Competitive and selective Csp3-Br versus Csp2-Br bond activation in palladium-catalysed Suzuki cross-coupling: An experimental and theoretical study of the role of phosphine ligands

Article abstract of DOI:10.1002/chem.201001113

Phosphine ligands have been demonstrated to have an effect on reactivity and selectivity in the competitive intramolecular palladium-catalysed Suzuki-Miyaura coupling of dibromo sulfoxide 1a possessing two different hybridised electrophilic carbons. It was found that the bromine bond to the sp³-hybridised carbon is selectively replaced in the presence of unhindered phosphines such as PPh₃ or xantphos. The use of hindered phosphine ligands such as P(o-tol)₃ and P(1-naphthyl)₃ reversed the selectivity, conducting the cross-coupling at the Csp ²-Br. Identical trends were observed in external competition experiments carried out with bromomethyl sulfoxide and different substituted bromoarenes. DFT and DFT/MM calculations showed that the selectivity observed is mainly due to the different facility of the ligands to dissociate. Bisphosphine catalysts favour coupling at the sp³ carbon, whereas monophosphine catalysts prefer the sp² carbon. Customised catalysts: Proper selection of the phosphine ligand in the palladium catalyst enables a highly intra- and intermolecular selective Suzuki-Miyaura coupling at either Csp ³-Br or Csp²-Br (see figure). DFT and DFT/MM calculations show that the selectivity observed is mainly due to the different facilities of the ligands for dissociation. Copyright

Full text of DOI:10.1002/chem.201001113

DOI: 10.1002/chem.201001113
3
2
ꢀ
ꢀ
Competitive and Selective Csp Br versus Csp Br Bond Activation in
Palladium-Catalysed Suzuki Cross-Coupling: An Experimental and
Theoretical Study of the Role of Phosphine Ligands
Cristian Mollar,[a] Maria Besora,[b] Feliu Maseras,*[b, c] Gregorio Asensio,*[a] and
Mercedes Medio-Simꢀn[a]
Dedicated to Professor Josꢀ Barluenga on the occasion of his 70th birthday
Abstract: Phosphine ligands have been
demonstrated to have an effect on re-
activity and selectivity in the competi-
tive intramolecular palladium-catalysed
Suzuki–Miyaura coupling of dibromo
sulfoxide 1a possessing two different
hybridised electrophilic carbons. It was
found that the bromine bond to the
sp3-hybridised carbon is selectively re-
placed in the presence of unhindered
phosphines such as PPh3 or xantphos.
The use of hindered phosphine ligands
such as P(o-tol)3 and P(1-naphthyl)3 re-
versed the selectivity, conducting the
cross-coupling at the Csp2-Br. Identical
trends were observed in external com-
petition experiments carried out with
bromomethyl sulfoxide and different
substituted bromoarenes. DFT and
DFT/MM calculations showed that the
selectivity observed is mainly due to
the different facility of the ligands to
AHCTUNGTRENNUNG
dissociate.
favour coupling at the sp3 carbon,
whereas monophosphine catalysts
prefer the sp2 carbon.
Bisphosphine
catalysts
Keywords: cross-coupling · density
functional calculations · ligand ef-
fects · palladium · reaction mecha-
nisms · selectivity
Introduction
However, selective transformations, despite being of high in-
terest for organic synthesis, have been much less explored in
comparison. Regio- and stereoselectivity, in the context of
palladium-catalysed cross-coupling reactions, are mainly, but
not exclusively,[2] derived from the different reactivities of
the electrophiles towards the oxidative addition step, in
most cases the turnover-limiting step in the catalytic cycle.[3]
In this way, the largely trans-selective monosubstitution, in
the Pd-catalysed cross-coupling of 1,1-dihalo-1-alkenes, is at-
tributed to steric effects exerted by carbon substituents,
which favour oxidative addition to the palladium atom trans
to carbon substituents in the b-position.[4] Unfortunately, dis-
ubstitution is sometimes an undesirable side reaction. On
the other hand, the regioselective monoarylation of unsym-
metrical dihaloarenes[5] is based on the well-known differen-
ces in the leaving group aptitude[6] of halides and pseudoha-
lides. Even with bromoaryl triflates, the selective displace-
ment of either bromide or triflate is possible through the ap-
propriate selection of the palladium ligands.[2,7] Recently, a
computational study analysed the ligand control of the re-
gioselectivity with respect to two Csp2 positions (chloro and
triflate).[8] It was found that a monophosphine catalyst
Palladium-catalysed coupling of aryl or vinyl halides or
pseudohalides with orgamometallics represents a consolidat-
ed method for achieving the formation of C C bonds.
[1]
ꢀ
[a] C. Mollar, Prof. Dr. G. Asensio, Dr. M. Medio-Simꢀn
Department of Quꢁmica Orgꢂnica
Universidad de Valencia
Avda Vicent Andres Estelles s/n
46100-Burjassot (Spain)
Fax : (+34)96-354-49-39
E-mail: gregorio.asensio@uv.es
[b] Dr. M. Besora, Prof. Dr. F. Maseras
Institute of Chemical Research of Catalonia (ICIQ)
Avgda Paꢃsos Catalans 16
43007-Tarragona, Catalonia (Spain)
Fax : (+34)97-792-02 31
E-mail: fmaseras@iciq.es
[c] Prof. Dr. F. Maseras
Unitat de Quꢁmica Fꢁsica
Universitat Autꢄnoma de Barcelona, Edifici Cn.
Universitat Autꢄnoma de Barcelona
08193 Bellaterra, Catalonia (Spain)
ꢀ
would promote the coupling at the C Cl bond, whereas a
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001113.
ꢀ
bisphosphine catalyst would promote the coupling at the C
13390
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13390 – 13397

FULL PAPER
OTf bond. A more sophisticated example of selective cou-
pling at a desired position has recently been provided in the
palladium-catalysed cross-coupling of dibromobenzenes with
Grignard reagents, in which the reaction takes place at the
less favourable site because of the binding properties of the
phosphine ligand.[9] The mechanism and details of the pro-
cess have been studied by several computational groups.[10]
The effect of substituents in the aromatic ring accelerating
(electron acceptors) or retarding (electron donors) the oxi-
dative addition of palladium complexes into carbon–halide
bonds was established many years ago,[11] but it is still an
active field of work.[12] In addition, the hybridisation of the
carbon atom that undergoes the oxidative addition also
plays a role in the relative reactivity. The slow oxidative ad-
dition of alkyl halides to palladium and the favoured com-
peting b-hydride elimination reaction are probably the main
reasons for the lack of success with this kind of substrate
until relatively recently, when Fu[13] conducted diverse palla-
dium-catalysed coupling reactions successfully in unactivat-
ed alkyl halides using sterically hindered electron-rich phos-
phines. These ligands have been postulated to be particular-
ly effective for couplings, because the steric demand facili-
tates dissociation to a monophosphine adduct, to which the
substrate undergoes rapid oxidative addition as a result of
the electron richness of the phosphine.[14] Even in the case
of alkyl electrophiles activated by a neighbouring carbonyl
group (a-halocarbonyl compounds), special conditions and
ligands are usually required to achieve the cross-coupling re-
action, despite the lack of b-hydrogen, which circumvents
the b-hydride elimination.[15] Therefore, all precedents con-
cerning differences in reactivity between aryl/vinyl halides
and alkyl halides in palladium-catalysed cross-coupling reac-
tions indicate a high reactivity for the former and a low re-
activity for the latter (Scheme 1), with the exception of
Scheme 2. Selective palladium-catalysed Suzuki–Miyaura coupling.
In this paper we show that the presence of a sulfoxide
group in the a-position with respect to the sp3 carbon acti-
3
ꢀ
vates the sp C Br bond enough to compete with aryl bro-
mides under ligand control, and allows a better general un-
derstanding of the oxidative addition process.
Results and Discussion
Internal competition experiments were performed with di-
bromo sulfoxide 1a possessing two different hybridised elec-
trophilic carbons. For this substrate, we found that the
proper selection of the phosphine ligand present in the pal-
ladium catalyst enables a competitive Suzuki–Miyaura reac-
tion, which can be directed with high selectivity to either
the sp3 or sp2 carbon atoms (Scheme 2). Monoarylated com-
pounds were the only products formed without contamina-
tion with the double arylation derivatives, which were not
detected in any case. As far as we are aware, the results re-
ported herein are the first examples of a palladium-cata-
lysed cross-coupling reaction being accomplished with a
high degree of selectivity in the aryl or the alkyl electrophil-
ic centre through a simple change in the catalyst ligands. Ini-
tially, we applied the same reaction conditions that we deter-
mined previously as optimal for analogous a-bromo sulfox-
ides (THF, CsF, 658C).[17] Under these conditions, the cross-
coupling reaction of the sulfoxide derivative 1a with boronic
acid 2a, catalysed by Pd0/PPh3, took place almost exclusively
at Csp3 to give the monoarylated product 3aa in moderate
Scheme 1. Relative reactivity of aryl and alkyl halides in palladium-cata-
lysed cross-coupling reactions.
benzyl halides.[16] Despite the striking difference in the reac-
tivity associated with the hybridisation of the carbon in the
electrophilic partner, there have been no systematic studies
exploring and exploiting this significant selectivity in palladi-
um-catalysed cross-coupling reactions. As continuation of
our research into palladium-catalysed Suzuki–Miyaura reac-
tions with a-bromomethyl sulfoxides,[17] we have carried out
yield. Both [PdACHTUNTRGENN(UG PPh3)4] and PdACHTUNTGREN(NUNG OAc)2/PPh3 (Table 1, en-
tries 1 and 2) were efficient catalysts for the reaction, which
was general for a range of boronic acids 2a–f (Table 1, en-
tries 1–7). Aryl boronic acids 2a–d afford a-arylated sulfox-
ides 3aa–ad (Table 1, entries 1–5) in moderate yields, where-
as the sterically hindered 2e (Table 1, entry 6) and the elec-
tron-poor boronic acid 2 f (Table 1, entry 7) provided the
corresponding cross-coupling products in a slightly lower
a study to account for the relative reactivity of (SO)sp3 and
2
ꢀ
sp C Br bonds. Both internal and external competition ex-
periments have been performed (Scheme 2).
Chem. Eur. J. 2010, 16, 13390 – 13397
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13391

F. Maseras, G. Asensio et al.
Table 1. Competitive palladium-catalysed Suzuki coupling at Csp3.
Table 2. Competitive palladium-catalysed Suzuki-coupling at Csp2.
Entry
Method[a]
R1
4 (yield [%])
Entry
Method[a]
R1
3 (yield [%])
4 (yield [%])
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
D
D
E
E
E
E
E
E
E
E
E
F
F
F
F
F
F
F
F
F
4-MeOC6H4
3-MeOC6H4
C6H5
ab (69)[b]
ac (65)[c]
aa (92)
ab (99)
ac (89)
ad (94)
ae (80)
af (86)
ag (94)
ah (96)
ai (0)
aa (76)
ab (85)
ac (75)
ad (62)
ae (50)
af (72)
ag (40)
ah (61)
ai (0)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
A
B
A
A
A
A
A
A
A
A
C
C
C
C
C
C
C
C
C
C6H5
C6H5
aa (50)
aa (48)
ab (69)
ac (60)
ad (50)
ae (40)
af (37)
ag (0)
aa (4)
aa (4)
ab (5)
ac (5)
ad (2)
ae (7)
af (3)
–
–
–
–
–
–
–
–
–
–
–
–
4-MeOC6H4
3-MeOC6H4
4-BrC6H4
2-MeC6H4
4-CF3C6H4
2-thienyl
3-thienyl
Me
4-MeOC6H4
3-MeOC6H4
2-MeC6H4
4-BrC6H4
4-CF3C6H4
2-thienyl
3-thienyl
Me
ah (0)
ai (0)
C6H5
aa (90)
ab (91)
ac (89)
ad (23)
ae (83)
af (90)
ag (88)
ah (95)
ai (0)
C6H5
4-MeOC6H4
3-MeOC6H4
4-BrC6H4
2-MeC6H4
4-CF3C6H4
2-thienyl
3-thienyl
Me
4-MeOC6H4
3-MeOC6H4
2-MeC6H4
4-BrC6H4
4-CF3C6H4
2-thienyl
3-thienyl
Me
[a] Method A: [Pd
G
N
[a] Method D: Pd
N
G
ACHTUNGTRENNUNG(o-tol)3
(1:1), 3 h; Method F: PdACHTUGNTRENNNUG
Method C: Pd(OAc)2/xantphos (1:1), 5 h.
ACHTUNGTRENNUNG
(31%). [c] 3ac (27%).
yield. Thus, the steric and electronic properties of the boron-
ic acids affected the yields of the corresponding products 3.
However, the selectivity was almost complete. Heteroaryl
(Table 1, entries 8 and 9) and alkyl boronic acids (Table 1,
entry 10) were non-reactive under the assayed conditions.
Owing to the high selectivity but moderate activity of the
precedent catalyst, we next examined complexes with bis-
monophosphines were used as palladium ligands. Pd
ACHTUNGTRENNUNG(OAc)2/
PN
ACHTUNGTRENNUNG
thyl)3 (Table 2, entries 12–20) serve as efficient catalysts for
the cross-coupling process at Csp2, affording the correspond-
ing biaryl derivatives 4. PACTHNUTRGNE(UNG o-tolyl)3 was seen to be the best
ligand, leading to the coupled derivatives 4aa–ai (Table 2,
entries 3–11) with excellent yields for a broad range of aryl
boronic acids. Once more, the selectivity depended exclu-
sively on the palladium ligands, and was the same for elec-
tron-rich, electron-poor aryl, and heteroaryl boronic acids.
To demonstrate substrate specificity, external competition
experiments were also performed. To this end, we designed
a series of experiments in which mixtures of sulfoxide 1b
and bromoarenes 5a–i with representative substitution pat-
terns were allowed to react with boronic acids 2b or 2 f con-
taining either an electron-rich or an electron-poor aromatic
ring (Tables 3 and 4). A very satisfactory intermolecular se-
lectivity was also found in these competitions. The reaction
of mixtures of bromo sulfoxide 1b and bromoarenes 5 con-
taining an electron-donor (5a–e) or electron-acceptor sub-
stituent (5 f–i) with boronic acids in the presence of Pd-
ACHTUNGTRENNUNGphosphines. Of the bisphosphines tested [2,2’–bis(diphenyl-
phosphino)–1,1’–binaphthyl (binap), 1,1’-bis(diphenylphos-
phino)ferrocene (dppf), 1,2-bis(diphenylphosphino)ethane
(dppe), 1,3-bis(diphenylphosphino)propane (dppp)] only
(9,9-dimethyl-9H-xanthene-4,5-diyl)bis(diphenylphosphine)
(xantphos) was active. Like PPh3, xantphos drives the reac-
tion to Csp3, but now the selectivity was complete and the
reactivity was considerably enhanced. Both yields and reac-
tion times were remarkably improved for all the boronic
acids tested (Table 1, entries 11–13 and 15–18), with the ex-
ception of the sterically hindered boronic acid 2e (Table 1,
entry 14). Thus, the cross-coupling products arising from the
parent compound 2a and different substituted boronic acids
were obtained with excellent yields (83–91%). Even hetero-
aryl boronic acids 2g and 2h (Table 1, entries 17 and 18),
which were inert under the [Pd
A
(o-tolyl)3 gave exclusively biaryls 6 from the Csp2–
ACHUTGTNNERNUG(OAc)2/PACHTUTGNRENNUGN
coupled with high yields (85 and 95%) by Pd
G
Csp2 bromoarene–boronic acid cross-coupling (Table 3).
Bromo sulfoxide 1b was recovered unchanged in the experi-
ments. Then, complete ligand-induced selectivity was ach-
ieved, which did not depend on the electronic character of
the substituents at the coupled partner.
phos. On the other hand, the use of monophosphine ligands
of high basicity and bulkiness reversed the selectivity
(Table 2). So, the relatively bulky trialkyl monophosphine
PCy3 showed a preference for the sp2 electrophilic carbon,
although selectivity was only moderate (Table 2, entries 1
and 2). The trend was amplified when encumbered triaryl
As expected, the reverse selectivity, that is, the formation
of the Csp3–Csp2 cross-coupling products, was achieved in
13392
www.chemeurj.org
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13390 – 13397

Palladium-Catalysed Suzuki Cross-Coupling
FULL PAPER
Table 3. Pd
tition experiments.
ACHTUNGTRENNUNG(OAc)2/PACHTUNGTRENNUNG(o-tol)3-catalysed Suzuki coupling. External compe-
as CF3 (Table 4 entry 8), the coupling occurs with complete
Csp3–Csp2 bond-forming selectivity.
The selectivity observed between sp2 and sp3 carbons due
to ligand change was explored by computational means. The
oxidative addition of 1a to the palladium catalyst with the
PPh3 (method A) and P(1-naphthyl)3 (method F) ligands
was studied. PPh3 was chosen as an example of a less hin-
dered phosphine, producing mainly the activation of the
Csp3 product, and P(1-naphthyl)3 was used as an example of
a hindered phosphine with similar electronic effects produc-
ing only the Csp2-activated product. The combination of two
metal complexes and two activation types yields four path-
ways to investigate. This number was doubled because we
also considered the possibility of monophosphine or bis-
Entry
2
5 (R2)
6 (yield [%])
1
2
3
4
5
6
7
8
9
f
f
f
b
b
f
f
f
b
f
b
f
a (H)
af (84)
bf (46)
cf (38)
db (40)
eb(46)
ef (46)
ff (75)
gf (92)
hb (48)
hf (66)
ib(60)
b (4-OMe)
c (2-Me)
d (4-Me)
e (2-NH2)
e (2-NH2)
f (4-Cl)
g (4-CF3)
h (4-CN)
h (4-CN)
i (3-CN)
i (3-CN)
AHCTUNGTRENNUNG
10
11
12
each of the eight computed pathways are collected in
Table 5. The species are labelled as Mnth, in which M is the
if (84)
Table 5. Relative enthalpies [kcalmolꢀ1] of the adducts and transition
states in the eight computed pathways for the 1a+PdACHTNUGTRNEG(UN PR3)n reaction.
the competitive experiment when PACHTUNRGTNEUNG(o-tolyl)3 was substituted
by xantphos as the phosphine ligand in the palladium cata-
lyst (Table 4). The ligand control of the selectivity was also
general, with the single exception of the coupling of bro-
moarenes containing a CN group (Table 4, entries 9–12).
PPh3
P(1-naphthyl)3
Csp3
Csp2
A2add3
A2ts3
A1add3
A1ts3
A2add2
A2ts2
A1add2
A1ts2
7.1
17.3
ꢀ5.7
ꢀ0.9
9.0
20.5
ꢀ4.4
ꢀ1.6
F2add3
F2ts3
F1add3
F1ts3
F2add2
F2ts2
F1add2
F1ts2
5.6
18.6
ꢀ3.2
1.7
7.1
Table 4. Pd
ACHTUNGTRENNUNG(OAc)2/xantphos-catalysed Suzuki coupling. External compe-
23.3
ꢀ0.7
0.4
tition experiments.
phosphine: A for P
number of phosphines involved: 1 for Pd
Pd(PR3)2; t is the nature of the species: add stands for
adduct before the transition state, and ts for the transition
state; and h shows the hybridisation of the activated carbon:
3 for sp3 carbon and 2 for sp2. Values are enthalpies in solu-
tion (DEsol) relative to the reactants (PdACTHUNGTRNENU(G PR3)1,2 plus 1a) at
infinite separation.
Table 5 shows the energies and transition states for the re-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Entry
2
5 (R2)
3 (yield [%])
AHCTUNGTRENNUNG
1
2
3
4
5
6
7
8
9
f
f
f
b
b
f
f
f
b
f
b
f
a (H)
bf (89)
bf (63)
bf (62)
bb (80)
bb(59)
bf (65)
bf (64)
bf (74)
bb (40)[a]
bf (37)[b]
bb(59)[c]
bf (54)[d]
b (4-OMe)
c (2-Me)
d (4-Me)
e (2-NH2)
e (2-NH2)
f (4-Cl)
g (4-CF3)
h (4-CN)
h (4-CN)
i (3-CN)
i (3-CN)
ꢀ
action of each complex with the two possible C Br bonds.
A clear pattern emerges in that the sp3 carbon is preferred
for the A2ts3/2 (17.3 vs. 20.5 kcalmolꢀ1) and F2ts3/2 (18.6
vs. 23.3 kcalmolꢀ1) systems, whereas the sp2 carbon is pre-
ferred for the A1ts2/3 (ꢀ1.6 vs. ꢀ0.9 kcalmolꢀ1) and F1ts2/3
(0.4 vs. 1.7 kcalmolꢀ1) complexes. The nature of the pre-
ferred product does not correlate directly with the type of
ligand (A vs. F), but with the coordination number (1 vs. 2).
This may seem at odds with the experimentally observed de-
pendence of product nature with the type of ligand. Howev-
er, the ligand type and the coordination number are corre-
lated.
10
11
12
[a] 6hb (37%). [b] 6hf (32%). [c] 6ib (25%). [d] 6if (22%).
The 3- and 4-bromobenzonitriles are highly activated sub-
strates in the oxidative addition step,[10] and the formation
of a mixture of the Csp3–Csp2 and the Csp2–Csp2 cross-cou-
pling products takes place in this case. In fact, the Csp3–Csp2
coupling still predominates, but the selectivity is low. How-
ever, we should note the broad scope of the ligand control
of the selectivity by xantphos, (Table 4 entries 1–8) as even
with bromoarenes bearing electron-withdrawing groups such
The bisphosphine complexes can dissociate one of the
phosphine ligands to form the monophosphine complex.
This PdACHTUNGRTNEUN(G PR3) complex is more reactive, but its formation de-
Chem. Eur. J. 2010, 16, 13390 – 13397
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13393

F. Maseras, G. Asensio et al.
pends on the equilibrium described in Equation (1). Bulky
phosphines displace the equilibrium to the right.[10,18]
over F2ts3. F1ts2 is located just 6.7 kcalmolꢀ1 above the
PdP(1-naphthyl)3
+ 1a, resulting in a stabilisation of
11.9 kcalmolꢀ1 over F2ts3. These results show that for this
ligand the activation of the sp2 carbon is favoured through
phosphine dissociation,[19] again in agreement with experi-
ment.
PdðPR3Þ2 ! PdðPR3Þ þ PR3
ð1Þ
The computed dissociation enthalpy energy of one phos-
phine from the bisphosphine complex is 20.4 kcalmolꢀ1 for
These theoretical results seem to correlate well with the
experimentally observed behaviour of other phosphines.
The other monophosphines experimentally considered
PdACHTUNGTRENNUNG
(PPh3)2 and 6.3 kcalmolꢀ1 for Pd(P(1-naphthyl)3)2. It is
important to note that this is a rough estimation. These en-
ergies are usually method-dependent, solvent and other mol-
ecules in the reaction media can play a role, as well as en-
tropy effects, which are difficult to estimate accurately. The
dissociation free energies in solution, which contain the en-
tropic correction from ideal gas statistical thermodynamics,
(PCy3, PACHTNUGRTNEUNG(o-tol)3) behave like P(1-naphthyl)3, as they are
bulkier than PPh3. They also provide a straightforward ex-
planation for the behaviour of the xantphos ligand, which,
being bisphosphine, probably remains as such, and presents
a similar behaviour to PdACHTNUGTRNEUG(N PPh3)2. The lack of reactivity of
are 7.3 kcalmolꢀ1 for Pd
ACHTUNGTRENNUNG
other bisphosphines (dppp, dppf, dppe) remains unex-
plained, but it could be related to some step of the catalytic
cycle other than the oxidative addition.[10,20]
Pd(P(1-naphthyl)3)2. As expected, the inclusion of entropic
effects favours the dissociation, although it is probably over-
estimated. We think, however, that the difference between
both ligands is significant, and is furthermore easily ex-
plained by the different steric bulk.
The remaining question relates to why monophosphine
complexes activate the sp2 centres preferentially and bis-
AHCTUNGTRENNUNG
phosphine complexes react more easily with sp3 centres.
Putting together the four most favoured pathways in
Table 5 and the estimated phosphine dissociation energies
reported above, the profiles in Scheme 3 can be obtained.
This might be unexpected, as typical SN2-like transition
ꢀ
states with Pd O interactions seem to be more sterically
hindered than concerted ones, and monophosphine com-
plexes are less hindered overall than bisphosphine species.
The structures for the optimised transition states of the pre-
ferred pathways, A2ts3 and F1ts2, are shown in Figure 1.
Scheme 3. Computed enthalpy profiles of the oxidative addition step for
each of the two ligands considered.
For the case of the PPh3 ligand, competition between the
pathways going through A2ts3 and A1ts2 has to be consid-
ered (Scheme 3a). The estimated cost of the dissociation of
the phosphine is 3.1 kcalmolꢀ1 higher than A1ts3. Then, the
Csp3 activation with the bisphosphine system is preferred
over the Csp2 activation with the monophosphine catalyst,
in agreement with experiment.
For the P(1-naphthyl)3 system, the comparison is between
pathways through transition states F1ts2 and F2ts3
(Scheme 3b). The low dissociation energy of the Pd(P(1-
naphthyl)3)2 (6.3 kcalmolꢀ1) makes the dissociation favoured
Figure 1. Optimised geometries of the transition states for the lowest
energy pathways A2ts3 and F1ts2. Selected distances are given in ꢆ.
13394
www.chemeurj.org
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13390 – 13397

Palladium-Catalysed Suzuki Cross-Coupling
FULL PAPER
These transition states are representative of SN2 and con-
certed mechanisms. The two possibilities, as well as alterna-
tive conformations, were tested for each of the systems con-
sidered, and only the lowest in energy are reported here.
A2ts3 corresponds to an SN2-like mechanism with an
almost linear Pd-C-Br arrangement. This is representative of
all the transition states found for the activation of the sp3
carbon in this work. Remarkably, the bending of the P-Pd-P
bond angle is minimal in this structure, although this possi-
bility was considered in the starting point of our geometry
optimisations. In contrast, the F1ts2 transition state corre-
sponds to a concerted process, with simultaneous formation
In order to get a deeper insight, we decomposed the rela-
tive energy of the transition state with respect to the reac-
tants in distortion and interaction terms. This same tech-
nique was applied recently by Schoenebeck and Houk to an-
ꢀ
alyse the results of the competitive oxidation of C Br and
[8]
ꢀ
C OTf bonds at palladium centres. This type of analysis is
also known as activation strain model,[22] and it is essentially
the first step of other well-known energy decomposition
analyses.[23]
For the four cases involving PPh3 (A1ts2, A1ts3, A2ts2,
A2ts3), the relative potential energies of the transition
states in the gas phase were decomposed in distortion
ꢀ
ꢀ
of the Pd C (2.037 ꢆ) and Pd Br (2.632 ꢆ) bonds. This is
also the case for the other sp2 carbon activations. Interest-
ingly, the interaction of the sulfoxide oxygen and palladium
seems quite weak in A2ts3, with a distance of 3.080 ꢆ. The
approach of oxygen to palladium seems hindered by the
presence of the two phosphine ligands. For the monophos-
phine case, the distance is much shorter. For instance, in
A1ts3 (see Figure 2), it is 2.162 ꢆ. In spite of this, sp3 activa-
tion is preferred only for bisphosphine systems. Therefore,
this palladium–oxygen interaction, which had been suggest-
ed by us as critical in the oxidative addition of a-bromo sulf-
oxides,[21] does not seem to play a key role in the current se-
lectivity problem.
(DEdist; from the substrate 1a and the catalyst Pd
and interaction (DEint) terms by computing the energy of
the separate fragments Pd(Ph3)n and 1a at the geometry of
ACHTUNGTRENNUNG(PPh3)n)
AHCTUNGTRENNUNG
the transition state. The results are summarised in Table 6.
Table 6. Decomposition of gas phase potential energies (DE°gas) in distor-
tion (DEdist) and interaction (DEint) terms. Energies [kcalmolꢀ1] are rela-
tive to 1a+PdACHTUNTRGNEUNG(PPh3)n.
DE°gas
DEdist Total (Pd
G
DEint
A1ts2
A1ts3
A2ts2
A2ts3
ꢀ12.2
ꢀ10.7
11.6
22.6 (2.3+20.3)
23.3 (0.9+22.4)
38.2 (17.7+20.5)
35.3 (4.6+30.7)
ꢀ34.8
ꢀ34.0
ꢀ26.6
ꢀ24.8
10.5
The analysis is carried out on gas-phase energies for simplic-
ity; the values follow the same trends as the enthalpy values
in solution discussed in the rest of the work. For the mono-
phosphine species, A1ts2 and A1ts3, distortion and interac-
tion terms contribute to a similar extent to the energy differ-
ence between the transition states. The activation of the sp2
carbon is favoured by 0.7 in terms of distortion, and by 0.8
in terms of interaction, resulting in a total difference of
1.5 kcalmolꢀ1 between the transition states. A different sit-
uation is observed for the bisphosphine species A2ts2 and
A2ts3. The interaction term still favours sp2 activation by
1.8 kcalmolꢀ1 (ꢀ26.6 vs. ꢀ24.8 kcalmolꢀ1), but this is more
than compensated by the distortion term, strongly in favour
of sp3 activation by 3.9 kcalmolꢀ1 (35.3 vs. 38.2 kcalmolꢀ1).
Thus, the key lies in the distortion terms for the bisphos-
phine systems.
Further decomposition of the distortion term in the com-
ponents for each reactant is clarifying. The main difference
is in the metal complex side. DEdist of the catalyst is 17.7 kcal
molꢀ1 for A2ts2 and just 4.6 kcalmolꢀ1 for A2ts3.
The concerted transition state A2ts2 needs to distort the
PdACTHUNTGRNENUG(PPh3)2 fragment more than the SN2 transition state
A2ts3. In A2ts2 (Figure 2), the concerted cleavage of the
2
ꢀ
Csp Br bond tends to occupy two coordination sites on pal-
ladium, and this forces the P-Pd-P angle to close down to
120.48, from the approximately linear value in the reactant.
In A2ts3 (Figure 1), only the carbon coordinates to the
metal centre because of the SN2 nature of the transition
state. The P-Pd-P angle remains as large 160.28, and the dis-
tortion energy is much smaller.
Figure 2. Optimised geometries of the transition states for the competing
pathways A1ts3 and A2ts2. Selected distances are given in ꢆ.
Chem. Eur. J. 2010, 16, 13390 – 13397
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13395

F. Maseras, G. Asensio et al.
thyl substituent, and was treated with the UFF method.[29] Frequency cal-
culations of all the minima and transition states were performed to con-
firm their nature. Single-point calculations with the polarisable continu-
um model (PCM)[30] were used to model the solvent effects (tetrahydro-
furan e=7.58). Gaussian 03 default options were chosen, but individual
spheres were placed on all hydrogen atoms to obtain a more accurate
cavity. All energies reported in the text correspond to enthalpies, with
solvation effects included unless otherwise stated. Some free energies in
solution are presented in the text; these energies were computed apply-
ing the free-energy corrections in the gas phase to solvent energies ob-
tained from the PCM calculation. Free energies in the gas phase are pro-
vided in the Supporting Information.
Conclusion
We have shown that the selection of the phosphine–palladi-
um catalyst allows the achievement of total selectivity in
cross-coupling with bifunctional substrates with Csp3 and
Csp2 electrophilic carbon atoms. Similar results were ob-
tained in the case of external competitive experiments. It is
worth noting that there are no precedents of selective palla-
dium-catalysed cross-coupling reactions of this type, in
which substitution is selective at either the sp3- or sp2-hybri-
dised carbon. The observed differences in the behaviour of
xantphos are also remarkable if compared with other bis-
ACHTUNGTRENNUNGphosphines, as is its efficiency at promoting cross-coupling
Acknowledgements
reactions involving Csp3 electrophilic centres, which could
lead to further applications in this challenging field.
Financial support by the Spanish Direcciꢀn General de Investigaciꢀn
Cientꢁfica y Tꢈcnica (CTQ2007—65720 and CTQ2008–06866-CO2–02),
Consolider Ingenio 2010 (CSD2006–0003 and CSD2007–00006) and the
ICIQ foundation is gratefully acknowledged. We gratefully acknowledge
SCSIE (Universidad de Valencia) for access to the instrumental facilities.
M. B. and C.M. thank the Spanish Ministerio de Ciencia e Innovacion for
a “Juan de la Cierva” grant and a “PTA” contract, respectively.
The computational results show that the key to selectivity
is the coordination number in the catalyst. Bisphosphine cat-
alysts favour the activation of the sp3 carbon, while mono-
phosphine catalysts favour the activation of the sp2 carbon.
In the concerted transition state for sp2 activation, the sub-
strate occupies more space in the palladium coordination
sphere, an optimal arrangement for monophosphine cata-
lysts. The case of sp3 carbons is different, because in the SN2
transition state the substrate occupies only one position in
the palladium coordination sphere, thus fitting better with
bisphosphine catalysts.
[1] a) Transition Metal for Organic Synthesis Vol 1, (Eds: M. Beller, C.
Bolm) Wiley-VCH, Weinheim, 2004; b) J. Tsuji Transition Metal Re-
agents and Catalysts Wiley, Chichester, 2000.
[2] G. Espino, A. Kurbangalieva, J. M. Brown, Chem. Commun. 2007,
1742–1744.
[3] A. H. Roy, J. F. Hartwig, J. Am. Chem. Soc. 2003, 125, 13944–13945.
[4] R. Riveiros, L. Saya, J. Pꢈrez Sestelo, L. A. Sarandeses, Eur. J. Org.
Chem. 2008, 1959–1966 and references therein.
[5] For some recent examples, see: a) J.-F. Wei, J. Jiao, J.-J. Feng, J. Lv,
X.-R. Zhang, X.-Y. Shi, Z.-G. Chen, J. Org. Chem. 2009, 74, 6283–
6286; b) W. Ren, M. Yamane, J. Org. Chem. 2009, 74, 8332–8335;
c) X. Zhang, A. Liu, W. Chen, Org. Lett. 2008, 10, 3849–3852; d) G.
Zhou, P. Ting, R. Aslanian, J. J. Piwinski, Org. Lett. 2008, 10, 2517–
2520; e) S. R. Borhade, S. B. Waghmode , Tetrahedron Lett. 2008, 49,
3423–3429; f) J. S. Siddle, A. S. Batsanov, M. R. Bryce, Eur. J. Org.
Chem. 2008, 2746–2750.
[6] L. S. Hegedus, B. C. G. Sçderberg Transition Metal in the Synthesis
of Complex Organic Molecules University Science Books, Sausalito,
California, 2010.
[7] T. Kamikawa, T. Hayashi, Tetrahedron Lett. 1997, 38, 7087–7090.
[8] F. Schoenebeck, K. N. Houk, J. Am. Chem. Soc. 2010, 132, 2496–
2497.
Experimental Section
Palladium-catalysed competitive Suzuki–Miyaura reaction with 1-bromo-
4-(bromomethylsulfinyl)benzene (1a), general procedure: A mixture of
sulfoxide 1a (0.04 mmol), boronic acid 2 (0.8 mmol), CsF (1.6 mmol), Pd-
ACHTUNGTRENNUNG(OAc)2 (0.04 mmol) and the appropriate phosphane ligand (0.04 mmol)
in THF (10 mL) was added to a flask fitted with a reflux condenser and
stirred at 658C under nitrogen. After the appropriate time, the mixture
was cooled to room temperature, quenched with water (10 mL), and ex-
tracted with diethyl ether (2ꢇ15 mL) and dichloromethane (3ꢇ15 mL).
The combined organic extracts were dried with Na2SO4 and evaporated.
The crude material was purified by flash column chromatography.
External competition experiments with bromomethyl sulfoxide 1b and
aryl bromides 5, typical procedure: A dry and N2-flushed Schenk flask
was charged with bromomethyl sulfoxide 1b (0.4 mmol), aryl bromide 5
[9] S. Ishikawa, K. Manabe, Angew. Chem. 2010, 122, 784–787; Angew.
Chem. Int. Ed. 2010, 49, 772–775.
(0.4 mmol), CsF (0.8 mmol), PdACTHNUTRGNE(NUG OAc)2 (0.04 mmol) and the appropriate
[10] a) V. P. Ananikov, D. G. Musaev, K. Morokuma, Eur. J. Inorg.
Chem. 2007, 5390–5399; b) A. A. C. Braga, G. Ujaque, F. Maseras,
Organometallics 2006, 25, 3647–3658; c) A. A. C. Braga, N. H.
Morgon, G. Ujaque, A. Lledos, F. Maseras, J. Organomet. Chem.
2006, 691, 4459–4466; d) A. A. C. Braga, N. H. Morgon, G. Ujaque,
F. Maseras, J. Am. Chem. Soc. 2005, 127, 9298–9307; e) L. J. Goos-
sen, D. Koley, H. Hermann, W. Thiel, Chem. Commun. 2004, 2141–
2143; f) J. P. Stambuli, C. D. Incarvito, M. Buhl, J. F. Hartwig, J. Am.
Chem. Soc. 2004, 126, 1184–1194.
phosphine ligand (0.04 mmol) in THF (10 mL). When the stirred mixture
reached 658C, boronic acid 2 (0.4 mmol) was added in one portion. After
the appropriate time, the mixture was cooled to room temperature,
quenched with water (10 mL), and extracted with diethyl ether (2ꢇ
15 mL) and dichloromethane (3ꢇ15 mL). The combined organic extracts
were dried with Na2SO4 and evaporated under reduced pressure. The
crude material was purified by flash column chromatography.
Computational details: Method
A and monophosphine species on
method F were treated with quantum mechanics (QM); bisphosphine
species on method F (F2add3, F2ts3, F2add2, and F2ts2), were treated
with quantum mechanics/molecular mechanics (QM/MM), namely
ONIOM.[24] QM calculations and QM partitions were carried out with
the B3LYP[25] DFT functional as implemented in Gaussian 03.[26] The
[11] A. Jutand, A. Mosleh, Organometallics 1995, 14, 1810–1817.
[12] Z-B. Dong, G. Manolikakes, L. Shi, P. Kochel, H. Mayr, Chem. Eur.
J. 2010, 16, 248–253.
[13] a) J. H. Kirchhoff, M. R. Netherton, I. D. Hill, G. C. Fu, J. Am.
Chem. Soc. 2002, 124, 13662–13663; b) I. D. Hills, M. R. Netherton,
G. C. Fu, Angew. Chem. 2003, 115, 5927–5930; Angew. Chem. Int.
Ed. 2003, 42, 5749–5752; c) J. Zhou, G. C. Fu, J. Am. Chem. Soc.
2003, 125, 12527–12530; d) J. Zhou, G. C. Fu, J. Am. Chem. Soc.
2003, 125, 14726–14727; e) M. R. Netherton, G. C. Fu, Angew.
Chem. 2002, 114, 4066–4068; Angew. Chem. Int. Ed. 2002, 41, 3910–
basis sets used were the standard split-valence polarised 6–31+GACHTNUTGRNEUNG
(d,p)[27]
for P, S, O, Br, C, and H, except for C and H atoms of the phosphine li-
gands for which 6–31G(d) was used. SDD valence basis sets with the as-
sociated effective core potentials were used for palladium.[28] The MM
part of the ONIOM calculations consisted of the outer ring of each naph-
13396
www.chemeurj.org
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13390 – 13397

Palladium-Catalysed Suzuki Cross-Coupling
FULL PAPER
3912; f) J. H. Kirchhoff, C. Dai, G. C. Fu, Angew. Chem. 2002, 114,
2025–2027; Angew. Chem. Int. Ed. 2002, 41, 1945–1947; g) M. R.
Netherton, C. Dai, K. Neuschꢉtz, G. C. Fu, J. Am. Chem. Soc. 2001,
123, 10099–10100.
phens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys. Chem.
1994, 98, 11623–11627.
[26] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. J. A. Montgom-
[14] A. F. Littke, G. C. Fu, Angew. Chem. 2002, 114, 4350–4386; Angew.
Chem. Int. Ed. 2002, 41, 4176–4211.
[15] a) M. Sato, N. Miyaura, A. Suzuki, Chem. Lett. 1989, 1405–1408;
b) L. J. Gooben, Chem. Commun. 2001, 669–670; c) Y-Z Duan, M-
Z. Deng, Tetrahedron Lett. 2003, 44, 3423–3426; d) X-X. Liu, M-Z.
Deng, Chem. Commun. 2002, 622–623.
[16] a) S. M. Nobre, A. L. Monteiro, Tetrahedron Lett. 2004, 45, 8225–
8228; b) S. Chowdhury, P. E. Georghiou, Tetrahedron Lett. 1999, 40,
7599–7603.
[17] a) N. Rodrꢁguez, A. Cuenca, C. Ramꢁrez de Arellano, M. Medio-
Simꢀn, G. Asensio, Org. Lett. 2003, 5, 1705–1708; b) N. Rodrꢁguez,
A. Cuenca, C. Ramꢁrez de Arellano, M. Medio-Simꢀn, D. Peine, G.
Asensio, J. Org. Chem. 2004, 69, 8070–8076; c) N. Rodrꢁguez, C.
Ramꢁrez de Arellano, G. Asensio, M. Medio-Simꢀn, Chem. Eur. J.
2007, 13, 4223–4229.
ACHTUNGERTNeNUNG ry, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J.Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[18] a) T. E. Barder, S. D. Walker, J. R. Martinelli, S. L. Buchwald, J. Am.
Chem. Soc. 2005, 127, 4685–4696; b) F. Barrios-Landeros, B. P.
Carrow, J. F. Hartwig, J. Am. Chem. Soc. 2009, 131, 8141–8154.
[19] The slower reaction rate observed when the amount of bulky phos-
phine used was >30% mol correlates well with the hypothesis of a
monophosphine complex catalyst.
[27] a) M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S.
Gordon, D. J. Defrees, J. A. Pople, J. Chem. Phys. 1982, 77, 3654–
3665; b) P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28,
213–222; c) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys.
1972, 56, 2257–2261.
[20] a) C. Sicre, A. A. C. Braga, F. Maseras, M. M. Cid, Tetrahedron
2008, 64, 7437–7443; b) E. Zuidema, P. van Leeuwen, C. Bo, Orga-
nometallics 2005, 24, 3703–3710.
[21] C. Gourlaouen, G. Ujaque, A. Lledos, M. Medio-Simon, G. Asensio,
F. Maseras, J. Org. Chem. 2009, 74, 4049–4054.
[22] F. M. Bickelhaupt, J. Comput. Chem. 1999, 20, 114–128.
[23] a) K. Kitaura, K. Morokuma, Int. J. Quantum Chem. 1976, 10, 325–
340; b) T. Ziegler, A. Rauk, Theor. Chim. Acta 1977, 46, 1–10.
[24] a) F. Maseras, K. Morokuma, J. Comput. Chem. 1995, 16, 1170–
1179; b) S. Dapprich, I. Komaromi, K. S. Byun, K. Morokuma, M. J.
Frisch, THEOCHEM 1999, 461–462, 1–21.
[28] D. Andrae, U. Haussermann, M. Dolg, H. Stoll, H. Preuss, Theor.
Chim. Acta 1990, 77, 123–141.
[29] A. K. Rappꢈ, C. J. Casewit, K. S. Colwell, W. A. Goddard III, W. M.
Skiff, J. Am. Chem. Soc. 1992, 114, 10024–10035.
[30] a) M. Cossi, V. Barone, B. Mennucci, J. Tomasi, Chem. Phys. Lett.
1998, 286, 253–260; b) S. Miertus, J. Tomasi, Chem. Phys. 1982, 65,
239–245; c) B. Mennucci, J. Tomasi, J. Chem. Phys. 1997, 106, 5151–
5158; d) E. Cances, B. Mennucci, J. Tomasi, J. Chem. Phys. 1997,
107, 3032–3041.
[25] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. T. Lee,
W. T. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789; c) P. J. Ste-
Received: April 27, 2010
Published online: October 7, 2010
Chem. Eur. J. 2010, 16, 13390 – 13397
ꢅ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13397