Influence of ligand polarizability on the reversible binding of O 2 by trans-[Rh(X)(XNC)(PPh3)2] (X = Cl, Br, SC6F5, C2Ph; XNC = Xylyl isocyanide). Structures and a kinetic study

Article abstract of DOI:10.1021/ic800478g

The complexes trans-[Rh(X)(XNC)(PPh₃)₂] (X = Cl, 1; Br, 2; SC₆F₅, 3; C₂Ph, 4; XNC = xylyl isocyanide) combine reversibly with molecular oxygen to give [Rh(X)(O ₂)(XNC)(PPh₃)₂] of which [Rh(SC ₆F₅)(O₂)(XNC)(PPh₃)₂] (7) and [Rh(C₂Ph)(O₂)(XNC)(PPh₃)₂] (8) are sufficiently stable to be isolated in crystalline form. Complexes 2, 3, 4, and 7 have been structurally characterized. Kinetic data for the dissociation of O₂ from the dioxygen adducts of 1-4 were obtained using ³¹P NMR to monitor changes in the concentration of [Rh(X)(O ₂)(XNC)(PPh₃)₂] (X = Cl, Br, SC ₆F₅, C₂Ph) resulting from the bubbling of argon through the respective warmed solutions (solvent chlorobenzene). From data recorded at temperatures in the range 30-70°C, activation parameters were obtained as follows: ΔH^? (kJ mol^-1): 31.7 ± 1.6 (X = Cl), 52.1 ± 4.3 (X = Br), 66.0 ± 5.8 (X = SC₆F₅), 101.3 ± 1.8 (X = C₂Ph); ΔS^? (J K^-1 mol^-1): -170.3 ± 5.0 (X = Cl), -120 ± 13.6 (X = Br), -89 ± 18.2 (X = SC ₆F₅), -6.4 ± 5.4 (X = C₂Ph). The values of ΔH^? and ΔS^? are closely correlated (R² = 0.9997), consistent with a common dissociation pathway along which the rate-determining step occurs at a different position for each X. Relative magnitudes of ΔH^? are interpreted in terms of differing polarizabilities of ligands X.

Full text of DOI:10.1021/ic800478g

Inorg. Chem. 2008, 47, 8696-8703
Influence of Ligand Polarizability on the Reversible Binding of O2 by
trans-[Rh(X)(XNC)(PPh3)2] (X ) Cl, Br, SC6F5, C2Ph; XNC ) Xylyl
Isocyanide). Structures and a Kinetic Study
Laurence Carlton,* Lebohang V. Mokoena, and Manuel A. Fernandes
Molecular Sciences Institute, School of Chemistry, UniVersity of the Witwatersrand, Johannesburg,
Republic of South Africa
Received March 17, 2008
The complexes trans-[Rh(X)(XNC)(PPh3)2] (X ) Cl, 1; Br, 2; SC6F5, 3; C2Ph, 4; XNC ) xylyl isocyanide) combine
reversibly with molecular oxygen to give [Rh(X)(O2)(XNC)(PPh3)2] of which [Rh(SC6F5)(O2)(XNC)(PPh3)2] (7) and
[Rh(C2Ph)(O2)(XNC)(PPh3)2] (8) are sufficiently stable to be isolated in crystalline form. Complexes 2, 3, 4, and 7
have been structurally characterized. Kinetic data for the dissociation of O2 from the dioxygen adducts of 1-4
were obtained using 31P NMR to monitor changes in the concentration of [Rh(X)(O2)(XNC)(PPh3)2] (X ) Cl, Br,
SC6F5, C2Ph) resulting from the bubbling of argon through the respective warmed solutions (solvent chlorobenzene).
From data recorded at temperatures in the range 30-70 °C, activation parameters were obtained as follows: ∆Hq
(kJ mol-1): 31.7 ( 1.6 (X ) Cl), 52.1 ( 4.3 (X ) Br), 66.0 ( 5.8 (X ) SC6F5), 101.3 ( 1.8 (X ) C2Ph); ∆Sq
(J K-1 mol-1): -170.3 ( 5.0 (X ) Cl), -120 ( 13.6 (X ) Br), -89 ( 18.2 (X ) SC6F5), -6.4 ( 5.4 (X )
C2Ph). The values of ∆Hq and ∆Sq are closely correlated (R2 ) 0.9997), consistent with a common dissociation
pathway along which the rate-determining step occurs at a different position for each X. Relative magnitudes of
∆Hq are interpreted in terms of differing polarizabilities of ligands X.
Introduction
binding ability is shown by the complexes trans-[Ir(Cl)(CO)(P-
Ph3)2]2 and trans-[Rh(Cl)(CO)(PPh3)2]:3 the former combines
reversibly with O2, while the latter does not react. The
influence of ligand polarizability on O2 binding is shown by
the complexes [Rh(Cl)(PPh3)3] and [Rh(NCBPh3)(PPh3)3]:
the former binds O2 (irreversibly)4 while the latter does not.5
Our interest in transition metal-O2 interactions arises from
work directed toward a fuller understanding of ligand
polarizability effects. In the present study we examine the
role of the anionic ligand in promoting oxygen binding by
the complexes trans-[Rh(X)(XNC)(PPh3)2] (X ) Cl, Br,
SC6F5, C2Ph; XNC ) xylyl isocyanide).
The ability of a transition metal to combine with molecular
oxygen is related to the presence of a vacant coordination
site, the accessibility of a higher oxidation state, the electron
density on the metal and the polarizability of the metal,
governed in part by the polarizability of attached ligands.
The relative ease with which suitable conditions can be
achieved is reflected in the wide range of known O2
complexes.1 The influence of metal polarizability on oxygen
*
To whom correspondence should be addressed. E-mail:
laurence.carlton@wits.ac.za.
(1) (a) Choy, V. J.; O’Connor, C. J. Coord. Chem. ReV. 1972, 9, 145. (b)
Valentine, J. S. Chem. ReV. 1973, 73, 235. (c) Henrici-Olive´, G.; Olive´,
S. Angew. Chem., Int. Ed. Engl. 1974, 13, 29. (d) Vaska, L. Acc. Chem.
Res. 1976, 9, 175. (e) Collman, J. P. Acc. Chem. Res. 1977, 10, 265.
(f) Jones, R. D.; Summerville, D. A.; Basolo, F. Chem. ReV. 1979,
79, 139. (g) Gubelman, M. H.; Williams, A. F. Struct. Bonding (Berlin)
1983, 55, 1. (h) Niederhofer, E. C.; Timmons, J. H.; Martell, A. E.
Chem. ReV. 1984, 84, 137. (i) Karlin, K. D.; Gultneh, Y. Prog. Inorg.
Chem. 1987, 35, 219. (j) Dickman, M. H.; Pope, M. T. Chem. ReV.
1994, 94, 569. (k) Busch, D. H.; Alcock, N. W. Chem. ReV. 1994, 94,
585. (l) Butler, A.; Clague, M. J.; Meister, G. E. Chem. ReV. 1994,
94, 625. (m) Momenteau, M.; Reed, C. A. Chem. ReV. 1994, 94, 659.
(n) Kitajima, N.; Moro-oka, Y. Chem. ReV. 1994, 94, 737. (o) Pecoraro,
V. L.; Baldwin, M. J.; Gelasco, A. Chem. ReV. 1994, 94, 807.
Experimental Section
Materials. [Rh(Cl)(PPh3)3] and [Rh(Br)(PPh3)3] were prepared
by the method of Wilkinson,6 and [Rh(H)(PPh3)4] was prepared
by the method of Robinson.7 Xylyl isocyanide (2,6-dimethylphenyl
isocyanide) was purchased from Fluka, and pentafluorophenyl thiol
(2) Vaska, L. Acc. Chem. Res. 1968, 1, 335.
(3) Vallarino, L. J. Chem. Soc. 1957, 2287.
(4) Bennett, M. J.; Donaldson, P. B. Inorg. Chem. 1977, 16, 1581.
(5) Carlton, L.; Weber, R. Inorg. Chem. 1996, 35, 5843.
8696 Inorganic Chemistry, Vol. 47, No. 19, 2008
10.1021/ic800478g CCC: $40.75  2008 American Chemical Society
Published on Web 09/06/2008

ReWersible Binding of O2 by trans-[Rh(X)(XNC)(PPh3)2]
and phenylacetylene were purchased from Aldrich Chemicals.
Dichloromethane was distilled from P2O5. Diethyl ether, 1,2-
dichloroethane, hexane, benzene, chlorobenzene, and toluene were
dried over calcium hydride. All operations except those involving
molecular oxygen were performed under an argon atmosphere.
Preparation of trans-[Rh(Cl)(XNC)(PPh3)2]·(0.5CH2Cl2) (1). A
mixture of [Rh(Cl)(PPh3)3] (0.050 g, 0.054 mmol) and xylyl
isocyanide (0.007 g, 0.054 mmol) was dissolved in dichloromethane
(3 mL) at room temperature. After a few minutes, during which
time a color change to orange-red was observed, the solution was
concentrated, treated with hexane, and allowed to stand for 20 h at
room temperature to give the product as a yellow microcrystalline
8.18-8.12 (mult, 12H, Ar), 7.06-6.88 (mults, 20H, Ar), 6.84-6.77
(mult, 3H, Ar), 6.62 (t (J ) 7.3), 1H, XNC Ar-4), 6.51 (d (J )
7.3), 2H, XNC Ar-3,5),1.60 (s, 6H, XNC Me); 31P NMR (toluene
300 K) δ 35.82 (d J(103Rh, 31P) 144 Hz); 103Rh NMR (toluene 300
K) δ -687 ppm. Anal. Calcd for C59H50NP2Rh: C, 75.56; H, 5.37;
N.1.49. Found: C, 75.95; H, 5.86; N, 1.67.
Reaction of 1 with O2. On shaking a solution of 1 in dichlo-
romethane with O2 [Rh(Cl)(O2)(XNC)(PPh3)2] (5) was obtained
(72% conversion measured by 31P NMR). IR (dichloromethane)
ν(O2) 894 (m, br) cm-1; 1H NMR (chloroform 300 K) δ 7.69-7.63
(mult, 12H, Ar), 7.36-7.22 (mults, 18H, Ar), 7.02 (t (J ) 7.5),
1H, XNC Ar-4), 6.89 (d (J ) 7.5), 2H, XNC Ar-3, 5), 1.99 (s, 6H
XNC Me); 31P NMR (dichloromethane 300 K): δ 28.14 (d J(103Rh,
31P) 91 Hz); 103Rh NMR (dichloromethane 300 K) δ 2130 ppm.
Attempts to isolate 5 yielded only 1.
1
powder. Yield 0.039 g (84%). H NMR (chloroform 300 K) δ
7.81-7.73 (mult, 12H, Ar), 7.26-7.20 (mult, Ar, 18 H), 6.79 (t (J
) 7.8), 1H, XNC Ar-4), 6.67 (d (J ) 7.8), 2H, XNC Ar-3,5), 1.61
(s, 6H, XNC Me); 31P NMR (dichloromethane 300 K) δ 30.80 (d
J(103Rh, 31P) 135 Hz); 103Rh NMR (dichloromethane 300 K) δ -261
ppm. Anal. Calcd for 1: C, 63.57; H, 4.69; N, 1.63. Found: C, 64.06;
H, 4.53; N, 1.69.
Reaction of 2 with O2. On shaking a solution of 2 in toluene
with O2 [Rh(Br)(O2)(XNC)(PPh3)2] (6) was obtained (100%
conversion measured by 31P NMR). IR (dichloromethane) ν(O2)
1
894 (m, br) cm-1; H NMR (benzene 300 K) δ 8.08-8.00 (mult,
Preparation of trans-[Rh(Br)(XNC)(PPh3)2]·(0.5CH2Cl2) (2). A
mixture of [Rh(Br)(PPh3)3] (0.100 g, 0.103 mmol) and xylyl
isocyanide (0.014 g, 0.108 mmol) was dissolved in dichloromethane
(4 mL) at room temperature. After a few minutes, during which
time a color change to dark red was observed, the solution was
centrifuged to remove a small quantity of black powder, concen-
trated, treated with hexane, and allowed to stand for 20 h at room
temperature to give the product as a yellow crystalline powder.
Yield 0.059 g (62%). 1H NMR (benzene 300 K) δ 8.14-8.02 (mult,
12H, Ar), 7.05-6.86 (mults, 18H Ar), 6.60 (t (J ) 7.3), 1H, XNC
Ar-4), 6.48 (d (J ) 7.3), 2H, XNC Ar-3,5), 1.62 (s, 6H, XNC Me);
31P NMR (toluene 300 K) δ 31.40 (d J(103Rh, 31P) 134 Hz); 103Rh
12H, Ar), 7.02-6.93 (mults, 18H, Ar), 6.61 (t (J ) 7.5), 1H, XNC
Ar-4), 6.49 (d (J ) 7.5), 2H, XNC Ar-3,5), 1.96 (s, 6H XNC Me);
31P NMR (toluene 300 K): δ 28.30 (d J(103Rh, 31P) 92 Hz); 103 Rh
NMR (toluene 300 K) δ 1831 ppm. Attempts to isolate 6 yielded
only 2.
Preparation of [Rh(SC6F5)(O2)(XNC)(PPh3)2]·(tol). (7). A
solution of 3 (0.009 g, 9 mmol) in toluene (1 mL) was shaken with
O2 and allowed to stand for 5 d at -20 °C to give the product as
dark brown crystals. Yield 0.007 g (71%). IR (dichloromethane)
1
ν(O2) 884 (m, br); H NMR (benzene 300 K) δ 7.98-7.88 (mult,
12H, Ar), 7.05-6.95 (mults, 18H, Ar), 6.60 (t (J ) 7.7), 1H, XNC
Ar-4), 6.48 (d (J ) 7.7), 2H, XNC Ar-3,5), 2.02 (s, 6H, XNC Me);
31P NMR (toluene 300 K) δ 25.42 (d J(103Rh, 31P) 94 Hz); 103Rh
NMR (toluene 300 K)
δ -325 ppm. Anal. Calcd for
C
45.5H40ClBrNP2Rh: C, 62.03; H, 4.58; N, 1.59. Found: C, 62.33;
NMR (toluene 300 K)
δ 1601 ppm. Anal. Calcd for
H, 5.00; N, 1.84. In addition to 2, a small quantity of red-brown
crystals of [Rh(Br)3(XNC)(PPh3)2](CH2Cl2) (2b) was also obtained
from solutions allowed to stand for more than 1 d.
C58H47F5NO2P2RhS: C, 64.39; H, 4.38; N, 1.29. Found: C, 64.96;
H, 4.48; N, 1.27.
Preparation of [Rh(C2Ph)(O2)(XNC)(PPh3)2] (8). A solution
of 4 (0.010 g, 11 mmol) in benzene (1 mL) was shaken with O2
and allowed to stand at room temperature for 3 d to give the product
as a pink microcrystalline powder. Yield 0.006 g (63%). IR
(dichloromethane) ν(O2) 858 (m, br); 1H NMR (benzene 300 K) δ
8.16-8.07 (mult, 12H, Ar), 7.10-6.90 (mults, 23H, Ar), 6.64 (t (J
) 7.6), 1H, XNC Ar-4), 6.51 (d (J ) 7.6), 2H, XNC Ar-3,5), 1.87
(s, 6H, XNC Me); 13C NMR (dichloromethane 300 K) acetylide δ
112.28 (dt {2J(Rh,C) 8.2, J(P,C) 2.4} Rh-Cꢀ), 100.65 (dt {1J(Rh,C)
47.3, J(P,C) 15.9} RhCR); 31P NMR (toluene 300 K) δ 31.11 (d
J(103Rh, 31P) 92 Hz). 103Rh NMR (toluene 300 K) δ 1010 ppm.
Anal. Calcd for C53H44NO2P2Rh: C, 71.38; H, 4.97; N, 1.57. Found:
C, 71.09; H, 5.07; N, 1.50.
NMR Spectroscopy. Spectra were recorded on a Bruker DRX
400 spectrometer equipped with a 5 mm triple resonance inverse
probe with a dedicated 31P channel and extended decoupler range
operating at 161.98 MHz (31P) and 12.65 MHz (103Rh). Two-
dimensional 103Rh-31P spectra were obtained using the pulse
sequence π/2(31P)-1/[2J(103Rh,31P)]-π/2(103Rh)-τ-π(31P)-τ-π/2(103Rh)-
1/[2J(103Rh,31P)]-Acq (31P).8 Further details are given in ref 9.
Chemical shifts were referenced to the generally accepted standards
of 85% H3PO4 and 3.16 MHz (103Rh).10 The chemical shift of
H3PO4 (with toluene-d8 external lock, temperature 300 K) corre-
Preparation of trans-[Rh(SC6F5)(XNC)(PPh3)2]·(0.5 Et2O)
(3). A stirred suspension of [Rh(H)(PPh3)4] (0.150 g, 0.130 mmol)
in Et2O (3 mL) at room temperature was treated first with C6F5SH
(0.030 g, 0.150 mmol) followed by, after 1-2 min when a clear
orange solution had been obtained, xylyl isocyanide (0.017 g, 0.130
mmol). After a further 5 min, stirring was discontinued and the
solution was allowed to stand for 20 h at room temperature to give
the product as yellow crystals. Yield 0.096 g (74%). 1H NMR
(benzene 300 K) δ 8.00-7.92 (mult, 12H, Ar), 7.02-6.88 (mult,
18H, Ar), 6.55 (t (J ) 7.5), 1H, XNC Ar-4), 6.43 (d (J ) 7.5), 2H,
XNC Ar-3,5), 1.60 (s, 6H, XNC Me); 31P NMR (toluene 300 K) δ
32.53 (dt J(103Rh, 31P) 139, J(31P, 19F) 3.1 Hz); 103Rh NMR (toluene
300 K) δ -366 ppm. Anal. Calcd for C53H44F5NO0.5P2RhS: C,
63.99; H, 4.46; N, 1.41. Found: C, 64.11; H, 4.20; N, 1.47.
Preparation of trans-[Rh(C2Ph)(XNC)(PPh3)2]·(C6H6) (4).
[Rh(H)(PPh3)4] (0.048 g, 0.042 mmol) and PhC2H (0.010 g 0.10
mmol) were allowed to combine (with stirring) at room temperature
in benzene (2 mL), distilled under vacuum from CaH2 into the
reaction flask. Within 2-3 min a dark red solution was obtained.
After stirring for ∼10 min, xylyl isocyanide (0.0055 g, 0.042 mmol)
was added and stirring was continued for a further 15 min. The
solution was then concentrated, treated with hexane, and allowed
to stand at room temperature for 2 d to give the product as yellow
crystals. Yield 0.022 g (56%). 1H NMR (benzene 300 K) δ
(8) Bax, A.; Griffey, R. H.; Hawkins, B. L. J. Magn. Reson. 1983, 55,
301.
(6) Osborn, J. A.; Wilkinson, G. Inorg. Synth. 1967, 10, 67.
(7) Ahmad, N.; Levison, J. J.; Robinson, S. D.; Uttley, M. F. Inorg. Synth.
1974, 15, 58.
(9) Carlton, L. Magn. Reson. Chem. 2004, 42, 760.
(10) Kidd, R. G.; Goodfellow, R. J. In NMR and the Periodic Table; Harris,
R. K.; Mann, B. E., Eds.; Academic Press: London, 1978; pp 244-249.
Inorganic Chemistry, Vol. 47, No. 19, 2008 8697

Carlton et al.
Table 1. Crystal Data and Structure Refinement for 2, 3, 4, and 7
[Rh(Br)(XNC)(PPh3)2]
[Rh(SC6F5)(XNC)(PPh3)2]
[Rh(C2Ph)(XNC)(PPh3)2]
[Rh(SC6F5)(O2)
(CH2Cl2) (2)
(0.5 Et2O) (3)
(C6H6) (4)
(XNC)-(PPh3)2](tol) (7)
empirical formula
formula weight
temperature (K)
wavelength (Å)
crystal system
C46H41BrCl2NP2Rh
923.46
173(2)
0.71073
monoclinic
C53H44F5NO0.5P2RhS
994.80
173(2)
C59H50NP2Rh
937.85
173(2)
0.71073
triclinic
j
P1
C58H47F5NO2P2RhS
1081.88
173 (2)
0.71073
monoclinic
0.71073
triclinic
j
space group
P21/m
P1
P21/c
unit cell dimensions
a ) 12.1403(10) Å
b ) 12.9180(10) Å
c ) 12.3376 (9) Å
R ) 90°
a ) 13.3062 (16) Å
b ) 13.5965 (16) Å
c ) 15.0540 (18) Å
R ) 69.599 (2)°
ꢀ ) 77.881 (2)°
γ ) 64.300(2)°
2294.6(5)
a ) 12.481(2) Å
b ) 13.973(2) Å
c ) 15.361 (3) Å
R ) 94.931 (3)°
ꢀ ) 100.906(3)°
γ ) 113.722(3)°
2368.8(7)
a ) 12.9718(16) Å
b ) 14.3791 (15) Å
c ) 26.873 (3) Å
R ) 90°
ꢀ ) 91.262 (4)°
γ ) 90°
ꢀ ) 91.497 (4)°
γ ) 90°
volume (Å3)
1934.4(3)
5010.8(10)
Z
2
2
2
4
density (calculated) (Mg/m3)
absorption coefficient (mm-1
crystal size (mm3)
1.585
1.731
1.440
0.547
1.315
0.468
1.434
0.509
)
0.26 × 0.24 × 0.03
2.28 to 28.27
-15 e h e 16
-17 e k e 17
-14 e l e 16
10584
0.38 × 0.26 × 0.09
1.45 to 27.00
-16 e h e 13
-17e k e 17
-18 e l e 19
14651
0.26 × 0.20 × 0.07
1.37 to 27.00
-15 e h e 15
-17e k e 15
-19 e l e 19
16013
0.17 × 0.16 × 0.06
1.52 to 27.00
-16 e h e 12
-18 e k e 18
-32 e l e 34
21943
θ range for data collection (deg)
index ranges
reflections collected
independent reflections
completeness to θmax (%)
max. and min. transmission
data/restraints/parameters
goodness-of-fit on F2
4986 [R(int) ) 0.0520]
9845 [R(int) ) 0.0240]
10284 [R(int) ) 0.0480]
10680 [R(int) ) 0.0768]
99.6
98.4
99.4
97.5
0.9481 and 0.6741
4986/3/261
1.083
0.9524 and 0.8192
9845/36/597
1.010
0.9721 and 0.8942
10284/1/490
1.022
0.9636 and 0.9196
10680/0/619
0.898
final R indices [I > 2σ (I)]
R1 ) 0.0557
wR2 ) 0.01579
R1 ) 0.0785
wR2 ) 0.1697
3.598 and -2.449
R1 ) 0.0310
wR2 ) 0.0721
R1 ) 0.0496
wR2 ) 0.0785
0.389 and -0.394
R1 ) 0.0534
wR2 ) 0.1368
R1 ) 0.0923
wR2 ) 0.1565
1.169 and -0.732
R1 ) 0.0463
wR2 ) 0.0804
R1 ) 0.1092
wR2 ) 0.0936
0.656 and -0.728
R indices (all data)
largest diff. peaks (e Å-3
)
sponds to a frequency of 161.975488 MHz in a field (9.395 T) in
which the protons of tetramethylsilane (TMS, in toluene at 300 K)
resonate at 400.130011 MHz.
sufficient to give a ∼0.02 M solution with 0.5 mL of solvent},
was placed in an NMR tube; the solvent, 0.5 mL chlorobenzene
containing 5 drops of toluene-d8, was added under argon and
warmed to dissolve the solid. The solution was then cooled, and
oxygen gas was admitted into the tube which was shaken (10-20
s) until complete conversion of the starting material to the dioxygen
complex (72% in the case X ) Cl) was obtained, as shown by the
31P spectrum. A sealed capillary (length ∼10 cm, external diameter
∼1.5 mm) containing Me3PO4 was then admitted into the tube for
the purpose of calibrating the 31P signal (peak height). A 31P
reference spectrum (time zero) was then recorded. The NMR tube
was then transferred to a bath maintained at a constant temperature
in the range 30-70 °C. The tube was immersed to a depth of ∼
5-6 cm, that is, to a level exceeding that of the contents.
Simultaneously, by means of a capillary, argon was bubbled through
the solution at a predetermined rate of 3.2 ( 0.1 mL min-1. After
a measured interval, which varied in duration from 10 s to 5 min
depending upon the complex and the temperature, the heating and
bubbling of argon were simultaneously discontinued and the
reaction was rapidly quenched in an ice-water mixture. During
this time (10-15 s) argon was allowed to flow into the NMR tube
which was then sealed prior to the recording of a 31P spectrum
(128 scans). This process, from transfer to the bath to the recording
of a spectrum, was repeated a number (in the range 5-12) of times.
Where necessary, allowance was made for small losses of solvent
by evaporation at higher temperatures.
X-ray Structure Determination. Intensity data were collected
at -100 °C on a Bruker SMART 1K CCD area detector diffrac-
tometer with graphite-monochromated Mo KR radiation (50 kV,
30 mA). The collection method involved ω-scans of width 0.3°.
Data reduction was carried out using the program SAINT+11 and
face-indexed absorption corrections were made using the program
XPREP.11 The crystal structures (Table 1) were solved by direct
methods using SHELXTL.12 Non-hydrogen atoms were first refined
isotropically followed by anisotropic refinement using full-matrix
least-squares calculation based on F2 using SHELXTL. Hydrogen
atoms were first located in the difference map and then positioned
geometrically and allowed to ride on their respective parent atoms.
Diagrams were generated using SHELXTL and PLATON.13
In the case of [Rh(C2Ph)(XNC)(PPh3)2].(C6H6) (4) both the
phenylacetylide ligand and the cocrystallized benzene are disor-
dered. Each was refined over two positions with site occupancies
set to 0.5; the refinement of the phenylacetylide ligand, however,
can not be regarded as optimal. Attempts to improve crystal quality
by using dichloromethane, toluene, or diethylether as the growth
medium were unsuccessful. In addition crystal shattering occurred
at temperatures below -100 °C, preventing the collection of
improved data.
Kinetic Study. Reactions were carried out in a 5 mm NMR tube
and changes in concentration monitored by 31P NMR. In each
experiment the starting material {trans-[Rh(X)(XNC)(PPh3)2],
The 31P signal heights were calibrated against the Me3PO4 signal
for each spectrum and converted into concentrations by comparison
with the initial spectrum (t ) 0) in the series. A first order rate
constant for the dissociation of O2 was obtained from a plot of
ln{concentration of [Rh(X)(O2)(XNC)(PPh3)2]} against time. No
significant difference in rate was found on changing the solvent to
(11) SAINT+, Version 6.02 (includes XPREP); Bruker AXS, Inc.: Madison,
WI, 1999.
(12) SHELXTL, Version 5.1; Bruker AXS, Inc.: Madison, WI, 1999.
(13) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7.
8698 Inorganic Chemistry, Vol. 47, No. 19, 2008

ReWersible Binding of O2 by trans-[Rh(X)(XNC)(PPh3)2]
Figure 1. Diagram of trans-[Rh(Br)(XNC)(PPh3)2] (2). Thermal ellipsoids
are shown at the 50% probability level. Hydrogens and a cocrystallized
CH2Cl2 are omitted.
1,2-dichloroethane or of a decrease (10%) in the rate of bubbling
of argon. Using the Eyring equation {ln(k1h/kBT) ) -(∆Hq/R)(1/
T) + ∆Sq/R; k1 ) observed rate, kB ) Boltzmann constant, h )
Planck’s constant, R ) gas constant} a plot of ln(k1h/kBT) versus
1/T gave the dissociation enthalpy and entropy from the gradient
and intercept, respectively.
Figure 2. Diagram of trans-[Rh(SC6F5)(XNC)(PPh3)2] (3). Thermal
ellipsoids are shown at the 50% probability level. Hydrogens and a
cocrystallized Et2O are omitted.
Results and Discussion
Complexes. The complexes trans-[Rh(X)(XNC)(PPh3)2]
(X ) Cl, 1; Br, 2; SC6F5, 3; C2Ph, 4; XNC ) 2,6-
dimethylphenyl isocyanide, xylyl isocyanide) are obtained
in moderate to good yield by the reaction of [Rh(X)(PPh3)3]
(X ) Cl, Br) with XNC and by the reaction of [Rh(H)-
(PPh3)4] with RH (R ) C6F5S, C2Ph) followed by XNC. On
slowly crystallizing 2 (X ) Br) small quantities of
[Rh(Br)3(XNC)(PPh3)2] (2b) are obtained. In solution com-
plexes 1-4 bind dioxygen reversibly, but only in the case
of 3 and 4 can O2 adducts {[Rh(SC6F5)(O2)(XNC)(PPh3)2]
(7) and [Rh(C2Ph)(O2)(XNC)(PPh3)2] (8)} be isolated. Com-
plexes 3 and 4 are air-sensitive in the solid state, 4
significantly more so than 3. The oxygenated complexes
[Rh(Cl)(O2)(XNC)(PPh3)2] (5) and [Rh(Br)(O2)(XNC)-
(PPh3)2] (6) in solution give up O2 readily, being ∼50%
reconverted to 1 and 2 on bubbling argon for 20 s and 3
min, respectively, at 27 °C. For 7 and particularly 8 the
process is much slower and heating is required.
Structures. The structures of complexes trans-[Rh(X)(X-
NC)(PPh3)2], 2 (X ) Br), 3 (X ) SC6F5), and 4 (X ) C2Ph),
which exhibit slight distortions from a square planar geom-
etry, are shown in Figures 1-3, respectively. Selected bond
lengths and angles are given in Table 2. The most noteworthy
feature of the data is a change in the Rh-C(1) (isocyanide)
bond length, which increases in the order X ) Br < SC6F5
< C2Ph. When the range is extended to include the complex
trans-[Rh(Cl)(XNC)(PiPr3)2], reported by Jones and Hes-
sell,14 the sequence of Rh-isocyanide carbon bond lengths
Figure 3. Diagram of trans-[Rh(C2Ph)(XNC)(PPh3)2] (4). Thermal el-
lipsoids are shown at the 30% probability level. Hydrogens and a
cocrystallized benzene are omitted.
(in Å) is as follows: 1.830(5) (X ) Cl), 1.861(7) (X ) Br),
1.882(2) (X ) SC6F5), 1.903(4) (X ) C2Ph). This result
would appear to arise from the operation of the well-known
trans influence,15 a ground-state effect by which metal-ligand
interactions are optimized. The polarization process discussed
below (which occurs only when the metal interacts with a
(15) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord. Chem. ReV. 1973,
10, 335.
(14) Jones, W. D.; Hessell, E. T. Organometallics 1990, 9, 718.
Inorganic Chemistry, Vol. 47, No. 19, 2008 8699

Carlton et al.
Table 2. Selected Bond Lengths (Å) and Angles (deg) for 2, 3, and 4
[Rh(Br)(XNC)(PPh3)2](CH2Cl2) (2)
[Rh(SC6F5)(XNC)(PPh3)2](0.5 Et2O) (3)
[Rh(C2Ph)(XNC)(PPh3)2](C6H6) (4)
Rh-P(1)
2.3384(18)
2.3349(18)
1.861(7)
1.168(9)
1.403(8)
2.3227(6)
2.3349(6)
1.882(2)
1.167(3)
1.401(3)
2.4025(6)
1.766(2)
87.22(6)
97.23(2)
175.98(2)
88.95(6)
86.55(2)
174.97(6)
113.87(8)
178.4(2)
2.3063(11)
2.3166(11)
1.903(4)
1.184(5)
1.389(5)
2.030(4)
1.298(7)
91.81(12)
86.83(13)
175.61(4)
91.36(12)
90.25(13)
175.3(2)
164.8(5)
177.6(3)
Rh-P(2)
Rh-C(1)
C(1)-N
N-C(2)
Rh-Aa
2.5196(8)
A-Bb
P(1)-Rh-C(1)
P(1)-Rh-Aa
P(1)-Rh-P(2)
P(2)-Rh-C(1)
P(2)-Rh-Aa
C(1)-Rh-Aa
88.1(2)
91.98(5)
176.04(6)
87.9(2)
91.98(5)
179.9(2)
Rh-A-Ba b
,
Rh-C(1)-N
179.6(6)
a A ) bromine (2), sulfur (3), carbon C (71) (4). b B ) carbon C (71) (3), carbon C (72) (4).
Table 3. Selected Bond Lengths(Å) and Angles (deg) for
[Rh(SC6F5)(O2)(XNC)(PPh3)2] (tol) (7)
Rh-P(1)
2.3737(10)
2.3624(10)
1.939(4)
1.165(4)
1.408(4)
2.4226(9)
1.771(4)
2.044(2)
2.018(2)
1.436(3)
83.38(10)
91.40(3)
88.77(7)
P(1)-Rh-O(2)
P(1)-Rh-P(2)
P(2)-Rh-C(1)
P(2)-Rh-S
P(2)-Rh-O(1)
P(2)-Rh-O(2)
C(1)-Rh-O(1)
C(1)-Rh-O(2)
C(1)-Rh-S
S-Rh-O(1)
S-Rh-O(2)
O(1)-Rh-O(2)
Rh-O(1)-O(2)
Rh-O(2)-O(1)
95.45 (7)
167.89(3)
89.99(10)
98.95(3)
85.90(8)
87.72(7)
120.65(12)
162.01(12)
91.74 10)
147.37(8)
106.24(7)
41.39(9)
Rh-P(2)
Rh-C(1)
C(1)-N
N-C(2)
Rh-S
S-C(71)
Rh-O(1)
Rh-O(2)
O(1)-O(2)
P(1)-Rh-C(1)
P(1)-Rh-S
P(1)-Rh-O(1)
68.35(14)
70.26(14)
Scheme 1. X ) Cl (1,5), Br (2,6), SC6F5 (3,7), C2Ph (4,8); XNC )
xylyl isocyanide; P ) PPh3; solvent, chlorobenzene
Table 4. Kinetic and Activation Parameters for the Dissociation of O2
from [Rh(X)(O2)(XNC)(PPh3)2] (X ) Cl, 5; Br, 6; SC6F5, 7; C2Ph, 8;
XNC ) xylyl isocyanide) in Chlorobenzene
a
Figure 4. Diagram of [Rh(SC6F5)(O2)(XNC)(PPh3)2] (7). Thermal ellipsoids
are shown at the 50% probability level. Hydrogens and a cocrystallized
toluene are omitted.
b
ligand X T/°C
k/s-1
R2
∆H‡/kJ mol-1 ∆S‡/J K mol-1
Cl
30
40
50
60
30
40
50
60
30
40
50
60
40
50
70
0.028
0.042
0.068
0.093
0.0041
0.0069
0.0135
0.029
0.9951 31.7 ( 1.6
-170.3 ( 5.0
reagent, e.g. O2) can be regarded as occurring independently
of this.
The structure of the dioxygen complex [Rh(SC6F5)(O2)-
(XNC)(PPh3)2] (7) is shown in Figure 4 and selected bond
lengths and angles are given in Table 3. The geometry is
best described as distorted octahedral, the distortions arising
from constraints imposed by the O-O bond. Data for 7 are
very similar to data for [Rh(CN)(O2)(XNC)(PPh3)2],16 [Rh-
(SPh)(O2)(PMe3)3],17 and [Rh(Cl)(O2)(PPh3)3].4 The O-O
bond length in 7 lies midway between the values found for
the Cl and SPh complexes, and for all four complexes there
is a close contact between one (or both) of the oxygen atoms
and a phenyl hydrogen. In the case of 7 the interaction (2.60
Å) is between O(1) and the hydrogen on C(6) of the
isocyanide phenyl of a neighboring molecule.
Br
0.986
52.1 ( 4.3
66.0 ( 5.8
-119.5 ( 13.6
-88.8 ( 18.2
-6.4 ( 5.4
SC6F5
C2Ph
0.00065 0.985
0.0015
0.0028
0.0081
0.00004 0.9997 101.3 ( 1.8
0.00013
0.0013
a R2 ) correlation coefficient for the plot (not shown) of ln(k1h/kBT) vs
(1/T). b ∆G‡ (300 K)/kJ mol-1: Cl, 82.8; Br, 88.0; SC6F5, 92.6; C2Ph, 103.2.
Scheme 1 activation parameters (Table 4) were obtained
which show appreciable variation according to ligand X. For
the complex showing the fastest rates of dissociation (X )
Cl) a fairly small ∆Hq (31.7 ( 1.6 kJ mol-1) is found together
with a large negative ∆Sq (-170.3 ( 5.0 J K-1 mol-1) while
for the complex showing the slowest rates (X ) C2Ph) a
Kinetic Study. From rate constants measured at temper-
atures in the range 30-70 °C for the reaction shown in
(16) Cˆırcu, V.; Fernandes, M. A.; Carlton, L. Polyhedron 2002, 21, 1775.
(17) Osakada, K.; Hataya, K.; Yamamoto, T. Inorg. Chem. 1993, 32, 2360.
8700 Inorganic Chemistry, Vol. 47, No. 19, 2008

ReWersible Binding of O2 by trans-[Rh(X)(XNC)(PPh3)2]
The work of Williams and co-workers19 has provided a
broad perspective of enthalpy-entropy relationships in
weakly bonded systems. A part of this account is summarized
briefly here. In a weak molecular association, where the
binding enthalpy is much lower than that of typical covalent
bonds, the potential energy well corresponding to the
associated state is shallow. Such a molecular association has
much residual motion (unlike the associated state occupying
a deep well) since energy levels corresponding to large
amplitude vibrations close to the lip of the well will be more
highly populated than when the well is deep. Weak associa-
tions with small binding enthalpies can give rise to remark-
ably small losses of entropy (upon binding), and the limiting
entropy, corresponding to the maximum possible loss of
motion, is only likely to be achieved in the case of a very
large binding enthalpy (a deep potential well).
Results reported in the present study can be interpreted in
this context. If, in a weakly associated system (such as a
metal-O2 complex), where the loss of entropy relative to
the isolated components (here the metal complex and free
molecular oxygen) is not large, the route to dissociation lies
through a small number of vibrational modes (or even a
single vibrational mode), then the transfer of vibrational
energy from the many modes (corresponding to the high
density of populated states close to the lip of the potential
well) to the very much smaller number of modes that can
lead to dissociation will involve the creation of order out of
relative disorder. A negative entropy can be expected for
such a process, with magnitude related to the shallowness
of the well and to the density of states close to the lip. For
more strongly associated systems (metal-O2 complexes
having higher ∆Hq) with a deeper potential well the
opportunities for vibrational motion will be more restricted
and the redistribution of vibrational energy that will be
required prior to dissociation will be proportionally less,
involving a smaller (less negative) activation entropy.
Williams19b draws attention to the number of studies that
have shown a linear relationship between the binding
enthalpy and the entropy change in the formation of weakly
bonded molecular associations.
Figure 5. Plot of ∆Hq vs ∆Sq for the dissociation of O2 from
[Rh(X)(O2)(XNC)(PPh3)2] (X ) Cl, 5; Br, 6; SC6F5, 7; C2Ph, 8) in
chlorobenzene.
much larger ∆Hq is found (101.3 ( 1.8 kJ mol-1) associated
with a ∆Sq close to zero (-6.4 ( 5.4 J K-1 mol-1). A plot
of ∆Hq versus ∆Sq (Figure 5) shows an excellent correlation
indicating that, to a high probability, the rate-determining
steps for the dissociation of O2 from the four complexes
[Rh(X)(O2)(XNC)(PPh3)2] correspond to successive positions
along a single generalized reaction coordinate. In such a
model the rate-determining step for the chloro complex
appears early on the reaction coordinate, where the molecular
geometry has undergone few distortions, and the rate-
determining step for the acetylide complex appears much
later on the reaction coordinate where substantial distortion
has occurred. The finding of a negative activation entropy
for a reaction involving the evolution of a gas is unusual
and would require a highly ordered transition state (see
below) but is not unprecedented. For example, an entropy
of activation of -41.8 J K-1 mol-1 has been reported by
Vaska and co-workers18 for the dissociation of O2 from
[Ir(NCO)(O2)(PPh3)2(CO)] (see below). A correlation be-
tween enthalpy and entropy of activation is frequently
encountered in weakly bonded interactions.19 A possible
explanation for the results shown in Figure 5 is as follows.
The requirement of a highly ordered transition state is met
by a situation in which the metal-ligand bonds all vibrate
in unison, that is, they simultaneously stretch and simulta-
neously shrink, with the result that electron density in the
metal-dioxygen bond(s) is depleted to the greatest possible
extent when the metal-ligand bonds experience maximum
extension. The differing contributions of electron density
from ligands X, in particular contributions arising from
electron density normally residing on the ligand (polariz-
ability effects), will then govern the position of the rate-
determining step on the reaction coordinate.
A possible precedent for a rhodium-dioxygen complex
in which the O2 has high residual motion can be found in a
recent report by Milstein and co-workers.20 The complex
[{Me2C6H(CH2PtBu2)2}Rh(η2-O2)] has an unusually short
O-O bond (1.365(18)Å), a value of J(103Rh, 31P) of 146
Hz, indicative more of Rh(I) than Rh(III), and readily gives
up O2 in exchange for CO. The authors interpret this evidence
in terms of an interaction of the type
RhI---O|
O
with the implication that the O2 ligand has the ability to rotate
in the manner of a propeller.
Activation parameters reported by Vaska and co-workers
for the dissociation of O2 from the complexes [Ir(X)(O2)-
(18) Vaska, L.; Chen, L. S.; Senoff, C. V. Science 1971, 174, 587.
(19) (a) Williams, D. H.; Searle, M. S.; Westwell, M. S.; Gerhard, U.;
Holroyd, S. E. Phil. Trans. R. Soc. Lond. Ser. A. 1993, 354, 11. (b)
Searle, M. S.; Westwell, M. S.; Williams, D. H. J. Chem. Soc., Perkin
Trans. 2 1995, 141. (c) Williams, D. H.; Westwell, M. S. Chem. Soc.
ReV. 1998, 27, 57.
(20) Frech, C. M.; Shimon, L. J. W.; Milstein, D. HelV. Chim. Acta 2006,
89, 1730.
Inorganic Chemistry, Vol. 47, No. 19, 2008 8701

Carlton et al.
indicate the presence of a good σ donor and/or poor π
acceptor (but see also ref 26). The ordering of the anionic
ligands derived from the data in Figure 6 can, however, be
readily accounted for qualitatively in terms of their relative
polarizability.
The ease or difficulty with which the electron density on
a coordinatively unsaturated metal atom can be attracted to
an approaching electrophile (a process in which the metal
becomes electronically polarized) will reflect contributions
both from the metal and from the ligands. The metal atom
can be regarded as having a certain intrinsic polarizability
which increases on moving down a group of the Periodic
Table and on which are superimposed effects arising from
variations in oxidation state and of ligands. A ligand may
be polarizable by virtue of having a polarizable donor atom
(Cl, Br, S) or as a consequence of having a multiply bonded
π electron system (CN, C2Ph), the polarizability being
inversely proportional to the π-acidity of the ligand. Such a
ligand can be regarded as possessing stored electron density
that can be given up to the metal on demand, that is, when
constrained to do so as a result of an interaction between
the metal and an electrophile. The availability of this stored
electron density will only be poorly reflected in spectroscopic
measures of (relative) electron density on the metal in its
ground-state but would be expected to become apparent in
a situation involving electronic distortion of the metal
induced by an external reagent such as O2.
From this perspective the data in Figure 6 show, perhaps
unsurprisingly, that iridium is intrinsically more polarizable
than rhodium and that polarizable ligands enhance the
binding of O2 to the metal. In view of the high correlation
between ∆Hq and ∆Sq, strongly indicating a common
pathway for the dissociation process, it becomes meaningful
to attribute differences in these parameters to differences in
polarizabilities of both the metal and the ligands. Although
the contributions of the metal and the ligands are not entirely
separable, it is possible from Figure 6 to conclude that the
ligands show increasing polarizability in the order NCO <
N3 < Cl < Br < SC6F5 < C2Ph, that their contribution to
the polarizability of the respective Rh and Ir complexes is
proportionately greater in the case of rhodium than of iridium,
and that the influence on the activation parameters of changes
in the polarizability of the anionic ligand is, to a good
approximation, equivalent to that of changes in the intrinsic
polarizability of the metal (i.e., of exchanging one metal for
the other). This means that within the limits imposed by the
present data the activation parameters are insensitive to the
source of the electron density that is given up by the complex
to the oxygen, that is, to the relative contributions made by
the metal and by the anionic ligand. This is by no means an
obvious result, and further work will help to establish how
widely applicable it is likely to be.
Figure 6. Plot of of ∆Hq vs ∆Sq for the dissociation of O2 from
[Rh(X)(O2)(XNC)(PPh3)2] (X ) Cl, 5; Br, 6; SC6F5, 7; C2Ph, 8) and from
[Ir(X)(O2)(CO)(PPh3)2] (X ) NCO, N3, Cl, Br, I) in chlorobenzene. Data
for the iridium complexes are taken from ref 18.
(CO)(PPh3)2] (X ) NCO, N3, Cl, Br, I)18 follow a similar
pattern to those given here for [Rh(X)(O2)(XNC)(PPh3)2] (X
) Cl, Br, SC6F5, C2Ph) and are included in Figure 6 which
shows that the plot of ∆Hq against ∆Sq for the rhodium
complexes can be extended to include the iridium complexes
(the data for the Ir iodide complex appear somewhat
anomalous). The two series of complexes would therefore
appear to share a common dissociation mechanism. Figure
6 shows that ∆Hq increases for the rhodium complexes in
the sequence Cl < Br < SC6F5 < C2Ph and for the iridium
complexes in the sequence NCO < N3 < Cl < Br with ∆Hq
for the Cl- and Br-containing iridium complexes having
significantly higher values than for the rhodium counterparts.
Taking into account both series of complexes (Rh, Ir) an
ordering of the anionic ligands according to their ability to
enhance the binding of O2 is NCO < N3 < Cl < Br < SC6F5
< C2Ph. For the rhodium complexes trans-[Rh(X)(X-
NC)(PPh3)2], a stable O2 adduct is obtained with X ) SC6F5
and with C2Ph, (and not with Cl or Br). A stable O2 adduct
is also found for X ) CN16 indicating that an appropriate
position for the CN ion in the above sequence is to the right
of Br (i.e., Cl < Br < CN). The sequence of anionic ligands
given above clearly reflects variations in an electronic
property of the ligands X but does not match the ordering
(taken in either direction) of the spectrochemical (Br < Cl
< NCO < CN)21 or nephelauxetic (CN ∼ Cl < Br)22 series;
neither is there agreement with the ordering of νco for the
complexes trans-[Rh(X)(CO)(PPh3)2] (C2Ph < Cl, Br <
SC6F5)23-25 where a low wavenumber is generally taken to
Conclusion
(21) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry, 3rd ed.;
J. Wiley: London, 1972; p 577.
Activation parameters for the dissociation of O2 from the
complexes [Rh(X)(O2)(XNC)(PPh3)2] show that the influence
(22) Mason, J. Chem. ReV. 1987, 87, 1299.
(23) Cetinkaya, B.; Lappert, M. F.; McMeeking, J.; Palmer, D. E. J. Chem.
Soc., Dalton Trans. 1973, 1202.
(25) Carlton, L. J. Organomet. Chem. 1992, 431, 103.
(26) McAuliffe, C. A.; Pollock, R. J. Organomet. Chem. 1974, 77, 265.
(24) Vaska, L.; Peone, J. J. Chem. Soc., Chem. Commun. 1971, 418.
8702 Inorganic Chemistry, Vol. 47, No. 19, 2008

ReWersible Binding of O2 by trans-[Rh(X)(XNC)(PPh3)2]
of ligand X in stabilizing the O2 adduct increases in the order
Cl < Br < SC6F5 < C2Ph. A high correlation between ∆Hq
and ∆Sq, in a plot which can be extended to include work
of earlier researchers using complexes of iridium, can be
taken as strong evidence of a common dissociation pathway
for complexes [M(X)(O2)(L)(PPh3)2] (M ) Rh, L ) XNC;
M ) Ir, L ) CO) and a single reaction coordinate along
which the rate-determining step occurs at a different position
(indicated by the magnitude of ∆Hq) for each relevant
combination of M, X, and L. The comparison of properties
of ligands X in terms of their respective contributions to ∆Hq
(for given M and L) is then practicable. The influence of
ligand X on ∆Hq can be related to its polarizability, that is,
to the presence of electron density that in the unoxygenated
complex resides on the ligand but which, in the presence of
O2, is given up to the metal to facilitate the metal-O2
interaction. It is proposed that the sequence NCO < N3 <
Cl < Br < SC6F5 < C2Ph represents a scale of increasing
polarizability of the anionic ligand.
Acknowledgment. We thank the University of the Wit-
watersrand for financial support and Professor H.M. Marques
for useful discussion.
Supporting Information Available: Crystallographic data in
CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC800478G
Inorganic Chemistry, Vol. 47, No. 19, 2008 8703