Competitive Aryl-Iodide vs Aryl-Aryl Reductive Elimination Reactions in Pt(IV) Complexes: Experimental and Theoretical Studies

Article abstract of DOI:10.1021/ja077258a

The platinum(IV) complex trans-(dmpe)Pt^IV(Ar)₂I ₂ (2, dmpe = 1,2-dimethylphosphinoethane, Ar = 4-FC₆H ₄) rapidly reacts, upon moderate heating in solution under ambient light, via two distinct pathways: isomerization to the corresponding cis-isomer (3) and Ar-I reductive elimination to give (dmpe)-Pt^II(Ar)I (4). Complex 3 undergoes, upon prolonged heating at high temperatures, an exclusive Ar-Ar reductive elimination reaction to give (dmpe)Pt^III₂. Experimental and DFT studies showed that the 2-to-3 isomerization proceeds via three pathways: photochemical or thermal phosphine chelate opening and a mechanism involving cleavage of the Pt-I bond. The isomerization reaction is significantly slowed down but not stopped in the absence of light or in the presence of an excess of tetra-n-butylammonium iodide. On the other hand, the Ar-I reductive elimination from 2 proceeds via the Pt^δ+-I ^δ_ ion pairlike transition state. Use of the rigid dmpe analogue 1,2-dimethylphosphinobenzene (dmpbz) as the ligand shuts down the chelate ring-opening isomerization pathway and enables faster Ar-I reductive elimination thus making the latter reaction the major reaction route for the dmpbz supported trans-diiodo Pt(IV) complex 8.

Full text of DOI:10.1021/ja077258a

Published on Web 12/15/2007
Competitive Aryl-Iodide vs Aryl-Aryl Reductive Elimination
Reactions in Pt(IV) Complexes: Experimental and Theoretical
Studies
Anette Yahav-Levi,† Israel Goldberg,† Arkadi Vigalok,*,† and Andrei N. Vedernikov*,‡
School of Chemistry, The Sackler Faculty of Exact Sciences, Tel AViV UniVersity,
Tel AViV 69978, Israel, and Department of Chemistry and Biochemistry,
UniVersity of Maryland, College Park, Maryland 20742
Received September 19, 2007; E-mail: avigal@post.tau.ac.il; avederni@umd.edu
Abstract: The platinum(IV) complex trans-(dmpe)PtIV(Ar)2I2 (2, dmpe ) 1,2-dimethylphosphinoethane, Ar
) 4-FC6H4) rapidly reacts, upon moderate heating in solution under ambient light, via two distinct
pathways: isomerization to the corresponding cis-isomer (3) and Ar-I reductive elimination to give (dmpe)-
PtII(Ar)I (4). Complex 3 undergoes, upon prolonged heating at high temperatures, an exclusive Ar-Ar
reductive elimination reaction to give (dmpe)PtIII2. Experimental and DFT studies showed that the 2-to-3
isomerization proceeds via three pathways: photochemical or thermal phosphine chelate opening and a
mechanism involving cleavage of the Pt-I bond. The isomerization reaction is significantly slowed down
but not stopped in the absence of light or in the presence of an excess of tetra-n-butylammonium iodide.
On the other hand, the Ar-I reductive elimination from 2 proceeds via the Ptδ+-Iδ- ion pairlike transition
state. Use of the rigid dmpe analogue 1,2-dimethylphosphinobenzene (dmpbz) as the ligand shuts down
the chelate ring-opening isomerization pathway and enables faster Ar-I reductive elimination thus making
the latter reaction the major reaction route for the dmpbz supported trans-diiodo Pt(IV) complex 8.
oxidation chemistry.5 In both cases, the reactions are believed
to proceed via the SN2-type attack of a halide at a Pt(IV)-
Introduction
Examples of a reductive elimination reaction of a carbon-
halogen bond from a late transition metal center are scarce
compared with the significantly more common carbon-oxygen,
carbon-nitrogen, or carbon-carbon bond forming reactions.1
Usually, the C-X bond formation is less favorable thermody-
namically than its microscopic reverse, C-X oxidative addition
to a low-valent metal center. Elimination of aryl halides from
bulky three coordinate palladium(II) complexes is a rare case
of such a reaction that is driven by the steric congestion at the
metal center.2,3 Several examples of C-X reductive elimination
from soluble high-oxidation state complexes were also reported.
Reductive elimination of iodomethane was observed as one of
two competing reactions of (L2)Pt(R)Me2(I),4 and chloromethane
was obtained along with methanol in the Shilov methane
coordinated methyl group.4,6,7 Very recently, the reductive
elimination of iodomethane from a Rh(III) center that does not
involve an SN2-type mechanism was reported.8 With sp2- and
sp-hybridized carbon ligands, the C-X reductive elimination
is normally unobserved, especially if the possibility of a more
rapid C-C or C-H reductive elimination exists.9,10 Thus far,
C(sp2)-X reductive elimination was only reported for the
chelating carbon ligands and those involving acylic sp2-
carbons.11-13 It was also proposed in a Pd-catalyzed intramo-
(5) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879.
(6) (a) Luinstra, G. A.; Wang, L.; Stahl, S. S.; Labinger, J. A.; Bercaw, J. E.
J. Organomet. Chem. 1995, 504, 75. (b) Luinstra, G. A.; Labinger, J. A.;
Bercaw, J. E. J. Am. Chem. Soc. 1993, 115, 3004.
(7) A similar mechanism was recently proposed for C(sp3)-O reductive
elimination in a Pd system: Liu, G.; Stahl, S. S. J. Am. Chem. Soc. 2006,
128, 7179.
† Tel Aviv University.
(8) Frech, C. M.; Milstein, D. J. Am. Chem. Soc. 2006, 128, 12434.
(9) Vinyl complexes: (a) Ananikov, V. P.; Mitchenko, S. A.; Beletskaya, I.
P. Russ. J. Org. Chem. 2002, 38, 636. (b) Ananikov, V. P.; Musaev, D.
G.; Morokuma, K. Organometallics 2001, 20, 1652. (c) Ananikov, V. P.;
Musaev, D. G.; Morokuma, K. J. Am. Chem. Soc. 2002, 124, 2839. (d)
Ananikov, V. P.; Musaev, D. G.; Morokuma, K. Organometallics 2005,
24, 715. (e) C(sp2)-C(sp2) reductive elimination is significantly more facile
than C(sp3)-C(sp3) reductive elimination: Low, J. J.; Goddard, W. A. J.
Am. Chem. Soc. 1986, 108, 6115.
‡ University of Maryland.
(1) For general references, see: (a) Collman, J. P.; Hegedus, L. S.; Norton, J.
R.; Finke, R. G. Principles and Applications of Organotransition Metal
Chemistry; University Science Books: Sausalito, CA, 1987. (b) Atwood,
J. D. Inorganic and Organometallic Reaction Mechanisms, 2nd ed. VCH
Publishers: New York, 1997.
(2) (a) Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125, 13944. (b)
Roy, A. H.; Hartwig, J. F. Organometallics 2004, 23, 1533.
(3) A recently reported mechanism of Ar-F reductive elimination from a Pd-
(II) center for the aryl ligand bearing a strong electron-accepting nitro group
((a) Yandulov, D. V.; Tran, N. T. J. Am. Chem. Soc. 2007, 129, 1342) has
been contested: (b) Grushin, V. V.; Marshall, W. J. Organometallics 2007,
26, 4997.
(4) (a) Goldberg, K. I.; Yan, J.; Winter, E. L. J. Am. Chem. Soc. 1994, 116,
1573. (b) Goldberg, K. I.; Yan, J.; Breitung, E. M. J. Am. Chem. Soc. 1995,
117, 6889. (c) Hughes, R. P.; Overby, J. S.; Lam, K.-C.; Incarvito, C. D.;
Rheingold, A. L. Polyhedron 2002, 21, 2357.
(10) Alkynyl complexes: Fuhrmann, G.; Debaerdemaeker, T.; Bauerle, P. Chem.
Commun. 2003, 948.
(11) van Belzen, R.; Elsevier, C. J.; Dedieu, A.; Veldman, N.; Spek, A. L.
Organometallics 2003, 22, 722.
(12) Alsters, P. L.; Engel, P. F.; Hogerheide, M. P.; Copijn, M.; Spek, A. L.;
van Koten, G. Organometallics 1993, 12, 1831.
(13) Acyl-X reductive elimination is a key step in the Monsanto methanol
carbonylation process: Dekleva, T. W.; Forster, D. AdV. Catal. 1986, 34,
81.
9
724
J. AM. CHEM. SOC. 2008, 130, 724-731
10.1021/ja077258a CCC: $40.75 © 2008 American Chemical Society

Competitive Elimination Reactions in Pt(IV) Complexes
A R T I C L E S
Scheme 1
Scheme 2
lecular electrophilic aromatic halogenation,14 although the
reductive elimination step in a Pd(IV) system thus far eluded
direct observation. To our knowledge, only one case of an Ar-I
reductive elimination from a Pt(IV) center was reported in 1969
and remained the sole example of such a reaction in an aromatic
system.15,16
(II) diaryl complex 1 results in the clean formation of the trans-
oxidative addition complex 2 (Scheme 2).17 This complex can
be easily isolated by precipitation with pentane and is reasonably
stable at room temperature to be fully characterized by multi-
nuclear NMR techniques. It was also possible to obtain deep
red crystals of 2, suitable for single-crystal X-ray diffraction
analysis, by rapid cooling of its CH2Cl2-toluene (1:1 ratio)
solution at -30 °C. The X-ray structure of 2 (Figure 1a) shows
a nearly symmetrical positioning of the pairs of equivalent
ligands in the octahedral Pt(IV) complex. The Pt-I1 and Pt-
I2 distances of 2.6614(5) and 2.6742(5) Å, respectively, are
significantly shorter than the Pt-I distances in the isomeric cis-
complex 3 (vide supra), where the iodo ligands are located trans
to phosphine and aryl ligands that are characterized by a stronger
trans-influence compared to iodide.19
We recently reported that the competitive C-X and C-C
reductive elimination reactions take place upon the addition of
X2 (X ) I, Br) to a series of diaryl Pt(II) complexes bearing
chelating diphosphine ligands.17 The C-X reductive elimination
was proposed to occur from the cationic pentacoordinate Pt-
(IV) intermediate A with the halide ligand occupying the axial
position trans to an empty coordination site (Scheme 1). Such
a cation is formed during the SN2-type X2 oxidative addition to
a square planar Pt(II) center.18 It also can be formed reversibly
via the X- dissociation from a trans-oxidative addition product.
Since most of the oxidative addition products, Pt(IV) complexes,
are too unstable to be isolated in pure form, no detailed
mechanistic studies of C-X reductive elimination could be
performed. The only common chelating diphosphine ligand that
gave stable Pt(IV) complexes upon I2 oxidative addition was
dmpe (1,2-dimethylphosphinoethane). Here we describe two
isomeric complexes cis- and trans-(dmpe)Pt(Ar)2I2 and their
dmpbz (dmpbz ) 1,2-dimethylphosphinobenzene) analogue that
show surprisingly different behavior in C-X reductive elimina-
tion.
Upon heating in benzene or toluene at 50 °C in an oil bath
for several hours, complex 2 cleanly converted to the yellow
cis-isomer 3, thus confirming that the trans-isomer 2 is the
kinetic product of oxidative addition of I2 to square planar Pt-
(II) complex 1. Complex 3 has significantly lower solubility in
nonpolar solvents than 2 and precipitates readily from the
aromatic solvents. Although the crystal structure of 3 was
reported (Figure 1b),17 it was not discussed in detail. In addition
to longer Pt-I bonds (2.7077(5) and 2.7848(5) Å for the iodo
ligands trans to the phosphine and aryl, respectively) when
compared with 2, the P1-Pt bond of 2.2989(16) Å (trans to
the iodine) is significantly shorter than the P2-Pt bond (trans
to the aryl group) 2.3892(16) Å. The latter is similar to those
observed for 2, which again correlates well with the trans-
influences of the three ligands at Pt. Interestingly, very little
difference is observed in the Pt-Caryl distances in 3 (2.093(6)
and 2.107(6) Å, for aryl groups trans to iodine and phosphorus
atoms, respectively). Also, the P1-Pt-P2 bond angles for
complexes 2 and 3, of 84.48(7)° and 85.59(6)°, respectively,
are very similar.
Results and Discussion
Synthesis of the Diiodo dmpe Pt(IV) Complexes 2 and 3.
We recently showed that addition of I2 to the square planar Pt-
(14) (a) Fahey, D. R. J. Organomet. Chem. 1971, 27, 283. (b) Kalyani, D.;
Dick, A. R.; Anani, W. Q.; Sanford, M. S. Tetrahedron 2006, 62, 11483-
11498. (c) Kalyani, D.; Dick, A. R.; Anani, W. Q.; Sanford, M. S. Org.
Lett. 2006, 8, 2523-2526.
(15) (a) Ettorre, R. Inorg. Nucl. Chem. Lett. 1969, 5, 45. (b) In our hands, the
reaction appeared to be more complex giving large amounts of a byproduct.
(16) We very recently observed facile Ar-I reductive elimination that occurs
upon the reaction of Pd(II) complexes with XeF2: Kaspi, A.; Yahav-Levi,
A.; Goldberg, I.; Vigalok, A. Inorg. Chem., published online Dec. 4, 2007
http://dx.doi.org/10.1021/ic701722f.
(17) Yahav-Levi, A.; Goldberg, I.; Vigalok, A. J. Am. Chem. Soc. 2006, 128,
8710.
(18) Kubota, M.; Boegeman, S. C.; Keil, R. N.; Webb, C. G. Organometallics
1989, 8, 1616.
C-C Elimination from Complex 3. Heating complex 3 at
85 °C in toluene resulted in a slow reaction to cleanly give
(19) For a recent review, see: Coe, B. J.; Glenwright, S. J. Coord. Chem. ReV.
2000, 203, 5.
9
J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008 725

A R T I C L E S
Yahav-Levi et al.
Scheme 4
mechanism in which the cationic intermediate 5 and/or ion pairs
[5]+,I- form in DMF (Scheme 5).
Five-coordinate cation 5 (A ) I), resulting from the iodide
dissociation from 2, is expected to be prone to isomerization
to, more stable thermodynamically, cation 6 (A ) I). Lower
stability of 5 vs 6 may originate from an unfavorable trans-
influence exerted by both aryl groups on the Pt-P bonds. In
turn, the trans-influence of the iodo ligand in 6 is much
weaker.19 Indeed, attempts to generate 5 (presumably, along its
DMF- or THF-solvate) in the form of a tetrafluoroborate salt
by reacting 2 with AgBF4 in THF or DMF resulted in the
instantaneous formation of cation 6 (A ) BF4), without
noticeable C-I reductive elimination. Complex 6 (A ) BF4)
was also prepared by reacting 3 with AgBF4. Based on these
observations, we suggest that 5 and 6 are potential intermediates
involved in the isomerization of 2 to 3 in polar DMF.
The fact that both 4 and 3 formed from 2 in DMF at
comparable rates could be assigned to the involvement of both
ion 5 and derived ion pairs [5]+,I- that are prone to C-I
elimination rather than isomerization, in contrast to 5 itself. The
involvement of an ion-pair-like transition state instead of ion
pairs [5]+,I- could not be excluded either.
Figure 1. X-ray crystal structure of complexes 2 (a) and 3 (b).
Scheme 3
Upon heating, 6 produced the biaryl (4-FC6H4)2 and (dmpe)-
Pt(I)BF4 (t1/2 ) 1.5 h in DMF at 60 °C) (Scheme 5). The latter
complex was prepared independently from (dmpe)Pt(I)2 and
AgBF4. The rate of the biaryl elimination from 6 was faster
than that from 3 under the same conditions (t1/2 ) 25 h in DMF
at 85 °C).
products of C-C reductive elimination, the biaryl (4-FC6H4)2
and (dmpe)Pt(I)2 (30% conversion after 25 h, Scheme 3). Clean
aryl-aryl reductive elimination from 3 was also observed in
more polar solvents, such as THF or DMF; the reaction rates
were faster than that using toluene (50% conversion after 25 h
at 85 °C in DMF). The biaryl elimination in DMF was inhibited
in the presence of 10 equiv of tetra-n-butyl ammonium iodide
(TBA-I) (30% conversion after 40 h at 85 °C) so suggesting
that the reaction might proceed via a five-coordinate cationic
intermediate resulting from iodide ion dissociation from 3. In
aromatic hydrocarbon solutions, where formation of free ions
is unlikely, the C-C elimination could proceed via ion pairs
and/or via the five-coordinate transient resulting from the chelate
ring opening, (η1-dmpe)PtAr2I2.
C-I Elimination vs Thermal Isomerization of Complex
2. Surprisingly, when DMF or THF, solvents more polar than
benzene or toluene, were used, trans-diiodo complex 2 formed
products of C-I reductive elimination, (dmpe)Pt(Ar)I (4) and
4-FC6H4I, along with the cis-diiodo complex 3 (Scheme 4).
Since 3 is essentially stable under the reaction conditions,
we initially attributed the formation of 4 and 4-FC6H4I to the
Photochemical and Thermal Isomerization of Complex 2
to Complex 3. Although the general reactivity patterns of 2 in
polar solvents were reproducible, the ratio between 3 and 4
varied noticeably even when the reaction was run under
seemingly identical conditions. Even more unexpectedly, heating
samples of 2 in DMF inside the NMR probe gave 3/4 ratios
that were different from those obtained in an oil bath under the
bright light of a 100 W incandescent lamp at the same
temperatures (see Table 1 for rate constants of isomerization,
k
isom, and reductive elimination reaction, kRE, entries 1 and 2).
Very slow formation of 4 and the iodoarene was also observed
in nonpolar benzene when the reaction was set up inside an
NMR probe (entry 5), although normally no such products were
observed in this solvent when the reaction was set up under
light in an oil bath (entry 6). Influence of an external magnetic
field aside, this observation suggested that the reactivity of 2
depends on the presence of light. Addition of 1 equiv of BHT
did not result in significant changes in the reactivity, suggesting
that no free radicals are formed in the reaction. Interestingly,
addition of 10 equiv of TBA-I to a reaction mixture in DMF
exposed to an ambient light slowed down the rate of disap-
9
726 J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008

Competitive Elimination Reactions in Pt(IV) Complexes
A R T I C L E S
Scheme 5
Scheme 6
Table 1. Kinetic Data for Conversion of Complexes 2 and 8 under Various Reaction Conditions including the Presence (Oil Bath, 100 W
Incandescent Lamp) and Absence of Light (inside an NMR Probe)
light
conditions
and TBA-I
t1
kisom
×
103,
kRE
×
103,
/
2
1
1
compd
solvent
t,
°C
(cis-PtIV/PtII ratio)
min-
min-
1
2
3
4
5
6
7
8
9
2
DMF
60
dark
light
light with TBA-I
dark with TBA-I
dark
80 min (1:4)
25 min (2:1)
60 min (1:2.5)
80 min (1:4)
9 h (9:1)
2.5 h (1:0)
4 h (1:0)
30 min (0:1)
30 min (0:1)
6 h (0:1)
1.7
18
3.3
1.7
1.2
4.6
2.9
0
0
0
0
0
7
9
8
7
0.1
0
0
23
23
1.9
9.2
173
2
8
benzene
DMF
60
50
light
light with TBA-I
dark
light
dark
dark
dark
10
11
12
30
40
60
75 min (0:1)
4 min (0:1)
pearance of 2 more than twice (entry 3), affecting kisom but not
kRE and leaving the Ar-I reductive elimination process as the
major reaction pathway. Similarly, compared to the reaction
setup in benzene under light (entry 6) the addition of 10 equiv
of TBA-I decreased the rate constant kisom noticeably (entry 7).
At the same time, for the dark reaction in DMF the addition
of 10 equiv of TBA-I did not affect the magnitude of kisom and
proposed for Pt(IV) iodo complexes bearing a chelating ethylene
diamine ligand. A thermal ring-opening was also reported in
the reductive elimination chemistry of Pt(IV), not assisted by
light.22
Finally, the C-I reductive elimination in both benzene and
DMF might proceed via ion pairs or an ion-pair-like transition
state.
Synthesis of the Diiodo dmpbz Pt(IV) Complex 8: C-I
Reductive Elimination. To verify that the diphosphine chelate
opening might be responsible for the thermal and/or photo-
chemical trans-to-cis rearrangement process, we prepared a more
rigid analogue of 1, (dmpbz)Pt(Ar)2 complex 7 (dmpbz ) 1,2-
k
RE; both C-I elimination and isomerization of 2 to 3 were
observed (entries 1 and 4).
Remarkably, as it follows from Table 1, the rate constant of
C-I elimination from 2 in DMF, kRE, remained approximately
unchanged in the presence of light or TBA-I, whereas the rate
constant of isomerization of 2 to 3, kisom, was affected by both
TBA-I and light. Importantly, the magnitude of kisom in DMF
remained significant even in the dark suggesting that at least
two mechanisms of isomerization were operative, a photochemi-
cal and a thermal one.
We suggest that in the cases of both thermal and photochemi-
cal isomerization reaction the mechanisms include conversion
of 2 to low-coordinate transients. Light might facilitate a ligand
dissociation from a Pt center via either loosening the Pt-I bond20
or promoting a chelate ring-opening.21 The latter process was
(20) (a) Bryan, S. A.; Dickson, M. K.; Roundhill, D. M. J. Am. Chem. Soc.
1984, 106, 1882. (b) Bryan, S. A.; Dickson, M. K.; Roundhill, D. M. Inorg.
Chem. 1987, 26, 3878. (c) Cross, R. J.; Davidson, M. F. J. Chem. Soc.,
Dalton Trans. 1987, 139. (d) Appleton, T. G.; Berry, R. D.; Hall, J. R.;
Neale, D. W. J. Organomet. Chem. 1988, 342, 399. (e) Rich, R. L.; Taube,
H. J. Am. Chem. Soc. 1954, 76, 2608.
(21) (a) Kratochwil, N. A.; Guo, Z.; del Socorro, Murdoch, P.; Parkinson, J.
A.; Bednarski, P. J.; Sadler, P. J. J. Am. Chem. Soc. 1998, 120, 8253. (b)
Kratochwil, N. A.; Ivanov, A. I.; Patriarca, M.; Parkinson, J. A.;
Gouldsworthy, A. M.; del Socorro Murdoch, P.; Sadler, P. J. J. Am. Chem.
Soc. 1999, 121, 8193.
(22) (a) Crumpton-Bregel, D. M.; Goldberg, K. I. J. Am. Chem. Soc. 2003, 125,
9442. (b) Crumpton, D. M.; Goldberg, K. I. J. Am. Chem. Soc. 2000, 122,
962.
9
J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008 727

A R T I C L E S
Yahav-Levi et al.
Figure 3. Eyring plot of the 4-FC6H4I reductive elimination from 8 in
DMF.
Scheme 7
Figure 2. X-ray crystal structure of complexes 8 (a) and 9 (b).
dimethylphosphinobenzene),23 and reacted it with I2. The trans-
diiodo Pt(IV) complex 8 was obtained in a quantitative yield
(Scheme 6) and fully characterized, including X-ray diffraction
analysis. Importantly, the crystal structure of 8 showed nearly
identical bond lengths with those in 2 (Figure 2a). The P1-
Pt-P2 bond angles for both complexes are also very similar,
82.56(7)° vs 84.48(7)°, in 8 and 2, respectively. Mild heating
of this complex in DMF both in the dark (entry 8) and under
light (entry 9) gave exclusively the products of C-I reductive
elimination, 4-FC6H4I and Pt(II) complex 9 which was char-
acterized by X-ray diffraction analysis (Figure 2b). No effect
of light on this reaction was observed. Furthermore, addition
of 10 equiv of TBA-I did not affect the C-I reductive
elimination rate in DMF.
Therefore, as in the case of analysis of the reactivity of 2,
we can conclude that the C-I reductive elimination does not
involve formation of a dmpbz analogue of ion 5 as a result of
iodide dissociation from 8. These observations can be accounted
for if C-I elimination occurs from ion pairs or via an ion-pair-
like transition state as it was assumed for complex 2.
To shed light on the reaction mechanism, the activation
parameters of the C-I reductive elimination from 8 in DMF
were determined using the NMR technique (Table 1, entries 8,
10-12). Kinetic experiments showed that the reaction is first-
order in complex 8 with ∆H‡ ) 29.11 ((0.12) kcal/mol and
∆S‡ ) 15.98 ((0.46) cal/(mol‚K) (Figure 3). The relatively high
activation enthalpy might be indicative of the significant Pt-C
and Pt-I bond weakening in the transition state. The relatively
high positive value of the activation entropy also suggests the
disorder increase in the transition state. The activation param-
eters are consistent with the idea of direct elimination from a
six-coordinate Pt(IV) complex via an ion-pair-like transition state
if the charge separation in the transition state is not very different
from that in 8. In contrast, a significant charge separation
increase in the transition state might induce the solvent
electrostriction, cause the entropy loss, and invert the sign of
the C-I elimination activation entropy.24
Mechanisms of C-I Elimination and Isomerization of 2
and 8. It is important to emphasize the difference between the
mechanism of Ar-I reductive elimination from 2 or 8 and that
for the CH3-I reductive elimination reaction from diphosphine
Pt(IV) iodo trimethyl complexes reported earlier.4 In the latter
case, as well as the mechanistically similar CH3-OR reductive
elimination reactions,4,25 the presence of an external nucleophile
is essential to achieve an SN2-type attack at a Pt(IV)-coordinated
C(sp3) center. In complexes 2 and 8 the aryl ligand is not as
susceptible to external attacks by nucleophiles as the methyl
group in LPtMe3I. C-I reductive elimination from 2 and 8 can
be viewed as a microscopic reverse of the concerted oxidative
addition of aryl halides to a low valent metal center. In addition,
the presence of two aryl ligands allows for competition between
the Ar-I and Ar-Ar reductive elimination reactions, the
outcome of which is dependent on the metal complex geom-
etry.26
Our view of plausible mechanisms of competing isomeriza-
tion and C-I reductive elimination reaction involving complexes
2 and 8 is summarized in Scheme 7.
As it follows from previous discussion, the C-I elimination
in our systems is a thermal-only reaction which is not affected
(24) Tobe, M. L.; Burgess, J. Inorg. React. Mech., 2nd ed.; Prentice Hall: New
York, 2000.
(25) (a) Williams, B. S.; Goldberg, K. I. J. Am. Chem. Soc. 2001, 123, 2576.
(b) Williams, B. S.; Holland, A. W.; Goldberg, K. I. J. Am. Chem. Soc.
1999, 121, 252. (c) Vedernikov, A. N.; Binfield, S. A.; Zavalij, P. Y.;
Khusnutdinova, J. R. J. Am. Chem. Soc. 2006, 128, 82. (d) Khusnutdinova,
J. R.; Zavalij, P. Y.; Vedernikov, A. N. Organometallics 2007, 26,
3466.
(23) For the ligand synthesis see: Kang, Y. B.; Pabel, M.; Pathak, D. D.; Willic,
A. C.; Wild, S. B. Main Group Chem. 1995, 1, 89.
(26) Ruddick, J. D.; Shaw, B. L. J. Chem. Soc. A 1969, 2969.
9
728 J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008

Competitive Elimination Reactions in Pt(IV) Complexes
A R T I C L E S
by light and might proceed via an ion-pair-like transition state
in a DMF or benzene solution at elevated temperatures (Scheme
7, path a). The ability of a metal complex to undergo a reductive
elimination reaction is a function of a number of factors:
electron density of the metal center, ligand steric interactions,
the presence of coordination vacancies, the configuration of the
metal center.1 Since 2 and 8 are structural analogues, two latter
factors can be considered identical for both complexes. Note-
worthy, dmpbz is not only a more rigid ligand but also expected
to be a poorer electron donor compared to dmpe as follows
from the IR data for similar LNi(CO)2 complexes with L )
(Et2PCH2)2 and (Et2P)2-o-C6H4.27 Therefore, C-I elimination
from 8 containing the rigid electron poorer dmpbz is expected
to be faster than that from 2. This consideration is consistent
with our observations: C-I elimination from 8 in DMF at 60
°C is about 25 times more rapid than that from 2 under the
same experimental conditions (Table 1, entries 1 and 12).
Correspondingly, the ion-pair-like transition state for C-I
elimination is expected to be more reagent-like and therefore
lower in energy in the case of dmpbz complex 8.
The ability of an octahedral 18-electron complex to undergo
isomerization might be a function of its ability to produce a
five-coordinate transient.25a As experiments with complex 2 and
AgBF4 show, “free ion” 5, or more likely a loose ion pair or a
solvento complex formed by 5, is most probably involved in
an isomerization reaction in DMF. Since this reaction channel
is suppressed by additives such as TBA-I and is in the dark, it
might include a photochemical Pt-I bond ionization (Scheme
7, path b). Path b is higher in energy in less polar solvents and
most likely will be suppressed in nonpolar benzene. More
electron-poor complex 8 will be less prone to produce five-
coordinate cations compared to 2.
Another way to produce a five-coordinate transient involved
in an isomerization reaction is by chelate ring-opening in 2 or
8 (Scheme 7, path c). Since in the case of complex 2 the
isomerization is observed under both dark and light conditions
in DMF and benzene solutions, both thermal and photochemical
ring-opening paths might be operative here. In turn, photo-
chemical and thermal chelate ring-opening are expected to be
more facile in 2 compared to more rigid 8. Accordingly, in
contrast to 2, complex 8 demonstrated the absence of isomer-
ization behavior in DMF.
Based on the analysis given above, complex 8 is less prone
to isomerization than 2. In turn, 8 is more prone to C-I
elimination. Therefore, the ratio of products of C-I elimination/
isomerization products will be higher for 8 compared to 2 under
identical reaction conditions, which was observed experimen-
tally.
Figure 4. Standard Gibbs energy changes for 4-IC6H4F elimination from
complexes 2 and 8, kcal/mol.
(4) How competitive are isomerization and C-I elimination
reactions of five-coordinate transients resulting from the chelate
ring opening in complexes 2, 8 and from iodo ligand abstraction,
cation 5?
DFT Modeling: C-I Elimination from 2 and 8. A diagram
given in Figure 4 shows that the activation barrier for 4-fluor-
oiodobenzene elimination from complex 2 is 3.0 kcal/mol higher
for (TS(C-I)2 vs TS(C-I)8) and the reaction energy is 4.2 kcal/
mol lower compared to complex 8 stabilized by dmpbz. These
results are consistent with the observed difference in reactivity
of 2 and 8 in the C-I elimination reaction. According to data
in Table 1, under the same conditions, kRE for 8 is significantly
higher than kRE for 2 in DMF. The experimentally found Gibbs
activation energy for C-I elimination from 8 in DMF is 26.4
kcal/mol, a reasonable match of the calculated 27.9 kcal/mol.
As it follows from the calculated Pt-I2 bond length, 3.107 Å
in TS(C-I)8 vs 2.743 Å in 8, and 3.306 Å in TS(C-I)2 vs
2.740 Å in 2, the ion-pair character of the transition state TS-
(C-I)8 is less pronounced than that for TS(C-I)2. Consistent
with that result, a more electron-rich metal center in 2 should
acquire a greater positive charge in the course of Pt-I2 bond
ionization and elongation to eliminate the Ar-I. The results in
Figure 4 are consistent with available experimental observations
showing significantly higher reactivity of 8 in C-I elimination
compared to 2.
In support of the idea of an ion-pair-like character of TS-
(C-I)’s, it is worth comparing the experimentally determined
and DFT calculated values of entropy of activation of C-I
elimination from 8, ∆S *(8). The calculated gas-phase entropy
of activation is 4.1 cal/(K‚mol) which is positive, similar to the
experimentally found entropy of activation in DMF solution,
16.0 cal/(K‚mol). If elimination occurs from contact or solvent
separated ion pairs where ions experience strong interactions
with solvent, solvent electrostriction might diminish significantly
the magnitude and even change the sign of ∆S *(8) that is
expected to be positive in the absence of such effects.24 The
calculated dipole moments of 2, 8, TS(C-I)2, and TS(C-I)8
are very similar: 11.5 D (2), 11.2 D (TS(C-I)2), 12.4 D (8),
and 11.3 D (TS(C-I)8) suggesting that the ion-pair-like
character of these TS(C-I)’s is not very pronounced and that
solvent effects should not alter dramatically the reaction entropy
of activation.
Computational Modeling of Reactivity of 2, 5, and 8. To
get a better understanding of the behaviors of complexes 2 and
8 in reductive elimination and isomerization reactions, a DFT
study was undertaken. The primary questions to answer were
as follows:
(1) How facile is direct C-I elimination from 2 and 8?
(2) How facile is thermal chelate ring opening in complexes
2 and 8?
(3) How facile is photochemical chelate ring opening in
complexes 2 and 8?
(27) Chatt, J.; Hart, F. A. J. Chem. Soc. 1960, 1378.
9
J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008 729

A R T I C L E S
Yahav-Levi et al.
Figure 5. Standard Gibbs energy changes for thermal chelate ring opening
in complexes 2 and 8, kcal/mol.
DFT Modeling: Thermal Ring Opening in 2 and 8. A
chelate ring opening in complexes 2 and 8 to produce five-
coordinate η1-diphosphine species requires 23.6 kcal/mol (in-
termediate 10) and 28.3 kcal/mol (intermediate 11), respectively
(Figure 5). Assuming that the barrier of ring opening is
approximately the same as the reaction energy, the ring opening
can occur at a slow rate already at 25 °C for complex 2 with
Figure 6. Standard Gibbs energy changes for chelate ring opening and
3
isomerization in [2] and3[8], kcal/mol.
*
t1/2 ) 6.4 h which corresponds to ∆G298 ) 23.6 kcal/mol,
according to the Eyring equation, whereas for complex 8 a
comparable rate is expected at 80 °C (t1/2 ) 8.9h) only.
Therefore, as compared to 8, 2 is much more efficient in
producing five-coordinate transient (η1-L)PtAr2I2 via thermal
chelate ring opening.
DFT Modeling: Photochemical Ring Opening in 2 and
8. Time-dependent DFT calculations allowed us to determine
the energy and light wavelength necessary for the lowest-energy
electronic excitation in 2 and 8,28 39.3 kcal/mol (727 nm) for
the dmpe complex and 37.8 kcal/mol (757 nm) for the dmpbz
complex. In both complexes the HOMO is a π-bonding orbital
localized mostly on 12 carbon atoms of two 4-fluorophenyl
ligands, whereas the LUMO is the σ-I-Pt-I antibonding orbital.
Correspondingly, the triplet species 3[2] and 3[8] having singly
occupied σ-I-Pt-I antibonding orbitals are characterized by
significantly elongated Pt-I bonds. For instance, the Pt-I bond
Figure 7. Standard Gibbs energy changes for Ar-I elimination from 10,
kcal/mol.
3[8] (Figure 6). Thus, the light-induced chelate ring opening in
2 and 8 may be an effective pathway to isomerization of these
species. Considering this, we assume that the light intensity is
sufficient to make these reactions fast enough.
DFT Modeling: Isomerization and C-I Elimination
Reactions of Five-Coordinate Transient 10. The calculated
barrier for the C-I reductive elimination from five-coordinate
transient 10 is 19.7 kcal/mol (Figure 7; TS(C-I)10), too high
to allow for the thermal chelate ring opening of 2 (23.6 kcal/
mol; Figure 5) with subsequent C-I elimination in a thermal
reaction at temperatures not exceeding 85 °C used in this work.
All our attempts to locate the transition state for isomerization
of 10 to 12 failed, presumably because of the relative flatness
of the corresponding potential energy surface. Still, available
computational estimates of the energy of species with the
I-Pt-I angle varying between ∼180° (10) and ∼90° (12)
suggest that the reaction barrier does not exceed 6 kcal/mol,
low enough to allow for thermal isomerization of 2 via five-
coordinate transient 10. This suggestion is consistent with our
experimental observations of the ability of 2 to undergo
isomerization in the dark.
DFT Modeling: Isomerization and C-I Elimination
Reactions of Cations 5 and 6. Complete removal of the iodide
anion from 2 which leads to cation 5 will increase electron
deficiency of the metal and, consistent with the results of our
computational analysis of C-I elimination (Figure 4), will lower
the barrier of C-I reductive elimination from the Pt(IV) center.
Indeed, the barrier height calculated for the gas-phase cation 5
3
1
length in [2] is 3.200 Å vs 2.740 Å in [2]. Assuming that
excited-state species should be long-lived enough to exhibit
chelate ring opening or other reactivity, we considered as such
the lowest energy triplet states of 2 and 8, 3[2] and 3[8],
respectively. These triplets could be produced as a result of spin
interconversion of singlet excited states of complexes 2 and 8.
3
3
The ring opening in [2] and [8] followed by another spin
interconversion might lead then to singlet transients 10 and 11,
respectively. Altogether, these photoassisted transformations
might constitute an alternative path to produce five-coordinate
intermediates 10 and 11.
3
3
In fact, for both [10] and [11] the stationary point corre-
sponding to 3[trans-(η1-L)PtAr2I2] could not be located. Instead,
all our geometry optimization attempts led to isomeric products,3-
[cis-(η1-L)PtAr2I2], 3[12] for dmpe and 3[13] for dmpbz.
3
3
Therefore, the ring opening in [2] and [8] was accompanied
with I-Pt-I angle change and isomerization. The energy
required for the chelate opening in 3[2] to produce five-
coordinate triplet 3[cis-(η1-dmpe)PtAr2I2] is 10.7 kcal/mol only,
whereas 3[cis-(η1-dmpe)PtAr2I2] is 17.6 kcal/mol in the case of
(28) JAGUAR, version 7.0; Schrodinger Inc.: Portland, OR, 2007.
9
730 J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008

Competitive Elimination Reactions in Pt(IV) Complexes
A R T I C L E S
computational model illustrated in Figure 8, we cannot predict
the relative height of barriers for C-I elimination and isomer-
-
ization of cation 5 associated with BF4 or DMF in DMF
solutions. Solvent effect might also be important in the
photochemically induced formation of 5 (path b, Scheme 7),
which could not be properly assessed. At the same time, our
calculations of the reactivity of the more stable cation 6 are
fully consistent with our experimental observations. Only C-C
bond formation can take place in this case, although the
difference in the kinetic barriers for both reductive elimination
reactions is not large.
Conclusion
In conclusion, we demonstrated that aryl-halide reductive
elimination can occur in Pt(IV) complexes under very mild
conditions. Although the complete mechanistic picture is
complicated, it is clear that the complex geometry plays a crucial
role in determining the reaction outcome. The less thermody-
namically stable trans-complex 2 can undergo a facile Ar-I
reductive elimination reaction. Maintaining this configuration
is essential for the formation of a carbon-halogen bond, as the
more stable cis-isomer 3 gives exclusively the products of C-C
reductive elimination, albeit under much harsher conditions.
Unlike the Ar-I elimination, which takes place in a single path,
the competing trans/cis isomerization reaction can proceed via
several pathways, thermal and photochemical. Thus, solvent,
light, and choice of ligand can influence the competition. The
use of the rigid and electron-poorer dmpbz analogue of dmpe
allowed for significant discrimination of these competing
reactions by rate, with the Ar-I reductive elimination becoming
the major reaction path. The kinetic analysis, supported by the
DFT calculations, showed that this reaction occurs via an ion-
pair-like, late transition state. Noteworthy, while the formation
of an aryl-iodide bond gives relatively little energy gain in
these systems, especially compared with the competitive Ar-
Ar reduction elimination (-14.8 kcal/mol vs -40.8 kcal/mol),
the reaction is kinetically driven and irreversible.
Figure 8. Standard Gibbs energy changes for isomerization, C-C and C-I
reductive elimination from cations 5 and 6, kcal/mol.
is 1.6 kcal/mol only (Figure 8; TS(C-I)5). With the activation
barrier for isomerization of 5 to 6 being 8.4 kcal/mol (TS-
(isom)5), the C-I elimination is a much faster reaction. No low
barrier C-C reductive elimination from cation 5 can be expected
since that reaction would lead to a nonplanar (dmpe)PtI(ArI)+
complex with the iodide ligand in the axial position, a highly
unfavorable configuration for a d8 metal complex.
In turn, 6 which is 10.9 kcal/mol more stable than 5, based
on our DFT calculations, can undergo fast C-C elimination
with a Gibbs activation energy of 8.6 kcal/mol (TS(C-C)6)
and slower C-I elimination with a higher activation barrier of
9.9 kcal/mol (TS(C-I)6). With regard to the difference in the
barrier heights, the gas-phase values given in Figure 8 should
be used with caution when applied to the analysis of reactivity
of [5]BF4 in solution. The presence of a counterion significantly
increases the barrier of the C-I elimination reaction, as follows
from the comparison of data in Figures 4 and 8. In support of
this statement, using DFT modeling we found that the barrier
for C-I elimination from a contact ion pair 5(κ1-F-BF4) is 10.7
kcal/mol, about 9 kcal/mol higher than that for 5. Similar is the
effect of a coordinated polar solvent molecule, such as DMF.
The DFT calculated barrier for C-I elimination from solvento-
complex 5(κ1-O-DMF)+ is 9.5 kcal/mol, an about 8 kcal/mol
increase compared to 5. In addition, ion association and solvent
coordination in DMF solution would affect the rate of isomer-
ization of 5 to 6. Its correct computational analysis would require
consideration of solvent effects which was not done in this work.
Therefore, because of the presence of strong solvent-/counte-
rion-cation interactions that are not taken into account in our
Acknowledgment. This work was supported by a Tel Aviv
University Internal Grant to A.V. and by the US-Israel
Binational Science Foundation.
Supporting Information Available: Experimental and com-
putational details (PDF); X-ray crystal data for complexes 2, 3,
8, and 9 (CIF). This material is available free of charge via the
Internet at http://pubs.acs.org.
JA077258A
9
J. AM. CHEM. SOC. VOL. 130, NO. 2, 2008 731