Platinum-catalyzed phenyl and methyl group transfer from tin to iridium: Evidence for an autocatalytic reaction pathway with an unusual preference for methyl transfer

Article abstract of DOI:10.1021/ja710365x

Platinum complexes have been found to catalyze the transfer of σ-bound ligands to the Ir center in Cp^*(PMe₃)IrCl₂ (C^p* = η⁵-C₅Me₅) from Bu₃SnPh and Ph_xSnMe<

Full text of DOI:10.1021/ja710365x

Published on Web 01/17/2008
Platinum-Catalyzed Phenyl and Methyl Group Transfer from Tin to Iridium:
Evidence for an Autocatalytic Reaction Pathway with an Unusual Preference
for Methyl Transfer
Stuart E. Smith,† Jennifer M. Sasaki,† Robert G. Bergman,*,† Joseph E. Mondloch,‡ and
Richard G. Finke‡
Department of Chemistry, UniVersity of California, Berkeley, California 94720-1460, and Department of Chemistry,
Colorado State UniVersity, Fort Collins, Colorado 80523
Received November 15, 2007; E-mail: rbergman@berkeley.edu
Scheme 1. Representative Mechanism for Metal-Catalyzed
Cross-Coupling Reactions
Transition-metal-catalyzed cross-coupling reactions (cycle A in
Scheme 1) have enjoyed wide application in organic synthesis and
are now extensively used for the efficient formation of carbon-
carbon and carbon-heteroatom bonds.1 The synthesis of these
complexes is believed to occur via intermediates containing metal-
heteroatom and metal-carbon bonds.1 In principle, analogous
coupling reactions could be devised in which a metal complex could
be used to catalyze the formation of a bond between a different
metal center and a carbon or heteroatom (cycle B in Scheme 1),
generating products potentially analogous to the above proposed
intermediates. The ability to catalytically synthesize these types of
complexes could have a wide impact on research areas in which
such complexes are employed or are postulated as intermediates,
especially if mechanistic insight can be gained to guide new
synthetic developments.
However, the application of transition-metal-catalyzed coupling
reactions to the synthesis of organometallic complexes remains
relatively rare. A few recent examples have come from the
laboratories of Lo Sterzo2 and Komiya.3 In both cases, it has been
proposed that insertion of a transition metal catalyst, M1, into a
transition metal complex M2-X bond (X ) halide, H) plays a key
role in the overall transformation. The M2-M1-X intermediate is
analogous to the R1-M1-X intermediate implicated in typical
organic cross-coupling reactions (Scheme 1).4 This M2-M1-X
intermediate is then assumed to react via a transmetalation/insertion
followed by reductive elimination, yielding the organometallic
product.
It seems likely that many organometallic reactions that are
catalyzed by a second metal complex remain to be discovered. To
investigate this, we have explored the possibility that Stille-type
reactions can be developed to promote the metal-catalyzed transfer
of an alkyl or aryl group from a tin reagent to the iridium center in
Cp*(PMe3)IrCl2 (1) (Cp* ) η5-C5Me5; eq 1). The resulting aryl/
alkyl chloride complexes are important precursors to catalysts for
H/D exchange5 and for the hydration of alkenes6 and C-H bond
activation.7 We have found that such reactions can be catalyzed
by coordinatively unsaturated platinum complexes. However, the
mechanism is much more complicated than the type of routes
proposed previously.2,3
Our initial experiments targeted the transmetalation of a phenyl
group from Bu3SnPh (2) to 1 to generate [Ir]PhCl (3) ([Ir] ) Cp*-
(PMe3)Ir; eq 1), resulting in the formation of a new iridium-carbon
bond. A variety of potential transition metal catalysts were examined
to effect the desired cross-coupling reaction (Table 1). The most
promising complexes were platinum-based. While catalysts derived
from Pt(dba)2 gave the best yields,8 we chose to use Pt(PtBu3)2 (4)
for our subsequent studies due to its solubility and the ability to
1
monitor 4 by both H and 31P NMR.
In an effort to lower the reaction temperature, the use of Me3-
SnPh (5a) was explored with the hope that decreased steric repulsion
would increase the reaction rate. Surprisingly, use of this compound
did not furnish 3. Instead, exclusive methyl transfer was observed
at 75 °C, yielding [Ir]MeCl (6). The selectivity for methyl transfer
contrasts with the preferential aryl transfer from tin reagents
observed in organic cross-coupling reactions,9 as well as with the
organometallic reactions studied by Lo Sterzo where selective
alkynyl transfer from R3Sn-R′ (R ) Bu, Me, R′ ) alkyne) is
observed.10 Tin reagents bearing both Ph and alkyl groups com-
monly transfer the Ph rather than the Me group in electrophilic
reactions.11 In our system, only methyl transfer is observed with
the tin reagents Me2SnPh2 (5b) and MeSnPh3 (5c), in addition to
SnMe4 (5d), where only methyl transfer is possible (Table 2).
In the absence of the Pt catalyst 4, a slow background reaction
is observed. However, in this thermal process, selective phenyl
transfer occurs at 105 °C from 5c, while both phenyl and methyl
transfers occur from 5a (Table 2).12 At 75 °C, these reactions are
extremely sluggish. Thus, 4 is crucial to controlling the rate and
chemoselectivity of the reaction.
Our first evidence for mechanistic complexity appeared when
reaction monitoring revealed that all of the cross-coupling reactions
studied showed the gradual formation of a new Pt-H complex,
identified as (PtBu3)2Pt(H)(Cl) (7). Since the source of the HCl in
this complex was unclear, we sought to determine whether the
presence of adventitious acid could be responsible for the catalytic
† University of California, Berkeley.
‡ Colorado State University.
9
10.1021/ja710365x CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 1839-1841
1839

C O M M U N I C A T I O N S
Table 1. Preliminary Examination of Transition-Metal-Catalyzed
Carbon-Iridium Cross-Coupling
Scheme 2. Proposed Mechanism
catalyst
(mol %)
ligand
temp
C)
time
(h)
(mol %)
(
°
% 3a
Pt(PPh3)4 (9)
NAd
135
135
135
105
135
135
130
135
130
26
26
26
23
26
26
22
21
22
0
0
0
<2
6
9
20
40
48
(COD)PtCl2 (30)
[(COD)IrCl]2 (20)
(CH3CN)2PdCl2c (15)
(COD)2Ni (55)
(COD)2Ni (36)
Pt(PtBu3)2 (9)
P(tBu)3 (45)
P(tBu)3 (57)
NA
P(tBu)3 (80)
NA
NA
NA
P(Mes)3 (9)
Pt(dba)2b (15)
Pt(dba)2 (9)
a Yield of product measured by 31P NMR. b Pt black formed. c Reaction
performed in DMF, and Pd black formed almost immediately. d NA ) not
applicable.
reaction of the product SnMe3Cl (8) with the surface hydroxyls of
the glass, and only silylation is effective in inhibiting this process.
The effective removal of adventitious acid allowed us to initiate
kinetic studies of the purely Pt-catalyzed reaction. Early experiments
revealed a novel feature of the reaction: autocatalysis (vide infra).
A plot of the loss of 1 versus time did not give simple kinetic
behavior but, instead, demonstrated the presence of an induction
period, which implies that 4 is not the active catalyst. Potential
detailed mechanisms that might explain this induction period are
(1) the generation16 of colloidal Pt0, (2) the presence of radical
intermediates,17 and (3) direct reaction of one of the products with
the catalyst precursor, 4.18
Table 2. Platinum-Catalyzed Sn-Me Transfer to 1
time
tin complex
(h)
% 1a
% 6a
Me3SnPh (5a)b
Me2SnPh2 (5b)
MeSnPh3 (5c)c
Me4Sn (5d)
25
46
140
29
18
36
57
15
74
59
24
85
Evidence against the generation of a colloidal Pt0 catalyst was
obtained by performing control reactions in the presence of excess
Hg0.19 No inhibition of the Me transfer reaction is seen, arguing
against colloidal Pt0 as the active catalyst. Similarly, evidence
against a radical reaction was obtained by performing a reaction in
the presence of the radical initiator AIBN (43 mol %) and a separate
reaction in the presence of the radical inhibitor BHT (120 mol %).
In both experiments, the induction period remained and there was
no dramatic alteration in the reaction rate. Hence, it appears unlikely
that the reaction proceeds through radical intermediates. The third
potential mechanism was tested by addition of the stoichiometric
products to the reaction mixture. Addition of 6 did not affect the
induction period, but the addition of 10 mol % of SnMe3Cl, 8,
substantially reduced it. Thus, it appears that reaction of 8 with 4
is required to form the active catalyst (Scheme 2).20 Addition of 1
equiv of free P(tBu)3 to a reaction mixture containing 10 mol % of
8 and 12 mol % of 4 severely inhibits the reaction, indicating that
dissociation of P(tBu)3 is also required for catalysis to ensue.21
A mechanism consistent with our currently available data is
shown in Scheme 2. We propose that the induction period results
from the slow direct reaction of 1 with 5d, producing 6 and 8.
Once 8 forms, a second reaction occurs with the precatalyst 4 to
produce a key species within the catalytic cycle, 9a, eventually
leading to autocatalysis and the signature sigmoidal curve of
autocatalytic kinetics (Figure 1). Included in the Supporting
Information is an explanation of how the proposed mechanism
accounts for the observed autocatalysis and a fit to an idealistic
autocatalytic model, which proved to be the key to discovering the
proposed mechanism. All of these data provide compelling evidence
for an autocatalytic mechanism. Additionally, we have been able
to fit the more detailed mechanism proposed in Scheme 2 directly
to the experimental data as shown in Figure 1.
a Yield of product measured by 31P NMR. b Reaction performed in THF-
d8. c Catalyst was generated in situ from Pt(dba)2 and P(tBu)3.
transformation of 1 to 6 by carrying out a control experiment with
deliberately added acid. This revealed that HCl, in the absence of
Pt, is capable of catalytically converting 1 to 6 with tin complexes
5a and 5d. However, with tin complex 5b, both phenyl and methyl
transfers occur, and only phenyl transfer occurs with 5c.13 This
differentiates the acid-catalyzed from the metal-catalyzed pathway.
In order to prevent complications resulting from the acid-catalyzed
background, reactions catalyzed by 4 only were performed in the
presence of base to scavenge adventitious acid. Amine bases were
found to shut down the HCl-catalyzed transmetalation reaction.
However, the resulting ammonium salts can apparently still
protonate 4, as hydride 7 was formed in the presence of the amine
bases.14 All other stronger bases examined (e.g., Verkade’s base,15
KOtBu, sodium 2-methylbutan-2-oxide, (dimethylamino)trimeth-
yltin(IV), sodium malononitrile) reacted with either 1, the solvent,
or 4, rendering them incompatible. Chloride is not responsible for
the conversion of 1 to 6 since it inhibits the transmetalation:
reactions performed in the presence of 1 equiv of (Ph3P)2N+Cl-
did not produce any product.
Several experiments were carried out to establish the source of
adventitious acid and remove it effectively. Basification of the
glassware with 1 M NaOH, filtration of all reactants but 4 through
basic alumina(I), and recrystallization of catalyst 4 from hexanes
failed to stop the formation of 7. It was ultimately discovered that
silylation of the glassware prior to reaction prevented the production
of 7. Our tentative conclusion is that generation of HCl occurs via
9
1840 J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008

C O M M U N I C A T I O N S
Supporting Information Available: Experimental details for the
reaction of 1 with tin complexes and for the curve-fitting are available.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Metal-catalyzed Cross-coupling Reactions, 2nd ed.; de Meijere, A.,
Diederich, F., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim,
Germany, 2004.
(2) Angelucci, F.; Ricci, A.; Masi, D.; Bianchini, C.; Lo Sterzo, C.
Organometallics 2004, 23, 4105.
(3) Kuramoto, A.; Nakanishi, K.; Kawabata, T.; Komine, N.; Hirano, M.;
Komiya, S. Organometallics 2006, 25, 311.
(4) For recent reviews, see: (a) Frisch, A. C.; Beller, M. Angew. Chem., Int.
Ed. 2005, 44, 674. (b) Suzuki, A. Chem. Commun. 2005, 4759. (c)
Christmann, U.; Vilar, R. Angew. Chem., Int. Ed. 2005, 44, 366. (d)
Espinet, P.; Echavarren, A. M. Angew. Chem., Int. Ed. 2004, 43, 4704.
(e) Ca´rdenas, D. J.; Echavarren, A. M. New J. Chem. 2004, 28, 338. (f)
Amatore, C.; Jutand, A. Acc. Chem. Res. 2000, 33, 314. (g) Amatore, C.;
Jutand, A. J. Organomet. Chem. 1999, 576, 254.
(5) (a) Yung, C. M.; Skaddan, M. B.; Bergman, R. G. J. Am. Chem. Soc.
2004, 126, 13033. (b) Skaddan, M. B.; Yung, C. M.; Bergman, R. G.
Org. Lett. 2004, 6, 11.
(6) (a) Woerpel, K. A.; Bergman, R. G. J. Am. Chem. Soc. 1993, 115, 7888.
(b) Ritter, J. C. M.; Bergman, R. G. J. Am. Chem. Soc. 1997, 119, 2580.
(7) Arndtsen, B. A.; Bergman, R. G. Science 1995, 270, 1970.
(8) There is some confusion in the literature regarding the Pt:dba ratio for
this complex. The Pt complex we synthesized matched the expected
elemental analysis for “[Pt(dba)2]n” and is thus reported as Pt(dba)2. (a)
Lewis, L. N.; Krafft, T. A.; Huffman, J. C. Inorg. Chem. 1992, 31, 3555.
(b) Moseley, K.; Maitlis, P. M. J. Chem. Soc., Dalton Trans. 1974, 169.
(c) Moseley, K.; Maitlis, P. M. J. Chem. Soc., Chem. Commun. 1971,
982.
(9) Labadie, J. W.; Stille, J. K. J. Am. Chem. Soc. 1983, 105, 6129.
(10) (a) Crescenzi, R.; Lo Sterzo, C. Organometallics 1992, 11, 4301. (b) Viola,
E.; Lo Sterzo, C.; Crescenzi, R.; Frachey, G. J. Organomet. Chem. 1995,
493, 55.
Figure 1. Concentration versus time data and curve-fit to the mechanism
in Scheme 2. Conditions: 1 (0.0278 M) with 5d (0.0445 M) and 4 (0.0050
M). The reaction was performed in dioxane-d8 at 74 °C and monitored by
1
1H NMR. There is some overlap of the H NMR signals for 1, 6, and 4,
which is the reason for the relatively large amount of scatter about the curve-
fit for 1 and 4. Rate constants for the fit are k1 ) 1.0 × 10-3 M-1 min-1
,
k2 ) 2.0 × 10-8 M-1 min-1, k-2 ) 580 M-1 min-1, k3 = 1.0 × 108 M-1
min-1, k4 ) 3.8 × 105 M-1 min-1. Additional fits and rate constants
documenting the autocatalysis are provided in the Supporting Information.
The autocatalysis and sigmoidal kinetics in Figure 1 are of
considerable fundamental interest. Specifically, they represent a
process that is much more common in nature22 but, at present,
infrequently identified in relatively simple characterizable chemical
systems,22 in which a slower (sometimes even undesired or
“wrong”) reaction22 evolves chemically to a faster, “smarter”
process. This can be considered as a primitive type of chemical
evolution,22 in this case evolution of a superior, faster catalytic
process.
(11) For example, it has been shown that the relative rate of Ph:Me cleavage
by mercury(II) halides is ca. 400-500:1. Abraham, M. H.; Sedaghat-
Herati, M. R. J. Chem. Soc., Perkin Trans. 2 1978, 729.
(12) We believe that the background reaction proceeds by slow ionization of
1, yielding [Ir]Cl+, followed by transmetalation of the Sn complexes to
this cationic unsaturated intermediate. The ratio of Ph:Me transfer will
then depend on the Sn complex employed.
(13) We believe that HCl protonates the Sn complex, generating the corre-
sponding Sn-Cl complex, which then acts as a Lewis acid and assists in
the ionization of 1, thus facilitating the background reaction.
(14) Addition of anilinium chloride to 4 causes the quantitative formation of
7. During further investigations into the HCl source, it was found that 4
slowly decomposes to 7 in CD2Cl2; thus, the ethereal solvents dioxane
and THF were utilized instead. Varying the amount of added Hu¨nig’s
base (1 and 11 equiv vs 1) revealed that the rates remained qualitatively
constant, confirming that the observed catalysis results from 4 and not
from HCl. No colloidal Pt0 is observed in these reactions, the only
observable Pt species being 4 and 7 by NMR. The use of 7, in the absence
of 4, produces 6 only after a substantial induction period, indicating that
7 is not the active catalyst.
(15) Tang, J.; Dopke, J.; Verkade, J. G. J. Am. Chem. Soc. 1993, 115, 5015.
(16) (a) Besson, C.; Finney, E. E.; Finke, R. G. J. Am. Chem. Soc. 2005, 127,
8179. (b) Begum, R. A.; Chanda, N.; Ramakrishna, T. V. V.; Sharp, P. S.
J. Am. Chem. Soc. 2005, 127, 13494.
(17) Yin, C.; Finke, R. G. Inorg. Chem. 2005, 44, 4175.
(18) Lesutis, H. P.; Gla¨ser, R.; Liotta, C. L.; Eckert, C. A. Chem. Commun.
1999, 2063.
(19) Whitesides, G. M.; Hackett, M.; Brainard, R. L.; Lavalleye, J. P. M.;
Sowinski, A. F.; Izumi, A. N.; Moore, S. S.; Brown, D. W.; Staudt, E.
M. Organometallics 1985, 4, 1819.
(20) A control experiment in which a co-catalytic amount of 8 was added to
a reaction mixture devoid of 4 proceeds at a much slower rate than when
4 is present, demonstrating that the catalysis does not result solely from
8.
(21) Addition of 55 mol % of N(hex)4Cl to a reaction mixture containing 10
mol % of 8 and 4 also inhibits the reaction. It is currently unknown whether
chloride dissociation from 8 is required prior to reaction with 4 or whether
the chloride is simply binding an unsaturated intermediate. Current studies
are focused on addressing these possibilities.
(22) Due to space limitations, the references documenting these points are
available in the Supporting Information.
In conclusion, an autocatalytic mechanism appears to be opera-
tional for this metal-catalyzed organometallic reaction, wherein a
slow direct background reaction of 1 with 5d is necessary to form
8. Once 8 is formed, a reaction with the catalyst precursor 4 occurs
to yield an active catalytic species 9a. Additionally, even when
the tin reagent bears Sn-Ph bonds, selective Sn-Me transfer
occurs. Overall, this system is a rare example of a metal-catalyzed
metal-carbon bond-forming reaction in which the catalyst is
produced in an autocatalytic reaction. A key to the reaction is that
the catalytic intermediate Pt(Me)(PtBu3)SnMe3 is superior to SnMe4
in transferring a Me group to 1. One presumes that this feature
will extend to other M-Cl (M ) transition metal) systems, although
additional work will be required to elucidate the more general
aspects of the mechanism proposed in Scheme 2. Current work is
aimed at providing additional information on the reaction mecha-
nism, synthesizing the putative catalytic intermediates, understand-
ing the unusual selectivity for Me transfer, and applying this method
to the synthesis of metal-heteroatom bonds.
Acknowledgment. This work was supported by NSF Grant
CHE-0345488 and a Eugene Cota-Robles Fellowship to S.E.S.
Work with the A f B and A + B f 2B autocatalytic model and
associated curve-fitting studies done at Colorado State University
were supported by DOE Grant DE-FG02-03ER15453.
JA710365X
9
J. AM. CHEM. SOC. VOL. 130, NO. 6, 2008 1841