Rearrangement of iridabenzvalenes to iridabenzenes and/or η5-cyclopentadienyliridium(I) complexes: Experimental and computational analysis of the influence of silyl ring substituents and phosphine ligands

Article abstract of DOI:10.1021/om700336d

Lithium-halogen exchange of either (Z)-1-phenyl-2-trimethylsilyl-(5a) or (Z)-l,2-bis(trimethylsilyl)3-(2-iodovinyl)cyclopropene (5b) and addition to either Vaska's or Vaska-type complexes generated iridabenzvalenes (9, 14, 17), iridabenzenes (10, 18), and/or cyclopentadienyl complexes (11, 15, 19), depending on both the substituents on the C₅ framework and the phosphine ligands on Ir. Specifically, the reaction of 5a with Vaska's complex afforded a mixture of 9, 10, and 11. Heating this mixture to 75 °C converted 9 and 10 to 11. NMR studies at 75°C showed that samples of 9 isomerize to 11 in high yield and generate regioisomeric iridabenzene 12 as an intermediate. The reaction of 5b with Vaska's complex produced benzvalene 14 as the sole product. Complex 14 transformed completely to cyclopentadienyliridium complex 15 at 75°C with no benzene intermediate detectable by NMR spectroscopy. The reaction of cyclopropene 5a with Vaska-type complexes containing alkylphosphines of varying cone angles yielded only benzvalene complexes, which either rearranged or decomposed depending upon the extent of heating. A hybrid-DFT computational study was carried out to investigate reactivity differences between phenyl and trimethylsilyl iridabenzvalenes, regioselective rearrangement of 9, and the unexpected stability/instability of 14/16. These calculations rationalize the sometimes contradictory experimental results.

Full text of DOI:10.1021/om700336d

Organometallics 2007, 26, 3957-3968
3957
Rearrangement of Iridabenzvalenes to Iridabenzenes and/or
η5-Cyclopentadienyliridium(I) Complexes: Experimental and
Computational Analysis of the Influence of Silyl Ring Substituents
and Phosphine Ligands§
He-Ping Wu,† Daniel H. Ess,‡,# Seren Lanza,† Timothy J. R. Weakley,† K. N. Houk,#
Kim K. Baldridge,*,‡ and Michael M. Haley*,†
Department of Chemistry, UniVersity of Oregon, Eugene, Oregon 97403-1253, Institute of Organic
Chemistry, UniVersity of Zu¨rich, Winterthurerstrasse 190, CH-8057 Zu¨rich, Switzwerland, and Department
of Chemistry and Biochemistry, UniVersity of California, Los Angeles, California 90095
ReceiVed April 5, 2007
Lithium-halogen exchange of either (Z)-1-phenyl-2-trimethylsilyl- (5a) or (Z)-1,2-bis(trimethylsilyl)-
3-(2-iodovinyl)cyclopropene (5b) and addition to either Vaska’s or Vaska-type complexes generated
iridabenzvalenes (9, 14, 17), iridabenzenes (10, 18), and/or cyclopentadienyl complexes (11, 15, 19),
depending on both the substituents on the C5 framework and the phosphine ligands on Ir. Specifically,
the reaction of 5a with Vaska’s complex afforded a mixture of 9, 10, and 11. Heating this mixture to 75
°C converted 9 and 10 to 11. NMR studies at 75 °C showed that samples of 9 isomerize to 11 in high
yield and generate regioisomeric iridabenzene 12 as an intermediate. The reaction of 5b with Vaska’s
complex produced benzvalene 14 as the sole product. Complex 14 transformed completely to
cyclopentadienyliridium complex 15 at 75 °C with no benzene intermediate detectable by NMR
spectroscopy. The reaction of cyclopropene 5a with Vaska-type complexes containing alkylphosphines
of varying cone angles yielded only benzvalene complexes, which either rearranged or decomposed
depending upon the extent of heating. A hybrid-DFT computational study was carried out to investigate
reactivity differences between phenyl and trimethylsilyl iridabenzvalenes, regioselective rearrangement
of 9, and the unexpected stability/instability of 14/16. These calculations rationalize the sometimes
contradictory experimental results.
Introduction
Metallabenzenes are six-membered metallacycles analogous
to benzene where one CH unit has been replaced by an
isoelectronic transition metal fragment (MLn).1 Originally
postulated to be stable species by Thorn and Hoffman in their
seminal paper in 1979,2 the first unambiguous isolation of a
metallabenzene was reported by Roper et al. just three years
later.3 In the subsequent two decades since their initial proposal,
about 30 varieties of metalla-aromatic molecules were synthe-
sized and/or characterized. Although most of the metallacycles
prepared were isolated examples, a majority exhibited structural
and spectroscopic properties normally associated with aromatic
systems, such as ring planarity, delocalized bonding, and
deshielded proton NMR chemical shifts.1
encompassing new synthetic methods5 and new metal centers,6
are now available. New aromatic metallacycle topologies7 and
constitutional isomers8 have been isolated and characterized.
In 1999 we introduced a new route to directly access the
(4) (a) Landorf, C. W.; Haley, M. M. Angew. Chem., Int. Ed. 2006, 45,
3914-3936. (b) Wright, L. J. Dalton Trans. 2006, 1821-1827.
(5) Inter alia: (a) Gilbertson, R. D.; Weakley, T. J. R.; Haley, M. M. J.
Am. Chem. Soc. 1999, 121, 2597-2598. (b) Xia, H. P.; He, G. M.; Zhang,
H.; Wen, T. B.; Sung, H. H. Y.; Williams, I. D.; Jia, G. C. J. Am. Chem.
Soc. 2004, 126, 6862-6863. (c) Chin, C. S.; Lee, H. Chem.-Eur. J. 2004,
10, 4518-4522. (d) Alvarez, E.; Paneque, M.; Poveda, M. L.; Rendon, N.
Angew. Chem., Int. Ed. 2006, 45, 474-477. (e) Hung, W. Y.; Zhu. J.; Wen,
T. B.; Yu, K. P.; Sung, H. H. Y.; Williams, I. D.; Lin, Z. Y.; Jia, G. C. J.
Am. Chem. Soc. 2006, 128, 13742-13752.
(6) (a) Jacob, V.; Weakley, T. J. R.; Haley, M. M. Angew. Chem., Int.
Ed. 2002, 41, 3470-3473. (b) Zhang, H.; Xia, H. P.; He, G. M.; Wen, T.
B.; Gong, L.; Jia, G. C. Angew. Chem., Int. Ed. 2006, 45, 2920-2923. (c)
Zhang, H.; Feng, L.; Gong, L.; Wu, L.; He, G.; Wen, T.; Xia, H.
Organometallics 2007, 26, 2705-2713.
(7) (a) Wen, T. B.; Zhou, Z. Y.; Jia, G. Angew. Chem., Int. Ed. 2001,
40, 1951-1954. (b) Paneque, M.; Posadas, C. M.; Poveda, M. L.; Rendon,
N.; Salazar, V.; On˜ate, E.; Mereiter, K. J. Am. Chem. Soc. 2003, 125, 9898-
9899. (c) Clark, G. R.; Johns, P. M.; Roper, W. R.; Wright, L. J.
Organometallics 2006, 25, 1771-1777. (d) Bierstedt, A.; Clark, G. R.;
Roper, W. R.; Wright, L. J. J. Organomet. Chem. 2006, 691, 3846-3852.
(e) Clark, G. R.; Lu, G.-L.; Roper, W. R.; Wright, L. J. Organometallics
2007, 26, 2167-2177.
Research into aromatic metallacycles has undergone a major
expansion since 2000.4 A wide variety of new metallabenzenes,
* Corresponding authors. Experimental section: E-mail: haley@
uoregon.edu. Fax: 541-346-0487. Phone: 541-346-0456. Computational
section: E-mail: kimb@oci.unizh.ch. Fax: 41-1-635-6812. Phone: 41-1-
635-4201.
§ Metallabenzenes and Valence Isomers 9. For Part 8, see ref 9b.
† University of Oregon.
‡ University of Zu¨rich.
# University of California, Los Angeles.
(1) Bleeke, J. R. Chem. ReV. 2001, 101, 1205-1227.
(2) Thorn, D. L.; Hoffman, R. NouV. J. Chim. 1979, 3, 39-45.
(3) (a) Elliott, G. P.; Roper, W. R.; Waters, J. M. J. Chem. Soc., Chem.
Commun. 1982, 811-813. See also: (b) Elliott, G. P.; McAuley, N. M.;
Roper, W. R. Inorg. Synth. 1989, 26, 184-189. (c) Rickard, C. E. F.; Roper,
W. R.; Woodgate, S. D.; Wright, L. J. J. Organomet. Chem. 2001, 623,
109-115.
10.1021/om700336d CCC: $37.00 © 2007 American Chemical Society
Publication on Web 07/03/2007

3958 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
Scheme 1
Scheme 2
Ester 6b was prepared according to the literature15 via Rh
catalysis because the Cu modification with bis(trimethylsilyl)-
acetylene instead yielded the pyrazole [3+2] cycloadduct as the
major product. Reduction of the esters with DIBAL-H gave
compounds 7a and 7b,15b which were subsequently oxidized
with Dess-Martin periodinane16 to produce 8a and 8b.15b Wittig
reaction with Ph3PdCHI17 furnished ligands 5a,b with high
stereoselectivity (>15:1 Z:E).
Reaction of Ligand 5a with Vaska’s Complex. The reaction
of cyclopropene 5a with (PPh3)2Ir(CO)Cl is shown in Scheme
2.13 Lithium-halogen exchange at low temperature followed
by addition of Vaska’s complex and warming to ambient
temperature resulted in a mixture of products, which after
chromatography gave 9, 10, and 11 in a 10:2:3 ratio. Complexes
9 and 11 could be isolated independently by treating the mixture
with MeI (removing 10 and 11 as polar salts18) and by heating
the mixture (Vide infra), respectively; nevertheless, samples of
10 always contained significant amounts of either 9 or 11. All
compounds were identified by NMR spectroscopic data, which
are similar to those of diphenyl analogue 2 and related structures.
The asymmetry of benzvalene 9 was indicated by two sets of
downfield proton resonances attributed to the PPh3 ligands and
two different C resonances (δ 75.8, 55.5 ppm) of the complexed
cyclopropenyl double bond. It is noteworthy that iridacycle 9
forms as the major product of this reaction, whereas the
corresponding diphenyl-substituted molecule from Vaksa’s
complex (4a, where the phosphines are PPh3) is never observed.
In an attempt to prepare a pure sample of 10, we heated a
C6D6 solution of 9 at 75 °C for 24 h. Although valence
metallabenzene manifold using suitably substituted 3-vinylcy-
clopropenes (e.g., 1a).5a Lithium-halogen exchange and sub-
sequent addition to various metal complexes have resulted in
the preparation of a variety of iridabenzenes (e.g., 2a)9 and
platinabenzenes (e.g., 3),6a,10 as well as the formation and
characterization of iridabenzvalenes (e.g., 4a)9,11 and rhoda-
benzvalenes (e.g., 4b),12 the first metallabenzene valence
isomers. We extended this novel synthetic route to cyclopropene
ligands with alkyl substituents (e.g., 1b) and found that
iridabenzvalene rearrangement to the corresponding benzene was
regioselective (e.g., 2b).9b To expand this chemistry further, we
present herein the results from the reaction of mono- and bis-
(trimethylsilyl)-substituted cyclopropenes with Vaska’s and
Vaska-type complexes, the former briefly described in our initial
communication.13 The reactions generate the corresponding
iridabenzvalene, iridabenzene, and/or η5-cyclopentadienyliridium
complexes depending upon the exact ring substituents, phos-
phine ligands, and reaction conditions. In addition, we have
performed a detailed DFT computational analysis to help explain
the unusual experimental results.
1
isomerization to yield 10 was the expected transformation, H
NMR analysis of the reaction mixture showed exclusive and
essentially quantitative formation of 11. This result is surprising
given the ease with which benzvalene 4 can be converted to
the corresponding iridabenzene.11 Closer inspection of the
thermal chemistry of these complexes showed multiple trans-
formations. Analysis of a C6D6 solution of 9 heated at 75 °C
for 1 h revealed, in addition to unreacted 9 and a small amount
Results and Discussion
Cyclopropene Ligand Synthesis. The straightforward prepa-
ration of the (Z)-1,2-disubstituted-3-iodovinylcyclopropene ligands
5a,b is shown in Scheme 1. The reaction of phenyl(trimethyl-
silyl)acetylene with ethyl diazoacetate, which was modified by
using Cu powder instead of a Rh complex as the catalyst,
afforded ester 6a in an improved 30% yield (lit.14 11% yield).
1
of 11, the characteristic H NMR signals for an iridabenzene.
These resonances, however, were at 10.85 and 8.78 ppm and
did not belong to 10 (10.45 and 8.14 ppm), thus suggesting the
formation of a different iridabenzene (12, Scheme 3). The
assignment of 12 as the regioisomer depicted in Scheme 3 is
based on 1H-29Si gHMQC NMR experiments, which indicated
that the para proton at 8.78 ppm is coupled to a silicon nucleus.
Careful monitoring of the reaction showed that the concentration
of 12 peaked after 8-10 h and then steadily decreased.
(8) Barrio, P.; Esteruelas, M. A.; On˜ate, E. J. Am. Chem. Soc. 2004,
126, 1946-1947.
(9) (a) Gilbertson, R. D.; Lau, T. L. S.; Lanza, S.; Wu, H. P.; Weakley,
T. J. R.; Haley, M. M. Organometallics 2003, 22, 3279-3289. (b) Wu, H.
P.; Weakley, T. J. R.; Haley, M. M. Chem.-Eur. J. 2005, 11, 1191-1200.
(10) Landorf, C. W.; Jacob, V.; Weakley, T. J. R.; Haley, M. M.
Organometallics 2004, 23, 1174-1176.
(11) Gilbertson, R. D.; Weakley, T. J. R.; Haley, M. M. Chem.-Eur. J.
2000, 6, 437-441.
(15) (a) Kohn, D. W.; Chen, P. J. Am. Chem. Soc. 1993, 115, 2844-
2848. (b) Maier, G.; Hoppe, M.; Reisenauer, H. P.; Kruger, C. Angew.
Chem., Int. Ed. Engl. 1982, 21, 437.
(16) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277-
7287.
(12) Wu, H.-P.; Weakley, T. J. R.; Haley, M. M. Organometallics 2002,
21, 4320-4322.
(13) Wu, H.-P.; Weakley, T. J. R.; Haley, M. M. Organometallics 2002,
21, 2824-2826.
(17) (a) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 2173-2174.
(b) Bestmann, H. J.; Rippel, H. G.; Dostalek, R. Tetrahedron Lett. 1989,
30, 5261-5262.
(14) (a) Mu¨ller, P.; Gra¨nicher, C. HelV. Chim. Acta 1993, 76, 521-534.
(b) Dolgil, I. E.; Okonnishnikova, G. P.; Nefedov, O. M. Akad. Nauk SSSR,
Ser. Khim. 1979, 4, 822-829.
(18) Wang, D.; Angelici, R. J. Inorg. Chem. 1996, 35, 1321-1331.

Rearrangement of IridabenzValenes to Iridabenzenes
Organometallics, Vol. 26, No. 16, 2007 3959
Scheme 3
Scheme 4
Benzvalene 9 was completely consumed after 15 h, with total
conversion of 9 (and 12) to 11 after 24 h at 75 °C in 89%
isolated yield. Heating the purified mixture of 9-11 afforded
similar results. After 1 h, the resonances of 10 had disappeared
and been replaced by signals for 12. The intensity of the signals
for 9 and 11 had decreased and increased, respectively. The
disappearance of 9 and 12 progressed as described above upon
further heating. After 24 h, conversion to 11 (with concomitant
formation of PPh3) was complete with a color change from red-
brown to yellow. Attempts to retard formation of 11 (and thus
potentially increase amounts of 10 and/or 12) by addition of
excess PPh3, known to inhibit phosphine dissociation in irida-
benzenes,19 were not successful. The computational results (Vide
infra) will show that benzene 12 is the preferred regioisomer;
thus, the origins of 10 are uncertain.
Figure 1. Molecular structure of diiridium complex 13. The
thermal ellipsoids are drawn at the 30% probability level. Selected
bond lengths (Å) and angles (deg): Ir1-Ir2 2.744(1), Ir1-I 2.792-
(1), Ir1-P1 2.283(4), Ir1-C1 2.066(13), Ir1-C5 2.027(13), Ir1-
C6 1.922(15), Ir2-P2 2.290(3), Ir2-C16 1.831(16), Ir2-C1
2.287(13), Ir2-C2 2.358(12), Ir2-C3 2.271(11), Ir2-C4 2.337-
(12), Ir2-C5 2.397(13), Si-C2 1.915(14), C1-C2 1.415(17), C2-
C3 1.442(17), C3-C4 1.447(17), C4-C5 1.389(17); I-Ir1-P1
93.73(9), I-Ir1-C1 156.0(3), I-Ir1-C5 86.3(4), I-Ir1-C6 85.4-
(4), P1-Ir1-C1 110.2(3), P1-Ir1-C5 104.4(4), P1-Ir1-C6 91.2-
(4), C1-Ir1-C5 89.4(5), C5-Ir1-C6 162.9(5), P2-Ir2-C16
89.0(4), Ir1-C1-C2 129(1), C1-C2-C3 121(1), C2-C3-C4
126(1), C3-C4-C5 125(1), C4-C5-Ir1 128.4(9).
To “simplify” the preparation of 11, the crude mixture of
9-11 obtained after reaction workup was heated in benzene at
75 °C for 24 h (Scheme 4). Rapid chromatography gave 11 as
the major product (ca. 40%) along with a new red byproduct,
complex 13 (ca. 5%), which is moderately air-stable as a solid
at 0 °C. Spectroscopic data indicated that 13 contained two
iridium atoms with different coordination spheres: 13C NMR
data exhibited two lower field resonances at 179.96 and 174.82
ppm, attributed to two CO ligands. As then expected, two
carbonyl bands at 2026 and 1950 cm-1 were observed in the
IR spectrum. The 31P NMR spectrum showed two sets of PPh3
Ir2 is bound to CO and PPh3 ligands as well as η6-bound to the
iridabenzene backbone, with an average Ir2-Cring distance of
2.33 Å. The iridabenzene ring is nearly planar with an average
deviation of 0.039 Å, which is slightly larger than for free
iridabenzene 2a (0.024 Å).5a Ir1-C1 and Ir1-C5 (2.066(13)
and 2.027(13) Å) as well as the C-C bond distances in the
ring (ranging from 1.389 to 1.447 Å) are very close to each
other, respectively, but are slightly longer than those in
uncoordinated 2a (1.334-1.410 Å).5a These data nonetheless
suggest that coordinated iridabenzene 13 possesses a delocalized
aromatic π-system. The torsion angle of Ir1-C1-C2-C3 (13°)
is much larger than that in iridabenzene 2a (-2°) and is likely
caused by steric repulsion between the bulky SiMe3 and Ph
substituents. The longer Ir1-C6 bond distance (1.922(15) Å)
versus Ir2-C16 (1.831(16) Å) may be attributed to stronger
back-bonding from Ir2 to the CO ligand and reflects the electron
difference of the coordination environments around both Ir
centers.
The structure of complex 13 shows that the positions of the
SiMe3 and Ph groups are the same as those in iridabenzene
regioisomer 12, the most likely precursor of 13. Attempts to
improve the yield of 13 by fine-tuning conditions, such as adding
excess Vaska’s complex or excess iodide to the crude reaction
mixture before heating, were unsuccessful. While it is not
entirely clear how the [Ir(PPh3)(CO)] and [Ir(CO)I(PPh3)]
fragments in complex 13 are generated, a likely explanation is
that during the heating step in which it forms, iridabenzene 12
is intercepted by an adventitious Ir(I) fragment (possibly via
11); subsequent phosphine replacement with I- results in a
neutral, isolable species.
1
resonance peaks at 17.00 and 11.06 ppm. H NMR spectra
showed the three CH resonances at 6.07, 5.81, and 5.16 ppm,
significantly upfield compared with the corresponding CH
proton resonances (10.85 and 8.78 ppm, with one peak masked
by the Ph multiplets) in iridabenzene 12. These latter data are
very common for η6-coordinated metallabenzene complexes1,19
and thus suggested that 13 is a dinuclear complex in which the
iridabenzene is coordinated to a second iridium fragment.
The exact structure of 13 (Figure 1) was confirmed by X-ray
diffraction of crystals obtained from Et2O/hexane overnight at
-30 °C. Selected bond lengths and bond angles are listed in
Figure 1. The crystal structure of 13 confirms our expectations
of two iridium centers, with an Ir-Ir bond distance of 2.743(1)
Å, comparable to previously determined Ir-Ir bond lengths (ca.
2.70-2.72 Å).20 Complex 13 can be formulated as a zwitterionic
species with 18 electrons on both iridium centers.21 Ir1 in the
iridabenzene ring is bound to CO, PPh3, and I ligands, whereas
(19) Bleeke, J. R.; Behm, R; Xie, Y.-F.; Chiang, M. Y.; Robinson, K.
D.; Beatty, A. M. Organometallics 1997, 16, 606-623.
(20) (a) Angoletta, M.; Ciant, G.; Manassero, M.; Sansoni, M. J. Chem.
Soc., Chem. Commun. 1973, 789-790. (b) Rausch, M. D.; Gastinger, R.
G.; Hardner, S. A.; Brown, R. K.; Wood, J. S. J. Am. Chem. Soc. 1977, 99,
7870-7876. (c) Browning, J.; Dixon, K. R.; Meanwell, N. J. Inorg. Chim.
Acta 1993, 213, 171-175.
Reaction of Ligand 5b with Vaska’s Complex. If switching
from 1 to 5a in iridabenzene synthesis led to such a more
complex reaction manifold, we were very curious to see how
inclusion of two SiMe3 substituents on 5b would alter the
reactivity with Vaska’s complex. Under identical conditions,
(21) A similar zwitterionic structure was proposed for a ruthenabenzene
η6-coordinated to a second Ru fragment; see: Lin, W.; Wilson, S. R.;
Girolami, G. S. J. Chem. Soc., Chem. Commun. 1993, 284-285.

3960 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
Scheme 5
the reaction produced iridabenzvalene 14 exclusively (Scheme
5), as evidenced by the 1H NMR spectrum of the crude mixture.
Compound 14 was purified by recrystallization in Et2O/hexane
overnight at -30 °C. It is stable for months as a solid and in
solution at room temperature under an N2 atmosphere. Complex
14 is in fact the most stable PPh3-containing benzvalene yet
produced, requiring heating in C6D6 solution at 75 °C for 20 h
for complete conversion to cyclopentadienyl complex 15; only
13% transformation was observed after 6 h at 50 °C. Surpris-
ingly, no iridabenzene intermediate was ever observed in the
1H NMR spectra during heating. While it is reasonable to assume
that arene 16 forms by isomerization of 14 based on the previous
studies from our lab,9,11 16 must be short-lived and thus
rearranges readily to 15. Interestingly, all attempts to protiode-
silylate pure 9 or 14, thus yielding either a mono- or unsubsti-
tuted metallabenzene backbone, were unsuccessful.
Figure 2. Molecular structure of iridabenzvalene 14. The thermal
ellipsoids are drawn at the 30% probability level. Selected bond
lengths (Å) and angles (deg): Ir-P1 2.367(1), Ir-P2 2.381(1), Ir-
C1 2.173(4), Ir-C2 2.189(4), Ir1-C5 2.074(4), Ir-C6 1.882(5),
Si1-C1 1.858(4), Si2-C2 1.941(4), O-C6 1.149(5), C1-C2
1.446(6), C1-C3 1.539(6), C2-C3 1.543(6), C3-C4 1.457(6),
C4-C5 1.330(5); P1-Ir-P2 106.47(4), P1-Ir-C1 105.0(1), P1-
Ir-C5 83.8(1), P1-Ir-C6 99.1(1), P2-Ir-C2 109.2(1), P2-Ir-
C5 93.4(1), P2-Ir-C6 92.2(1), C1-Ir-C2 38.7(2), C5-Ir-C6
172.6(2), Ir-C1-Si1 134.0(2), Ir-C1-C2 71.2(2), Ir-C1-C3
96.5(2), Si-C1-C2 130.4(4), C2-C1-C3 62.2(3), Ir-C2-Si2
140.8(2), Ir-C2-C1 70.0(2), Ir-C2-C3 95.8(3), Si2-C2-C1
137.4(3), Si2-C2-C3 120.9(3), C1-C2-C3 61.9(3), C1-C3-
C2 56.0(3), C1-C3-C4 117.8(4), C2-C3-C4 117.0(4), C3-C4-
C5 115.2(4), Ir-C5-C4 112.3(3).
1
The H NMR spectrum of 14 showed the typical peaks for
an iridabenzvalene skeleton at 6.47, 5.84, and 2.68 ppm, with
the upfield shift of the H3 resonance attributed to the influence
of the two SiMe3 groups (δH3 3.08 ppm in 9). The CS-symmetry
of the complex was confirmed by NMR: one set of Ph peaks
for the PPh3 ligand was observed in the proton spectrum, and
the 31P NMR spectrum showed only one resonance at 3.09 ppm.
The IR spectrum possessed a strong CO absorption band at 1973
cm-1. The structure of complex 14 was further confirmed by
single-crystal X-ray diffraction; the molecular structure and
selected bond lengths and angles are given in Figure 2. The
X-ray data confirm a symmetrical plane defined by Ir-C3-
C4-C5, the midpoint between C1-C2, and the CO group.
Complex 14 has a trigonal bipyramidal configuration similar
to diphenyl analogue 4a.9,11 The C1-C2 bond distances in 14
(1.446 Å) and 4a (1.447 Å) are essentially identical. Addition-
ally, the C5-Ir-C6 bond angle (172.6°) is smaller than in 4a
(178.9°), while the Ir-C1 and Ir-C2 bond distances (2.173,
2.189 Å for 14) are slightly longer (2.146, 2.143 Å for 4a),
data reflecting the steric repulsion between the bulky SiMe3
groups and the PPh3 ligands.
Scheme 6
Table 1. Reaction of Lithiated 5a with Vaska-Type
Complexes and Subsequent Heating of 17a-e in C6D6
Solution at 75 °C
Reaction of Ligand 5a with Vaska-Type Complexes.
Reaction of cyclopropene 5a with Vaska-type complexes
containing alkylphosphines with different cone angles was next
yield of
17 (%)
cone
heating
1
entry
PR3
angle (deg)a time (h) ratio 17:18:19
studied. H NMR spectra of crude reaction mixtures indicated
17a
17b
17c
17d
50
85
42
59
PMe3
PMe2Ph
PEt3
118
122
132
136
24
24
24
24
90
24
24
100:0:0
>99:trace:0
100:0:0
that benzvalenes 17a-e are the sole isolable product (Scheme
6 and Table 1). Unlike 9, complexes 17a-e underwent
significant decomposition on attempted chromatographic puri-
fication. Instead, treatment of the mixture with Bu4NF (in which
unreacted or protiodeiodonated 5a could be desilylated and the
resultant unstable, volatile cyclopropene byproducts removed)
afforded 42-85% yields of 17a-e as pale yellow oils, which
were stable at room temperature. The 1H NMR spectra of 17a-e
showed the characteristic resonance of the sp3 CH group at
3.12-3.46 ppm, and the CO absorption band in the IR spectra
PMePh2
59:23:18
0:0:100
17e
9
63
Pi-Bu3
PPh3
143
145
0:0:100b
0:0:100c
a Reference 22. b No 18e was observed. c Ratio of 9:12:11.
to depend upon the cone angle of the phosphine ligands22 and
the temperature applied. The distribution of complexes 17, 18,
and 19 in the thermal interconversion is summarized in Table
was in the range 1986-1965 cm-1
.
The thermal stabilities of iridabenzvalenes 17a-e were
investigated, and their rearrangement and decomposition found
(22) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.

Rearrangement of IridabenzValenes to Iridabenzenes
Organometallics, Vol. 26, No. 16, 2007 3961
Scheme 7
1. Whereas complex 17a was stable at 75 °C for 24 h, at
100 °C in toluene-d8 it gradually decomposed with trace
formation of iridabenzene 18a and, more importantly, with no
evidence of formation of 19a. Iridabenzvalenes 17b,c exhibited
similar behavior. Conversely, thermolysis of 17e afforded
cyclopentadienyl 19e cleanly without intermediate iridabenzene
16e being observed, analogous to the transformation of 14 f
15. Benzvalene 17d was the sole species to give a mixture of
complexes after 24 h at 75 °C, yielding a 59:23:18 ratio of 17d,
18d, and 19d, and the only system to generate an iridabenzene
in an appreciable amount. Conversion to exclusively 19d was
accomplished by protracted heating (90 h). It is doubtful that
increasing the cone angle sterics by 4° from PEt3 to PMePh2 is
solely responsible for inducing the rearrangement of the d series.
As we have previously observed,9 the electron-donating ability
of the phosphine ligands plays a significant role in benzvalene
stability. PEt3 with three alkyl groups is considerably more
donating than PMePh2; thus, the more electron-rich Ir center in
17c back-bonds more efficiently and stabilizes the benzvalene.
Similarly, the cone angle of Pi-Bu3 is very close to that of PPh3,
yet the reaction with (Pi-Bu3)2Ir(CO)Cl produced only 17e at
room temperature, in stark contrast to the three products
generated in the reaction with Vaska’s complex (9-11) under
the same conditions.
iridium metal. This was done for several reasons. First, the scope
and primary purpose of this study was to investigate electronic
effects exerted by phenyl and trimethylsilyl groups. Second, the
previous section showed convincingly that the well-known
electronic effects of PR3 groups are more important than steric
influences. Last, calculations with triphenylphosphine ligands
are still computationally unfeasible for full Hessian characteriza-
tion (∼1000 basis functions). Constrained optimizations (freez-
ing breaking/forming bonds) on test cases were performed and
showed no significant steric interaction with triphenylphosphine
groups in the transition states.
In 2004, Martin and co-workers published a series of
introductory computational studies into metallabenzene chem-
istry using the MPW1K density functional.27 Five significant
results were proposed and are summarized here. (1) Metalla-
benzenes are formed from metallabenzvalene intermediates
through Ir-C and C-C bond breaking. The alternative route
through a Dewar-metallabenzene species is significantly higher
in energy. (2) Metallabenzenes rearrange through a carbene
migration reaction with a three-atom-centered asymmetric (two
nonequivalent Ir-C bond lengths) transition state that couples
the C-C bond by pinching two CH groups together. To
accommodate the C-C bond formation, the Ir metal center and
ligands are distorted from planarity. (3) The C-C coupling is
followed by phosphine ligand loss to generate the η5-cyclopen-
tadienyl complex in a generally exergonic process. (4) Metal
ligands may alter the barrier height and stability of metalla-
benzenes. For example, changing from (C5H5Ir)(PH3)3 to (C5H5-
Computational Studies
In this section we report a computational investigation for
the determination of the stability of phenyl- and trimethylsilyl-
substituted iridabenzvalenes and iridabenzenes. We employed
the hybrid B3LYP functional23 with the LANL2DZ24 (ECP)
basis set for geometry optimizations and thermal corrections,
and the 6-311+G(2d,p)/LANL2DZ basis set for energies as
implemented in the Gaussian 03 suite of programs.25 Stuttgart-
Dresden effective core potentials resulted in nearly identical
geometries.26 All stationary points were verified as minima or
first-order saddle points by full calculation of the Hessian and
a harmonic frequency analysis. Our model system employs
phosphine ligands in place of triphenylphosphine ligands on the
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
Revision B.05; Gaussian, Inc.: Pittsburgh, PA, 2003.
(26) Fuentealba, P.; Preuss, H.; Stoll, H.; von Szentpaly, L. Chem. Phys.
Lett. 1982, 89, 418-422.
(27) (a) Iron, M. A.; Martin, J. M. L.; van der Boom, M. E. J. Am. Chem.
Soc. 2003, 125, 13020-13021. (b) Iron, M. A.; Lucassen, A. C. B.; Cohen,
H.; van der Boom, M. E.; Martin, J. M. L. J. Am. Chem. Soc. 2004, 126,
11699-11710. (c) See also: Zhu, J.; Jia, G.; Lin, Z. Organometallics 2007,
26, 1986-1995.
(23) The B3LYP functional has been successful for a variety of
organometallic reactions; see for example: (a) Strassner, T.; Taige, M. A.
J. Chem. Theory Comput. 2005, 1, 848-855. (b) Oxidative addition: de
Jong, G. T.; Bickelhaupt, F. M. J. Chem. Theory Comput. 2006, 2, 322-
335. (c) de Jong, G. T.; Bickelhaupt, F. M. J. Phys. Chem. A 2005, 109,
9685-9699. (d) de Jong, G. T.; Geerke, D. P.; Diefenbach, A.; Bickelhaupt,
F. M. Chem. Phys. 2005, 313, 261-270. (e) de Jong, G. T.; Geerke, D. P.;
Diefenbach, A.; Sola`, M.; Bickelhaupt, F. M. J. Comput. Chem. 2005, 26,
1006-1020.
(24) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.

3962 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
Figure 3. B3LYP minima for iridabenzvalenes 22A-E (compounds 4a, 9, 14).
Ir)(PH3)2Cl2 lowers the ∆Gq from 44.4 kcal mol-1 to 35.2 kcal
Table 2. B3LYP/6-311+G(2d,p)/LANL2DZ(ECP) Reaction
Energies for Iridabenzvalenes 22A-E and Activation
Energies for Rearrangement to Iridabenzenesa
mol-1. (5) Substitution on the hydrocarbon ring para to the
metal fragment also affects the barrier height. A Hammett plot
for para-NH2-, OMe-, Cl-, COOH-, and NO2-substituted (C5H5-
Ir)(PH3)3 activation energies against σ+ values indicated a
reductive elimination mechanism, since electron-donating groups
raise the barrier height for carbene migration, while electron-
withdrawing groups lower the barrier height.
The purpose of the current hybrid-DFT study is threefold:
(1) explore the reactivity differences between phenyl- and
trimethylsilyl-substituted iridabenzvalenes and iridabenzenes; (2)
investigate the regioselective rearrangement to the meta-
(trimethylsilyl) iridabenzene 12 over the ortho-(trimethylsilyl)
iridabenzene 10; and (3) investigate the unexpected stability of
iridabenzvalene 14 and instability of the bis(trimethylsilyl)-
iridabenzene 16. To this end, the stationary points along the
minimum energy reaction pathway from iridacyclopropene
complex 20 to the cyclopentadienyliridium complex 27 for
A-E, where R1 and R2 ) H, Ph, or SiMe3, have been
investigated (Scheme 7).
species
R1, R2
∆Eb
∆Hc
∆Gc
22A
23A
22B
23B
22C
23C
22D
23D
22E
23E
H, H
H, H
Ph, Ph
Ph, Ph
SiMe3, Ph
SiMe3, Ph
SiMe3, Ph
Ph, SiMe3
SiMe3, SiMe3
SiMe3, SiMe3
-13.9
-14.2
-11.2
29.7 [15.8]
-5.8
28.0 [13.8]
-6.5
27.4 [16.2]
-3.1
21.9 [16.1]
-6.0
20.2 [13.7]
-6.8
21.1 [18.0]
-3.8
28.8 [22.8]
-6.0
27.4 [20.6]
-6.8
28.1 [24.3]
-3.8
27.7 [21.7]
-5.8
26.1 [19.3]
-6.5
27.5 [23.7]
-1.3
35.2 [29.4]
33.5 [27.0]
32.5 [31.2]
a Values relative to the iridium vinylcyclopropene complexes 20A-E
are bracketed. Reaction energies are italicized (kcal mol-1). b ZPE-
corrected. c Corrections done at 298 K.
22B-E compared to 22A is due to longer Ir-C bond lengths,
which range from 2.171 to 2.182 Å compared to 2.147 Å for
22A, caused by the repulsion of the metal center and ligands
with the phenyl and trimethylsilyl groups.
The reaction involving unsubstituted iridium σ-vinyl complex
20A (R1, R2 ) H) involves a symmetric transition state 21A,
with an activation free energy of 7.4 kcal mol-1, that leads to
the corresponding iridabenzvalene 22A. In the cases of phenyl-
and trimethylsilyl-substituted R1 and R2 groups, the potential
energy surfaces for formation of the iridacyclopropane Ir-C
bonds are flat with no well-defined transition state (see
Supporting Information). The B3LYP iridabenzvalene structures
22A-E are shown in Figure 3. The ∆Grxn for 22A, B, C/D,
and E are -11.2, -3.1, -3.8, and -1.3 kcal mol-1, respectively
(Table 2), indicating relatively weak η2-interactions for 22B-
E. The approximately 7 kcal mol-1 drop in reaction energy for
As discussed previously, Iron et al. showed that the minimum
energy pathway for transformation of iridabenzvalene to irida-
benzene involves transition state 23, where the Ir-C2 and C1-
C3 bonds break along with twisting the C2 atom of the
iridacyclopropane ring, to give the planar iridabenzene 24
(Scheme 7). Figure 4 shows the calculated transition states 23B-
E. The activation barriers corresponding to these transition states
are reported in Table 2. For the unsubstituted benzvalene 23A,
the activation free energy is 27.4 kcal mol-1. The 6.3 kcal mol-1
lowering of the ∆Gq for the bis(phenyl)benzvalene transition
state 23B (21.1 kcal mol-1) compared to 23A is due to a less
stable iridabenzvalene complex 22B. Barrier heights relative to

Rearrangement of IridabenzValenes to Iridabenzenes
Organometallics, Vol. 26, No. 16, 2007 3963
Figure 4. Transition states for iridacyclopropane ring opening of iridabenzvalenes.
the respective iridium vinylcyclopropene complexes for 23A
and 23B are very close in energy (15.8 and 16.1 kcal mol-1).
Benzvalene 22C/D has the possibility to rearrange by
breaking either the Ir-C1 or Ir-C2 bonds. Transition state 23C
reveals the ortho-SiMe3 metallabenzene, whereas 23D forms
the meta-SiMe3 metallabenzene. The ∆Gq values for 23C and
23D are 28.1 and 27.5 kcal mol-1, respectively, about 6-7 kcal
mol-1 higher than 23B. Although 23D is only favored by 0.6
kcal mol-1, there is a substantial difference in its geometry
compared to 23C. The Ir-C2 bond length in 23D is 2.460 Å,
much shorter than 2.595 Å for 23C. There is also a substantial
difference in C1-C3 partial bond lengths, 2.334 versus 2.269
Å for 23D and 23C, respectively. The selectivity between 23C
and 23D shows preference for breakage of the Ir-C bond with
the SiMe3 attached due to the ability of this group to stabilize
through σ-donation. The extent of this effect is evident from
the charges developed in the transition state. For 23C, the C1
and C2 atoms have Mulliken charges of -0.56e and +0.78e.
The magnitudes of these charges are much larger than for 23D,
which are +0.25e and -0.29e for C1 and C2 atoms, respec-
tively, where the trimethylsilyl group stabilizes the developing
positive charge at the C2 atom position.
Figure 5. Transition state for C-C bond coupling of (C5H5Ir)-
(PH3)2CO.
charges, -0.12e and -0.60e for atoms C1 and C2.28 Beyond
the electronic effects of two SiMe3 groups, there is also increased
steric repulsion between these groups in the transition state. In
22E, the nearest Si-Me/Si-Me interaction is 2.45 Å, while in
23E, this decreases to 2.29 Å.
We have also investigated the reactivity and substituent effects
for the C-C bond coupling, referred to as a carbene migratory
insertion, of 24A-E to their corresponding η1- and η5-
cyclopentadienyliridium complexes 26 and 27. Figure 5 shows
transition state 25A for the unsubstituted iridabenzene (C5H5-
Ir)(PH3)2CO. The activation free energy is 41.8 kcal mol-1 and
The barrier for rearrangement involving transition state 23E
is 32.5 kcal mol-1, 11.4 kcal mol-1 higher than that involving
the bis(phenyl)-substituted 23B transition state. This large
difference in activation free energy is likely the reason for the
observed stability of compound 14. The geometry of 23E is
similar to 23C, with a long Ir-C2 partial bond length of 2.564
Å and C1-C3 partial bond length of 2.291 Å. The σ-donation
of two trimethylsilyl groups, due to the low electronegativity
of the silicon atom, is apparent from the transition state atomic
(28) (a) Basindale, A. R.; Eaborn, C.; Walton, D. R. M.; Young, D. J. J.
Organomet. Chem. 1969, 20, 49-56. (b) Cook, M. A.; Eaborn, C.; Walton,
D. R. M. J. Organomet. Chem. 1970, 24, 293-299. (c) Campanelli, A. R.;
Domenicano, A.; Ramondo, F. J. Phys. Chem. A 2003, 107, 6429-6440.
(d) Schormann, M.; Garratt, S.; Bochmann, M. Organometallics 2005, 24,
1718-1724.

3964 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
Figure 6. Transition states 25B-D for C-C bond coupling of iridabenzenes 24B-D.
Table 3. B3LYP/6-311+G(2d,p)/LANL2DZ(ECP) Activation
and Reaction Energies, Enthalpies, and Free Energies for
C-C Coupling of 24A-Ea
the ∆Grxn is -24.7 kcal mol-1 for rearrangement to the η5-
complex. 25A shows the C1-C5 bond coupling via an Ir-C1-
C5 three-centered transition state. The C1-C5 distance of
iridabenzene 24A is 2.751 Å and decreases to 2.225 Å in 25A.
Transition state 25A is asymmetric, with two very different Ir-C
bond lengths, 1.935 and 2.242 Å for the Ir-C1 and Ir-C5
bonds, respectively. As the reaction progresses toward the
transition state from 24A, which has two equivalent Ir-C bond
lengths of 1.992 Å, the degeneracy is lifted and it becomes
evident that the C5 position is more appropriately classified as
a vinyl anion, while the C1 position is better classified as a
metal-carbon double bond. This is also evident from the large
dihedral angle difference between D(H-C1-C2-H) and D(H-
C5-C4-H), 71.4° and -8.6°, respectively. The torquing motion
at the C1-H center aligns the C1 p-orbital with the C5 vinyl
anion orbital in the plane of the metallabenzene ring; therefore,
π-interactions at C1 will directly influence bond formation,
while C2-C5 substitution will affect bond formation indirectly.
Figure 6 shows the transition structures (25B-E) for C-C
bond coupling of 24B-E. Table 3 reports the B3LYP activation
and reaction energetics for the reactions involving these
transition states. The ∆Gq for bis(phenyl) 25B is 28.5 kcal
mol-1, a 13.3 kcal mol-1 lower barrier than 25A with no
substitution. The ∆∆Gq for 25B compared to 25A is larger than
the sum of single Ph substitution effects (see Supporting
Information), indicating a synergetic lowering of the barrier due
to ground state destabilization of iridabenzene 24B from adjacent
Ph groups and because the Ph groups are twisted compared to
the metallabenzene ring system in the bis(phenyl) iridabenzenes.
The torquing of the C1-Ph group in the transition state allows
the C2-Ph group to be conjugated with the C2-C5 hydrocarbon
ring fragment. This is reflected in the intermediate C2-CB bond
species
R1, R2
∆Eb
∆Hc
∆Gc
25A
27A
25B
27B
25C
27C
25D
27D
25E
27E
H, H
H, H
Ph, Ph
Ph, Ph
SiMe3, Ph
SiMe3, Ph
Ph, SiMe3
Ph, SiMe3
SiMe3, SiMe3
41.5 [1.6]
41.4 [1.0]
41.8 [4.0]
-14.3 [-54.2] -14.1 [-54.5] -24.7 [-62.5]
28.5 [8.6] 28.4 [8.1] 28.5 [12.3]
-21.9 [-41.8] -21.7 [-42.1] -32.6 [-48.8]
22.8 [15.2] 23.0 [14.7] 22.1 [18.5]
-31.9 [-24.9] -31.6 [-24.5] -43.3 [-46.9]
27.6 [12.9] 27.7 [12.3] 27.1 [16.6]
-52.5 [-24.9] -52.2 [-24.5] -63.5 [-46.9]
15.6 [17.4] 15.8 [16.6] 15.6 [22.7]
SiMe3, SiMe3 -39.6 [-37.9] -39.4 [-38.6] -49.9 [-42.8]
a Reaction energies are italicized. Values in brackets are energetics
compared to the iridium vinylcyclopropene species 20A-E (kcal mol-1).
c
b Values are ZPE-corrected. Corrections done at 298 K.
length of 1.481 Å in 25B. The torquing motion also allows the
full effect of the Ph group at the C1 center, which has a larger
influence on bond formation than substitution at the C2 position.
Computational analyses for iridabenzenes 24C and 24D reveal
the barriers for C-C bond coupling are quite different. In
agreement with the experimental results for 10 and 12, transition
structure 25C has a barrier of 22.1 kcal mol-1, while 25D has
a barrier of 27.1 kcal mol-1. This is consistent with the C1
position being directly involved with the C-C bond coupling
process and SiMe3 group having in general a larger effect on
the barrier than Ph. The largest geometric difference between
25B and 25C/D is the dihedral angle D(CB-C2-C3-C4),
which is 160.6°, due to repulsion from the adjacent SiMe3 group,
compared to 180.0° for 25B.
The lowest barrier for rearrangement was computed for bis-
(trimethylsilyl) 25E. The ∆Gq value is only 15.6 kcal mol-1
,

Rearrangement of IridabenzValenes to Iridabenzenes
Organometallics, Vol. 26, No. 16, 2007 3965
their respective η5-cyclopentadienyliridium complexes 27. These
potential energy surfaces clearly show the dramatic reactivity
difference between trimethylsilyl and phenyl groups at R1 and
R2. The largest difference in these surfaces is the (relatively)
shallow minimum associated with iridabenzene 24E on the blue
surface (as discussed above).
Conclusions
The reaction of 1-phenyl-2-trimethylsilyl- or 1,2-bis(trimethyl-
silyl)-substituted 3-iodovinylcyclopropenes with Vaska’s and
Vaska-type complexes produced either iridabenzenes, iridaben-
zvalenes, and/or cyclopentadienyliridium complexes, depending
on both the substituents on the C5 framework and the phosphine
ligands on the Ir center. Compared to previous results with the
diphenylcyclopropene analogue, inclusion of SiMe3 groups
stabilizes the iridabenzvalene and correspondingly destabilizes
the iridabenzene when utilizing PPh3 ligands. Switching to more
electron-rich phosphines afforded iridabenzvalenes that were
stable for extended periods at 75 °C. Hybrid-DFT calculations
along the minimum energy pathway for conversion of iridavi-
nylcyclopropenes 20A-E to cyclopentadienyliridium complexes
27A-E revealed that phenyl- and trimethylsilyl-substituted
benzvalenes form relatively weak π-complexes due to steric
repulsions, and the barrier for rearrangement to iridabenzene is
controlled by electronic effects in the transition state and not
the degree of π-complexation. Trimethylsilyl-substituted ben-
zvalenes have larger barriers for rearrangement than Ph-
substituted species. For iridabenzvalene 9(22C/D), there is a
preference for breakage of the Ir-C bond with the SiMe3
attached due to charge stabilization. For the C-C bond coupling
carbene migration of iridabenzene to the cyclopentadienyliridium
complex, ortho-trimethylsilyl- and ortho,meta-bis(trimethylsi-
lyl)iridabenzenes have low barriers for rearrangement, due to
direct electronic effects at the ortho position. Also, the bis-
(trimethlysilyl)iridabenzene has significant steric repulsion
between adjacent groups, leading to a nonplanar geometry, that
is relieved in the transition state, making this a very facile
rearrangement process. Future work will focus on the preparation
of irida-aromatics with different substituents, as well as reaction
of cyclopropene ligands 5a,b with additional transition metal
complexes. The results of these studies will be reported shortly.
Figure 7. Nonplanar minimum for bis(trimethylsilyl) 24E.
Experimental Section
(Iodomethyl)triphenylphosphonium iodide,17 1,1,1-triacetoxy-
1,1-dihydro-1,2-benziodoxol-3(1H)-one,16 the Vaska-type com-
plexes,29 and cyclopropene derivatives 6b-8b15 were prepared
according to literature procedures. All other chemicals were
purchased from commercial suppliers and used as received. Column
chromatography was performed on Whatman reagent grade silica
gel (230-400 mesh). Manipulation of organometallic reagents was
carried out using either a Vacuum Atmospheres inert atmosphere
glovebox or standard Schlenk techniques. THF, Et2O, and hexanes
were distilled from Na/benzophenone and degassed by three freeze/
pump/thaw cycles prior to use. Toluene and benzene were distilled
from LiAlH4 and degassed by three freeze/pump/thaw cycles prior
to use. NMR spectra were recorded on a Varian Unity-INOVA 300
spectrometer at ambient temperature. 1H, 13C, and 31P NMR spectra
were acquired at 299.95, 75.43, and 121.42 MHz, respectively.
Chemical shifts for 1H and 13C NMR spectra are reported in parts
per million (δ) downfield from tetramethylsilane using the residual
Figure 8. Free-energy potential surfaces (kcal mol-1) for the
rearrangements of 22A (black), 22B (red), and 22E (blue).
indicating a very facile reaction (and thus why we never observe
16). Similar to phenyl substitution, the barrier for 25E is lower
than expected on the basis of individual effects of meta or ortho
SiMe3 groups, since the ground state is destabilized due to
adjacent bulky SiMe3 groups. In fact, the repulsion is so large
that the lowest energy geometry of 24E is nonplanar (Figure
7).27c The SiMe3 groups have a dihedral angle (Si-C1-C2-
Si) of 36.9°, and the C2 carbon is slightly pyramidalized to
2.2°. This nonplanar structure for 24E is 2.1 kcal mol-1 lower
in energy than the planar geometry. Progression toward the
transition state 25E maximizes the relief of this repulsive
interaction, with a dihedral angle D(Si-C1-C2-Si) of 81.4°.
In the bis(trimethylsilyl) cyclopentadienyl complex 27E this
dihedral decreases to 4.4° to allow for anion delocalization.
Finally, Figure 8 depicts a composite of the potential energy
surfaces for conversion of benvalenes 22A, 22B, and 22E to
(29) (a) Collman, J. P.; Sears, C. T., Jr.; Kubota, M. Inorg. Synth. 1973,
14, 92-93. (b) Burk, M. J.; Crabtree, R. H. Inorg. Chem. 1986, 25, 931-
932.

3966 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
solvent signal (CDCl3: 1H 7.26, 13C 77.00; C6D6: 1H 7.16, 13C
128.39) as an internal standard. The 31P NMR spectrum is
referenced relative to external H3PO4 or PPh3. Coupling constants
are reported in hertz. FT-IR spectra were recorded using a Nicolet
Magna 550 FT-IR spectrometer. Elemental analyses were performed
by Robertson Microlit Laboratories, Inc.
129.55, 129.51, 128.95, 114.8, 76.17, 26.93, -0.94. IR (Et2O): ν
1847 (cyclopropene CdC) cm-1
.
(Z)-3-(2-Iodoethenyl)-1,2-bis(trimethylsilyl)-1-cyclopropene (5b).
As described above for the preparation of compound 5a, (iodom-
ethyl)triphenylphosphonium iodide (1.23 g, 2.32 mmol) was treated
with NaN(SiMe3)2 (1 M in THF, 2.1 mL, 2.1 mmol) and HMPA
(0.7 mL), respectively. Wittig reaction of the resulting ylide with
aldehyde 8b (0.55 g, 2.58 mmol) gave 5b (0.25 g, 35%) as a
colorless oil after chromatography on silica (hexanes). The pure
material was stored similarly to 5a. 1H NMR (CDCl3): δ 5.75 (d,
J ) 7.4 Hz, 1H), 5.45 (t, J ) 7.4 Hz, 1H), 2.05 (d, J ) 7.5 Hz,
1H), 0.20 (s, 18H). 13C NMR (CDCl3): δ 149.15, 134.75, 72.58,
25.00, -1.32.
Synthesis of Complexes 9-11. A flame-dried Schlenk flask (50
mL) was charged with cyclopropene 5a (0.30 g, 0.93 mmol) and
Et2O (20 mL) and cooled to -78 °C. To this was added BuLi (375
µL, 2.5 M in hexane, 0.93 mmol) at -78 °C over 5 min. After
stirring for 15 min at -78 °C, the resulting mixture was transferred
over 5 min by a canula to a -78 °C suspension of (PPh3)2Ir(CO)-
Cl (0.60 g, 0.76 mmol) in Et2O (5 mL). The mixture was
continuously stirred at -78 °C for 30 min and then slowly warmed
to room temperature over a 3 h period. The solvent was removed
under vacuum, and the residue was chromatographed over silica
gel (8:1 hexane/Et2O) to give a 10:3:2 mixture of complexes 9,
10, and 11 (329 mg, 47% combined yield). Benzvalene 9 could be
isolated in pure form by reacting a mixture in EtOAc (2 mL) with
excess MeI for 30 min, then adding hexanes (10 mL) and filtering
off the unwanted salts. Cyclopentadienyl complex 11 could be
isolated by refluxing the mixture in benzene for 24 h.
Ethyl 2-phenyl-3-trimethylsilyl-2-cyclopropenecarboxylate (6a).
A mixture of 1-phenyl-2-(trimethylsilyl)acetylene (60 g, 0.34 mol)
and Cu powder (1.2 g, 0.02 mol) was heated to 130 °C under N2.
A solution of ethyl diazoacetate (14 g, 0.12 mol) and cyclohexane
(15 mL) was added dropwise to the stirring suspension via syringe
at a rate of 1 mL/h. After addition was complete, the mixture was
heated an additional 1 h and then cooled. The excess 1-phenyl-2-
(trimethylsilyl)acetylene was distilled under vacuum. The resulting
residue was chromatographed to give 6a14 as a colorless oil (8.93
g, 30% based on ethyl diazoacetate). Spectroscopic data matched
those previously reported.14
3-Hydroxymethyl-1-phenyl-2-trimethylsilyl-1-cyclopropene (7a).
To a solution of 6a (2.2 g, 8.45 mmol) in THF (40 mL) cooled to
0 °C was added DIBAL-H (12.67 mL, 1 M in hexane, 12.67 mmol)
via syringe over 5 min. The solution was stirred at 0 °C until no
starting material was visible by TLC. The mixture was transferred
to a separatory funnel, and potassium sodium tartrate (Rochelle’s
salt, 2.75 g) and water (10 mL) were added. After shaking, the gel
formed was extracted with Et2O (3 × 25 mL). The combined
organics were dried (MgSO4) and the solvent removed in Vacuo.
Chromatography of the residue on silica (4:1 hexanes/Et2O) afforded
1
7a (1.52 g, 81%) as a colorless oil. H NMR (CDCl3): δ 7.44 (d,
J ) 8.4 Hz, 2H), 7.26 (t, J ) 7.5 Hz, 2H), 7.24 (t, J ) 7.8 Hz,
1H), 3.50 (t, J ) 4.7 Hz, 2H), 1.83 (t, J ) 4.7 Hz, 1H), 1.10 (br s,
1H), 0.15 (s, 9H). 13C NMR (CDCl3): δ 135.75, 130.33, 129.58,
129.27, 128.92, 115.63, 69.35, 21.99, -0.74. IR (Et2O): ν 3333
1
Iridabenzvalene 9. H NMR (C6D6): δ 7.63-7.55 (m, 6H),
7.40-7.31 (m, 8H), 7.04-6.82 (m, 21H), 6.71-6.63 (m, 1H, H4),
6.39-6.28 (m, 1H, H5), 3.08 (br s, 1H, H3), 0.05 (s, 9H, Si(CH3)3).
13C NMR (C6D6): δ 181.58 (t, J ) 6.0 Hz, CO), 151.24 (t, J )
14.1 Hz), 145.76 (br s), 144.15 (t, J ) 7.1 Hz), 137.25 (dd, J )
40.3, 4.0 Hz), 135.58 (dd, J ) 41.3, 10.7 Hz), 134.67 (d, J ) 10.1
Hz), 134.18 (d, J ) 10.1 Hz), 129.65 (s), 129.24 (s), 127.99 (d, J
) 9.1 Hz), 124.27 (s), 75.81 (d, J ) 58.4 Hz), 56.97 (s), 55.54 (d,
J ) 53.4 Hz), 0.40 (s). 31P NMR (C6D6): δ 23.01 (s). IR (Et2O):
ν 1983 (s, CO) cm-1. Anal. Calcd for C51H47IrOP2Si: C, 63.93;
H, 4.94. Found: C, 64.18; H, 5.35.
Iridabenzene 10. Partial 1H NMR (C6D6): δ 10.45 (dt, J ) 11.7,
11.4 Hz, 1H), 8.14 (dt, J ) 7.3, 7.0 Hz, 1H). The remaining signals
overlapped with the large quantities of 9 and 11, which proved
impossible to remove to any significant degree.
η5-Cyclopentadienyliridium 11. 1H NMR (C6D6): δ 7.80-7.65
(m, 8H), 7.08-6.91 (m, 12H), 5.40 (dd, J ) 2.3, 1.8 Hz, 1H), 4.67
(dd, J ) 2.3, 1.8 Hz, 1H), 4.54 (ddt, J ) 2.6, 2.3, 0.9 Hz, 1H),
0.28 (s, 9H, Si(CH3)3). 13C NMR (C6D6): δ 178.96 (d, J ) 18.1,
CO), 138.26 (s), 137.40 (d, J ) 12.1 Hz), 134.69 (d, J ) 12.1 Hz),
132.26 (s), 130.42 (d, J ) 2.0 Hz), 128.47 (d, J ) 2.0 Hz), 128.35
(s), 127.72 (s), 111.85 (d, J ) 7.1 Hz), 92.25 (d, J ) 4.0 Hz),
91.06 (s), 89.67 (s), 85.47 (s), 1.76 (s). 31P NMR (C6D6): δ 17.97
(s). IR (Et2O): ν 1936 (s, CO) cm-1. Anal. Calcd for C33H32-
IrOPSi: C, 56.96; H, 4.63. Found: C, 56.80; H, 4.58.
(OH) cm-1
.
2-Phenyl-3-trimethylsilyl-2-cyclopropenecarbaldehyde (8a). A
solution of 7a (1.51 g, 6.87 mmol) in CH2Cl2 (20 mL) was added
dropwise to a mixture of 1,1,1-triacetoxy-1,1-dihydro-1,2-benzio-
doxol-3(1H)-one (3.8 g, 9.36 mmol) and CH2Cl2 (40 mL), and the
resulting mixture was stirred for 30 min at room temperature. The
reaction was quenched with 1 N NaOH solution (70 mL) and the
organic phase extracted with Et2O (2 × 60 mL). The organic layer
was dried (MgSO4) and the solvent removed in Vacuo. Chroma-
tography of the residue on silica (6:1 hexanes/Et2O) gave 8a (1.05
1
g, 71%) as a colorless oil. H NMR (CDCl3): δ 8.77 (d, J ) 7.8
Hz, 1H), 7.55 (d, J ) 7.5 Hz, 2H), 7.47 (t, J ) 7.2 Hz, 2H), 7.45
(t, J ) 7.2 Hz, 1H), 2.56 (d, J ) 7.5 Hz, 1H), 0.35 (s, 9H). 13C
NMR (CDCl3): δ 205.62, 130.49, 130.20, 129.14, 127.98, 125.25,
109.08, 34.11, -1.29. IR (Et2O): ν 1690 (s, CdO) cm-1
.
(Z)-3-(2-Iodoethenyl)-1-phenyl-2-(trimethylsilyl)-1-cyclopro-
pene (5a). (Iodomethyl)triphenylphosphonium iodide (3.47 g, 6.54
mmol) was dried at 100 °C under vacuum for 30 min prior to use.
The phosphonium salt was suspended in THF (15 mL) under an
N2 atmosphere, NaN(SiMe3)2 (1 M in THF, 5.95 mL, 5.95 mmol)
was added via syringe, and the mixture was stirred for 5 min to
give a yellow solution. The reaction was cooled to -78 °C, and
HMPA (2.0 mL) was added and stirred for 10 min. Aldehyde 8a
(1.55 g, 7.17 mmol) in THF (5 mL) was then added, and the reaction
was stirred until no 8a was detected by TLC (ca. 30 min). The
mixture was filtered through a short silica column and washed with
Et2O until no product was eluted further. The solvent was removed
in Vacuo and the residue chromatographed on silica gel (hexanes)
to give 5a (1.35 g, 58%) as a colorless oil. The material was stored
at -30 °C, and a few crystals of hydroquinone were added to neat
5a to inhibit decomposition. 1H NMR (CDCl3): δ 7.38 (d, J ) 6.9
Hz, 2H), 7.27 (t, J ) 7.2 Hz, 2H), 7.25 (t, J ) 7.2 Hz, 1H), 5.78
(d, J ) 7.5 Hz, 1H), 5.55 (t, J ) 7.5 Hz, 1H), 2.42 (d, J ) 7.5 Hz,
1H), 0.15 (s, 9H). 13C NMR (CDCl3): δ 146.96, 132.30, 129.64,
Thermal Conversion of 9 to 11 via 12. Benzvalene 9 (100 mg,
0.104 mmol) was heated in C6D6 for 24 h at 75 °C. Chromatography
over silica (6:1 hexane/Et2O) gave complex 11 (62 mg, 89% yield)
as bright yellow crystals. If the reaction was stopped after 10 h,
chromatography of the mixture gave Cp complex 11 and a 3:2
mixture of 11 and 12, the latter for spectral analysis.
Iridabenzene 12. 1H NMR (C6D6): δ 10.85 (ddt, J ) 11.7, 10.8,
1.2 Hz, 1H, H5), 8.78 (dt, J ) 6.8, 6.2 Hz, 1H, H3), 7.65-7.26
(m, 15H), 7.12-6.87 (m, 21H), 0.28 (s, 9H, Si(CH3)3). 13C NMR
(C6D6): δ 188.93 (s), 188.00 (t, J ) 6.0 Hz), 171.27 (t, J ) 4.0
Hz), 145.32 (t, J ) 8.0 Hz), 141.37 (dd, J ) 16.1, 4.0 Hz),

Rearrangement of IridabenzValenes to Iridabenzenes
Organometallics, Vol. 26, No. 16, 2007 3967
137.15 (d, J ) 5.0 Hz), 134.68 (s), 132.70 (s), 129.97 (s), 128.95
(s), 128.24 (s), 126.56 (s), 124.35 (s), 3.67 (s). 31P NMR (C6D6):
δ 17.13 (s).
Hz), 58.06 (d, J ) 54.4 Hz), 55.46 (t, J ) 4.0 Hz), 21.40 (dd, J )
28.2 Hz, J ) 4.0 Hz), 19.87 (dd, J ) 29.2 Hz, J ) 3.0 Hz), 0.49
(s). 31P NMR (C6D6): δ 29.24 (d, J ) 37.6 Hz), 28.52 (d, J )
37.6 Hz). IR (Et2O): 1971 cm-1
.
Diiridium Complex 13. Reaction of cyclopropene 5a (0.30 g,
0.93 mmol) with BuLi (2.5 M in hexane, 0.375 µL, 0.93 mmol)
and then with (Ph3P)2Ir(CO)Cl (0.60 g, 0.93 mmol) in Et2O at
-78 °C gave a mixture that was slowly warmed to room
temperature over a 3 h period. Workup was as described above for
the preparation of 9-11. The crude residue was redissolved in
benzene (2 mL) and heated for 24 h at 75 °C. Chromatography
over silica (6:1 hexane/Et2O) yielded complexes 11 (0.26 g, 39%)
and 13 (0.02 g, 4%). Red crystals of 13 were obtained by
1
Iridabenzvalene 17b: 85% yield. H NMR (C6D6): δ 7.60 (d,
J ) 7.3 Hz, 2H, o-H Ph), 7.31-6.97 (m, 14H), 6.60-6.48 (m, 1H,
H5), 3.11-3.14 (br s, 1H, H3), 1.63 (d, J ) 7.91 Hz, 3H), 1.46 (d,
J ) 8.2 Hz, 3H), 1.36 (d, J ) 8.2 Hz, 3H), 1.19 (d, J ) 8.1 Hz,
3H), 0.24 (s, 9H, Si(CH3)3). 13C NMR (C6D6): δ 179.89 (t, J )
7.0 Hz, CO), 148.17 (br s), 146.27 (t, J ) 8.1 Hz), 142.73 (t, J )
15.1 Hz), 141.61 (dd, J ) 35.2 Hz, J ) 7.0 Hz), 140.42 (dd, J )
34.2 Hz, J ) 6.0 Hz), 129.59 (d, J ) 9.1 Hz), 129.14 (d, J ) 9.1
Hz), 128.93 (d, J ) 2.0 Hz), 128.65 (d, J ) 2.0 Hz), 128.30 (s),
127.56 (s), 128.15 (s), 127.66 (s, 2C), 124.18 (s), 73.98 (dd, J )
54.4 Hz, J ) 6.0 Hz), 57.62 (dd, J ) 48.3 Hz, J ) 3.0 Hz), 55.97
(t, J ) 4.0 Hz), 21.40 (dd, J ) 28.2 Hz, J ) 7.0 Hz), 19.04 (dd,
J ) 30.2 Hz, J ) 7.0 Hz), 17.81 (dd, J ) 29.2 Hz, J ) 7.0 Hz),
15.38 (dd, J ) 28.2 Hz, J ) 5.0 Hz), 0.48 (s). 31P NMR (C6D6):
δ 41.13 (d, J ) 36.9 Hz), 40.66 (d, J ) 36.9 Hz). IR (Et2O): 1980
1
recrystallization in Et2O/hexane overnight at -30 °C. H NMR
(C6D6): δ 10.91 (d, J ) 8.8 Hz, 1H), 7.92-7.80 (m, 7H), 7.66-
7.55 (m, 6H), 7.18-7.10 (m, 6H), 7.09-6.99 (m, 9H), 6.98-6.83
(m, 5H), 6.12-6.02 (m, 1H), 5.86-5.76 (m, 1H), 5.20-5.12 (m,
1H), -0.07 (s, 9H). 13C NMR (C6D6): 179.96 (d, J ) 7.6 Hz,
CO), 174.82 (d, J ) 15.1 Hz, CO), 160.01 (s), 152.12 (s), 142.97-
(dd, J ) 21.4 Hz, 6.6 Hz), 136.90 (s), 136.12 (s), 135.69 (s), 135.29
(t, J ) 5.0 Hz),134.83 (d, J ) 12.1 Hz), 134.33 (d, J ) 10.1 Hz),
130.61 (s), 125.89 (s), 130.03 (s), 128.52 (s), 127.80 (s), 126.21
(s), 125.89 (s), 125.00 (s), 109.69 (s), 98.65 (s), 95.72 (s), 2.05 (s).
31P NMR (C6D6): δ 17.00 (d, J ) 8.5 Hz), 11.06 (d, J ) 8.5 Hz).
IR (Et2O): ν 2026, 1950 (s, CO) cm-1. Anal. Calcd for C52H47-
IIr2O2P2Si: C, 47.99; H, 3.33. Found: C, 47.75; H, 3.47.
cm-1
.
1
Iridabenzvalene 17c: 42% yield. H NMR (C6D6): δ 7.66 (d,
J ) 7.3 Hz, 2H, o-H Ph), 7.23 (t, J ) 7.6 Hz, 2H, m-H Ph), 7.02
(t, J ) 7.3 Hz, 1H, p-H Ph), 6.88-6.81 (m, 1H, H4), 6.68-5.59
(m, 1H, H5), 3.09 (br s, 1H, H3), 1.84-1.38 (m, 12H), 0.98-0.84
(m, 9H), 0.78-0.66 (m, 9H), 0.30 (s, 9H, Si(CH3)3). 13C NMR
(C6D6) δ 180.68 (t, J ) 7.5 Hz, CO), 149.25 (dd, J ) 6.0, 4.0 Hz),
145.69 (t, J ) 7.1 Hz), 139.92 (t, J ) 15.1 Hz), 127.82 (s), 127.79
(s), 124.10 (s), 72.11 (dd, J ) 62.4, 4.0 Hz), 55.46 (t, J ) 3.0 Hz),
54.72 (dd, J ) 57.4, 4.5 Hz), 21.72 (d, J ) 26.2 Hz), 20.46 (d, J
) 25.2 Hz), 8.20 (s), 7.52 (s), 0.47 (s). 31P NMR (C6D6): δ -19.30
Synthesis of Benzvalene 14. Cyclopropene 5b (0.17 g, 0.5
mmol) was reacted with BuLi (2.5 M in hexanes, 0.22 mL, 0.55
mmol) and (Ph3P)2Ir(CO)Cl (0.47 g, 0.6 mmol) as described above
for the preparation of 9. The mixture was filtered through silica
and washed with small amounts of Et2O. The filtrate was
concentrated and hexane was added. The resulting solution was
stored at -30 °C overnight to give 14 (0.39 g, 40%) as pale yellow
crystals. 1H NMR (C6D6): δ 7.61-7.55 (m, 10H), 7.08-6.90 (m,
20H), 6.51-6.44 (m, 1H, H5), 5.84 (ddt, J ) 8.2, 7.9, 2.6 Hz, 1H,
H4), 2.71-2.66 (m, 1H, H3), 0.22 (s, 18H, Si(CH3)3). 13C NMR
(C6D6): δ 180.92 (t, J ) 6.0 Hz, CO), 152.92 (t, J ) 14.1 Hz),
144.40 (t, J ) 9.1 Hz), 137.25 (t, J ) 22.2 Hz), 134.66 (dd, J )
5.0 Hz), 129.6 (s), 127.75 (t, J ) 5.0 Hz), 62.70 (m), 54.22 (t, J )
4.0 Hz), 1.72 (s). 31P NMR (C6D6): δ -3.09 (s). IR (Et2O): ν
1973 (s, CO) cm-1. Anal. Calcd for C48H51IrOP2Si2: C, 60.42; H,
5.39. Found: C, 60.22; H, 5.39.
Synthesis of η5-Cyclopentadienyliridium 15. A solution of 14
(100 mg, 0.105 mmol) in benzene (5 mL) was heated at 75 °C for
24 h. The mixture was cooled and the solvent removed in Vacuo.
Chromatography of the residue on silica (6:1 hexane/Et2O) gave
complex 15 (67 mg, 93%) as yellow crystals. 1H NMR (C6D6): δ
7.84-7.77 (m, 6H), 7.09-6.94 (m, 9H), 5.12 (d, J ) 4.3 Hz, 2H),
4.73-4.70 (m, 1H), 0.44 (s, 18H, Si(CH3)3). 13C NMR (C6D6): δ
178.24 (d, J ) 18.1, CO), 137.69 (d, J ) 57.4 Hz), 134.28 (d, J )
12.1 Hz), 129.95 (d, J ) 2.0 Hz), 127.86 (s), 95.61 (s), 93.48 (s),
93.41 (s), 87.93 (s), 2.03 (s). 31P NMR (C6D6): δ 17.84 (s). IR
(Et2O): ν 1940 (s, CO) cm-1. Anal. Calcd for C30H36IrOPSi2: C,
52.07; H, 5.24. Found: C, 51.91; H, 5.34.
(d, J ) 37.1 Hz), 21.32 (d, J ) 37.1 Hz). IR (Et2O): 1966 cm-1
.
1
Iridabenzvalene 17d: 59% yield. H NMR (C6D6): δ 7.57-
7.48 (m, 5H), 7.33-7.23 (m, 4H), 7.18-6.86 (m, 17H), 6.83-
6.73 (m, 1H), 3.46 (br s, 1H), 1.73 (d, J ) 7.8 Hz, 3H,), 1.55 (d,
J ) 8.2 Hz, 3H,), 0.20 (s, 9H, Si(CH3)3). 13C NMR (C6D6): δ
180.19 (t, J ) 6.0 Hz, CO), 147.37 (dd, J ) 6.0, 3.0 Hz), 145.87
(t, J ) 8.0 Hz), 144.69 (t, J ) 14.5 Hz), 138.58 (m, 2C), 133.15
(d, J ) 11.1 Hz), 132.55 (d, J ) 10.1 Hz), 132.18 (d, J ) 12.1
Hz), 131.96 (d, J ) 10.8 Hz), 129.45 (dd, J ) 28.2, 2.0 Hz), 129.39
(d, J ) 17.1 Hz), 128.20 (s), 128.00 (s), 127.83 (s), 124.15 (s),
76.90 (dd, J ) 65.5 Hz, J ) 3.0 Hz), 58.79 (dd, J ) 57.4, 4.0 Hz),
56.91 (t, J ) 4.0 Hz), 18.54 (d, J ) 31.2 Hz), 16.87 (d, J ) 33.2
Hz), 0.84 (s). 31P NMR (C6D6): δ -23.27 (d, J ) 39.7 Hz), -24.76
(d, J ) 39.7 Hz). IR (Et2O): 1986 cm-1
.
1
Iridabenzvalene 17e: 63% yield. H NMR (C6D6): δ 7.68 (d,
J ) 7.3 Hz, 2H), 7.23 (d, J ) 7.5 Hz, 2H), 7.04 (t, J ) 7.5 Hz,
1H), 6.92-6.78 (m, 2H), 3.12 (br s, 1H), 2.22-1.93 (m, 12H),
1.80-1.72 (m, 6H), 1.18-0.88 (m, 36H), 0.34 (s, 9H, Si(CH3)3).
13C NMR (C6D6): δ 181.52 (t, J ) 6.5 Hz, CO), 148.97 (dd, J )
6.6, 4.6 Hz), 145.81 (t, J ) 8.1 Hz), 141.01 (t, J ) 15.1 Hz), 127.91
(m, 2C), 124.06 (s), 72.79 (d, J ) 63.5 Hz), 55.93 (br s), 54.81
(dd, J ) 59.4, 3.8 Hz), 40.90 (d, J ) 21.1 Hz), 39.91 (d, J ) 21.1
Hz), 25.80 (d, J ) 7.0 Hz), 25.63 (s), 0.68 (s). 31P NMR (C6D6):
δ 17.84 (d, J ) 35.1 Hz), 18.78 (d, J ) 35.1 Hz). IR (Et2O): 1965
General Preparation of Benzvalenes 17a-e. Reaction of
cyclopropene 5a with BuLi, then with (R3P)2Ir(CO)Cl in Et2O at
-78 °C, gave a mixture that was slowly warmed to room
temperature over a 3 h period. Workup was as described above for
the preparation of 9. The resulting mixture was treated with Bu4-
NF (1 M in THF, 2 equiv) and stirred for 5 h. Filtration and removal
of the volatiles in Vacuo gave 17a-e as pale yellow oils.
cm-1
.
Thermolysis of Benzvalene 17a. Complex 17a (20 mg, 0.034
mmol) was dissolved in benzene-d6 (0.5 mL) and heated at 75 °C
for 24 h. 1H NMR spectroscopy showed no changed. Switching to
toluene-d8 and heating at 100 °C for 24 h afforded a ∼40:1 mixture
of 17a and 18a, along with significant decomposition. Partial data
for 18a: 1H NMR (C7D8) δ 11.08 (ddt, J ) 12.7, 10.0, 1.2 Hz,
H5), 8.52 (ddt, J ) 8.1, 5.9, 1.2 Hz, H3), 7.70 (t, J ) 8.9 Hz, H4).
1
Iridabenzvalene 17a: 50% yield. H NMR (C6D6): δ 7.60 (d,
J ) 7.3 Hz, 2H, o-H Ph), 7.22 (t, J ) 7.6 Hz, 3H, m-H Ph), 7.04-
6.94 (m, 2H, p-H Ph and H4), 6.45-6.36 (m, 1H, H5), 3.12 (br s,
1H, H3), 1.31 (d, J ) 8.8 Hz, 9H), 1.08 (d, J ) 8.8 Hz, 9H), 0.27
(s, 9H, Si(CH3)3). 13C NMR (C6D6): δ 179.74 (t, J ) 7.0 Hz, CO),
148.81 (br s), 146.62 (t, J ) 8.1 Hz), 141.13 (t, J ) 15.1 Hz),
128.18 (s), 127.56 (s), 127.53 (s), 123.97 (s), 74.97 (d, J ) 62.4
Thermolysis of Benzvalene 17b. Complex 17b (20 mg, 0.028
mmol) was dissolved in benzene-d6 (0.5 mL) and heated at 75 °C
1
for 24 h. H NMR spectroscopy showed the solution to be a 20:1

3968 Organometallics, Vol. 26, No. 16, 2007
Wu et al.
mixture of 17b and 18b. Partial data for 18b: 1H NMR (C6D6) δ
11.28 (q, J ) 11.3 Hz, H5), 8.72 (q, J ) 8.5 Hz, H3), 7.91 (t, J )
8.6 Hz, H4).
X-ray Structure Determinations. Data were collected on an
Enraf-Nonius CAD-4 Turbo diffractometer using Mo KR radiation
(λ ) 0.71073 Å), graphite monochromator, T ) 296 K, and scan
mode ω-2θ. Pertinent crystallographic data and refinement
parameters are given below. Structure refinement (C atoms aniso-
tropic, H atoms riding) was accomplished with the teXsan program
suite (version 1.7 for SGI workstations). The X-ray structures have
been deposited with the Cambridge Crystallographic Data Centre
as supplementary publication nos. CCDC-219153 (13) and CCDC-
219154 (14). Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge CB21EZ, UK
(fax: (+44) 1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
Thermolysis of Benzvalene 17c. Complex 17c (20 mg, 0.03
mmol) was dissolved in benzene-d6 (0.5 mL) and heated at 75 °C
1
for 24 h. H NMR spectroscopy showed no change.
Thermolysis of Benzvalene 17d. Complex 17d (20 mg, 0.024
mmol) was dissolved in benzene-d6 (0.5 mL) and heated at 75 °C
1
for 24 h. H NMR spectroscopy showed the solution to be a 10:
4:3 mixture of 17d, 18d, and 19d, along with partial decomposition.
Partial characterization data for 18d: 1H NMR (C6D6) δ 11.18 (q,
J ) 10.5 Hz, H5), 8.91 (q, J ) 6.6 Hz, H3), 8.05 (t, J ) 8.8 Hz,
H4), 1.53 (d, J ) 13.2 Hz), 0.34 (s).
Crystal structure of 13: C52H47IIr2O2P2Si‚2CH2Cl2, Mr
)
The solution was heated further at 75 °C until NMR showed no
remaining 17d or 18d (total time: 90 h). The solvent was removed
and the residue was purified by chromatography on silica (6:1
1475.2, red block, 0.10 × 0.18 × 0.20 mm, monoclinic, space group
P21/a, a ) 18.558(3) Å, b ) 15.874(3) Å, c ) 20.342(2) Å, â )
112.65(2)°, V ) 5530(2) Å3, Z ) 4, Fcalc ) 1.772 g cm-3, µ )
56.9 cm-1, F(000) ) 2840, 2θmax ) 44°, 6327 independent
reflections scanned, 4573 reflections in refinement [I g σ(I)], 541
parameters, R ) 0.046, Rw ) 0.049.
1
hexane/Et2O) to give 19d (8.4 mg, 55%) as yellow crystals. H
NMR (C6D6): δ 7.72-7.67 (m, 2H), 7.65-7.53 (m, 4H), 7.13-
7.00 (m, 9H), 5.41-5.37 (m, 1H), 4.79-4.75 (m, 1H), 4.62-4.58
(m, 1H), 2.03 (d, J ) 9.7 Hz, 3H), 0.29 (s, 9H, Si(CH3)3). 13C
NMR (C6D6): δ 179.11 (d, J ) 18.2 Hz, CO), 139.57 (m), 136.94
(s), 132.79 (s), 132.63 (s), 132.47 (s), 132.31 (s), 132.03 (s), 129.80
(m), 128.11 (s), 127.27 (s), 111.66 (s), 91.88 (s), 89.62 (s), 89.32
(s), 84.01(s), 21.52 (d, J ) 42.3 Hz), 1.38 (s). 31P NMR (C6D6):
δ -4.85 (s). IR (Et2O): ν 1937 (s, CO) cm-1. Anal. Calcd for
C28H30IrOPSi: C, 53.06; H, 4.77. Found: C, 53.22; H, 4.89.
Thermolysis of Benzvalene 17e. Complex 17e (20 mg, 0.024
mmol) was dissolved in benzene-d6 (0.5 mL) and heated at 75 °C
for 24 h. The solvent was removed, and the residue was purified
by chromatography on silica (6:1 hexane/Et2O) to give 19e (7.0
Crystal structure of 14. C48H51IrOP2Si2, Mr ) 954.3, pale
yellow prisms, 0.12 × 0.14 × 0.27 mm, triclinic, space group P1h,
a ) 11.958(3) Å, b ) 13.111(2) Å, c ) 14.535(4) Å, R ) 95.28-
(3)°, â ) 112.65(2)°, γ ) 95.25(2)°, V ) 2199.5(10) Å3, Z ) 2,
Fcalc ) 1.441 g cm-3, µ ) 32.05 cm-1, F(000) ) 964, 2θmax
)
50°, 7812 independent reflections scanned, 7024 reflections in
refinement [I g σ(I)], 488 parameters, R ) 0.033, Rw ) 0.038.
Acknowledgment. We thank the National Science Founda-
tion (CHE-0075246 and -0647252) and the Schweizerischer
Nationalfonds for support of this research. D.H.E. acknowledges
the NSF IGERT program for a fellowship (DGE-0114443).
1
mg, 46%) as yellow crystals. H NMR (C6D6): δ 7.69-7.74 (m,
2H), 7.02-7.21 (m, 3H), 5.46 (br s, 1H), 5.11 (br s, 1H), 4.83 (br
s, 1H), 2.05-2.25 (m, 6H), 1.62-1.68 (m, 3H), 0.98-1.10 (m,
18H), 0.32 (s, 9H, Si(CH3)3). 13C NMR (C6D6): δ 179.43 (d, J )
17.1 Hz, CO), 137.19 (s), 131.97 (s), 128.05 (s), 127.10 (s), 111.51
(d, J ) 6.2 Hz), 91.52 (s), 88.27 (s), 86.47 (s), 82.37 (s), 40.48 (d,
J ) 33.2 Hz), 26.20 (s), 25.16 (s), 1.29 (s). 31P NMR (C6D6): δ
-3.61 (s). IR (Et2O): ν 1937 (s, CO) cm-1. Anal. Calcd for C27H44-
IrOPSi: C, 51.00; H, 6.97. Found: C, 50.82; H, 6.78.
Supporting Information Available: X-ray structure cif files
for compounds 13 and 14; Cartesian coordinates and absolute
electronic energies for structures 21-25 and 27. This material is
available free of charge via the Internet at http://pubs.acs.org.
OM700336D