Acetonitrile accelerated stereoselective C-H bond activation of the cationic Ir(I) complex having the PN/CH3 heterochelate hybrid ligand

Article abstract of DOI:10.1246/cl.2001.300

The cationic iridium(I) complex having a heterochelate lig-and, [Ir(cod)(PN/CH₃)]PF₆ {PN/CH₃ = o-Ph₂PC₆H₄CH(CH₃)-OCH₂(C₅H₄N-2)} (3) was prepared and characterized. Complex 3 underwent stereoselective facile activation of the benzylic C-H bond in CH₃CN at 50 °C to produce the oxidative addition product, a hydrido alkyl complex fac-4 after displacement of COD by 2 molecules of CH₃CN. The initially generated fac-4 isomerized gradually to give the geometrical isomer selectively mer-4 in CH₃CN.

Full text of DOI:10.1246/cl.2001.300

300
Chemistry Letters 2001
Acetonitrile Accelerated Stereoselective C–H Bond Activation of the Cationic Ir(I) Complex
Having the PN/CH3 Heterochelate Hybrid Ligand
Yasutaka Kataoka,* Kazuko Shizuma, Tsuneaki Yamagata, and Kazuhide Tani*
Department of Chemistry, Graduate School of Engineering Science, Osaka University, Toyonaka, Osaka 560-8531
(Received January 15, 2001; CL-010034)
The cationic iridium(I) complex having a heterochelate lig-
but after 69 h at 60 °C in CDCl3 complex 3 decomposed com-
and, [Ir(cod)(PN/CH3)]PF6 {PN/CH3 = o-Ph2PC6H4CH(CH3)-
OCH2(C5H4N-2)} (3) was prepared and characterized. Complex
3 underwent stereoselective facile activation of the benzylic C–H
bond in CH3CN at 50 °C to produce the oxidative addition prod-
uct, a hydrido alkyl complex fac-4 after displacement of COD by
2 molecules of CH3CN. The initially generated fac-4 isomerized
gradually to give the geometrical isomer selectively mer-4 in
CH3CN.
pletely to give a complex mixture.
Efficient and capable ligands for metal-mediated or -cat-
alyzed stereoselective organic syntheses are considered to be able
to coordinate strongly to a central metal supplying a rigid reac-
tion environment.1 If the ligand has too much flexibility in the
coordination to the metal, the rigid environment collapsed and
satisfactory stereoselectivity in the reaction could not be attained.
Intentional introduction of flexibility to the metal complexes by
using a hemilabile ligand, however, can lead to disclose new
remarkable features.2 We have recently paid much attention to
flexibility of a ligand which could stabilize different types of
coordination modes for its organometallic complexes. For exam-
ple, we designed the hemilabile PN hybrid ligand (1) which was
able to act as a P-N bidentate, a P-O-N tridentate or a P-C-N tri-
dentate ligand3 and showed that the flexibility of the PN ligand
was an important factor for isolation of both Ir(I) and Ir(III) com-
plexes before and after C–H bond activation.4 Herein, we pre-
pare another hemilabile ligand, a methyl-substituted PN ligand
(PN/CH3 ligand) (2),5 and show interesting behavior, acetonitrile
accelerated stereoselective C–H bond activation of its iridium
complex.
The 1H NMR of complex 3 in CD3CN at 30 °C showed very
broad signals different from that in CD2Cl2, though the spectrum
at –20 °C showed sharp signals assignable to the two diastere-
omers of complex 3 (67 : 33), indicating that rapid interconversion
between the two diastereomers was occurring without dissociation
of COD at 30 °C (Scheme 1).8 At 50 °C, the color of the CD3CN
solution of complex 3 gradually changed from orange to pale yel-
Reaction of [IrCl(cod)]2 with two equiv of ligand 2 in the
presence of excess AgPF6 in ethanol gave the cationic complex
[Ir(cod)(PN/CH3)]PF6 (3) as orange powders in 86% yield.6 The
31P NMR spectrum in CD2Cl2 showed that complex 3 was com-
prised of two diastereomers (major : minor = 67 : 33). Air-stable
orange crystals of one diastereomer suitable for an X-ray analysis
were obtained directly from the reaction mixture. An ORTEP
drawing of complex 3a having a S*c,S*pl configuration is shown
in Figure 1.7 Different from [Ir(cod)(PNn=1)]PF6, previously
reported by us,4 no obvious interaction between the iridium and
the oxygen atom [Ir···O = 4.152 Å] exists in a solid state. The
crystals of complex 3a used for the X-ray analysis gave again the
same equilibrium mixture of the two diastereomers in CD2Cl2
(major : minor = 67 : 33). Complex 3 was stable in CDCl3 or
CD2Cl2 at room temperature or even at 50 °C for 18 h in CDCl3,
1
low. The reaction was monitored by H NMR as well as 31P
NMR in CD3CN. Sharp signals in the 1H NMR assignable to the
C–H bond activation product 4 and free COD appeared accompa-
nying several signals for small amounts of unidentified products.
After 1 day at 50 °C, complex 3 disappeared completely and the
Ir(III) hydrido alkyl complex 4 was produced as a mixture of two
geometrical isomers (major-4 : minor-4 = 94 : 6). The structure of
the major isomer was fully characterized to be fac-4 by elemental
analysis as well as spectral data and was confirmed by an X-ray
analysis (Figure 2).9 The PN/CH3 ligand acts as a P-C-N triden-
tate ligand and coordinates to the Ir center with a facial manner.
The 1H NMR signal of the proton at the 6-position of the pyridine
ring in the PCN/CH3 ligand10 of fac-4 showed a characteristic
shape for that of the cis-coordinated PN ligand,3 which is con-
Copyright © 2001 The Chemical Society of Japan

Chemistry Letters 2001
301
48, 233 (1999).
sistent with the facial coordination of the PCN/CH3 ligand. The
1H and 31P NMR indicated that fac-4 was gradually converted fur-
ther to minor-4, which can be tentatively assigned to a geometrical
isomer mer-4.11 After 17 days at 50 °C mer-4 became the major
isomer (fac-4 : mer-4 = 10 : 90). From the stereochemical consid-
eration, fac-4 and mer-4 were exclusively derived from the isomer
3a. We could not obtain the concrete evidence that the C–H bond
activation of the diastereomer 3b proceeded under these condi-
tions.
3
a) M. Yabuta, S. Nakamura, T. Yamagata, and K. Tani, Chem. Lett.,
1993, 323. b) K. Tani, M. Yabuta, S. Nakamura, and T. Yamagata, J.
Chem. Soc., Dalton Trans., 1993, 2781. c) K. Tani, S. Nakamura, T.
Yamagata, and Y. Kataoka, Inorg. Chem., 32, 5398 (1993). d) Y.
Kataoka, Y. Tsuji, O. Matsumoto, T. Ohashi, T. Yamagata, and K. Tani,
J. Chem. Soc., Chem. Commun., 1995, 2099.
4
5
Y. Kataoka, Y. Imanishi, T. Yamagata, and K. Tani, Organometallics,
18, 3563 (1999).
The PN/CH3 ligand could be prepared by a similar method to that of the
PNn=1 ligand from 1-(o-diphenylphosphinophenyl)ethanol and 2-picolyl
chloride. 2 (racemic compound): 1H NMR (CDCl3, 30 °C) δ 1.37 (d, J
= 6.3 Hz, 3H), 4.35 (d, J = 13.2 Hz, 1H), 4.43 (d, J = 13.2 Hz, 1H),
5.36–5.50 (m, 1H), 6.86–6.98 (m, 1H), 7.10–7.46 (m, 14H), 7.58–7.72
(m, 2H), 8.46–8.56 (m, 1H). 31P{1H}NMR (CDCl3, 30 °C) δ –17.0 (s).
Anal. (C26H24ONP) C, H, N.
6
3: Mp 183.0 °C (dec.). Both 3-major and 3-minor showed very similar
sets of 1H NMR which can well explain structures 3a or 3b, respective-
ly. Some representative NMR data: 1H NMR (CD2Cl2, 30 °C) 3-major
δ 1.56 (d, J = 7.0 Hz, 3H, –CHMeO–), 5.02–5.20 (m, 1H, –CHMe–),
5.10 (d, J = 14.1 Hz, 1H, –OCH2–py), 5.53 (d, J = 14.1 Hz, 1H,
–OCH2–py); 3-minor 2.38 (d, J = 6.6 Hz, 3H, –CHMeO–), 5.02–5.20
(m, 1H, –CHMe–), 4.68 (d, J = 15.5 Hz, 1H, –OCH2–py), 5.26 (d, J =
15.5 Hz, 1H, –OCH2–py). 31P{1H}NMR (CD2Cl2, 30 °C) δ 5.37 (s,
major), 7.46 (s, minor). Anal. Found: C, 48.12; H, 4.35; N, 1.75%.
Calcd for C34H36F6IrNOP2: C, 48.45; H, 4.31; N, 1.66%.
7
Crystal data for 3a: C34H36F6IrNOP2, fw = 842.78, Rigaku AFC7R, tri-
–
clinic, P1 (No. 2), a = 11.259(8), b = 14.618(9), c = 9.926(10) Å, α =
95.39(7), β = 89.97(9), γ = 95.60(5) ˚, V = 1619(2) Å3, Z = 2, T = 296(2)
K, Dcalcd = 1.729 Mg m–3, λ = 0.71069 Å, F(000) = 832, µ = 4.287
mm–1, Tmin = 0.2789, Tmax = 0.6057, 2θ(max) = 60 deg, 19598 measured
reflections, 9417 unique reflections, Rint = 0.0302, solution method
SIR97, refinement method SHELXL97, R(all data) = 0.0418, wR(all) =
0.0789, R(I > 2σ(I)) = 0.0279, wR(I > 2σ(I)) = 0.0737, GOF = 1.052,
∆/σ(max) = 0.001.
8
9
The 1H NMR of complex 3 in the presence of additional free COD (ca.
10 equiv) in CD3CN at 30 °C afforded the same broad signals besides
the signals of the added free COD.
fac-4: mp 78.0 °C (dec.). 1H NMR (CD3CN, 30 °C) δ –20.72 (d, J =
22.7 Hz, 1H), 1.78–2.14 (br, 6H), 2.09 (s, 3H), 3.64 (d, J = 15.8 Hz,
1H), 4.09 (d, J = 15.8 Hz, 1H), 6.78–7.98 (m, 17H), 8.64 (d, J = 5.5 Hz,
1H). 31P{1H}NMR (CD3CN, 30 °C) δ 27.4 (s), –143.2 (septet, J = 706
Hz, PF6). MS (FAB) m/z 590 (M+–PF6–2CH3CN). IR (nujol) 2280,
2178, 1606, 1111, 850, 838, 770, 750, 724 and 697 cm-1. Anal. Found:
C, 43.76; H, 3.44; N, 5.06%. Calcd for C30H30F6IrN3OP2: C, 44.12; H,
1
The H NMR of complex 3 in the presence of CH3CN (ca.
3.70; N, 5.14%. Crystal data: C30H30F6IrN3OP2, fw = 842.78, Rigaku
–
1.7 equiv) in CD2Cl2 indicated a new signal at δ 1.99 due to the Ir-
coordinated CH3CN of an intermediate complex in addition to the
signals of complex 3 and free CH3CN (δ 1.98), but did not show
any signals for free-COD. After 5 days at room temperature (ca.
20 °C), the C–H bond activation products 4 appeared (major-3 :
minor-3 : fac-4 : mer-4 = 48 : 26 : 20 : 6). The C–H bond activa-
tion, however, did not proceed at all in the absence of CH3CN
under the same conditions. Complex 3 was stable in a mixed sol-
vent of CD2Cl2 and CD3OD and deuterium incorporation into
complex 3 could not be detected; a plausible reversible C–H bond
activation did not occur in these solvent systems. The hypothesis
that dissociation of COD from complex 3 is the driving force for
this C–H bond activation is inconsistent with the fact that the rate
of the C–H bond activation in CD3CN was not affected by the
presence of excess COD. Coordination of CH3CN renders the
metal center more electron rich, which causes facile C–H bond
activation in only complex 3a.12
RAXIS-RAPID, triclinic, P1 (No. 2), a = 11.7843(4), b = 14.5811(6), c
= 9.4638(4) Å, α = 105.4882(14), β = 96.9441(15), γ = 97.2193(14) ˚, V
= 1534.25(10) Å3, Z = 2, T = 100(1) K, Dcalcd = 1.768 Mg m–3, λ =
0.71069 Å, F(000) = 800, µ = 4.521 mm–1, Tmin = 0.4592, Tmax
=
0.6958, 2θ(max) = 63 deg, 29369 measured reflections, 10149 unique
reflections, Rint = 0.0745, solution method SIR97, refinement method
SHELXL97, R(all data) = 0.0793, wR(all) = 0.0952, R(I > 2σ(I)) =
0.0488, wR(I > 2σ(I)) = 0.0871, GOF = 1.020, ∆/σ(max) = 0.000.
10 The PCN/CH3 ligand represents the PN/CH3 ligand acted as a P-C-N tri-
dentate ligand.
11 Although isolation of pure mer-4 has not been successful yet due to its
instability, the structure of mer-4 is determined by both the characteristic
shape of the 1H NMR signal for the proton at the 6-position of the pyri-
dine ring3 and the NOESY experiments. The proton at the 6-position of
the pyridine ring in mer-4 showed NOE only to the proton at the 5-posi-
tion of the pyridine ring, while that in fac-4 showed NOE to one of the
aromatic protons except the pyridine protons, in addition to the proton at
the 5-position of the pyridine ring. The relative positions of the hydrido
ligand and the two CH3CN ligands are also determined by NOESY
experiments. The Ir-H in mer-4 showed NOE to one of the α-methylene
protons of the pyridine ring, but not to the methyl group of the PN/CH3
ligand, while the Ir–H in fac-4 showed NOE to the methyl group, but
not to the α-methylene protons. 1H NMR of mer-4 (CD3CN, 30 °C) δ
–21.47 (d, J = 19.2 Hz, 1H), 1.10 (s, 3H), 1.78–2.14 (br, 6H), 4.66 (d, J
= 20.6 Hz, 1H), 4.71 (d, J = 20.8 Hz, 1H), 6.92–8.12 (m, 17H),
9.08–9.18 (m, 1H). 31P{1H}NMR (CD3CN, 30 °C) δ 21.4 (s), –143.2
(septet, J = 706 Hz, PF6).
This work was supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Science, Sports, and
Culture, Japan.
12 The accelerating effect by the coordination of CH3CN to the metal cen-
ter was also observed in the reaction of [Ir(cod)(PN)]PF6. For example,
C–H activation of [Ir(cod)(PN)]PF6 in CH3CN was almost completed at
50 °C for 3 h, while the C–H activation in CDCl3 proceeded only ca.
20% under the same conditions.
References and Notes
1
J. Seyden-Penne, “Chiral Auxiliaries and Ligands in Asymmetric
Synthesis,” John Wiley & Sons, New York (1995).
C. S. Slone, D. A. Weinberger, and C. A. Mirkin, Prog. Inorg. Chem.,
2