Catalytic adaptive recognition of thiol (SH) and selenol (SeH) groups toward synthesis of functionalized vinyl monomers

Article abstract of DOI:10.1021/ja210596w

An unprecedented sustainable procedure was developed to produce functionalized vinyl monomers H₂C=C(R)(FG) starting from a mixture of sulfur and selenium compounds as a functional group donor (FG = S or Se). The reaction serves as a model for e

Full text of DOI:10.1021/ja210596w

Article
pubs.acs.org/JACS
Catalytic Adaptive Recognition of Thiol (SH) and Selenol (SeH)
Groups Toward Synthesis of Functionalized Vinyl Monomers
Valentine P. Ananikov,*,† Nikolay V. Orlov,† Sergey S. Zalesskiy,† Irina P. Beletskaya,‡
Victor N. Khrustalev,§ Keiji Morokuma,∥ and Djamaladdin G. Musaev*,∥
†Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospect 47, Moscow, 119991, Russia
‡Lomonosov Moscow State University, Chemistry Department, Vorob’evy gory, Moscow, 119899, Russia
§Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Moscow, Russia
∥Cherry L. Emerson Center for Scientific Computation and Department of Chemistry, Emory University, Atlanta, Georgia 30322,
United States
S
* Supporting Information
ABSTRACT: An unprecedented sustainable procedure was developed to
produce functionalized vinyl monomers H2CC(R)(FG) starting from a
mixture of sulfur and selenium compounds as a functional group donor
(FG = S or Se). The reaction serves as a model for efficient utilization of
natural resources of sulfur feedstock in oil and technological sources of
sulfur/selenium. The catalytic system is reported with amazing ability to
recognize SH/SeH groups in the mixture and selectively incorporate them
into valuable organic products via wastes-free atom-economic reaction with
alkynes (HCCR). Formation of catalyst active site and the mechanism of
the catalytic reaction were revealed by joint experimental and theoretical study. The difference in reactivity of μ1- and μ2-type
chalcogen atoms attached to the metal was established and was shown to play the key role in the action of palladium catalyst. An
approach to solve a challenging problem of dynamically changed reaction mixture was demonstrated using adaptive tuning of the
catalyst. The origins of the adaptive tuning effect were investigated at molecular level and were found to be governed by the
nature of metal−chalcogen bond.
point of view of organic synthesis), possesses high energy
consumption, and generates a considerable amount of wastes. A
sustainable alternative of this multistep and inefficient
(energetically and environmentally) process is the direct use
of the desired thiols from the crude oil in organic synthesis.
Although the idea is well appreciated, its practical realization
requires solving several challenging tasks including the
separation of individual (desirable for the given organic
synthesis) thiols from a mixture of aromatic and aliphatic
thiols, sulfides, disulfides, thiophenes, etc. that exist in crude oil.
The selective recognition, capture, and transformation of
selenium analogues of thiols (i.e., selenols RSeH and their
derivatives) are another challenging task for the environment.
Indeed, in spite of a much lower content of selenols RSeH and
their derivatives, the handling of these species requires
considerable care due to their highly toxic properties and
dangerous influence on the environment. Material science
applications and microelectronics actively utilize various
selenium and sulfur compounds (RZH), thus driving selective
capture and removal of chalcogen derivatives in RSeH/RSH
mixtures is of the utmost importance.
1. INTRODUCTION
The capture, selective recognition and use of thiols, selenols,
and their derivatives in organic synthesis are monumental
scientific challenges with huge environmental and economical
advantages. These vital problems continue to be a focus of
scientific and technological communities. Indeed, organosulfur
compounds are the most abundant functionalized hydro-
carbons available in nature and present as contaminants in
crude oil. For environmental protection (to reduce critical SO2
emissions) and for cost-efficient industrial applications (sulfur
species are dangerous poisons of metal catalysts) sulfur removal
from crude oil must be performed prior to manufacturing any
gasoline, petrol, kerosene, or fuel products. Currently, the
petroleum refinery industry uses the rigorously developed
industrial hydrodesulfurization process for capture and removal
of a mixture of sulfur species from crude oil. This process
converts thiols to hydrogen sulfide, which is subsequently
transformed to elemental sulfur (Scheme 1).1−4 To highlight
the scale, as large as 68,000,000 t of sulfur were produced
worldwide in 2010 with the vast majority coming from
hydrocarbon processing plants.5
Elemental sulfur serves as a starting reagent for organic
synthesis through preparation of single thiols and their
subsequent transformation to organic products (Scheme 1).
However, such an overall process is highly inefficient (from the
Received: November 14, 2011
Published: April 4, 2012
© 2012 American Chemical Society
6637
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Scheme 1. Hydrodesulfurization (HDS) of Thiols in Crude Oil, Preparation of Elemental Sulfur, Conversion to a Single Thiol,
and Utilization in Organic Synthesis
Scheme 2. Direct Involvement of Thiols and Selenols in Organic Synthesis via Addition to Alkynes
Consequently, the goal of our research is to develop a
catalyst for selective recognition and capture of the RZH
derivatives (Z = S or Se) and their use as reagents for the
preparation of important organic products. For this purpose we
have chosen the addition of RZH (Z = S and Se) to the triple
bond of acetylenic hydrocarbons (Scheme 2).6 This addition
reaction has a Green Chemistry advantage, as it is 100% atom
efficient and does not produce wastes.7,8 It is important to
mention that simple metal salts, particularly Pd(OAc)2, can be
used as catalyst precursors to render the catalytic trans-
formation in a selective manner.9 The proposed reaction
utilizes easily available acetylenes as starting material and leads
to very useful functionalized vinyl derivatives. Among the
possible products 1−7, compound 1 is of substantial interest as
a functionalized monomer with significant opportunities as a
reagent in organic synthesis, including material and polymer
sciences.10−13 Important to note is that compound 1 is virtually
nontoxic or significantly less toxic compared to RZH (Z = S
and Se) species, and it does not have an awful odor typical for
RSeH and RSH species. This product is not sensitive to air in
contrast to RSeH (rapidly oxidized on contact with air) and
RSH (slowly oxidized and reacts on air to generate free
radicals). Thus, compound 1 is much easier to store and handle
and it provides excellent opportunities for industry and organic
chemistry.
The transition-metal-catalyzed atom-economic carbon−het-
eroatom bond formation via the addition reaction to acetylenic
hydrocarbons has been the subject of recent studies.6 Several
homogeneous and heterogeneous catalytic systems were
investigated for accomplishing the addition of thiols and
selenols to multiple carbon−carbon bonds.6,9,14 It was shown
that the metal-catalyzed procedure led to the formation of
Markovnikov-type product 1, while a mixture of anti-
Markovnikov derivatives 2 and 3 was formed in the reaction
without a metal catalyst. Formation of byproduct 4 from the
alkyne and two molecules of thiols was accompanied by
evolution of molecular hydrogen, and in situ hydrogenation led
to byproduct 5 (Scheme 2).15,16 Compounds 6 and 7 also can
be readily formed after reaction of the second molecule of RZH
with alkenes 1−3.17 Overall, several side reactions leading to
byproducts 2−7 may take place in the studied system. Thus,
achievement of high selectivity in the addition reaction (i.e.,
exclusive formation of 1) is a challenging task by itself beyond
selective capture of RSH/RSeH from the mixture.
In the present contribution we report a breakthrough
heterogeneous Pd-based catalytic system with unprecedented
ability for selective synthesis of product 1 starting with a
mixture of reagents, RSH/RSeH, and acetylenic hydrocarbons.
Importantly, two key factors, (i) selective capture of a reagent
from the mixture and (ii) highly selective transformation of
each reagent to vinyl monomer 1, were achieved within a single
catalyst. We have developed a procedure that shows high
efficiency and selectivity toward RSH and RSeH groups and
tolerance to oxygen species and water (Z = O), which may be
present in the initial reagents. Furthermore, for practical
purposes, the fundamental principles governing these reactions
that are vital for designing more efficient catalytic processes
were elaborated (experimentally and computationally) by
investigating the mechanisms and controlling factors of the
reaction of RZH (chalcogen derivatives, where Z = S and Se)
with alkynes catalyzed by the newly designed Pd-based catalyst
(Scheme 2).
2. RESULTS AND DISCUSSION
2.1. Catalytic Addition Reaction in Multicomponent
Mixtures. Addition of catalytic amounts of Pd(OAc)2 as a
catalyst precursor to a mixture of chalcogen reagents RZH (Z =
S and Se) and acetylenic hydrocarbons generates an active
catalyst, which promotes the formation of product 1. NMR
monitoring of the three-component mixture containing phenyl-
selenol, benzenethiol as aromatic thiol, and cyclohexyl thiol as
aliphatic thiol (PhSeH/PhSH/CySH = 1:1:1) and 2-methyl-3-
butyn-2-ol (3 equiv) as alkyne has revealed outstanding
selectivity. At the beginning, the reaction proceeds exclusively
as addition of PhSeH to the alkyne leading to product 1a
(Figure 1a). None of other components of the mixture (i.e.,
PhSH and CySH) were involved in the addition reaction until
complete utilization of PhSeH (Figure 1b). Furthermore,
compound 1a was the only product formed in the reaction
without any other byproduct 2−7.
After complete utilization of PhSeH and formation of
product 1a (40 min), benzenethiol PhSH was selected next
6638
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Figure 1. (a) Reaction of PhSeH, PhSH, and CySH with 2-methyl-3-butyn-2-ol, (b) NMR monitoring of the reaction starting with three-component
PhSeH/PhSH/CySH mixture, (c) NMR monitoring of the reaction starting with binary PhSH/CySH mixture and with PhSeH added after 2 h of the
reaction.
by the catalyst for addition to the alkyne that led to synthesis of
product 1b (Figure 1b). The latter reaction was completed in 4
h. At the third stage, which started after complete usage of
PhSH, the CySH was picked-up for reaction with the alkyne
that led to formation of product 1c (Figure 1b). At the second
and third stages of the transformation, again, compounds 1b
and 1c were the only products; no other byproduct, such as 2−
7, were obtained. Thus, the developed heterogeneous catalyst
exhibits exceptional selectivity to the three-component mixture
of PhSeH/PhSH/CySH = 1:1:1, and exclusively promotes the
formation of Markovnikov-type products 1a, 1b and 1c,
respectively. No other byproduct, including mixture of anti-
Markovnikov products, were obtained.
Under the same conditions the catalytic addition reaction
was also carried out for the binary mixtures PhSeH/PhSH,
PhSeH/CySH, and PhSH/CySH. In all three cases the same
order in the reactivity (and capture) was observed: PhSeH >
PhSH > CySH. It is essential to note that the order of
chalcogen species removal from the mixture matches their
6639
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
a
Table 1. Temperature and Yields for Reaction of Different Alkynes with PhSeH/PhSH/CySH Mixture
a
n
R′ = C(Me)2OH, CH2OMe, C4H9, CH(OH)CH3 for entries 1−4, respectively (Figure 1a).
toxicity; the more toxic is the component, the earlier the
removal from the mixture takes place.
In the above-presented experiments we used the equimolar
mixtures of chalcogen derivatives. In order to determine the
minimal amount of a mixture component that can be selectively
captured on a large background concentration of another
reactive species, we performed dynamic range tests. These tests
have shown that phenylselenol was completely extracted from
the PhSeH:PhSH binary mixture that led to the quantitative
yield of 1a in 2.0−2.0 × 10−3 M concentration range of PhSeH
and under constant background concentration of PhSH = 2.0
M. Even at 4 orders of magnitude difference in the
concentrations, CPhSeH = 2.0 × 10−4 M and CPhSH = 2.0 M,
the minor component of PhSeH addition was fully detected by
NMR after the reaction. This gives an excellent dynamic range
of at least 1:10000 for the developed catalytic system. The
studies with further decrease in the concentration of PhSeH
were not feasible because the sensitivity limit of the used
analytic method was already reached.
2.3. The Nature of the Catalytic Species. Our extensive
analysis of the nature of active catalyst generated after addition
of Pd(OAc)2 as a catalyst precursor to a mixture of chalcogen
reagents RZH (Z = S and Se) and acetylenic hydrocarbons
revealed heterogeneous nature of the studied system. Insoluble
metal-containing particles have been formed shortly after
reacting of Pd(OAc)2 with thiols or selenol. Filtering out the
metal-containing solids resulted in complete loss of the catalytic
activity, indicating a very minor (if any) metal-containing
species in solution.
The isolated catalyst was characterized using field-emission
scanning electron microscopy technique (FE-SEM). Approx-
imately round-shaped particles with sizes in the range of 30−60
nm were observed (Figure 2). The particles did not
agglomerate into the range of microscale sizes, and the edges
between the individual nanoparticles were clearly resolved
(Figure 2). The nanoparticles were self-assembled under simple
conditions in solution upon contact of Pd(OAc)2 with the
reagents: alkyne and RZH. Neither stabilizers/surfactants nor
any dedicated support were required to keep the nanosized
structural organization of the catalyst. Thus, the system
demonstrated outstanding practical advantages for easy and
convenient in situ catalyst formation without the need of
complicated and expensive catalyst preparation procedures
utilized for making supported heterogeneous catalysts.
Microanalysis of the precipitate formed at the first stage of
reaction in the three-component system PhSeH/PhSH/CySH
(Figure 1b) showed the initial catalyst composition to be
[Pd(SePh)2]n (entry 1, Table 2): no sulfur in the active catalyst
composition was detected. It should be pointed out that the
microanalysis gave overall information about bulk catalyst (both
core and surface). Therefore, we performed a separate analysis
The above presented results clearly demonstrate that,
although the reagents PhSeH, PhSH, and CySH were presented
as a mixture, the products 1a, 1b, and 1c, respectively, can be
obtained in a pure form. Comparison of experimental
conditions needed for full conversion, 40 min at 80 °C for
PhSeH, 4 h at 100 °C for PhSH, and 8 h at 120 °C for CySH,
clearly emphasizes high potential of the used Pd-based catalyst
to distinguish the reagents of similar nature.
The scope of the reaction was verified with different alkynes
and in all studied cases the stepwise reaction with excellent
selectivity was observed (Table 1). Only one of the
components of PhSeH/PhSH/CySH mixture was reacted at a
time with alkyne. Similar kinetic curves were observed for all
studied alkynes independent of substituent (R′) induced
differences in reactivity. The final yields of products 1a−1c
were in the range of 96−99% (Table 1).
In summary, the developed catalytic system has shown (i)
exceptional reagent-specific selectivity toward chalcogen
species; (ii) excellent selectivity and high product yield in the
addition reaction, without formation of any byproduct; and (iii)
easy availability of catalyst precursor with smart recovery after
the reaction (heterogeneous catalysis).
2.2. Dynamically Changed Mixture of Reagents and
Dynamic Range Test. As a crucial test of the designed
catalytic system, a dynamically changed mixture of reagents was
investigated (Figure 1c). The reaction was initiated with the
PhSH/CySH (1:1) two-component mixture. As it was
expected, at first, the selective capture of PhSH by the catalyst
gave product 1b upon reaction with the alkyne. During this
process (after 2 h) the PhSeH was added to the reaction
mixture. As a result, immediately after the PhSeH addition, the
formation of product 1b was stopped and the formation of the
product 1a was started without any visible induction period.
After complete utilization of PhSeH (∼3 h) the use of PhSH by
the catalytic system has resumed and led to production of 1b.
Finally, after complete usage of PhSH (totally ∼9 h), the
remaining component of the mixture, CySH, was consumed to
produce 1c. As already noted, no other byproducts were
detected in the course of the reaction. The dynamic nature of
the catalytic system implies possible application in flow
conditions, where a varying content of the mixture may be
expected. No additional manipulations with the catalytic system
were needed; the catalyst was adaptively tuned upon interaction
with reaction media and showed a rapid response reflecting the
content of the mixture. Remarkably, the order in the reactivity
and chalcogen species removal from the mixture, i.e. PhSeH >
PhSH > CySH, was preserved.
6640
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
groups occurs. Intriguingly, these overall experiments indicate
the SPh coordination to the catalyst surface to be more
preferable than the SCy groups coordination.
2.4. Comparison with Noncatalytic Conditions and
with the Homogeneous Catalytic System. In order to
demonstrate superior properties of the newly developed
catalytic system we have compared its performance in the
capture and addition of chalcogen species to alkyne with those
reactions performed under no metal catalyst and homogeneous
metal catalyst conditions. Our experiments under no metal
catalyst conditions did not lead to the formation of desired
(industrially) product 1; instead, the formation of a mixture of
compounds 2 and 3 was observed. These results well agree with
those reported for addition of thiols and selenols to alkynes
under radical and nucleophilic conditions.18,19
Experiments under the homogeneous metal catalyst conditions
were performed by using an excess of phosphine ligand (below
called as a Pd/PPh3 catalyst), which prevented precipitation of
metal particles and kept the metal species in the solution. 31P
1
and H NMR study of the nature of the active catalyst under
the homogeneous metal catalyst conditions identified the
formation of [Pd(ZR)2(PPh3)2] complexes (Z = S, Se) in
agreement with previous studies.20 The reaction of PhSeH (or
PhSH) as an individual reagent with the alkyne, performed
under the Pd/PPh3 catalyst conditions, resulted in a mixture of
vinylic products containing one and two ZPh groups (1−4; Z =
S and Se). The results are even worse for the two-component
PhSeH/PhSH (1:1) mixture that reacted with the alkyne under
the same Pd/PPh3 catalyst. For this reaction, we have found a
multicomponent mixture of products containing vinylic
compounds with one and two chalcogen groups (1−4), as
well as the alkene 5. Furthermore, the formation of compound
4, containing both SePh and SPh groups, was also observed. In
other words, addition of thiols and selenols (and related
systems) to alkyne under the homogeneous metal catalyst
conditions led to the formation of a mixture of various
products without desired selectivity.
Thus, homogeneous and heterogeneous catalytic systems
reported in this article contain the same structural unit
Pd(ZR)2; however, the heterogeneous Pd catalyst shows
superiority in reactivity and selectivity compared to the
homogeneous catalyst.
Having the above presented results in our hands, we
elucidate below the mechanism and the key factors controlling
selective capture and addition of chalcogen species to alkyne
exclusively via the formation of Markovnikov-type products. As
a first step, we have studied the structure and stability of the
catalyst [Pd(ZR)2]n for Z = S and Se, R = Me, and n = 1−9 by
utilizing computational tools.
Figure 2. FE-SEM of the [Pd(ZR)2]n catalyst particles isolated from
the reaction mixture (50000× magnification; 100-nm scale bar is
shown).
of the catalyst surface by utilizing EDS-SEM technique and
confirmed the presence of the SePh groups on the surface
(entry 1, Table 2). Uniform distribution of Se on the surface
was detected by recording X-ray maps in the EDS-SEM study.
In contrast, the active catalyst isolated after 3 h of reaction,
i.e. after the consumption of all selenols in the PhSeH/PhSH/
CySH mixture [this stage of the reaction corresponds to the
PhSH capture and addition (Figure 1b)], contained 3−6% of
sulfur, and the amount of selenium decreased. These findings
indicate that some part of the SePh groups in the original
[Pd(SePh)2]n catalyst has been replaced by SPh groups. The
corresponding EDS-SEM study confirmed the presence of the
SPh groups on the surface (entry 2, Table 2). Thus, these
experiments suggest an exchange of SePh groups located at the
surface of the original [Pd(SePh)2]n catalyst with SPh groups.
At the third stage of the reaction that corresponds to the
selective CySH capture from the mixture and 1c formation,
further modification of the surface was observed, and
attachment of the SCy groups took place (entry 3, Table 2,
and Figure 1, b and c).
On the basis of the presented data one can conclude that, at
the first stage of the reaction in the three-component system
PhSeH/PhSH/CySH (Figure 1b), the [Pd(SePh)2]n (entry 1,
Table 2) is the active catalyst. At the second and third stages of
this reaction corresponding to the selective PhSH and CySH
capture from the mixture, respectively, the exchange of SePh
groups (located at the surface of the catalyst) with SPh and SCy
2.5. Computed Structure and Stability of the [Pd-
(ZR)2]n Clusters, for Z = S and Se, R = Me, and n = 1−9.
Reaction of two RZH molecules with the precursor of the
a
Table 2. Microanalysis and EDS-SEM Study of the Catalytic Particles at Various Stages
c
c
microanalysis, %
EDS-SEM, %
b
entry
step
catalyst
Pd
Se
S
Pd
Se
S
1
2
3
A
B
C
[Pd(SePh)2]n
25.1
26.2
26.3
37.3
30.1
29.8
0.0
3.5
3.9
25.0
26.1
26.0
38.2
28.3
27.7
0.0
5.9
6.5
[Pdn(SePh)2n−x(SPh)x]
[Pdn(SePh)2n−x(SR)x]
a
Calculated values (%) [Pd(SePh)2]n: Pd 25.4; Se 37.7; C 34.5; H 2.4; [Pd(SPh)2]n: Pd 32.8; S 19.7; C 44.4; H 3.10; [Pd(SCy)2]n: Pd 31.6; S 19.0;
b
c
C 42.8; H 6.6. A - reaction with PhSeH, B - reaction with PhSH, C - reaction with CySH (see Figure 1b). Estimated error = 0.5%.
6641
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Figure 3. Optimized structures of [Pd(SR)2]n chains at B3LYP level for n = 1, 3, 6, and 8 (atom colors are shown for the mononuclear derivative).22.
Scheme 3. Formation of Polynuclear Metal Species from Pd(II) Salt
catalyst, Pd(OAc)2, leads to formation of Pd(ZR)2 complex
with a bent structure (see Figure 3; optimized geometries of all
reported structures in this paper are presented in the
Supporting Information). The overall reaction Pd(OAc)2 +
2HZR → Pd(ZR)2 + AcOH is calculated to be exothermic by
−50.0 (−63.3) kcal/mol (throughout this paper, the energy
values given without parentheses are for Z = S, while the values
given in parentheses are for Z = Se) on free energy surface and
expected to proceed with a small energy barrier. Under
experimental conditions this step of the reaction, i.e. the
substitution of the OAc ligand by RZH and release of free
Table 3. Calculated Pd−Z Bond Energy for Chain and Cyclic
a,b
[Pd(ZR)2]n Palladium Complexes (in kcal/mol)
chain complexes
cyclic complexes
ΔH/m ΔG/m
n
ΔH/m
ΔG/m
2
3
4
5
6
7
8
9
−6.9 (−7.4)
−9.1 (−9.5)
−4.8 (−5.3)
−6.4 (−6.9)
−7.0 (−7.6)
−7.4 (−7.8)
−7.6 (−8.1)
−7.7 (−8.1)
−7.9 (−8.3)
−7.9 (−8.4)
−
−7.9 (−8.7)
−
−5.1 (−6.0)
−7.9 (−8.4)
−8.4 (−8.9)
−8.8 (−9.3)
−8.7 (−9.2)
−8.9 (−9.4)
−8.8 (−9.2)
−9.9 (−10.3)
−10.5 (−10.8)
−10.7 (−11.1)
−11.0 (−11.3)
−11.1 (−11.5)
−11.2 (−11.6)
−11.0 (−11.5)
−11.7 (−12.1)
−12.1 (−12.5)
−12.1 (−12.4)
−12.2 (−12.5)
−12.1 (−12.5)
1
AcOH, was confirmed by H and 13C NMR measurements.
If the Pd(ZR)2 complex is formed in the absence of
phosphine or other strong ligands,21 then a lone-pair of the
coordinated chalcogen atom reacts with the second Pd center
to form a dimeric species (see Scheme 3). For the [Pd(ZR)2]2
dimer, the rhombic structure with two bridging (μ2, we also call
them “core” ligands) and two terminal (μ1, we also call them
“surface” ligands) ZR ligands is found to be energetically the
most stable. The calculated free energy of dimerization is −28.5
(−31.8) kcal/mol. The averaged Pd−Z bond energy (ΔG/m)
[as an oligomerization energy divided by number of Pd−Z
bonds, m, which is equal to 4n for cyclic and 4n − 2 for linear
complexes] is found to be 6.9 (7.4) kcal/mol in [Pd(ZR)2]2
dimer (see Table 3). Subsequent addition of the Pd(ZR)2 units
is expected to lead to the formation of the [Pd(ZR)2]n species,
which at a certain point (i.e., at certain n) would become
insoluble and precipitate. Indeed, our experimental studies of
the reaction of Pd(OAc)2 with a mixture of chalcogen
derivatives confirmed that this process does take place under
the studied reaction conditions, and the resultant metal
a
The values without parentheses correspond to Z = S, the values in
b
parentheses correspond to Z = Se. Averaged ΔH/m and ΔG/m
values are given, where ΔH and ΔG were calculated for the n
Pd(ZMe)2 → [Pd(ZMe)2]n reaction and m is the number of Pd−Z
bonds (m = 4n for cyclic and m = 4n − 2 for linear complexes,
respectively).
particles of general formula [Pd(ZR)2]n (see Table 2) are
insoluble. Our computational studies show that the clusters
with more than two Pd(ZR)2 units may have at least two stable
structureschain and cyclic (see Scheme 3 for reactions; and
Figures 3, 4 for optimized structures).
As seen in Table 3, the chain structure is more stable than
the cyclic one, only for n = 3, where the calculated averaged
Pd−Z bond free energies (ΔG/m) are 6.4 (6.9) and 5.1 (6.0)
kcal/mol, respectively. However, for n ≥ 4, the calculated ΔG/
m value become larger for the cyclic structure than for the chain
one, indicating the cyclic structures to be energetically more
6642
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Figure 4. Optimized structures of cyclic [Pd(SR)2]n complexes at B3LYP level with n = 3−6.22.
Scheme 4. Schematic Presentation of the Alkyne Insertion into Terminal Pd−(μ1-ZR) and Bridging Pd−(μ2-Z′R) Bonds (Z, Z′ =
S and Se) in Binuclear Complexes
favorable than chain structures. Importantly, the Pd−Z bond
energies show different trends for the chain and cyclic
complexes. In the cyclic complexes the Pd−Z energy has
converged at n = 6 to a value of 8.8 (9.3) kcal/mol and did not
change noticeably afterward (Table 3), whereas in the chain
complexes it monotonously increases and may finally reach a
value similar (or larger) to that for cyclic structures at large n’s.
The calculated ΔG/m value of Pd−Z bond is 7.9 (8.4) kcal/
mol at n = 9.
ZR ligands with alkyne. We believe that the obtained
knowledge on reactivity of μ2-ZR ligands in the chain structure
will be transferable to the cyclic [Pd(ZR)2]n compounds.
2.6. Difference in Reactivity of the μ1-(Terminal) and
μ2-(Bridging) ZR Groups. Calculations revealed critical
difference in reactivity of μ1-(terminal) and μ2-(bridging) ZR
(R = Me) groups of binuclear complex V (Scheme 4), a model
of active catalyst, with alkyne (i.e., C2H2).23 The first step of the
reaction is the alkyne coordination to one of the coordinatively
unsaturated Pd-centers to form a weakly bound π-complex
VI.24 In complex VI the coordinated alkyne ligand can react
either with μ1-ZR (via the VI→VII-TS→VIII pathway) or with
μ2-ZR ligand (via the VI→IX-TS→X pathway). At first, we
rationalize the difference in reactivity of the μ1- and μ2-
coordinated chalcogen groups ZR toward alkynes (i.e., either
surface or core ZR ligands, respectively; where Z = S and Se).
As seen in Table 4, the alkyne coordination to one of the
It is important to state that the ΔG/m values are found to be
0.3−0.5 kcal/mol larger for the Pd−Se bonds than the Pd−S
bonds (see Table 3). This indicates that in the three-
component mixture of PhSeH/PhSH/CySH = 1:1:1, selenol
will react first with Pd-centers before sulfur-containing species.
This finding is in a good agreement with our experimental data.
On the basis of these findings we conclude that at lower
numbers of n (but higher than n = 3) the most stable structure
of the [Pd(ZR)2]n cluster is its cyclic conformer, while at larger
numbers of n the chain structure may become energetically as
stable as or even more stable than the cyclic structure. It should
be emphasized that in cyclic structures all Pd centers are
equivalent and contain only μ2-ZR ligands. Alternatively, in
chain structures two different types of metal centers exist: two
unsaturated Pd-centers at the ends of the chain each with one
μ1-ZR ligand and one coordination vacancy, and n − 2
saturated Pd-centers with four μ2-ZR ligands each. Smaller
averaged bond energy (Table 3) in the case of chain structures
may be explained by the presence of such unsaturated centers.
Armed with the above presented knowledge on the structure
of [Pd(ZR)2]n compounds, below we intend to elucidate the
reaction mechanism and the factors controlling chalcogen
addition to alkyne. At first, we elaborate the reactivity of the
chain structure of [Pd(ZR)2]n compound because it has two
different chalcogen groups, terminal μ1-ZR and bridging μ2-ZR,
while the cyclic structure has only the bridging μ2-ZR ligands.
Consequently, the structural motif of chain structure provides
better opportunity to elucidate reactions of both μ1-ZR and μ2-
Table 4. Free Energy Parameters (ΔG, in kcal/mol) for the
Model Alkyne Insertion Reaction V + C2H2 (see Scheme 4),
Calculated at the B3LYP Level of Theory
sulfur core (Z′ = S)
selenium core (Z′ = Se)
point
Z = S
Z = Se
Z = S
Z = Se
ΔG1 V→VI
−1.4
5.9
−3.2
6.8
−1.8
7.8
−1.8
7.3
ΔG⧧ VI→VII-TS
2
ΔG2 VI→VIII
−19.7
20.5
5.5
−15.1
22.6
6.5
−17.6
21.8
−14.3
21.7
ΔG⧧ VI→ IX-TS
3
ΔG3 VI →X
10.0
10.0
coordinatively unsaturated Pd-centers of V to form weakly
bound π-complex VI is calculated to be 1.4 (1.8) kcal/mol
exothermic (based on the free energies).24 For Z = S, the
alkyne insertion into the terminal (i.e., surface) Pd−(μ1-SR)
bond, occurring via the VI→VII-TS→VIII pathway, requires
⧧
ΔG2 = 5.9 kcal/mol activation barrier at the transition state
6643
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
VII-TS and is exothermic by ΔG2 = −19.7 kcal/mol. Whereas
the alkyne insertion into the bridging (i.e., core) Pd−(μ2-SR)
bond, occurring via the VI→IX-TS→X pathway, requires a
larger n’s, the chain structure of [Pd(ZR)2]n becomes
energetically closer to the cyclic structure and is accessible for
alkyne addition reaction; (b) reactivity of the μ1-(terminal) and
μ2-(bridging) ZR groups (Z = S and Se, R = Me) are very
different: the terminal (or surface) μ1-ZR groups are
significantly more reactive than the bridging (or core) μ2-ZR
groups, and (c) the unique feature of the cyclic complexes is the
absence of the η1-SR groups. If our results on relative reactivity
of the μ1-SR and μ2-SR groups discussed above are correct,
then the cyclic complexes should be totally unreactive toward
alkynes and can be prepared and isolated, at least for smaller
n’s.
⧧
much larger, ΔG3 = 20.5 kcal/mol, energy barrier at the
transition state IX-TS, and is ΔG3 = 5.5 kcal/mol endothermic.
Similarly, for Z = Se, the alkyne insertion into the terminal Pd−
(μ1-SeR) bond occurring via the VI→VII-TS→VIII pathway
requires ΔG2⧧ = 7.3 kcal/mol activation barrier at VII-TS, and
is exothermic by ΔG2 = −14.3 kcal/mol. However, the alkyne
insertion into the bridging Pd−(μ2-SeR) bond occurring via the
⧧
VI→IX-TS→X pathway requires a much larger, ΔG3 = 21.7
kcal/mol, energy barrier at IX-TS, and is endothermic with
ΔG3 = 10.0 kcal/mol. Comparison of these energetics between
Z = S and Z = Se shows that for selenium compounds, the
calculated barriers are slightly larger and overall reactions are
less exothermic (or more endothermic) than those for sulfur
compounds. This trend is consistent with the difference in the
bonding energies of Pd−Se and Pd−S bonds: Pd−Se is slightly
stronger than Pd−S.
On the basis of these computational findings, we propose the
following plausible mechanism of the selective capture and
addition of various chalcogen groups to alkyne (Scheme 5).
Scheme 5. Proposed reaction mechanism of catalyst
formation and the plausible catalytic cycle
Thus, the insertion of alkyne into the Pd−(μ1-ZR) and Pd−
(μ2-ZR) bonds of [Pd(ZR)2]2 is dramatically different and
preferably proceeds via the insertion into Pd−(μ1-ZR) bond. As
seen in Scheme 4, in binuclear complex [Pd(ZR)2]2 (as well as
possibly in all other chain [Pd(ZR)2]n compounds) only one
ZR group in each Pd-center located at the ends of the chain is
reactive toward alkyne. With such a difference in reactivity of
the coordinated ZR groups in [Pd(ZR)2]n (chain compounds)
we should expect the formation of only vinylic product 1: the
formation of product 4 with two ZR groups is not feasible.25
On the basis of the above-presented analyses for the
[Pd(ZR)2]2 systems with the same ZR (either SR or SeR)
groups on the both core (μ2-) and surface (μ1-) positions, we
conclude that the weaker Pd−Z bond the smaller energy
barrier, and the more exothermic is alkyne insertion into Pd−Z
bond. Now, we wish to elucidate the effect of the nature of core
μ2-ZR groups to the reactivity of the surface μ1-ZR groups. As
seen in Table 4, replacement of all μ2-SR core ligands in
[Pd(SR)2]2 by μ2-SeR (i.e., S-to-Se replacement) has slightly
increased the calculated energy barriers and reduced exother-
micity of alkyne insertion into the Pd−(μ1-SR) bond. On the
contrary, replacement of all core μ2-SeR ligands in [Pd(SeR)2]2
by μ2-SR (i.e., Se-to-S replacement) has slightly reduced the
energy barrier and increased the exothermicity of the reaction of
alkyne insertion into the Pd−(μ1-SeR) bond. The obtained
effect of the nature of the core (μ2-ZR group) on the reactivity
of the surface Pd−(μ1-ZR) bond could be, again, explained in
terms of the Pd−Z bonding energy. Indeed, as we have
discussed above (see Table 3), the ΔG/m values for the Pd−Se
bond (averaged, 0.3−0.5 kcal/mol) are slightly stronger than
those for the Pd−S bond, which allows us to conclude that the
weaker bonding is observed in the core, and the easier alkyne
insertion involves the terminal Pd−(μ1-ZR) bond.
The catalytic cycle starts with the formation of active catalyst
[Pd(ZR)2]n from the Pd(OAc)2 precursor and chalcogen
derivatives. Formation of the metal organochalcogen particles
was confirmed by EDS-SEM and microanalysis (Table 2). For
smaller n’s, the metal organochalcogen particles contain only
inactive core (μ2-SR) ligands and, therefore, are inert and can
be isolated (see below). However, for larger n’s, the chain
structure of the [Pd(ZR)2]n compounds containing both
inactive μ2-ZR and reactive μ1-ZR groups becomes energetically
accessible. Within such an arrangement of the active catalyst
[Pd(ZR)2]n the further transformations leading to product 1
are alkyne coordination, alkyne insertion into the terminal Pd−
(μ1-ZR) bond, and protonolysis. Since the averaged Pd−SeR
bond energy is calculated to be slightly larger than the averaged
value for the Pd−SR bond, the catalytic system is first expected
to utilize the selenols before thiols, which is in a good
agreement with the above-presented experiments.
Further proof for this mechanism proposed on the basis of
the computational studies can be experimental evidence such as
preparation and investigation of reactivity of the cyclic
[Pd(ZR)2]n compounds.
2.7. Proof of Concept: Unreactive μ2-SR Groups.
Gratifyingly, unambiguous confirmation of the above proposed
mechanism was obtained by successful preparation of the cyclic
complex [Pd(SnHex)2]6 (8), the structure of which was
Furthermore, as seen in Table 4, if the terminal Pd−(μ1-ZR)
bond is not stronger than the core Pd−(μ2-ZR) bond, then it
has no effect on the reactivity of the alkyne insertion into the
core Pd−(μ2-ZR) (Z = S and Se) bonds. On the contrary,
replacement of stronger terminal Pd−(μ1-SeR) bonds by the
weaker Pd−(μ1-SR) bonds reduces both the barrier and the
endothermicity of the alkyne insertion into the core Pd−(μ2-
SR) bond.
Thus, from extensive computational studies, we have shown
that: (a) the energetically most favorable structure of the
[Pd(ZR)2]n compound for smaller n’s is a cyclic structure. For
6644
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Figure 5. Molecular structure of cyclic [Pd(SnHex)2]6 complex (8) determined by X-ray analysis (50% ellipsoids; side view and top view). One of
the two crystallographically independent molecules is presented (hydrogen atoms are omitted for clarity).
Scheme 6. Surface Groups Exchange and Alkyne Insertions in Multicomponent Mixtureddd
established by a single crystal X-ray diffraction analysis (Figure
5) in the solid state. Solution structure of the complex was
confirmed by 1H, 13C NMR and complete line assignment was
carried out using 1D selective TOCSY and 2D HSQC NMR
(see Supporting Information for details). Hexanuclear nature of
complex 8 in solution was confirmed by ESI-MS measure-
ments.
The crystal of 8 contained two crystallographically
independent molecules with similar geometries (see Supporting
Information for details). As seen in Figure 5, the Pd(II) thiolate
complex 8 has a tiara-like structure. In each cluster, the six
palladium atoms form a nearly planar hexagonal ring with the
average Pd···Pd distances of 3.112(2) and 3.115(2) Å, for the
two crystallographically independent molecules. Each palladium
atom is coordinated to four sulfur atoms with an approximately
6645
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
Figure 6. Calculated ΔG energy surface of the reaction in multicomponent mixture; for the sulfur core the numbers are in green, for the selenium
core the numbers are in blue (see Scheme 6 for structures).
square-planar configuration (the S−Pd−S bond angles vary
within 81.0(1)−98.9(1) and 81.4(1)−98.6(1)°, for the two
crystallographically independent molecules), and the average
Pd−S bond lengths are 2.313(4) and 2.320(4) Å, for the two
crystallographically independent molecules. The 12 sulfur
atoms form two S6 hexagons parallel to the central Pd6 ring
from both sides. The dihedral angles between the PdS4
coordination planes range within 115.19(6)−126.39(6) and
115.77(7)−126.53(7)°, for the two crystallographically in-
dependent molecules. The average S···S distances in S6 rings
are 3.489(6), 3.498(6) and 3.503(6), 3.493(6) Å, for the two
crystallographically independent molecules, indicating no S−S
bond. Such a triple-layered tiara-like structure is about 3.06 and
3.05 Å in height for the two crystallographically independent
molecules, with a symmetry close to D6h (see Supporting
Information for detailed structure data). The S−C bonds are
alternately directed either parallel or perpendicular to the S6
rings (Figure 5, right). Such an arrangement is attributed to the
steric repulsion and has been found in other analogous
[Pd(SR)2]6 (R = CH3,26a CH2CH3,26b (CH2)11CH3,26c
CH2COOCH326d) complexes. However, regardless of the S−
C bonds, the long terminal carbon chains are disposed roughly
parallel to each other (Figure 5). This disposition is apparently
explained by the crystal packing of molecules, which is stacked
along the b-axis.
alkyne. The calculations with selenium and sulfur cores were
carried out to model the experimental reaction showed in
Figure 1 b and c, respectively.
A seen in Scheme 6, the reaction leading to the formation of
sulfur product XI-a (left-hand-side of the Scheme 6 and Figure
6) starts from the complex V-a and proceeds via alkyne
insertion transition state VII-a-TS. In a similar pathway, the
reaction leading to the formation of selenium product XI-b
(right-hand-side of the Scheme 6 and Figure 6) starts from the
complex V-b and proceeds via alkyne insertion transition state
VII-b-TS.
Surprisingly, we have found that μ1-coordinated chalcogen
groups can exchange very easily at the catalyst active site (the
middle part of Figure 6 and the top part of Scheme 6).
Coordination of RSeH to V-a leads to the formation of
complex XII, followed by facile proton transfer by overcoming
small activation barriers ΔG⧧ = 0.7 and 5.2 kcal/mol for the
XII→XIII-TS step with sulfur and selenium cores, respectively.
Dissociation of RSH from the complex XIV furnishes complex
V-b with a selenium group at the catalyst active site. The
reverse process of interconversion of V-b to V-a is also possible
and involves coordination of RSH followed by dissociation of
RSeH. Small activation barriers ΔG⧧ = 6.2 and 6.7 kcal/mol for
the XIV→XIII-TS step were calculated with sulfur and
selenium cores, respectively. Thus, the exchange of μ1-
coordinated chalcogen groups at the catalyst active site is a
feasible process and should proceed fast. According to
thermodynamic and kinetic factors formation of complex XIV
with coordinated selenium group is more favorable.
Considering the overall energetics (Figure 6), the pathway to
produce selenium product XI-b requires overcoming the XIV→
VII-b-TS barrier ΔG⧧ = 10.9 and 9.8 kcal/mol for sulfur and
selenium cores, respectively. The transformation XIV→ XI-b is
exothermic with ΔG = −20.7 and −23.6 kcal/mol for sulfur
and selenium cores, respectively. The pathway to the sulfur
product requires overcoming the higher activation barrier
XIV→VII-a-TS, where ΔG⧧ = 17.8 and 16.6 kcal/mol for
sulfur and selenium cores, respectively. The transformation
XIV→ XI-a is less exothermic with ΔG = −15.3 and −18.0
kcal/mol for sulfur and selenium cores, respectively. Thus, in
total agreement with experimental findings RSeH should be
involved in the reaction first and should provide vinylic product
containing selenium functional group. The relative reactivity of
The symmetrical cyclic structure of 8 in solution was
1
confirmed by H and 13C NMR (see Supporting Information).
Importantly, it was found that complex 8 is totally inactive
toward alkynes. Addition of alkyne to 8 produces neither 1 nor
4, even after continuous heating at 100 °C. Thus, the synthesis
of the complex 8 and its inactivity with alkyne clearly confirms
our theoretical predictions on the stability of cyclic derivatives
for smaller n’s and unreactive nature of the μ2-SR groups.
2.8. Adaptive Tuning of the Catalyst. An important
issue, which we address in this paper, is the nature of adaptive
tuning of the catalyst (Scheme 6). Obviously, the relative
reactivity of terminal Pd−(μ1-ZR) bonds and the effect of the
core μ2-ZR ligands on their reactivity are not responsible for the
observed adaptive tuning of the catalyst. To obtain insight into
the nature of adaptive tuning we have constructed the potential
energy profile of the reaction (Figure 6). The studied system
mimics the experimental conditions when the mixture of
RSeH/RSH was used as the initial reagent in the reaction with
6646
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
RSH was found to be significantly lower, and the product with
the sulfur group will be formed only after consumption of
RSeH.
The above presented mechanistic findings have revealed the
Pd−(μ1-ZR) bond energy as the key factor to determine the
reactivity of the catalytic system. To model the three-
component PhSeH/PhSH/CySH mixture we have calculated
the Pd−(μ1-ZR) bond energies in the (μ1-ZR)Pd2(SeR)3
system (Scheme 7). Amazingly, the calculated Pd−(μ1-SCy),
a full agreement with theoretically predicted mechanism and
justified potential application for organic synthesis.
3. CONCLUSIONS
The reaction of a simple palladium salt, Pd(OAc)2, with a
mixture of SH/SeH compounds produced Pd(ZR)2 species,
which under controlled conditions self-assembled to metal−
organic structure with outstanding functional properties. Two
leading characteristics of the synthesized [Pd(ZR)2]n material
opened very promising application in catalysis: (i) the
[Pd(ZR)2]n molecules promote reaction with alkyne and
facilitate formation of new carbon-heteroatom bonds (C-Z)
in catalytic manner; (ii) binding of ZR groups to the metal
center differentiates the reactivity of chalcogens in the Pd−(μ1-
ZR) moiety and induces selectivity in the chemical trans-
formation of interest.
Scheme 7. Calculated Thermodynamic Parameters (ΔG,
kcal/mol) for Adaptive Recognition of PhSeH, PhSH, and
CySH Reagents in the Mixture
In conclusion, these findings made it possible to develop an
efficient and practical procedure for the preparation of sulfur-
and selenium-functionalized vinyl compounds starting from
alkynes and a mixture of SH/SeH derivatives as initial
compounds. The developed approach is of much importance
in view of abundant natural sources of sulfur compounds
feedstock and easy availability of alkynes.
High efficiency of the developed procedure and excellent
dynamic range in toxicity-ordered extraction from the mixture
can be of great importance for environmental protection
technologies, especially to remove SeH compounds. The
extraction procedure captures RSeH compounds into sub-
stantially less toxic vinylselenides suitable for practical
utilization as reagents in organic synthesis and material science.
The developed system is a rare example of more selective
transformation involving heterogeneous catalysts rather than
homogeneous catalysts based on similar metal complexes. The
mechanism of the catalytic reaction and the origin of high
selectivity exhibited in the reaction were established at the
molecular level; easy exchange of the ZR groups on the surface
of the catalyst (in contrast to the core of the catalyst) and
thermodynamic control by Pd−Z bond strength ensure the
desired structure of the catalyst active site. It is important to
note that catalyst design utilized in the present study does not
require any catalyst support; the catalyst was self-organized in
situ after the contact with reacting species.
Pd−(μ1-SPh), and Pd−(μ1-SePh) bond energies are in
complete agreement with the reaction selectivity ZR = PhSe
> PhS > CyS.
2.9. Extraction of RZH Derivatives from Multicompo-
nent Mixture of Chalcogen Species. Natural sources of
sulfur compounds in crude oil represent a heavy mixture,
which, in addition to thiols, contains disulfides, cyclic and
acyclic sulfides and thiophenes. Our theoretical calculations
have shown that substitution of RZ group in the catalyst active
center (Scheme 7) by any of the chalcogen derivatives (RSSR,
RSR, RSeSeR, C4H4S, C4H8S) is unlikely, since it is
energetically disfavorable by >20 kcal/mol. If this prediction
is correct, we should be able to extract thiols directly from such
heavy mixtures without a loss of selectivity due to influence of
other components.
To verify this prediction we have prepared a mixture of
chalcogen compounds containing 10 different derivatives
belonging to main classes mentioned above (Scheme 8).
4. EXPERIMENTAL SECTION AND METHODS
4.1. General Experimental Procedures. Unless otherwise noted,
the synthetic work was carried out under argon atmosphere. Reagents
were obtained from Acros and Aldrich and used as supplied (checked
by NMR before use).
Scheme 8. Extraction of RZH from a Mixture of Chalcogen
Species
All NMR measurements were performed using Bruker DRX-500
and Avance-600 spectrometers operating at 500.1/600.1 and 202.5/
243.0 MHz for 1H and 31P nuclei, respectively. The spectra were
processed on a Linux workstation using TopSpin software package. 1H
chemical shifts are reported relative to the corresponding solvent
signals used as internal reference, external H3PO4/H2O was used for
31P chemical shifts. Estimated errors in the yields determination by 1H
NMR <2%. The products 1, 2, and 3 were identified according to
published data.27 Scanning electron microscopy study was carried out
using Hitachi SU-8000 field emission electron microscope, and EDS-
SEM analysis was carried out using Hitachi SU-3400 electron
microscope.
4.2. Catalytic Addition Reactions in the Three-Component
PhSeH/PhSH/CySH Mixture. The alkyne (6.0 × 10−3 mol) was
added to Pd(OAc)2 (4.0 × 10−5 mol) and stirred at room temperature
until a dark-brown solution was formed (∼5 min). PhSeH (2.0 × 10−3
mol), PhSH (2.0 × 10−3 mol), and CySH (2.0 × 10−3 mol) were
Running the reaction with alkyne using developed catalytic
system has shown complete retain of selectivity. The product 1a
was formed first in 99% yield, followed by product 1b in 99%
yield and then by product 1c in 96% yield (Scheme 8). The
presence of the other sulfur and selenium compounds in the
mixture did not influence activity and selectivity of the catalytic
system. High tolerance of the developed catalytic system was in
6647
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
mixed in a separate vessel and were added to the stirred alkyne
solution. The stirring was continued for an additional 10 min, and the
solution became turbid. The reaction was carried out at 80−120 °C
under stirring. Aliquots of the reaction mixture were taken periodically
to the stirred alkyne solution at ∼5 °C (water/ice bath). The stirring
was continued for an additional 10 min, and the solution became
turbid. The reaction was carried out at 80−120 °C under stirring.
Aliquots of the reaction mixture were taken periodically to monitor the
1
1
reaction by H NMR.
to monitor the reaction by H NMR.
4.7. X-ray Structural Study of Complex 8. Data were collected
on a Bruker three-circle diffractometer equipped with a SMART 1K
CCD detector and corrected for absorption using the SADABS
program.28 Data reduction was performed using SMART2 and
SAINTPlus3 programs. The structure was solved by direct methods
and refined by full-matrix least-squares technique on F2 with
anisotropic displacement parameters for all the non-H atoms (for
details, see Supporting Information). The H atoms were placed in
calculated positions and refined within the riding model with fixed
isotropic displacement parameters (Uiso(H) = 1.5 Ueq(C) for the
CH3 groups and Uiso(H) = 1.2 Ueq(C) for the other groups). All
calculations were carried out using the SHELXTL program.29
Crystallographic data for the structure reported in this study have
been deposited with the Cambridge Crystallographic Data Centre as
supplementary CCDC.
4.8. Theoretical Calculations. Geometry and energy of the
reactants, intermediates, transition states, and products of the reactions
were calculated using the B3LYP hybrid density functional method30
in conjunction with standard 6-31G(d) basis set31 for H, C, O, S, and
Se and Lanl2dz basis set with effective core potentials32 for the metal
(denoted as B3LYP/6-31G(d) and Lanl2dz level). In previous studies
it was established that this level of theory reasonably describes the
energy and geometry parameters of the systems involving transition
metal complexes.33
4.3. Catalytic Addition Reactions in the Binary Mixtures
PhSeH/PhSH, PhSeH/CySH, and PhSH/CySH. The alkyne (4.0 ×
10−3 mol) was added to Pd(OAc)2 (4.0 × 10−5 mol) and stirred at
room temperature until a dark-brown solution was formed (∼5−10
min). PhSeH (2.0 × 10−3 mol) and PhSH (2.0 × 10−3 mol) were
mixed in a separate vessel and were added to the stirred alkyne
solution. The reaction was carried out at 80−120 °C under stirring.
Aliquots of the reaction mixture were taken periodically to monitor the
1
reaction by H NMR.
Reactions with PhSeH/CySH and PhSH/CySH binary mixtures
were carried out in analogous manner.
4.4. Catalyst Isolation from Reaction Mixture. The alkyne (1.8
× 10−2 mol) was added to Pd(OAc)2 (1.2 × 10−4 mol) and stirred at
room temperature until a dark-brown solution was formed (∼5 min).
PhSeH (6.0 × 10−3 mol), PhSH (6.0 × 10−3 mol), and CySH (6.0 ×
10−3 mol) were mixed in a separate vessel and were added to the
stirred alkyne solution. The stirring was continued for an additional 10
min, and the solution became turbid. The reaction was carried out at
80 °C under stirring for 25 min to obtain [Pd(SePh)2]n particles, 3 h
to obtain [Pd(SePh)2−‑m(SPh)m] ones. The precipitates were
separated by centrifugation (5 min, 5000 rpm), washed with hexanes
(4 × 5 mL), and dried in a vacuum. Dark-brown solids were obtained.
4.5. Preparation of Complex 8. Pd(OAc)2 (0.39 g, 1.34 mmol)
and 5 mL of toluene were placed into a 25-mL round-bottom flask
equipped with a magnetic stirrer bar. The suspension was stirred at r.t.
For all studied structures normal coordinate analysis was performed
to characterize the nature of the stationary points and to calculate
thermodynamic properties (298.15 K and 1 atm). Transition states
were confirmed with IRC (Intrinsic Reaction Coordinate) calculations
using the standard method.34 All calculations were performed without
any symmetry constraints using the Gaussian09 program.35
n
upon complete dissolution of the palladium acetate. C6H13SH (0.38
mL, 2.67 mmol) was added to the resulting brown solution. The
reaction mixture was stirred for 1.5 h at 80 °C. After the reaction was
finished, the crude product was purified twice via dry-column, vacuum
flash-chromatography (gradient elution with petroleum ether−ethyl-
acetate, 4:1 mixture). Purified product was obtained as an orange solid.
Yield: 0.32 g, 71%. Crystals suitable for X-ray diffraction experiments
were grown from toluene−petroleum ether, 1:1 mixture, at room
temperature.
ASSOCIATED CONTENT
■
S
* Supporting Information
1H and 13C NMR spectra of complex 8, line assignment of
complex 8 with TOCSY and HSQC NMR, X-ray structure
analysis and details on theoretical study. This material is
available free of charge via the Internet at http://pubs.acs.org.
1H NMR (500.1 MHz; CDCl3; δ, ppm; J, Hz): 3.03−3.11 (t, 2H),
2.80−2.89 (t, 2H), 2.10−2.24 (m, 2H), 2.00−2.10 (m, 2H), 1.67−1.78
(m, 2H), 1.55−1.65 (m, 2H), 1.31−1.54 (m, 8H), 0.91−1.09 (m, 6H).
13C{1H} NMR (125.8 MHz; CDCl3; δ, ppm): 35.53, 35.37, 33.55,
32.38, 31.87, 31.28, 29.22, 29.01, 23.09, 14.35, 14.32. HRMS(ESI):
Calculated for C72H156Pd6S12 [M + nH] = 2045.3185, found
2045.3180 (Δ = 1.2 ppm). Elemental analysis: Calculated for
C72H156Pd6S12 (%): C, 42.48; H, 7.69; Pd, 31.22; S, 18.81; found:
C, 42.56; H, 7.65; Pd, 31.12; S, 18.91.
AUTHOR INFORMATION
Corresponding Author
(V.P.A.) val@ioc.ac.ru; (D.G.M.) dmusaev@emory.edu
■
Notes
4.6. Dynamically Changed Mixture of Reagents and
Dynamic Range Test. The alkyne (2.0 × 10−3 mol) was added to
Pd(OAc)2 (2.0 × 10−5 mol) and stirred at room temperature until a
dark-brown solution was formed (∼5−10 min). PhSH (1.0 × 10−3
mol) and CySH (1.0 × 10−3 mol) were mixed in a separate vessel and
were added to the stirred alkyne solution. The stirring was continued
for additional 10 min, and a dark-brown suspension was formed. The
reaction was carried out for 2 h at 80 °C under stirring. Aliquots of the
reaction mixture were taken periodically to monitor the reaction by 1H
NMR. After 2 h PhSeH (1.0 × 10−3 mol) was added to the reaction
mixture. The reaction was continued for an additional 10 h at 80−120
°C under stirring. Aliquots of the reaction mixture were taken
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
V.P.A. acknowledges the Emerson Center Fellowship. The
work was supported in part by the Russian Foundation for
Basic Research (Project No. 10-03-00370), Grant MD-
4969.2012.3 and Programs of Division of Chemistry and
Material Sciences of RAS. We gratefully acknowledge NSF
MRI-R2 Grant (CHE-0958205) and the use of the resources of
the Cherry Emerson Center for Scientific Computation. D.G.M
gratefully acknowledge the NSF Center of Chemical Innovation
in Stereoselective C¬H Functionalization (CHE-0943980) for
support.
1
periodically to monitor the reaction by H NMR.
4.7. Catalytic Reaction in Multicomponent Mixture. The
alkyne (6.0 × 10−3 mol) was added to Pd(OAc)2 (4.0 × 10−5 mol) and
stirred at room temperature until a dark-brown solution was formed
(∼5 min). Ph2Se2 (5.0 × 10−4 mol), Ph2S2 (5.0 × 10−4 mol), Cy2S2
(5.0 × 10−4 mol), Bu2S2 (5.0 × 10−4 mol), Bu2S (5.0 × 10−4 mol),
thiophene (5.0 × 10−4 mol), and tetrahydrothiophene (5.0 × 10−4
mol) were mixed in a separate vessel, and the 0.34 mL aliquote of
PhSeH (1.0 × 10−3 mol), PhSH (1.0 × 10−3 mol) and CySH (1.0 ×
10−3 mol) was added to the mixture. The resulted mixture was added
REFERENCES
■
(1) Topsøe, H., Clausen, B. S., Massoth, F. E. Hydrotreating Catalysis;
Springer-Verlag: New York, 1996.
(2) Kabe, T., Ishihara, A., Qian, W. Hydrodesulfurization and
Hydrodenitrogenation; Wiley-VCH: New York, 1999.
6648
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649

Journal of the American Chemical Society
Article
(3) Song, C. Catal. Today 2003, 86, 211.
Buslaeva, T. M.; Romm, I. P. Russ. J. Coord. Chem. 2001, 27, 585.
(c) Yang, Z.; Smetana, A. B.; Sorensen, C. M.; Klabunde, K. J. Inorg.
Chem. 2007, 46, 2427. (d) Schneider, I.; Horner, M.; Olendzki, R. N.;
Strahle, J. Acta Crystallogr. 1993, C49, 2091.
(27) (a) Ananikov, V. P.; Orlov, N. V.; Beletskaya, I. P.
Organometallics 2006, 25, 1970. (b) Ananikov, V. P.; Orlov, N. V.;
Beletskaya, I. P. Organometallics 2007, 26, 740.
(28) Sheldrick, G. M. SADABS; Bruker AXS Inc.: Madison, WI, 1997.
(29) Sheldrick, G. M. SHELXTL-97, Version 5.10; Bruker AXS Inc.:
Madison, WI, 1997.
(30) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (c) Becke, A. D. J. Chem.
Phys. 1993, 98, 5648.
(4) Babich, I. V.; Moulijn, J. A. Fuel 2003, 82, 607.
(5) Sulfur Production Report, U.S. Geological Survey, January 2011;
http://minerals.usgs.gov/minerals/pubs/commodity/sulfur/mcs-
2011-sulfu.pdf.
(6) Metal-catalyzed Z−H additions: (a) Beletskaya, I. P.; Ananikov,
V. P. Chem. Rev. 2011, 111, 1596. (b) Bichler, P.; Love, J. A. Top.
Organomet. Chem. 2010, 31, 39. (c) Weiss, C. J.; Marks, T. J. Dalton
Trans. 2010, 39, 6576. (d) Ogawa, A. J. Organomet. Chem. 2000, 611,
463. (e) Kuniyasu, H.; Ogawa, A.; Sato, K.; Ryu, I.; Kambe, N.;
Sonoda, N. J. Am. Chem. Soc. 1992, 114, 5902. (f) Ishii, A.; Kamon, H.;
Murakami, K.; Nakata, N. Eur. J. Org. Chem. 2010, 1653. (g) Ishii, A.;
Nakata, N.; Uchiumi, R.; Murakami, K. Angew. Chem., Int. Ed. 2008,
47, 2661. (h) Nakata, N.; Yamaguchi, Y.; Ishii, A. J. Organomet. Chem.
2010, 695, 970.
(7) Sheldon, R. A. Pure Appl. Chem. 2000, 72, 1233.
(8) Anastas, P. T.; Warner, J. C. Green Chemistry; Theory and Practice;
Oxford University Press: New York, 1998.
(9) (a) Baeckvall, J.-E.; Ericsson, A. J. Org. Chem. 1994, 59, 5850.
(b) Kuniyasu, H.; Ogawa, A.; Sato, K.-I.; Ryu, I.; Sonoda, N.
Tetrahedron Lett. 1992, 38, 5525. (c) Kamiya, I.; Nishinaka, E.; Ogawa,
A. J. Org. Chem. 2005, 70, 696.
(31) (a) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971,
54, 724. (b) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys.
1972, 56, 2257.
(32) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.
(b) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (c) Hay, P. J.;
Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (d) Dunning, T. H., Jr.; Hay,
P. J. In Modern Theoretical Chemistry; Schaefer, H. F., III, Ed.; Plenum:
New York, 1976; Vol. 3, pp 1−28.
(33) (a) Cramer, C. J.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2009,
11, 10757. (b) Ananikov, V. P.; Musaev, D. G.; Morokuma, K. J. Mol.
Catal. A: Chem. 2010, 324, 104.
(34) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154.
(b) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(35) Frisch, M. J. ;et al.et al. Gaussian 09; Gaussian, Inc.: Pittsburgh
PA, 2009.
(10) Akelah, A.; Moet, A. Functionalized Polymers and Their
Applications; Chapman and Hall: New York, 1990.
(11) Paquette, L. A., Ed. In Sulfur-Containing Reagents; Wiley-
Blackwell: New York, 2010.
(12) McReynolds, M. D.; Dougherty, J. M.; Hanson, P. R. Chem. Rev.
2004, 104, 2239.
(13) Patai, S.; Rappoport, Z., Eds. In The Chemistry of Organic
Selenium and Tellurium Compounds; John Wiley and Sons: New York,
1986−1987; Vols. 1−2.
(14) (a) Ananikov, V. P.; Beletskaya, I. P. Dalton Trans. 2011, 40,
4011. (b) Ananikov, V. P.; Orlov, N. V.; Beletskaya, I. P.; Khrustalev,
V. N.; Antipin, M. Yu.; Timofeeva, T. V. J. Am. Chem. Soc. 2007, 129,
7252.
(15) Ananikov, V. P.; Malyshev, D. A.; Beletskaya, D. A.;
Aleksandrov, G. G.; Eremenko, I. L. J. Organomet. Chem. 2003, 679,
162.
(16) Ananikov, V. P.; Malyshev, D. A.; Beletskaya, I. P.; Aleksandrov,
G. G.; Eremenko, I. L. Adv. Synth. Catal. 2005, 347, 1993.
(17) Jin, Z.; Xu, B.; Hammond, G. B. Eur. J. Org. Chem. 2010, 168.
(18) Trofimov, B. A. Curr. Org. Chem. 2002, 6, 1121.
(19) Perin, G.; Lenardao, E. J.; Jacob, R. G.; Panatieri, R. B. Chem.
̃
Rev. 2009, 109, 1277.
(20) The concentration of phosphine ligands was adjusted to stabilize
the complexes in solution, and 31P NMR chemical shifts of the
complexes (δ = 98.8 ppm, 101.2 ppm) were identified according to
NMR analysis described earlier: Ananikov, V. P.; Zalesskiy, S. S.;
Kachala, V. V.; Beletskaya, I. P. J. Organomet. Chem. 2011, 696, 400.
(21) Phosphine or other strong ligands occupy vacancies in the
coordination sphere of the metal and prevent formation of polymeric
metal species.
(22) Complete set of calculations for n = 2, 3, 4, 5, 6, 7, 8, and 9
(chain structures) and n = 3, 4, 5, 6, 7, 8, and 9 (cyclic structures) with
both Z = S and Se were carried out; see Supporting Information for
details.
(23) Calculated energy surface of mononuclear complexes I−IV is
shown in Scheme S2 for comparison (see Supporting Information);
this pathway is not discussed in details here.
(24) Coordination of alkyne to the metal center was calculated to be
exothermic by −1 to −3 kcal/mol on the ΔG surface and by −11 to
−14 kcal/mol on the ΔE and ΔH surfaces. Thus, it can be considered
as a weakly bound π-complex on the ΔG surface. See Table S11 in the
Supporting Information for details.
(25) Note also, that reductive elimination involving chelate ligand
was not observed: Ananikov, V. P.; Mitchenko, S. A.; Beletskaya, I. P. J.
Organomet. Chem. 2000, 604, 290.
(26) (a) Spek, A. L. Cambridge Crystallographic Database; refcode
AFACUG, 2007. (b) Stash, A. I.; Perepelkova, T. I.; Noskov, Yu. G.;
6649
dx.doi.org/10.1021/ja210596w | J. Am. Chem. Soc. 2012, 134, 6637−6649