Why excess cyanide can be detrimental to Pd-catalyzed cyanation of haloarenes. Facile formation and characterization of [Pd(CN)3(H)] 2 and [Pd(CN)3(Ph)]2-

Article abstract of DOI:10.1021/ja066931d

Unexpected side reactions at the Pd center have been identified for Pd-catalyzed cyanation of haloarenes. In the presence of excess cyanide, these reactions can proceed with ease, interfering with every key step of the catalytic cycle and thereby disrupting the process. Mixed PPh₃/CN· Pd(0) species easily activate the C-N bond of Bu₄N⁺ to produce cleanly and selectively a novel hydride [(CN)₃PdH]² which is stable in the presence of extra CN^- but easily undergoes HCN reductive elimination under CN^--free conditions. While slow addition of 1 equiv of CN^- to [(Ph₃P)₂Pd(Ph)I] leads to PhCN formation, quick mixing of the same complex with a 4-fold excess of CN- does not produce PhCN but rather leads to [(CN)₃Pd(Ph)]^2-. Reductive elimination of PhCN from the latter is strongly inhibited by small quantities of free cyanide. Copyright

Full text of DOI:10.1021/ja066931d

Published on Web 12/14/2006
Why Excess Cyanide Can Be Detrimental to Pd-Catalyzed Cyanation of
Haloarenes. Facile Formation and Characterization of [Pd(CN)3(H)]2- and
[Pd(CN)3(Ph)]2-
Kerwin D. Dobbs, William J. Marshall, and Vladimir V. Grushin*
Central Research & DeVelopment, E. I. DuPont de Nemours and Company, Inc., Experimental Station,
Wilmington, Delaware 19880-0328
Received September 26, 2006; E-mail: vlad.grushin-1@usa.dupont.com
Palladium-catalyzed cyanation of haloarenes (eq 1) is one of the
most important modern methods for the introduction of the cyano
group into the aromatic ring.1 This reaction is extensively used in
the synthesis of pharmaceuticals, agrochemicals, dyes, and various
intermediates.1
Ar-X + CN- Pd catalyst8 Ar-CN + X-
(1)
Figure 1. 1H NMR signal from [(13CN)3Pd(H)]2- (trans-JC-H ) 64.0 Hz;
cis-JC-H ) 2.3 Hz).
Since the discovery2 of Pd- and Ni-catalyzed aromatic cyanation,
it has been widely recognized that the amount of free cyanide in
solution during the reaction must be carefully controlled. While
being one of the two reagents (eq 1), the cyanide ion in excess can
deactivate the catalyst, which leads to process termination.1 This
has been recently re-confirmed in a special study.3 Furthermore,
the same coupling constant JC-C ) 6.8 Hz. The formation of Bu3N
1
and 1-butene was confirmed by H and 13C NMR data. A sharp
singlet at -5.5 ppm was observed in the 31P NMR spectrum of the
reaction solution, pointing to full release of all of the PPh3 originally
bound to the starting Pd(0). While [Bu4N]+2‚1 is an oil at room
temperature, [PPN]+2‚1 obtained from [Bu4N]+2‚1 and [PPN]+ CN-
(PPN ) Ph3PNPPh3) by metathesis is a solid and was isolated (see
below).8
Zn(CN)2 and K4[Fe(CN)6]5 with weakly ionizable M-CN bonds
4
have been found advantageous for use in reaction 1, as compared
to the conventional ionic cyanide sources, KCN and NaCN.
Surprisingly, the origin of the detrimental effect of the CN- in
excess remains unclear, except it has been shown6 that [Bu4N]+ CN-
retards the ability of [(Ph3P)4Pd] to oxidatively add an aryl halide.
In this communication, we report that each key step of the catalytic
loop for the cyanation process (Scheme 1) can be affected by excess
cyanide in a startling manner.
Extra PPh3 strongly inhibited reaction 2. When run with [Pd] )
1.7 × 10-2 M and [13CN-] ) 6.9 × 10-2 M in THF-d8 at 25 °C,
reaction 2 proceeded to ca. 80 and 20% conversion (13C NMR)
after 6 h without and with extra PPh3 (6 equiv per Pd), respectively.
Full conversion was achieved faster for more dilute solutions,
certainly of relevance to the catalytic process (eq 1) that employs
low concentrations of Pd catalysts. Thus, the above-described
reaction without extra PPh3 reached full conversion (complete
discoloration) after 1 day. Under identical, rigorously dry conditions
but with twice the concentration of the reagents, full color loss
was observed only after 1 week. Added water (2 equiv per Pd)
neither had a dramatic effect on reaction 2 nor decomposed 1.9
No reaction between [Bu4N]+ X- (X ) I, Cl) and [(Ph3P)4Pd]
was observed under similar conditions. It is clear, however, that
the use of R4N+ X- as phase-transfer catalysts for reaction 11 might
easily damage the active Pd catalyst, as shown in eq 2. The
uncommon reactivity (eq 2) apparently originates from mixed
cyano-phosphine Pd(0) species. Our preliminary low-temperature
NMR (13C and 31P) and computational DFT studies indicated that
both [(Ph3P)3Pd(CN)]- and [(Ph3P)2Pd(CN)2]2- may be formed
upon treatment of [(Ph3P)4Pd]10 with the CN- (see Supporting
Information).11 Such anionic Pd(0) complexes are fully expected
to exhibit enhanced reactivity toward oxidative addition, especially
upon loss of one more phosphine ligand.12 It is conceivable12,13
that a C-N bond of the [Bu4N]+ oxidatively adds to a mixed
phosphine-cyanide Pd(0) species, followed by â-elimination and
ligand exchange to produce 1. Our computations, however, point
to significant localization of a negative charge on the N atom(s) of
[(Ph3P)nPd(CN)m]m- (see Supporting Information). Therefore, clas-
sical Hofmann elimination involving deprotonation at a â-position
of [Bu4N]+ should not be ruled out at this point. Both N-protonation
of coordinated cyanide14 and H-transfer from LnMCNH to form
Scheme 1. Accepted Simplified Mechanism for Reaction 1
We found that adding anhydrous [Bu4N]+ CN- to [(Ph3P)4Pd]
in dry THF resulted in immediate color change from yellow to tan,
after which the solution gradually faded to colorless. Analysis of
the reaction products revealed a remarkably selective, unexpected
transformation, as shown in eq 2.
The formation of the dianionic hydride [(CN)3Pd(H)]2- (1)7 was
established by running the experiment with [Bu4N]+ 13CN- and
observing the hydrido resonance as a doublet of triplets at -10.85
1
ppm (1H NMR, Figure 1). In the H{13C} NMR spectrum, this
resonance appeared as a singlet. The 13C NMR spectrum of 1-13C
displayed two signals, a doublet at 144.6 ppm (2C, mutually trans-
CN ligands) and a triplet at 148.0 ppm (1C, CN trans to H) with
9
30
J. AM. CHEM. SOC. 2007, 129, 30-31
10.1021/ja066931d CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
LnM(H)CN14b are well documented. A detailed mechanistic study
of reaction 2 is currently underway.
coordinate Pd(II) complexes.20 It is likely therefore that predisso-
ciation of one cyanide ligand from 1 and 2 is required for the
reductive elimination to occur. Extra CN- shifts the equilibria
toward nondissociated, four-coordinate 1 and 2 which are not prone
to reductive elimination under these conditions.
In conclusion, unexpected side reactions at the Pd center have
been identified for Pd-catalyzed cyanation of haloarenes (eq 1). In
the presence of excess cyanide, these reactions can proceed with
ease, interfering with the key steps of the catalytic loop (Scheme
1) and thereby disrupting the process.
Regardless of whether the cyanation (eq 1) is run with or without
[R4N]+ as a phase-transfer agent, excess CN- can poison the Pd
catalyst at Step 2 of the catalytic loop (Scheme 1). When an
equimolar amount of [Bu4N]+ CN- was slowly added to [(Ph3P)2-
Pd(I)Ph] in THF at 25 °C, an instantaneous reaction occurred to
produce PhCN and Pd(0). In sharp contrast, quick mixing of
[(Ph3P)2Pd(I)Ph] with a 4-fold excess of [Bu4N]+ CN- in THF
resulted in no PhCN formation but rather led selectively to stable
[(CN)3PdPh]2- (2); see Scheme 2.15 Similarly, the reaction of
[(Ph3P)2Pd(I)Ph] with 4-6 equiv of K13CN in a biphasic D2O/
benzene system cleanly afforded K2[(13CN)3PdPh] (aqueous phase;
1H and 13C NMR) and free PPh3 (organic phase; 31P NMR).16
Assuming the NMR detection limit at ca. 1%, the exclusive
formation of either PhCN or 2 (Scheme 2) indicates that rate
constants for the I-/CN- exchange, Ph3P/CN- exchange, and Ph-
CN reductive elimination decrease in the order kI/CN > kP/CN g kRE
× 102.
Acknowledgment. This is DuPont CR&D Contribution No.
8745. We thank Ms. Laurie A. Howe and Dr. D. Christopher Roe
for selected NMR experiments, and Drs. Albert L. Casalnuovo, Jo¨rg
Bru¨ning, and Stuart A. Macgregor for discussions.
Supporting Information Available: Experimental and computa-
tional details, NMR data (PDF), and X-ray analysis data (CIF) for
[PPN]+2‚2. This material is available free of charge via the Internet at
http://pubs.acs.org.
Scheme 2. Reactions of [(Ph3P)2Pd(I)Ph] with 1 Equivalent of
References
[Bu4N]+ CN- and with 4 Equivalents of [Bu4N]+ CN- in THF
(1) For a recent review, see: Sundermeier, M.; Zapf, A.; Beller, M. Eur. J.
Inorg. Chem. 2003, 3513.
(2) (a) Pd: Takagi, K.; Okamoto, T.; Sakakibara, Y.; Oka, S. Chem. Lett.
1973, 471. (b) Ni: Cassar, L. J. Organomet. Chem. 1973, 54, C57.
(3) Marcantonio, K. M.; Frey, L. F.; Liu, Y.; Chen, Y.; Strine, J.; Phenix, B.;
Wallace, D. J.; Chen, C.-y. Org. Lett. 2004, 6, 3723.
(4) Tschaen, D. M.; Desmond, R.; King, A. O.; Fortin, M. C.; Pipik, B.; King,
S.; Verhoeven, T. R. Synth. Commun. 1994, 24, 887.
(5) Schareina, T.; Zapf, A.; Beller, M. Chem. Commun. 2004, 1388.
(6) Sundermeier, M.; Zapf, A.; Mutyala, S.; Baumann, W.; Sans, J.; Weiss,
S.; Beller, M. Chem.sEur. J. 2003, 9, 1828.
(7) (a) A material has been formulated as a neutral, four-coordinate Pd(IV)
complex [HPd(CN)3], based on C,H-analysis data: Nozaki, K. U.S. Patent
3835123, 1974. (b) [HPt(CN)3]2- has been reported: Almeida, J. F.;
Pidcock, A. J. Organomet. Chem. 1981, 208, 273.
(8) X-ray analysis of [PPN]+2‚1 was inconclusive because of disorder.
(9) This stability of 1 is uncommon in the chemistry of palladium hydrides:
Grushin, V. V. Chem. ReV. 1996, 96, 2011.
(10) [(Ph3P)4Pd] undergoes full dissociation of one PPh3 in solution: Mann,
B. E.; Musco, A. J. Chem. Soc., Dalton Trans. 1975, 1673.
Like [Bu4N]+2‚1, [Bu4N]+2‚2 is an oil at room temperature but
[PPN]+2‚2 is a solid that was isolated and structurally characterized
(Scheme 2). Although cyanide is known to displace R3P in Pd
complexes,17 this remarkably facile, apparently irreversible R3P/
CN- ligand exchange has not received sufficient investigation in
previous studies of Pd-catalyzed aromatic cyanation (eq 1). We
found that not only PPh3 but even more strongly binding PCy3 is
easily and fully released from Pd in the presence of the CN-.
Treatment of [(Cy3P)2Pd(Cl)Ph] with excess of [Bu4N]+ CN- in
THF cleanly produced free PCy3 and 2 in quantitative yield.
Both dianions 1 and 2 were found to be stable in solution in the
presence of extra cyanide. However, freshly isolated well-shaped
white crystals of [PPN]+2‚1 turned grayish on the surface upon
drying under vacuum for only a few minutes.18 When this salt was
redissolved in MeCN-d3, the originally solid-free, colorless solution
turned cloudy and yellowish-gray within a minute. After 10 min,
the formation of a dark-brown precipitate and HCN (1H NMR: 4.2
ppm, s) was observed, pointing to reductive elimination of HCN
from 1. Similarly, 2 was found to be stable in aqueous or organic
(THF, MeCN) solutions containing extra cyanide even in small
amounts (ca. 10%) but readily underwent reductive elimination of
PhCN at room temperature when the solution was CN--free.19
Reductive elimination from Pd(II) is often promoted by ligand
dissociation, commonly proceeding from three- rather than four-
(11) The ∆G values (298 K) for the sequential stepwise PPh3 displacement
reactions of [(Ph3P)4Pd] with cyanide to [(Ph3P)3Pd(CN)]-, [(Ph3P)2Pd-
(CN)2]2-, [(Ph3P)Pd(CN)3]3-, and finally [Pd(CN)4]4- were computed at
-52, +23, +112, and +194 kcal mol-1, respectively.
(12) For detailed accounts of the pioneering work of Amatore, Jutand, and
co-workers in the field, see: Amatore, C.; Jutand, A. J. Organomet. Chem.
1999, 576, 254 and Acc. Chem. Res. 2000, 33, 314.
(13) For example, Suzuki coupling has been reported with [ArNMe3]+ as the
arylating electrophile: Blakey, S. B.; MacMillan, D. W. C. J. Am. Chem.
Soc. 2003, 125, 6046.
(14) For reviews, see: (a) Fehlhammer, W. P.; Fritz, M. Chem. ReV. 1993,
93, 1243. (b) Pombeiro, A. J. L. Inorg. Chem. Commun. 2001, 4, 585.
(15) The anion [(CN)3PdPh]2- has been reported to form from the redox
reaction of K2[Pd(CN)2]‚xNH3 with KCN and [Ph4P]+: Nast, R.; Buelck,
J.; Kramolowsky, R. Chem. Ber. 1975, 108, 3461.
(16) 31P NMR (C6H6, 25 °C, δ): -5.5 ppm (s, PPh3). NMR for
K2[(13CN)3PdPh] in the D2O phase (25 °C, δ): 1H: 7.1 (t, 1H, p-Ph); 7.2
(t, 2H, m-Ph); 7.5 (d, 2H, o-Ph). 13C: 145.4 (d, JC-C ) 7.3 Hz, 2C,
mutually trans-CN); 146.7 (t, JC-C ) 7.3 Hz, 1C, CN trans to Ph).
(17) For instance, cyanide has been used to displace BINAP mono-oxide from
its Pd complex: Gladiali, S.; Pulacchini, S.; Fabbri, D.; Manassero, M;
Sansoni, M. Tetrahedron: Asymmetry 1998, 9, 391.
(18) For this reason, no combustion analysis of [PPN]+2‚1 was attempted.
(19) Stirring [(Ph3P)2Pd(I)Ph] in benzene with D2O containing a substoichio-
metric amount (2.9 equiv; see Scheme 2) of K13CN led to extra cyanide-
free K2[(13CN)3PdPh] in the aqueous phase (13C NMR). The separated
clear D2O solution turned cloudy in ca. 30 min, and after 12 h, the
formation of an amorphous-looking, brown-black precipitate was ob-
served. The mixture was extracted with CD2Cl2 to detect Ph13CN in the
extract (13C NMR, δ: 119.0 (s) and GC-MS).
(20) See, for example: (a) Yamashita, M.; Hartwig, J. F. J. Am. Chem. Soc.
2004, 126, 5344. (b) Marshall, W. J.; Grushin, V. V. Organometallics
2003, 22, 1591. (c) Grushin, V. V. Organometallics 2000, 19, 1888.
JA066931D
9
J. AM. CHEM. SOC. VOL. 129, NO. 1, 2007 31