-state energies of 2 and
[Ir–F] were determined using cryostatic phosphorescence measure-
ments as 66.1 2.2 and 60.8 0.01kcalmol−1, respectively, suggest-
ing a slightly endergonic activation of 2 by the [Ir–F] photocatalyst,
a phenomenon that has been rarely described in the literature41. To
examine the underlying kinetics, nanosecond time-resolved tran-
sient absorption spectroscopy (ns-TAS) was performed using [Ir–F]
in the absence and presence of variable 2 concentrations reaching
9×10−2 M (Fig. 4c). The ns-TAS results for 387nm photoexcited
[Ir–F] are dominated by a rather broad 462nm maximum. Based
on literature reports, we ascribe the new transient to the excited
triplet metal-to-ligand charge-transfer/ligand-centred (MLCT/LC)
state of the [Ir–F] photocatalyst42. Its lifetime, which is 2.49 0.06µs
in the absence of any 2, is subject to a 2 concentration-dependent
shortening. Concomitant with quenching of the [Ir–F] 3*MLCT/LC
transients, we observed the formation of newly developing tran-
sient absorption features at 400 and 440nm, which we tenta-
tively assign to 3*2. Confirmation for our assignment came from
ns-TAS photosensitization experiments with Michler’s ketone
(4,4′-bis(dimethylamino) benzophenone)—a well-known organic
photosensitizer—rather than benzophenone (Supplementary
Fig. 8). Here, the same 400 and 440nm features emerged at the
expense of the Michler’s ketone triplet excited-state maxima at 415,
[Ru(phen)3](PF6)2
[Ru(bpy)3](PF6)2
b
100
1a
3a
80
60
40
20
0
0
20
40
60
80
100
Time (min)
Fig. 2 | Hypothesis-driven luminescence screening and reaction profile
of the chemoselective anti-Markovnikov disulfide–ene reaction.
a, Mechanism-based luminescence quenching studies of dimethyl disulfide
(2) and various photosensitizers reveal a correlation between quenching
and the photocatalysts’ triplet excited-state energy ET (for details on
luminescence quenching studies, see Supplementary Section 2.1). This
trend was also observed in the hydromethylthiolation of the exocyclic
alkene functionality of carvone (1a), presumably enabled by triplet–triplet
sensitization of dimethyl disulfide (2) by the respective photocatalyst.
All potentials are given in volts versus the saturated calomel electrode
(SCE) and were measured in acetonitrile. b, Hydromethylthiolation of the
exocyclic alkene functionality of carvone (1a) proceeds in a quick fashion
without an induction period. The starting material is completely consumed
within 90ꢀmin. The reaction profile was independently determined twice
with similar results.
43
512 and 695nm (ref. ). Turning to kinetic analysis of the TTEnT
we consider the following sequence:
kGSR,[Ir–F]
3*[Ir–F]⎯⎯⎯⎯⎯⎯⎯⎯⎯→[Ir–F]
(1)
kTTEnT
3*[Ir–F]+ 2 ⎯⎯⎯⎯⎯→⎯ [Ir–F]+ 3
(2)
*
2
By virtue of the intrinsic ground-state recovery rate (kGSR) of the
was successful, affording an E/Z mixture of 3aj and 3ak in 83% and 3*MLCT/LC state of the [Ir–F] photocatalyst, the observed rate con-
68% yield, respectively. The second methylthiolation event most stant (kobs) for the overall TTEnT is based on
probably occurs through reaction of the vinyl radical intermediate
kobs = kGSR + [2] ⋅ kTTEnT
(3)
with a ground-state disulfide molecule as part of a chain reaction.
The benzothiophene heterocycle 3al could be regioselectively cre-
ated when reacting diphenyl disulfide with 1-phenyl-prop-1-yne
Overall, a perfect linear relationship was confirmed in the
(73%)38. Variousdrugderivatives, incorporatingaprobenecid(3am), corresponding (kobs −kGSR) versus [2] plots with a slope of
a dehydrocholic acid (3an), an estrone (3ao) or an erucic acid (3ap) (1.16 0.01)×107 M−1 s−1 (Supplementary Fig. 11). The slope corre-
scaffold, could be methylthiolated in good to excellent yields. lates with the TTEnT rate constant kTTEnT, while a sizeable intercept
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
a
b
c
Additive
remaining
Relative product
yield
CF3
F
F
15
15
R4
R1
tBu
tBu
E
T = 60.8 kcal mol–1
S
N
Ir
R2
[Ir-F] (1.0 mol%)
R4
S
R2
R3
R1
N
N
+
F
F
S
R4
E1/2II/III* = +1.21 V
E1/2III*/IV = –0.89 V
DCE, r.t., 16 h
Blue LEDs
R3
H
N
N
N
(λmax= 455 nm)
Alkene/alkyne (1)
Disulfide (2)
3
0
0
(1.0 equiv.)
(2.0 equiv.)
41 examples
Yield
Yield 100%
100%
CF3
High (>66%)
Low (<34%)
Medium (66%–34%)
[Ir-F]
(PF6)
Average yield
d
O
O
O
S
O
HO
S
O
S
H
N
S
O
S
S
H
S
S
S
3b, 56%
3c, 48%
3d, 81%
3e, 83%
3m, 64%
3f, 80%
3g, 84%
3h, 93%
3i, 42%
3a, 74%
67% on 3.0 mmol scale
O
O
S
S
S
OH
S
O
S
S
F3C
O
S
O
S
N
O
3j, 56%
3k, 69%
3l, 75%
3n, 74%
3o, 71%
3p, 76%
O
S
S
O
S
O
O
O
O
S
N
N
S
S
S
O
S
O
S
N
S
O
3q, 67%
3r, 68%
3s, 74%
3t, 83%
3u, 57%
3v, 61%
3w, 85%
3x, 63%
OH
e
f
g
O
O
S
N3
S
S
O
N
S
H
O
S
O
S
3ac, 88%
3af, 41%
3ag, 69%
Cl
3y, 80%
3z, 72%
S
O
O
3ad, 84%
OH
OH
O
Me
HN
O
S
HN
S
OH
O
S
O
S
O
O
S
3aa, 63%
3ab, 0%
3ae, 89%
3ah, 80%
3ai, 71%
h
i
S
O
O
S
N
O
3am, 72%
Probenecid scaffold
3ao 73%
Estrone scaffold
H
H
S
3aj, 83% (E:Z = 40:60)
H
O
S
S
O
S
S
O
O
O
OH
3ak, 68% (E:Z = 32:68)
O
O
H
3ap, 62%
Erucic acid
S
S
3an, 84%
S
Dehydrocholic acid scaffold
Ph
H
H
O
O
3al, 73%
Fig. 3 | Scope of the disulfide–ene hydrothiolation of alkenes and alkynes. a, General reaction conditions used for scope preparation: alkene or alkyne
(0.3ꢀmmol, 1.0ꢀequiv.), disulfide (0.6ꢀmmol, 2.0ꢀequiv.), [Ir–F] (0.003ꢀmmol, 1.0ꢀmol%), DCE (0.1ꢀM), 455ꢀnm, r.t., 16ꢀh. b, Structure of the applied [Ir]-
based photocatalyst and its photophysical properties. All potentials are given in volts versus SCE and were measured in acetonitrile. c, Results of an
additive-based robustness screen corroborate a mild synthetic protocol. d, Hydromethylthiolation of diverse alkenes exhibiting different electronic and
steric properties. Substrates exhibiting more than one unsaturated carbon–carbon bond are exclusively hydrothiolated at the more electron-rich, sterically
more easily accessible alkene functionality. e, Energy transfer sensitization of diverse disulfides yielding the respective hydrothiolated products. f, Slight
modification of the reaction conditions gives access to arylthioethers (using acetone as solvent). g,i, Application of this protocol to functionalize drug
derivatives and bioconjugates. h, Hydrothiolation of alkynes allows the vic-di-methylthiolation or construction of sulfur-containing heterocycles. Data
are reported as percent isolated yield except 3e, 3i, 3l, 3u, 3af and 3aj, which are reported as 1H NMR yield using CH2Br2 as internal standard. Product
3v was obtained due to rearrangement. Hydrothiolation reactions and the robustness screen were performed once. See Supplementary Section 3.2 for
experimental details and characterization data. DCE, 1,2-dichloroethane.
suggests that the reverse reaction, namely rTTEnT (vide infra) defin- (Supplementary Fig. 12). A kTTEnT of (1.31 0.04)×107 M−1 s−1 is in
itively takes place. Independent confirmation for the TTEnT rate excellent agreement with the ns-TAS experiments. For determina-
constant was gathered in phosphorescence lifetime measurements tion of the aforementioned krTTEnT and the underlying equilibrium
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
a
b
c
6 × 105
3 × 105
∆OD / × 10–3 (a.u.)
3.0
2.2
2.0
1.8
1.6
1.4
1.2
1.0
344.18 15.3
100
10
1
0.1
0.01
2.5
2.0
1.5
1.0
0.5
0.0
Dimethyl disulfide
Carvone
1-octene
[Ir-F]
0
250 300 350 400 450 500
8 × 103
4 × 103
31.35 1.5
Carvone (1a)
0
250 300 350 400 450 500
4 × 102
1.0
0.8
0.6
0.4
0.2
0.0
2 × 102
0.9 0.6
3*MLCT/LC [Ir-F]
Dimethyl disulfide (2)
3*2
0
250 300 350 400 450 500
0.000 0.005 0.010 0.015 0.020
Concentration (mol dm−3
)
λ (nm)
400
500
600
700
800
900
d
e
λ (nm)
[Ir-F] (1.0 mol%), DCE (0.1 M)
S
S
S
+
S
S
S
∆OD × 10–3 (a.u.)
∆OD × 10–3 (a.u.)
Blue LEDs (λmax = 455 nm)
2
4
5
4.2
4.2
C1-octene = 1 × 10–4
M
r.t., 16 h
C1-octene = 0 M
100
100
10
1
Detected via
GC-MS
3.5
2.8
2.1
1.4
0.7
0.0
3.5
2.8
2.1
1.4
0.7
0.0
10
1
O
O
[Ir-F] (1.0 mol%)
deuterated solvent
0.1
0.01
0.1
+
Me2S2
H/D
0.01
Blue LEDs (λmax = 455 nm)
S
r.t., 16 h
400 500 600 700 800 900
λ (nm)
400 500 600 700 800 900
1a
2
3a
λ (nm)
d3-MeCN or CDCl3 = exclusively H
d4-MeOH or D2O = mostly D
f
S
or
S
S
S
S
S
S
S
8
Me2S2
R4 = Me in
nonpolar or
polar aprotic solvent
2
9
Me
S
R2
R3
R1
Hydrogen atom
abstraction
R4
R1
H
3*
IrIII (T1)
IrIII (S0)
R2
S
3
R2
R1
R4
R4
S
S
R3
+
Quantum yield
R4
S
R3
Energy transfer
activation
Φ = 0.71
2
6
1
7
R4
Solvent
S
R2
R3
R4 = Ph or
4 = Me in
polar protic solvent
R1
Hydrogen atom
abstraction
R
H
3
Fig. 4 | Proposed reaction mechanism of the photosensitized disulfide–ene reaction derived from diverse mechanistic studies. a, UV–vis absorption
spectroscopy of the individual reaction components shows that the photocatalyst is the only light absorber. b, Stern–Volmer quenching studies suggest
an energy transfer event between the excited-state photosensitizer and the disulfide (nꢀ=ꢀ6 independent samples (dimethyl disulfide (2) and carvone
(1a)) or nꢀ=ꢀ5 independent samples (1-octene) for the linear regression). c, Triplet–triplet sensitization of the disulfide was investigated using ns-TAS to
shed light on the reaction kinetics. d, A scrambling experiment utilizing two aliphatic disulfides points towards the existence of thiyl radical intermediates.
e, Depending on the reaction conditions, hydrogen atom abstraction can either take place from the alkyl disulfide or the solvent. f, Proposed reaction
mechanism of the overall transformation revealed by different spectroscopic as well as additional experiments. Experiments in a, b, d and e were all
performed once. See Supplementary Information for further details. OD, optical density.
constant KTTEnT we performed ns-TAS experiments with different
[Ir–F] to 2 ratios. For the
We took the (kobs-kGSR,[Ir–F])/[2] versus [[Ir–F]]/[2] plots to derive
kTTEnT as (5.26 0.23)×107 M−1 s−1, krTTEnT as (8.65 0.88)×106 M−1 s−1
and KTTEnT as 6.08 0.71. Notably, due to S–S bond weakening and
homolytic cleavage to afford thiyl radical intermediates (vide infra),
the thermodynamically favoured reverse energy transfer, which rein-
states the 3*MLCT/LC state of the [Ir–F] photocatalyst, is slowed down.
Having confirmed the triplet sensitization mechanism, the con-
secutive reaction steps were systematically investigated. To this end,
we performed a radical scrambling experiment to investigate the
presence of thiyl radical intermediates under our reaction condi-
3
3
*
*
2
(4)
MLCT∕LC([Ir–F]) + 2 ⇄ [Ir–F]+
equilibrium, (kobs-kGSR,[Ir–F]) for the overall TTEnT relates to44
(kobs−kGSR,[Ir–F]) = kTTEnT ⋅ [[Ir–F]] + krTTEnT ⋅ [2]
(5)
Division by the concentration of 2 results in the following linear tions. Irradiation of dimethyl disulfide (2) and dibutyl disulfide (4)
expression:
with visible light in the presence of the [Ir–F] photocatalyst resulted
in methyl butyl disulfide (5) as the major product. Very probably,
methyl butyl disulfide (5) formation involves thiyl radicals as inter-
mediates and their radical recombination or radical addition into
another disulfide molecule (Fig. 4d).
(kobs−kGSR,[Ir–F]
)
kTTEnT ⋅ [[Ir–F]]
[2]
(6)
=
+ krTTEnT
[2]
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
a
b
R4
[Ir-F] (1.0 mol%) or
alloxazine 10 (5.0 mol%)
Alloxazine
(10, 5.0 mol%)
O
O
S
R2
R2
R1
R4
S
S
R1
+
+
R3
R3
R4
S
S
DCE (0.1 M)
Blue LEDs
(λmax = 455 nm or 400 nm)
r.t., time
H
S
DCE, r.t., 16 h
Blue LEDs
(λmax = 400 nm)
3
Alkene 1
(1.0 equiv.)
Disulfide 2
(2.0 equiv.)
1a
2
3a
(1.0 equiv.)
(2.0 equiv.)
O
100
80
60
40
20
O
P
S
S
Si
S
O
S
Si
O
O
nBu
3a, 82% (74%)
3aq, 49% (0%)
3ar, 84% (41%)
3as, 62% (11%)
MeO
N
N
N
O
N
MeO
Me
O
O
NH
H
O
10
O
S
HN
S
T = 65.1 kcal mol–1
S
E
H
[Ir-F]
O
O
3au, 91% (21%)
Biotin scaffold
O
O
S
Alloxazine (10)
O
O
0
0
3at, 65% (0%)
Mycophenolic acid scaffold
20
40
60
80
100
Time (min)
c
O
O
O
O
O
O
O
O
H
O
H
O
mCPBA
S
S
S
H2O2, Sc(OTf)3
O
H
H
O
CH2Cl2
0 °C, 2 h
88%
CH2Cl2:EtOH (10:1)
r.t., 4 h
H
O
H
H
O
H
H
O
3an
11
Sulfoxide
12
Sulfone
O
O
O
95%
d
e
f
[Ir-F] (1.0 mol%)
biomolecule
Tris-HCl buffer, pH = 7.4 (0.2 M)
O
O
O
OH
O
BSA
RNase A
Me2S2
+
HN
S
Disulfide-ene
reaction
+
+
+
+
b
–
–
a
b
a
S
O
(kDa)
M
Blue LEDs
(λmax = 455 nm)
r.t., 24 h
3
3ai, 64%
1a
2
180
130
100
70
(1.0 equiv.)
(2.0 equiv.)
66.5 kDa
O
O
55
40
Entry Bio-additive
Conversion (%) Yield (%) Additive recovery
35
25
O
S
1
2
3
4
5
6
7
8
None
99
98
98
98
98
98
98
98
68
65
54
63
66
68
64
62
-
B
A
A
A
C
C
ND
H
L-glutathione (1.0 equiv.)
Bovine serum albumine (100 µM)
RNase A (100 µM)
ssDNA (5 µM)
3an, 73%
H
H
O
15
10
O
13.7 kDa
RNA (2.5 µM)
Total RNA (5.5 µg ml–1
)
O
O
O
NH
H
Cell lysate (1:10 vol/vol)
S
HN
S
• 20 biomolecule additives investigated
• No significant influence on yield
H
3au, 72%
• Overall good biomolecule preservation
Fig. 5 | improvement of the photocatalytic system based on mechanistic investigation, access to sulfoxides and sulfones by stepwise oxidation
of thioethers and biocompatibility screening of the disulfide–ene reaction. a, The previously gained mechanistic understanding has been used for
improvement of the catalytic system. By switching to transition-metal-free alloxazine photocatalyst 10, the reaction rate could be significantly improved.
b, The improved catalytic system allows the successful hydromethylthiolation of alkenes, which were low-yielding or not reactive using [Ir–F] as
photosensitizer ([Ir–F] yields are stated in parentheses). Reaction conditions: alkene (0.3ꢀmmol, 1.0ꢀequiv.), disulfide (0.6ꢀmmol, 2.0ꢀequiv.), alloxazine
(0.015ꢀmmol, 5.0ꢀmol%), DCE (0.1ꢀM), λmaxꢀ=ꢀ400ꢀnm, r.t., 16ꢀh. Data are reported as percent isolated yield except 3ar and 3as, which are reported as 1H
NMR yield using CH2Br2 as internal standard. c, The obtained methylthioether products can be stepwise oxidized to pharmaceutically important sulfoxide or
sulfone functionalities. d, The biocompatibility of the disulfide–ene reaction was investigated by additive-based screening of biomolecules and was performed
in cell lysate (Aꢀ=ꢀbiomolecule identified, no degradation observed; Bꢀ=ꢀbiomolecule identified, some degradation observed; Cꢀ=ꢀbiomolecule identified,
significant degradation observed; Dꢀ=ꢀno biomolecule identified, complete degradation observed, NDꢀ=ꢀnot determined, a conclusion is not possible).
Reduced form of L-glutathione was used. e, Qualitative analysis of biomolecules by UPLC–MS or gel electrophoresis after irradiation revealed that, in most
cases, no degradation occurred, as depicted by the SDS–polyacrylamide gel, indicating preservation of BSA and RNase A (100ꢀµMa and 10ꢀµMb) under the
reaction conditions. f, Various alkenes were also successfully hydromethylthiolated in cell lysate. Reaction profile (a), hydrothiolation reactions (b), follow-up
reactions (c) and biocompatibility screening experiments (d and f) were performed once. SDS–PAGE (e) was repeated twice with similar results.
Hydrogen atom abstraction in the α-position to the sulfur atom in different deuterated polar aprotic or nonpolar solvents, no deu-
of 2 by carbon-centred radical 7 could lead to the formation of the terium incorporation within the product structure was observed,
carbon-centred disulfide radical 8. This radical consecutively reacts suggesting that the hydrogen atom abstraction is exclusively taking
with a methylthiyl radical or another dimethyl disulfide (2) mol- place from the disulfide (Fig. 4e). In contrast, if the reaction was
ecule to yield trisulfide 9, which was isolated as a side product dur- performed in polar protic solvents, almost full deuterium incorpo-
ing the optimization process. When the reaction was performed ration in the product molecule was detected. Under these reaction
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
conditions, the polar protic solvent is the hydrogen atom source
(Supplementary Section 4.9).
Motivated by the importance of methylthioether scaffolds in
biological processes we investigated a biocompatible version of the
As we succeeded in monitoring the TTEnT from the [Ir–F] disulfide–ene reaction potentially suitable for applications in bio-
photocatalyst to dimethyl disulfide (2), we turned our attention to logical systems, such as labelling of biomolecules48. Using water
probe the final step in the disulfide–ene reaction, namely the reac- with a physiologically compatible Tris-HCl buffer (0.2M, pH7.4)
tion of 3*2 with 1-octene. In this context, the 400 and 440nm signa- as solvent, we probed the biocompatibility of the disulfide–ene
ture of 3*2 enabled the reaction to be followed spectroscopically. In reaction of carvone (1a) in the presence of 20 biomolecules in an
ns-TAS experiments with 1-octene in a concentration range from additive-based screening approach (Fig. 5d)39,40. Amino acids, sac-
3*
1×10−7 to 5×10−3 M, the concentration-dependent 2 decay gave charides, nucleosides, single-stranded DNA, RNA (short RNA and
rise to kreact =(1.38 0.07)×108 M−1 s−1. The final thioether prod- total RNA) and human cell lysate were among the tested biomol-
uct is, however, spectroscopically invisible in the wavelength range ecules. Gratifyingly, the yield of 3a was not affected by the presence
from 400 to 1600nm.
of any of the biomolecules. In most cases, ultra performance liquid
Based on screening, ns-TAS, phosphorescence lifetime quench- chromatography–mass spectrometry (UPLC–MS) and gel elec-
ing and radical scrambling experiments, the following mechanism trophoresis analyses documented that the biomolecules were not
is proposed (Fig. 4f). After excitation of [Ir–F], a TTEnT activation subject to degradation under the reaction conditions (Fig. 5e and
of dimethyl disulfide (2) leads to homolytic S–S bond cleavage, gen- Supplementary Section 8.2). In addition, the disulfide–ene reaction
erating corresponding thiyl radical 6. Similar to the classical thiol– between dimethyl disulfide (2) and a variety of olefins was tested in
ene reaction, the chemoselective anti-Markovnikov addition into the presence of human cell lysate, containing various endogenous
the most electron-rich alkene occurs in an irreversible and quick biomolecules (Fig. 5f). Notably, the desired hydromethylthiolated
fashion and gives rise to carbon-centred radical 7. If aliphatic disul- products were formed. Our results indicate that this transformation
fides and polar aprotic or nonpolar solvents are used, 7 abstracts a might be suitable for biological systems, and applications are cur-
hydrogen atom from another disulfide molecule 2, resulting in the rently being investigated in our laboratories.
formation of product 3 and trisulfide 9. In contrast, if aryl disul-
Due to its high atom economy, complete chemo- and regiose-
fides in acetone as solvent or aliphatic disulfides in polar protic sol- lectivity, biocompatibility and high synthetic yields, this transfor-
vents are reacted, the hydrogen atom abstraction event takes place mation might be classified as a click reaction, in analogy to the
between carbon-centred radical 7 and the solvent.
photoinitiated thiol–ene reaction49. Remarkably, the TTEnT-enabled
disulfide–ene reaction proceeds chemoselectively in the presence of
Applications and outlook. The mechanistic analysis revealed the thiols, providing ways and means to a complementary strategy of
occurrence of a Dexter-type TTEnT process, where physical contact constructing carbon–sulfur bonds (Supplementary Section 4.12).
between energy acceptor and energy donor is required. We there-
fore aimed for the improvement of our catalytic system by decreas- Conclusion
ing the steric bulkiness of the photosensitizer, on the one hand, and In conclusion, we have discovered a photosensitized disulfide–
increasing its triplet excited-state energy, on the other. Due to its ene click reaction that enables the chemo- and regioselective con-
planar shape and its triplet excited-state energy of 65.1kcalmol−1, struction of alkyl- and arylthioethers under remarkably mild and
organic alloxazine photocatalyst 10 emerged as an ideal candidate45. biocompatible reaction conditions. Our protocol allows the chemo-
The irradiation wavelength for this catalytic system was adjusted to and regioselective incorporation of methylthioether scaffolds into
400nm. The transition-metal-free alloxazine-derived photosensi- highly complex substrates, as proven by a broad substrate scope
tizer might allow transformations not suitable with the [Ir–F] cata- and the functionalization of bioconjugates or drug derivatives. The
lytic system due to different photo- and electrochemical properties. detailed study of the reaction mechanism, with special focus on the
We validated our hypothesis by running the benchmark reaction TTEnT sensitization process via TAS, has led to the development
between carvone (1a) and dimethyl disulfide (2) under reaction of an improved and more efficient transition-metal-free catalytic
conditions optimized for the [Ir–F]-sensitized protocol, but using system. We are convinced that the establishment of visible-light-
alloxazine 10. Importantly, the desired product was obtained in 62% mediated energy transfer activation, based on careful mechanistic
yield, proving the suitability of the designed catalytic system. After analysis and rational design, will open up a new field with versatile
optimizing the reaction conditions, we recorded the reaction pro- applications in organic synthesis, allowing hardly accessible bond
file of this transition-metal-free catalytic system and compared it to (dis)connections.
the [Ir–F]-based system. Due to the increase in triplet excited-state
energy and the decrease in steric bulkiness of the photosensitizer, Data availability. The data supporting the findings of this study are
TTEnT is favoured, and, in turn, the reaction is completed within available within the paper and its Supplementary Information, or
10min rather than 90min (Fig. 5a). All of the aforementioned facts from the corresponding authors upon reasonable request.
support our mechanistic hypothesis.
Received: 21 February 2018; Accepted: 29 May 2018;
Published: xx xx xxxx
With our optimized catalytic system in hand, we focused on
the hydromethylthiolation of substrates, which turned out to be
low-yielding or unreactive in the [Ir–F]-based catalysis (Fig. 5b).
Satisfyingly, the yields of some substrates were increased substan-
tially using our improved protocol, for example, hydromethylthi-
olation of the biotin derivative 3au was increased by 70% to 91%
isolated yield. Moreover, these new conditions enabled the hydro-
methylthiolation of substrates that could not be transformed using
the [Ir–F] photocatalyst. For example, methylthiolated phospho-
References
1. Trost, B. M. Selectivity: a key to synthetic efciency. Science 219,
245–250 (1983).
2. Wender, P. A. & Miller, B. L. Synthesis at the molecular frontier. Nature 460,
197–201 (2009).
3. Takahashi, N. et al. Reactive sulfur species regulate tRNA methythiolation
and contribute to insulin secretion. Nucleic Acids Res. 45, 435–445 (2017).
nate 3aq and the mycophenolic acid derivative 3at were obtained in 4. Kowalak, J. A. & Walsh, K. A. β-methylthio-aspartic acid: identifcation of
a novel posttranslational modifcation in ribosomal protein S12 from
49% and 65% yield, respectively.
Escherichia coli. Protein Sci. 5, 1625–1632 (1996).
Any of the alkylthioether products are oxidizable in a stepwise
5. Forouhar, F. et al. Two Fe–S cluster catalyse sulfur insertion by radical-SAM
and high-yield manner to the corresponding sulfoxides (11) or sul-
methylthiotransferases. Nat. Chem. Biol. 9, 333–338 (2013).
fones (12). Both are versatile and important structural motifs in
6. Stubbe, J. & van der Donk, W. A. Protein radicals in enzyme catalysis.
diverse pharmaceuticals or agrochemicals (Fig. 5c)46,47
.
Chem. Rev. 98, 705–762 (1998).
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

NATuRe CHeMISTRy
Articles
7. Dunbar, K. L., Scharf, D. H., Litomska, A. & Hertweck, C. Enzymatic carbon–
sulfur bond formation in natural product biosynthesis. Chem. Rev. 117,
5521–5577 (2017).
35. Welin, E. R., Le, C., Arias-Rotondo, D. M., McCusker, J. K. & MacMillan, D.
W. C. Photosensitized, energy transfer-mediated organometallic catalysis
through electronically excited nickel(II). Science 355, 380–385 (2017).
36. Hopkinson, M. N., Gómez-Suárez, A., Teders, M., Sahoo, B. & Glorius, F.
Accelerated discovery in photocatalysis using a mechanism-based screening
method. Angew. Chem. Int. Ed. 55, 4361–4366 (2016).
37. Teders, M., Gómez-Suárez, A., Pitzer, L., Hopkinson, M. N. & Glorius, F.
Diverse visible-light-promoted functionalizations of benzotriazoles inspired
by mechanism-based luminescence screening. Angew. Chem. Int. Ed. 56,
902–906 (2017).
38. Benati, L., Montevecchi, P. C. & Spagnolo, P. Free-radical reactions of
benzenethiol and diphenyl disulphide with alkynes. Chemical reactivity of
intermediate 2-(phenylthio)vinyl radicals. J. Chem. Soc. Perkin Trans. 1,
2103–2109 (1991).
39. Collins, K. D. & Glorius, F. A robustness screen for the rapid assessment of
chemical reactions. Nat. Chem. 5, 597–601 (2013).
8. Kolberg, M., Strand, K. R., Graf, P. & Andersson, K. K. Structure, function,
and mechanism of ribonucleotide reductases. Biochim. Biophys. Acta 1699,
1–34 (2004).
9. Fraústo da Silva, J. J. R. & Williams, R. J. P. Te Biological Chemistry of the
Elements (Oxford Univ. Press: New York, NY, 2011).
10. Damani, L. A. Sulphur-Containing Drugs and Related Organic Compounds
(Wiley: Chichester, 1989).
11. Dénès, F., Pichowicz, M., Povie, G. & Renaud, P. Tiyl radicals in organic
synthesis. Chem. Rev. 114, 2587–2693 (2014).
12. Posner, T. Beiträge zur Kenntniss der ungesättigten Verbindungen. II. Ueber
die Addition von Mercaptanen an ungesättigte Kohlenwasserstofe. Ber. Dtsch.
Chem. Ges. 38, 646–657 (1905).
13. Hoyle, C. E. & Bowman, C. N. Tiol–ene click chemistry. Angew. Chem. Int.
Ed. 49, 1540–1573 (2010).
40. Gensch, T., Teders, M. & Glorius, F. Approach to comparing the functional
group tolerance of reactions. J. Org. Chem. 82, 9154–9159 (2017).
41. Snyder, J. J., Tise, F. P., Davis, R. D. & Kropp, P. J. Photochemistry of alkenes. 7.
E–Z isomerization of alkenes sensitized with benzene and derivatives.
J. Org. Chem. 46, 3609–3611 (1981).
14. Stenzel, M. H. Bioconjugation using thiols: old chemistry rediscovered
to connect polymers with nature’s building blocks. ACS Macro Lett. 2,
14–18 (2013).
15. van Dijk, M., Rijkers, D. T. S., Liskamp, R. M. J., van Nostrum, C. F. &
Hennink, W. E. Synthesis and applications of biomedical and pharmaceutical
polymers via click chemistry methodologies. Bioconjug. Chem. 20,
2001–2016 (2009).
42. Lowry, M. S. et al. Single-layer electroluminescent devices and photoinduced
hydrogen production from an ionic iridium(III) complex. Chem. Mater. 17,
5712–5719 (2005).
16. Huynh, V. T., Chen, G., de Souza, P. & Stenzel, M. H. Tiol–yne and
thiol–ene ‘click’ chemistry as a tool for a variety of platinum drug delivery
carriers, from statistical copolymers to crosslinked micelles.
Biomacromolecules 12, 1738–1751 (2011).
17. Lowe, A. B. Tiol–ene ‘click’ reactions and recent applications in polymer and
materials synthesis. Polym. Chem. 1, 17–36 (2010).
18. Tyson, E. L., Ament, M. S. & Yoon, T. P. Transition metal photoredox
catalysis of radical thiol–ene reactions. J. Org. Chem. 78, 2046–2050 (2013).
19. Tyson, E. L., Niemeyer, Z. L. & Yoon, T. P. Redox mediators in visible light
photocatalysis: photocatalytic radical thiol–ene additions. J. Org. Chem. 79,
1427–1436 (2014).
20. Keylor, M. H., Park, J. E., Wallentin, C.-J. & Stephenson, C. R. J.
Photocatalytic initiation of thiol–ene reactions: synthesis of thiomorpholin-3-
ones. Tetrahedron 70, 4264–4269 (2014).
21. Chen, Y., Kamlet, A. S., Steinman, J. B. & Liu, D. R. A biomolecule-
compatible visible-light-induced azide reduction from a DNA-encoded
reaction-discovery system. Nat. Chem. 3, 146–153 (2011).
22. Huang, H., Zhang, G., Gong, L., Zhang, S. & Chen, Y. Visible-light-induced
chemoselective deboronative alkynylation under biomolecule-compatible
conditions. J. Am. Chem. Soc. 136, 2280–2283 (2014).
23. Nair, D. P. et al. Te thiol–Michael addition click reaction: a powerful and
widely used tool in materials chemistry. Chem. Mater. 26, 724–744 (2014).
24. Luo, Y.-R. Handbook of Bond Dissociation Energies in Organic Compounds
(CRC Press: Boca Raton, FL, 2003).
25. Zhao, J., Wu, W., Sun, J. & Guo, S. Triplet photosensitizers: from molecular
design to applications. Chem. Soc. Rev. 42, 5323–5351 (2013).
26. Atta, M. et al. Te methylthiolation reaction mediated by the radical-SAM
enzymes. Biochim. Biophys. Acta 1824, 1223–1230 (2012).
27. Stillwell, W. G. Methylthiolation: a new pathway of drug metabolism.
Trends Pharmacol. Sci. 2, 250–252 (1981).
43. Murov, S. L., Carmichael, I. & Hug, G. L. Handbook of Photochemistry 2nd
edition (Marcel Dekker, New York, 1993).
44. Guldi, D. M., Neta, P. & Asmus, K.-D. Electron-transfer reactions between C60
and radical ions of metaloporphyrins and arenes. J. Phys. Chem. 98,
4617–4621 (1994).
45. Mojr, V. et al. Tailoring favins for visible light photocatalysis: organocatalytic
[2+2] cycloadditions mediated by a favin derivative and visible light. Chem.
Commun. 51, 12036–12039 (2015).
46. Feng, M., Tang, B., Liang, S. H. & Jiang, X. Sulfur containing scafolds in
drugs: synthesis and application in medicinal chemistry. Curr. Top. Med.
Chem. 16, 1200–1216 (2016).
47. Devendar, P. & Yang, G. F. Sulfur-containing agrochemicals. Top. Curr. Chem.
375, 82 (2017).
48. Shieh, P. & Bertozzi, C. R. Design strategies for bioorthogonal smart probes.
Org. Biomol. Chem. 12, 9307–9320 (2014).
49. Kolb, H. C., Finn, M. G. & Sharpless, K. B. Click chemistry: diverse
chemical function from a few good reactions. Angew. Chem. Int. Ed. 40,
2004–2021 (2001).
Acknowledgements
The authors thank K. Gottschalk, L. Roling, S. Hüwel, W. Dörner and S. Wulff for
experimental and technical assistance and R. Honeker, L. Candish and Z. Nairoukh
for helpful discussions (all WWU Münster). This work was supported by the Deutsche
Forschungsgemeinschaft (Leibniz Award to F.G. and RE2796/6-1 to A.R.) and by the
Fonds der Chemischen Industrie (doctoral fellowship to L.A. and Dozentenpreis to
A.R.). M.T. thanks SusChemSys 2.0 for general support.
Author contributions
M.T., F.S.-K., A.G.-S., R.K. and F.G. designed, performed and analysed the catalytic
and mechanistic experiments. C.H., A.K. and D.G. designed, performed and analysed
transient absorption data and related spectroscopic mechanism studies. L.A. and M.T.
designed and performed the biocompatibility screening experiments. M.T., C.H., L.A.,
F.S.-K., A.R., D.G. and F.G. prepared the manuscript, with contributions
from all authors.
28. Ciamician, G. Te photochemistry of the future. Science 36, 385–394 (1912).
29. Prier, C. K., Rankic, D. A. & MacMillan, D. W. C. Visible light photoredox
catalysis with transition metal complexes: applications in organic synthesis.
Chem. Rev. 113, 5322–5363 (2013).
30. Twilton, J. et al. Te merger of transition metal and photocatalysis. Nat. Rev.
Chem. 1, 0052 (2017).
31. Skubi, K. L., Blum, T. R. & Yoon, T. P. Dual catalysis strategies in
photochemical synthesis. Chem. Rev. 116, 10035–10074 (2016).
32. Metternich, J. B. & Gilmour, R. Photocatalytic E→Z isomerization of alkenes.
Synlett 27, 2541–2552 (2016).
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-018-0102-z.
33. Lu, Z. & Yoon, T. P. Visible light photocatalysis of [2+2] styrene cycloadditions
by energy transfer. Angew. Chem. Int. Ed. 51, 10329–10332 (2012).
34. Blum, T. R., Miller, Z. D., Bates, D. M., Guzei, I. A. & Yoon, T. P.
Enantioselective photochemistry through Lewis acid-catalyzed triplet energy
transfer. Science 354, 1391–1395 (2016).
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to D.G. or F.G.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
NATuRe CHeMiSTRy | www.nature.com/naturechemistry

Corresponding author(s):
Dirk Guldi and Frank Glorius
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistical parameters
When statistical analyses are reported, confirm that the following items are present in the relevant location (e.g. figure legend, table legend, main
text, or Methods section).
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
An indication of whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistics including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND
variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Clearly defined error bars
State explicitly what error bars represent (e.g. SD, SE, CI)
Our web collection on statistics for biologists may be useful.
Software and code
Policy information about availability of computer code
No software was used for data collection.
Data collection
Data analysis
Regression analysis (linear regression) and processing of reaction profiles, luminescence and UV/vis spectra was performed using
Microsoft Excel 2013 or Origin 2016. UPLC-MS analysis was performed utilizing Bruker Compass DataAnalysis 4.4 for data analysis.
Transient absorption spectroscopy data was evaluated using the TIMP based GloTarAn program. GC-MS spectra were analysed using the
Agilent MSD Productivity ChemStation software. Electrochemistry data analysis was performed using Metrohm Autolab - Nova 1.10
software.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers
upon request. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
1

Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
All data is made available in the Supplementary Information. Raw data is available from the corresponding authors upon reasonable request.
Field-specific reporting
Please select the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences
Behavioural & social sciences
Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/authors/policies/ReportingSummary-flat.pdf
Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
No statistical methods were used to predetermine sample size. Verification of the reaction methodology was performed exhaustively by
scope preparation and by the performance of an additive-based robustness screen.
Sample size
No data were excluded from the analysis.
Data exclusions
Replication
Newly reported reactions were repeated and / or performed by several of the authors. All attempts at replication were succesful.
Spectroscopic data could be reproduced with similar results. Furthermore, the use of diverse spectroscopic techniques confirmed the
observed kinetic rate constants. During the whole study, no replication problems were observed.
In our study, an energy transfer enabled hydroalkylthiolation of diverse alkenes and alkynes is described. Consecutively, as part of the design,
mechanistic analysis, scope preparation and biocompatibility assesment, radomization is not relevant to this study (methodology paper).
Randomization
Blinding
The study describes the design and the mechanistic analysis of a novel hydroalkylthiolation methodology. Only in vitro samples were tested.
As part of the study, no blinding was necessary.
Reporting for specific materials, systems and methods
Materials & experimental systems
n/a Involved in the study
Unique biological materials
Antibodies
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
Eukaryotic cell lines
MRI-based neuroimaging
Palaeontology
Animals and other organisms
Human research participants
Eukaryotic cell lines
Policy information about cell lines
Cell lines were obtained from Sigma (HeLa).
Cell line source(s)
Cells lines were purchased and not authenticated.
Authentication
All cell lines used were tested negative for mycoplasma contamination.
Mycoplasma contamination
2

Commonly misidentified lines
(See ICLAC register)
No commonly misidentified cell lines were used.
3