Tuning the properties of the osmium nitrido group in TpOs(N)X2 by changing the ancillary ligand

Article abstract of DOI:10.1021/ic0260264

TpOs(N)(OAc)₂ (2) is formed upon reaction of TpOs(N)Cl₂ (1) with excess silver acetate (Tp = hydrotrispyrazolylborate). Treatment of 2 with protic acids HX gives the osmium(VI) complexes TpOs(N)X₂, where X = trifluoroacetate (TFA, 3), trichloroacetate (TCA, 4), tribromoacetate (TBA, 5), bromide (6), oxalate (X₂ = O₂C₂O₂, 7), or nitrate (ONO₂, 8). Cyclic voltammetry studies of 1-8 show irreversible reductions of Os^VI to Os^V, varying over a range of 0.63 V. Much smaller relative variations are observed in ¹⁵N NMR chemical shifts, v(Os≡N) stretching frequencies, and optical absorbances. Compounds 1-8 all react with PPh₃ by nucleophilic attack at the nitride ligand, yielding TpOs(NPPh₃)X₂. The reactions are accelerated by more electron withdrawing ligands X. The relative rates correlate with the peak reduction potentials although the effect is small: the rates vary by only 10² while the 0.63 V change in E_p,c corresponds to a change in equilibrium constant for electron transfer of ～10¹¹. Compounds 1-8 also react with triphenylboron, BPh₃, with formation of borylanilido complexes TpOs[N(Ph)BPh₂]X₂. However, rate constants for reactions with BPh₃ to yield Os^IV boryl-amido complexes do not in general correlate with one- electron-reduction potentials. This is likely due to the mechanism of the BPh₃ reactions being a two-step process and not simply nucleophilic attack.

Full text of DOI:10.1021/ic0260264

Inorg. Chem. 2003, 42, 605−611
Tuning the Properties of the Osmium Nitrido Group in TpOs(N)X by
2
Changing the Ancillary Ligand
Ahmad Dehestani, Werner Kaminsky,† and James M. Mayer*
Department of Chemistry, UniVersity of Washington, Box 351700, Seattle, Washington 98195-1700
Received September 11, 2002
TpOs(N)(OAc)2 (2) is formed upon reaction of TpOs(N)Cl2 (1) with excess silver acetate (Tp ) hydrotrispyra-
zolylborate). Treatment of 2 with protic acids HX gives the osmium(VI) complexes TpOs(N)X , where X )
2
trifluoroacetate (TFA, 3), trichloroacetate (TCA, 4), tribromoacetate (TBA, 5), bromide (6), oxalate (X ) O C O ,
2
2 2 2
7), or nitrate (ONO , 8). Cyclic voltammetry studies of 1−8 show irreversible reductions of OsVI to OsV, varying over
2
a range of 0.63 V. Much smaller relative variations are observed in 15N NMR chemical shifts, ν(OstN) stretching
frequencies, and optical absorbances. Compounds 1−8 all react with PPh3 by nucleophilic attack at the nitride
ligand, yielding TpOs(NPPh3)X . The reactions are accelerated by more electron withdrawing ligands X. The relative
2
rates correlate with the peak reduction potentials although the effect is small: the rates vary by only 102 while the
0.63 V change in Ep,c corresponds to a change in equilibrium constant for electron transfer of ∼1011. Compounds
1−8 also react with triphenylboron, BPh3, with formation of borylanilido complexes TpOs[N(Ph)BPh2]X . However,
2
IV
rate constants for reactions with BPh3 to yield Os boryl−amido complexes do not in general correlate with one-
electron-reduction potentials. This is likely due to the mechanism of the BPh3 reactions being a two-step process
and not simply nucleophilic attack.
Introduction
related tetraalkyl and 1,2-benzenedithiolate derivatives
[Os(N)R4]- and [Os(N)(bdt)2]- are electron rich and readily
add R+ to make imido compounds, due to electron-releasing
alkyl and thiolate ligands.3,4
We have previously described a number of reactions in
which TpOsVI(N)Cl2 (1) acts as an electrophile at the nitrido
ligand [Tp ) hydrotris(pyrazolyl)borate].5,6 For instance, 1
reacts rapidly with triphenylphosphine to give the phosphin-
iminato complex TpOsIV(NPPh3)Cl2.6 Triphenylboron un-
dergoes an unusual reaction with 1 to afford the borylanilido
compound [TpOsIV{N(Ph)BPh2}Cl2] (eq 2).5 This is an
example of C-N bond formation by carbanion addition to
the nitrido ligand.5b
Multiply bonded ligands, such as nitrido or oxo groups,
can react as electrophiles or nucleophiles. When they react
as electrophiles, the net result is usually two-electron
reduction of the metal center and formation of new chemical
bond(s) (eq 1). This coupling of redox activity and bond
LnMn+(X) + :Nu f LnM(n-2)+(X-Nu)
(1)
formation and the resulting penchant for two-electron
reactivity are among the reasons that such compounds are
so valuable as oxidants for organic compounds. A variety
of osmium(VI) nitrido compounds are known,1 and the
nitrido ligand can react as either an electrophile or a
nucleophile depending on the ligand environment. For
example, the nitride ligands of [OsVI(N)X4]- (X ) Cl or Br)
are electrophilic and subject to nucleophilic attack by
phosphines (PAr3) to afford the phosphiniminato complexes
Os(NPR3)X3(PAr3)2.2 In contrast, the nitride ligands of the
This report describes a systematic study of how changing
the chlorides to other ancillary ligands affects the reactivity,
* Author to whom correspondence should be addressed. E-mail: mayer@
chem.washington.edu.
† UW crystallography facility.
(3) Marshman, R. W.; Shapley, P. A. J. Am. Chem. Soc. 1990, 112, 8369.
(4) Sellman, D.; Wemple, M. W.; Donaubawer, W.; Heinemann, F. W.
Inorg. Chem. 1997, 36, 1379.
(1) Griffith, W. P. Coord. Chem. ReV. 1972, 8, 369.
(2) Griffith, W. P.; Pawson, D. J. Chem. Soc., Dalton Trans. 1973, 531.
10.1021/ic0260264 CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/28/2002
Inorganic Chemistry, Vol. 42, No. 2, 2003 605

Dehestani et al.
Scheme 1
spectroscopic, electrochemical, and structural properties of
TpOs(N)X2 compounds. A key theme of this work is to
explore the relationship between bond-forming reactivity and
one-electron redox properties, in this case measured by cyclic
voltammetry. The oxidizing power of reagents is often
quantified by electrochemical potentials, but these do not
simply relate to the inner-sphere multielectron processes
typical of metal-ligand multiple bonds such as those shown
in eqs 1 and 2.
Figure 1. ORTEP drawing of TpOs(N)(TFA)2 (3).
Table 1. Crystal and Structural Refinement Dataa
TpOs(N)TFA2 (3)
TpOs(N)(ONO2)2 (8)
empirical formula
fw
C13H10BF6N7O4Os
643.29
C9H10BN9O6Os
541.27
cryst syst
monoclinic
monoclinic
cryst size (mm)
space group
0.25 × 0.10 × 0.05
0.24 × 0.05 × 0.005
P21/c (No. 13)
P21/n (No. 14)
unit cell (Å, deg)
Results
a
b
c
17.608(2)
13.718(4)
19.390(4)
90
156.47(2)
90
1870.2(7)
4
2.285
6.917
130(2)
4786
1869
289
7.264(8)
13.237(11)
16.43(2)
90
93.26(5)
90
1578(3)
4
2.279
8.135
130(2)
4026
2174
235
Syntheses and Structures of TpOs(N)X2. TpOs(N)-
(OAc)2 (2) is prepared by refluxing excess silver acetate with
TpOs(N)Cl2 (1) in acetonitrile for 12 h. Column chroma-
tography on silica gel affords pure 2 in 70-90% yield.
Related silver metatheses were unsuccessful for AgF and
AgOTf (OTf ) triflate, OSO2CF3). Complex 2 is a valuable
precursor for the preparation of other substituted TpOs(N)-
X2 complexes. Treatment of acetone solutions of 2 with an
excess of the corresponding acid, HX, yields the air stable
compounds TpOs(N)(TFA)2 (3, TFA ) trifluoroacetate),
TpOs(N)(TCA)2 (4, TCA ) trichloroacetate), TpOs(N)-
(TBA)2 (5, TBA ) tribromoacetate), TpOs(N)Br2 (6), TpOs-
(N)(O2C2O2) (7, O2C2O2 ) oxalate), and TpOs(N)(ONO2)2
(8). The synthetic approach is depicted in Scheme 1. The
related fluoride complex could be generated (and the triflate
derivative from the fluoride complex plus MeOTf), but these
could not be obtained in pure form and were only character-
R
â
γ
vol (Å3)
Z
density (g/cm3, calcd)
abs coeff (mm-1
temp (K)
)
reflns collected
indep reflns
no. of params
final R, Rw (%)
GOF
6.41, 16.42
1.042
6.25, 13.41
0.852
a The radiation for both structures was Mo KR (λ ) 0.71073 Å).
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
TpOs(N)(TFA)2 (3) and TpOs(N)(ONO2)2 (8)
TpOs(N)(TFA)2 (3)
TpOs(N)(ONO2)2 (8)
Os(1)-N(1)
2.289(20)
2.024(19)
2.044(16)
2.030(17)
1.996(14)
1.602(20)
80.63(6)
82.84(6)
89.93(8)
174.16(8)
95.38(8)
92.9(8)
159.20(7)
164.55(6)
92.55(7)
89.71(7)
79.20(7)
81.86(6)
105.1(9)
102.47(8)
82.47(7)
2.219(19)
2.045(16)
2.091(16)
2.009(14)
2.019(14)
1.654(17)
80.93(7)
81.41(7)
88.25(7)
174.34(7)
96.36(7)
93.57(7)
161.66(7)
160.23(7)
86.78(6)
94.11(6)
80.87(6)
79.60(6)
101.56(6)
105.63(7)
84.86(6)
Os(1)-N(3)
1
ized by H NMR and mass spectrometry.
Os(1)-N(5)
Os(1)-O(1)
Single-crystal X-ray structures have been solved for the
trifluoroacetate and nitrate complexes 3 and 8. Crystal
refinement data are given in Table 1, and selected bond
lengths and angles are given in Table 2. ORTEP diagrams
are displayed in Figures 1 and 2. The structures are similar
to each other and to other Tp and nitrido complexes.5-7 The
typically short OstN triple bonds push the cis ligands away
[all N(7)-Os-X angles are >90°] and exerts a strong trans
influence.8 The geometric features of the Tp ligand also
contribute to the distorted octahedral structures, with Npz-
Os(1)-O(2)
Os(1)tN(7)
N(3)-Os(1)-N(1)
N(5)-Os(1)-N(1)
N(3)-Os(1)-N(5)
N(1)-Os(1)-N(7)
N(3)-Os(1)-N(7)
N(5)-Os(1)-N(7)
O(1)-Os(1)-N(3)
O(2)-Os(1)-N(5)
O(1)-Os(1)-N(5)
O(2)-Os(1)-N(3)
O(1)-Os(1)-N(1)
O(2)-Os(1)-N(1)
O(1)-Os(1)-N(7)
O(2)-Os(1)-N(7)
O(1)-Os(1)-O(2)
(5) (a) Crevier, T.; Mayer, J. M. Angew. Chem. 1998, 37, 1891-1892.
(b) Crevier, T.; Bennett, B. K.; Soper, J. D.; Bowman, J. A.; Dehestani,
A.; Hrovat, D. A.; Lovell, S.; Kaminsky, W.; Mayer, J. M. J. Am.
Chem. Soc. 2001, 123, 1059-1071.
(6) (a) Bennett, B. K.; Pitteri, S.; Pilobello, L.; Lovell, S.; Kaminsky,
W.; Mayer, J. M. J. Chem. Soc., Dalton Trans. 2001, 3489-3497.
(b) Crevier, T. J. Ph.D. Thesis, University of Washington, 1998. (c)
Bennett, B. K.; Saganic, E.; Lovell, S.; Kaminsky, W.; Samual, A.;
Mayer, J. M., submitted for publication.
Os-Npz angles less than 90°. As noted by Chang et al. and
confirmed here, MtN bond distances in [OsVI(N)L] do not
vary with the electron-donating properties of L.9
606 Inorganic Chemistry, Vol. 42, No. 2, 2003

Properties of the Osmium Nitrido Group in TpOs(N)X2
350 M-1 cm-1). There are also two intense bands in the UV
(ꢀ > 104 M-1 cm-1), at ∼260 and ∼240 nm. In the more
-
symmetrical complexes Os(N)X4 , Gray, Che, and co-
workers have assigned weak absorptions at 400-500 nm to
dxyfdπ* transitions.13 The dπ* orbitals are the antibonding
combination of dxz and dyz with the nitrido px and py.13c
Cyclic voltammograms of 1-8 show only irreversible
reduction waves, which are attributed to reduction of OsVI
to OsV. The Ep,c values vary from -1.20 to -1.83 V vs
Cp2Fe+/0. The values are given in Table 3, where the
compounds are listed in order of decreasing potentials. In
general, the more negative potentials are found for com-
pounds with the more electron donating substituents (acetate
and oxalate), and the most easily reduced compound is the
trifluoroacetate. While there is a trend within the carboxylate
complexes, overall the Ep,c values for TpOs(N)X2 do not
correlate with any simple parameter for X such as Hammett
σp parameters, Lever’s EL values,14 the PL parameters
developed by Chatt, Leigh, and Pickett,15 or the pKa of HX.
We have previously shown that for 1 and a range of related
d2 Tp-osmium and -rhenium complexes, the potentials for
oxidation to d1 species correlate well with Hammett σp
parameters but the reduction potentials do not.16
Figure 2. ORTEP drawing of TpOs(N)(ONO2)2 (8).
Table 3. Electrochemical and Spectroscopic Properties of TpOs(N)X2
Complexes
a
c
15
TpOs(N)X2
Ep,c
δ(15N)b
ν
OstN (νOst
)
N
λ
max (ꢀ)d
TFA (3)
TCA (4)
Br (6)
-1.20
-1.25
-1.28
-1.34
-1.36
-1.38
-1.40
-1.83
908
909
887
890
907
922
884
881
1080 (1042)
1081 (1042)
1068 (1031)
1066 (1031)
1082 (1044)
1059 (1024)
1070 (1034)
1087 (1050)
440 (380)
425 (350)
475 (400)
400 (420)
440 (300)
435 (330)
440 (310)
440 (314)
Cl (1)
TBA (5)
ONO2 (8)
O2C2O2 (7)
OAc (2)
a Irreversible Os(VI)/Os(V) peak potentials in V vs Cp2Fe+/0 in MeCN
at 0.1 V s-1 scan rate. b 15N NMR chemical shifts in ppm referenced to δ
Reactions with BPh3 and PPh3. Compounds 2-8 all react
with BPh3 to form borylamido complexes TpOs[N(BPh2)-
Ph]X2, as previously reported for 1.5 1H NMR spectra of
the Os(IV) borylamido products closely resemble that of the
dichloride derivative. The only exception is the acetate
product TpOs[N(BPh2)Ph](OAc)2 (9), which has C1 rather
than Cs symmetry. This is likely due to an interaction
between the boron and an acetate oxygen as shown in eq 3.
) 0.0 for aqueous 15NH4
. .
c In cm-1 d λmax/nm (ꢀ/M-1 cm-1) for the visible
+
band; UV bands given in the Experimental Section.
Spectroscopy and Electrochemistry of TpOs(N)X2. 1H
NMR spectra indicate that all of the complexes are diamag-
netic and have Cs symmetry, based on the pyrazole peaks of
the Tp ligand appearing in 2:1 ratios.10 15N-Enriched materi-
als, prepared from TpOs(15N)Cl2,5 were used to obtain 15N
NMR spectra. All the compounds show a single peak
between 821 and 922 ppm (Table 3), referenced to external
+
aqueous 15NH4 (δ ) 0). These values are similar to those
reported for other nitrido ligands on osmium(VI).11
The presence of the nitrido ligand is evidenced by the
strong sharp V(OstN) band between 1000 and 1100 cm-1
in the IR spectra of 1-8 (Table 3).12 15N substitution was
used to confirm this assignment for each compound because
of the many Tp ligand bands in the same spectral region.
All 14N/15N isotopic shifts agreed well with values calculated
from a diatomic harmonic oscillator model. For instance,
V(OstN) for 3 and 3-15N are 1080 and 1042 cm-1, consistent
with the calculated shift of 38 cm-1.
A similar but apparently weaker B-Cl interaction was found
in the structure of TpOs[N(BPh2)Ph]Cl2.5 Variable temper-
1
ature H NMR spectra of the trifluoroacetate derivative 3
showed Cs symmetry down to 213 K. In this case, either a
carboxylate interaction analogous to that in eq 3 is not present
or it is too weak to give a nonfluxional structure in the NMR.
Competition reactions were used to determine the relative
reactivities of the nitrido complexes (Scheme 2). Stock
solutions of two osmium complexes in CD2Cl2, with C6Me6
as an internal standard, were mixed, and a 1H NMR spectrum
UV-vis spectra of the complexes (Table 3) show a broad
band in the visible region between 400 and 475 nm (ꢀ ∼
(7) See, for instance: (a) Demadis, K. D.; El-Samanody, E.; Meyer, T.
J.; White, P. S. Polyhedron 1999, 18, 1587-1594. (b) Bright, D.; Ibers,
J. A. Inorg. Chem. 1969, 8, 709. (c) Collison, P. M.; Garner, C. D.;
Mabbs, F., E.; Salthouse, J. A.; King, T. J. J. Chem. Soc., Dalton
Trans. 1981, 1812-1819. (d) Hunt, S. L.; Shapley, P. A. Organo-
metallics 1997, 16, 4071-4076. (e) See also refs 8-11.
(8) Nugent, W. A.; Mayer, J. M. Metal-Ligand Multiple Bonds; Wiley-
Interscience: New York, 1988; Chapter 5.
(9) Chang, P. M.; Yu, W. Y.; Che, C. M.; Cheung, K. K. J. Chem. Soc.,
Dalton Trans. 1998, 3183.
(10) Abrams, M. J.; Davison, A.; Jones, A. G. Inorg. Chim. Acta 1984,
82, 125.
(13) (a) Hopkins, M. D.; Miskowski, V. M.; Gray, H. J. Am. Chem. Soc.
1986, 108, 6908-6911. (b) Chin, K.-F.; Cheung, K.-K.; Yip, H.-K.;
Mak, T. C. M.; Che, C. M. J. Chem. Soc., Dalton Trans. 1995, 657-
663. (c) Cowman, C. D.; Trogler, W. C.; Mann, K. R.; Poon, C. K.;
Gray, H. B. Inorg. Chem. 1976, 8, 1747-1751.
(14) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271-1285.
(15) (a) Butler, G.; Chatt, J.; Leigh, G. J.; Pickett, C. J. J. Chem. Soc.,
Dalton Trans. 1979, 113. (b) Chatt, J.; Kan, C. T.; Leigh, G. J.; Pickett,
C. J.; Stanley, D. R. J. Chem. Soc., Dalton Trans. 1980, 2023. (c)
Chatt, J. Coord. Chem. ReV. 1982, 43, 337.
(11) Stumme, M.; Preetz, W. Z. Anorg. Allg. Chem. 2000, 626, 1186-
1190.
(16) Bennett, B. K.; Crevier, T. J.; DuMez, D. D.; Matano, Y.; McNeil,
W. S.; Mayer, J. M. J. Organomet. Chem. 1999, 591, 96-103.
(12) Griffith, W. P. J. Am. Chem. Soc. 1965, 87, 3964.
Inorganic Chemistry, Vol. 42, No. 2, 2003 607

Dehestani et al.
Scheme 2
Table 4. Kinetic Results for Reactions of TpOs(N)X2 with BPh3 and
PPh3
k for BPh3
TpOs(N)X2
(kX/kTFA) for BPh3
(M-1 s-1
)
(kX/kTFA) for PPh3
TFA (3)
TCA (4)
Br (6)
[1]
4100
3300
8600
5300
2100
5200
4100
41
[1]
0.8
2.1
1.3
0.5
1.3
1.0
0.01
0.95
0.89
0.85
0.81
0.77
0.75
0.15
Figure 3. A plot of Y ){[2]o - [BPh3]o}-1 ln{[BPh3]o[2]/[BPh3][2]o} vs
time for the reaction of TpOs(N)(OAc)2 (2) + BPh3 in CH2Cl2 at 25 °C.
The reaction is 90% finished by 95 s.
Cl (1)
TBA (5)
ONO2 (8)
O2C2O2 (7)
OAc (2)
the other osmium complexes with BPh3 can be calculated
(Table 4).
was taken to confirm the initial concentrations. Addition of
BPh3 caused an immediate color change from light orange
to deep red, consistent with the formation of osmium(IV)
amido complexes. Final concentrations were determined by
1H NMR integration. Each of the 24 different competition
experiments was repeated three times, with a reproducibility
within 5%. Due to the rapid rates, we were concerned that
complete mixing was not obtained prior to reaction. There-
fore additional experiments were performed where solid BPh3
was placed at the top of a tilted NMR tube loaded with the
solution containing the osmium complexes, and then the
NMR tube was mixed vigorously and quickly. This procedure
gave the same results as above within experimental error.
The relative rate constants were determined from the initial
and final concentrations according to eqs 4 and 5.17a As a
-d[2]
dt
) k[2][BPh3]
[BPh3]o[2]
(6)
(7)
1
ln
) kt
(
)
[2]o - [BPh3]o [BPh3][2]o
Compounds 2-8 also react rapidly with PPh3 to make
osmium(IV) phosphiniminato complexes (eq 8). These
reactions appear to be very similar to the analogous reaction
with 1, for which the product has been structurally character-
ized.6 Competition studies between different TpOs(N)X2
compounds were run in the same manner as BPh3 described
above. The relative rates are given in Table 4.
[TpOsNX2]f
[TpOsNX2]i
k
[TpOsNY2]f
[TpOsNY2]i
ln
)
x ln
(4)
(
)
(
)
ky
1 - [TpOs{N(Ph)BPh2}X2]
Discussion
ln
)
(
)
[TpOsNX2]i
This study compares the spectroscopic, electrochemical,
and reactivity properties of a series of osmium nitrido
complexes (1-8). Cyclic voltammetry shows that each
compound undergoes an irreversible reduction, and that the
peak potentials Ep,c vary over a relatively large range of 0.63
V. While irreversible peak potentials are not necessarily
related to thermodynamic potentials, the relationship appears
to be fairly good in this case. Potentials for the carboxylate
complexes, for instance, fall in the order of the pKa values
for the monovalent acids: O2CMe , oxalate < O2CCBr3
< O2CCCl3 < O2CCF3. In general, a more electron with-
drawing ligand makes the osmium complex more electron
deficient and makes OsVI f OsV reduction easier. The good
correlation of Ep,c with rate constant described below further
suggests that the Ep,c values reflect properties of the TpOs-
(N)X2 complexes rather than the rates of decomposition of
the reduced forms.
k
1 - [TpOs{N(Ph)BPh2}Y2]
x ln
(5)
(
)
ky
[TpOsNY2]i
final check, the relative rate constants were found to be
transferable. For example, competition reactions between 3
and 1 or 6 gave kCl/kTFA and kBr/kTFA values of 1.3 and 2.1,
whose ratio of 0.62 is consistent with the directly determined
kCl/kBr of 0.65. The relative rates versus TpOs(N)(TFA)2 as
the standard complex are given in Table 4.
The kinetics of the reaction between BPh3 and 2 was
examined under second-order conditions in anhydrous CH2-
Cl2 by optical spectroscopy. The absorbance increased
throughout the observed spectrum, with the formation of 9
indicated by new peaks at λmax ) 270, 325, 365, and 450
nm. A simple bimolecular rate law (eq 6) is indicated by
the reasonably linear integrated rate law (eq 7, Figure 3).17b
Three independent runs gave kOAc ) 41 ( 10 M-1 s-1 at 25
°C. Using this rate constant and the relative rates from
competition experiments, the rate constants for reactions of
A variation of 0.63 V in redox potential corresponds to a
change in equilibrium constant for electron transfer of 1010.7
.
Given this large difference, the spectroscopic differences
between the compounds are quite small. The OstN stretch-
ing frequencies, 15N NMR chemical shifts, and optical d-d
(17) Weston, R.; Schwarz, H. Chemical Kinetics; Prentice Hall: Upper
Saddle River, NJ, 1972; (a) p 16, (b) p 10.
608 Inorganic Chemistry, Vol. 42, No. 2, 2003

Properties of the Osmium Nitrido Group in TpOs(N)X2
transition energies vary only slightly over the series. It seems
that changes in the X ligands do not change the nature of
the bonding and orbital energy gaps, but rather they shift
the whole manifold of orbitals up or down. The shifting up
of the whole manifold is observed in DFT calculations
comparing TpOs(N)Cl2 and the hypothetical TpOs(N)-
(CH3)2.5b Changing from Cl to Me raises the energy of all
orbitals, although not by the same amounts: the nitrido lone
pair rises 1.1 eV, the roughly nonbonding dxy HOMO rises
0.7 eV, and the π* LUMOs (dxz, dyz) rise 1.3 eV. The Cl to
Me change does affect the HOMO-LUMO gap, from 1.6
to 2.2 eV, in contrast to what is observed for compounds
1-8. This is presumably due to chloride vs methyl being a
much larger variation in σ- and π-donor abilities than is
found over the whole set of ligands in 1-8.
Electrochemical properties are often assumed to correlate
with reactivity. In a series of related molecules, such as 1-8,
the derivatives with more favorable potentials for reduction
(more positive potentials) should be, in general, more reactive
toward reduction by inner-sphere processes such as eqs 2,
3, and 8. Such a connection is intuitively reasonable, but
there is no strict thermochemical reason why rates of bond-
forming reactions should be related to redox potentials. The
connection between one-electron potentials and reactivity is
particularly tenuous for the two-electron inner-sphere redox
processes studied here.
For compounds 1-8, the relative rates of reaction are
remarkably insensitive to the reduction potentials. The
relative rates for reduction of Os(VI) to Os(IV) by BPh3 vary
by only 102, and the reactions with PPh3 vary by only a factor
of 7. These ranges are very small given that the range of
potentials corresponds to J1010 in equilibrium constant for
electron transfer as noted above. The Brønsted R values for
these reactions [the slope of log(krel) vs log(KET) plots (taking
log(KET) ) Ep,c/0.059 V)] are 0.2 for BPh3 and 0.08 for PPh3.
The small variation in rate constants for PPh3 + TpOs-
(N)X2 correlate quite well with Ep,c, as shown in the graph
of log(kX/kTFA) vs Ep,c, Figure 4A. The most easily reduced
complex, 3, reacts the fastest, and the most difficult to reduce,
2, reacts the slowest. However, relative rate constants for
reduction of TpOs(N)X2 by BPh3 do not correlate with Ep,c,
as shown in Figure 4B. The trifluoroacetate complex 3 is
more easily reduced than any other complex but reacts more
slowly than 1, 6, 7, and 8. Within the same type of ligand,
there appears to be a correlation; for instance, TpOs(N)Br2
is more easily reduced and reacts faster than TpOs(N)Cl2.
The relative rates TpOs(N)(TFA)2 > TpOs(N)(TBA)2 >
TpOs(N)(TBA)2 > TpOs(N)(OAc)2 follow the electrochemi-
cal potentials but only within the acetate series; the oxalate
and nitrate complexes do not fit this trend.
Figure 4. log(kX/kTFA) vs Ep,c for the reactions of TpOs(N)X2 with PPh3
(A) and with BPh3 (B) (for compound numbers see Table 4).
additions of various phosphines to [(tpy)Os(N)Cl2]+, and Lau
et al. have examined how PPh3 addition to the nitrido ligand
in Os(N)(L)(Cl) is affected by different substituents on the
salen ligand L.18b The BPh3 reactions, however, have been
found to proceed by a two-step mechanism: initial rapid
equilibrium binding of the Lewis-acidic boron to the nitrido
lone pair is followed by aryl migration from boron to nitrogen
(eq 9).5b As previously described, the rate for this process
depends on both Keq and kmig (rate ) Keqkmig[TpOs(N)X2]-
[BPh3]). Aryl groups with higher migratory aptitudes react
faster with 1 because of their higher kmig, and more Lewis-
acidic boranes react faster because of their higher Keq.5b
Related nitrido-borane adducts have been reported, includ-
ing Re(NBPh3)(Et2dtc)2(PMe2Ph)19 and TpOs[NB(C6F5)3]-
Ph2.5b The multistep nature of this mechanism means that
its rate is unlikely to respond simply to changes in the
electrochemical potential of the starting osmium compound.
Increased electron density at osmium favors the initial
binding but should retard the aryl migration step.
The reactions of BPh3 and PPh3 both involve reduction
of Os(VI) to Os(IV), but only one correlates with one-
electron potentials. This is likely due to differences in the
reaction mechanisms. PPh3 addition to nitrido ligands most
likely occurs by direct attack of the nucleophilic phosphorus
on the nitrogen, at the M-N π* LUMO, as we have
suggested for 15b and as others have proposed in related
systems.18 For instance, Meyer et al. have examined18c
Conclusions
Changing the ancillary ligand X from halide to various
carboxylates to nitrate changes the properties of the osmium-
(18) (a) Reference 8, p 245. (b) Wong, T. W.; Lau, T. C.; Wong, W. T.
Inorg. Chem. 1999, 38, 6181-6186. (c) Demadis, K. D.; Bakir, M.,
Klesczewski, B. G.; Williams, D. S.; White, P. S.; Meyer, T. J. Inorg.
Chim. Acta 1988, 270, 511-526.
(19) Ritter, S.; Abram, U. Inorg. Chim. Acta 1995, 231, 245.
Inorganic Chemistry, Vol. 42, No. 2, 2003 609

Dehestani et al.
was stripped to dryness, yielding orange 2 (0.190 g, 0.355 mmol,
87%). H NMR (CD2Cl2): 6.16 (t), 7.57 (d), 7.58 (d) (each 1H,
(VI) nitrido complexes TpOs(N)X2. The irreversible poten-
tials for reduction to Os(V) vary substantially, by 0.63 V,
but the IR, 1H and 15N NMR, and UV-vis spectra are little
affected. Thus, the substituents seem to move the whole
orbital manifold without strongly affecting the spectroscopic
properties of the OstN triple bond. The compounds are all
reduced to osmium(IV) by PPh3, to form phosphiniminato
complexes TpOs(NPPh3)X2, and by BPh3, yielding boryl-
anilido complexes TpOs[N(BPh2)Ph]X2. Similar rate con-
stants are observed for all of the compounds, despite the large
range in reduction potential. The insensitivity of the two-
electron reactions to one-electron redox potential is note-
worthy. The relative rate constants for the PPh3 reactions
correlate well with the cathodic peak potentials, but those
for the BPh3 reactions do not, because of the multistep nature
of the mechanism for the latter reactions.
1
pz); 6.44 (t), 7.85 (d), 8.21 (d) (each 2H, pz); 2.23 (s) (6H, CH3).
ES/MS: 538 (M+). IR: 1087 cm-1 (νOst14N); νOst15N, 1050
cm-1 (calculated 1051 cm-1). UV/vis: 440 (314), 260 (7500), 230
(11000). Anal. Found: C, 29.35; H, 3.07; N, 18.20. Calcd for
C13H16BN7O4Os: C, 29.17; H, 3.01; N, 18.32.
TpOs(N)(TFA)2 (3). A solution of trifluoroacetic acid (144 µL,
1.87 mmol, 10 equiv) and 2 (0.100 g, 0.187 mmol) in 10 mL of
acetone was stirred for 1 h in air. The solvent solution was removed
in vacuo and the residue redissolved in 2 mL of CH2Cl2 and
chromatographed (silica gel, CH2Cl2/hexane (80/20) as the eluant).
An orange band was collected, and orange 3 (0.09 g, 0.140 mmol,
74% yield) was obtained by vapor diffusion of hexane into a CH2-
Cl2 solution. 1H NMR (CD2Cl2): 6.20 (t), 7.52 (d), 7.64 (d) (each
1H, pz); 6.54 (t), 7.95 (d), 8.15 (d) (each 2H, pz). ES/MS: 645
(M+). IR: 1081 cm-1 (νOst14N); νOst15N, 1042 cm-1 (calculated
1044 cm-1). UV/vis: 440 (265), 260 (7500), 230 (11000). Anal.
Found: C, 24.19; H, 1.47; N, 15.10. Calcd for C13H10BF6N7O4Os:
C, 24.27; H, 1.57; N, 15.24.
Experimental Section
Materials. All reactions were conducted under air unless stated
otherwise. Anhydrous hexane was distilled from Na/benzophenone,
while CH2Cl2 and CH3CN were distilled from CaH2. CD2Cl2 and
CD3CN were dried in the same manner and were stored over 4 Å
molecular sieves. TpOs(N)Cl2 and TpOs(15N)Cl2 were prepared
according to literature procedures.5b BPh3 (Aldrich) was sublimed
under static vacuum. Other reagents were purchased from Aldrich
and used without further purification.
Preparations of 4-10 all followed the procedure described
above for 3 using different acids. TpOs(N)(TCA)2 (4). Yield )
75%. H NMR (CD2Cl2): 6.17 (t), 7.63 (d), 7.68 (d) (each 1H,
1
pz); 6.55 (t), 7.95 (d), 8.20 (d) (each 2H, pz). ES/MS: 744 (M+).
IR: 1080 cm-1 (νOst14N); νOst15N, 1042 cm-1 (calcd 1043
cm-1). UV/vis: 425 (350), 260 (8400), 230 (12000). Anal. Found:
C, 21.20; H, 1.45; N, 13.12. Calcd for C13H10BCl6N7O4Os: C,
1
21.04; H, 1.36; N, 13.21. TpOs(N)(TBA)2 (5). Yield ) 63%. H
1
Instrumentation and Measurements. H NMR spectra were
NMR (CD2Cl2): 6.19 (t), 7.63 (d), 7.83 (d) (each 1H, pz); 6.54 (t),
7.95 (d), 8.21 (d) (each 2H, pz). ES/MS: 1010 (M+). IR: 1082
cm-1 (νOst14N); νOst15N, 1045 cm-1 (calcd 1044 cm-1). UV/
vis: 440 (300), 260 (11000), 230 (19000). Anal. Found: C, 15.53;
H, 1.00; N, 9.52. Calcd for C13H10BBr6N7O4Os: C, 15.48; H, 1.07;
N, 9.72. TpOs(N)Br2 (6). Yield ) 62%. 1H NMR (CD2Cl2): 6.03
(t), 7.47 (d), 7.52 (d) (each 1H, pz); 6.55 (t), 7.91 (d), 8.46 (d)
(each 2H, pz). ES/MS: 1010 (M+). IR: 1068 cm-1 (νOst14N);
νOst15N, 1032 cm-1 (calcd 1031 cm-1). UV/vis: 475 (400), 260
(10000), 240 (13000). Anal. Found: C, 19.00; H, 1.89; N, 16.67.
Calcd for C9H10BBr2N7Os: C, 18.73; H, 1.75; N, 16.99. TpOs-
(N)(O2C2O2) (7). Yield ) 63%. 1H NMR (CD2Cl2): 6.19 (t), 7.63
(d), 7.83 (d) (each 1H, pz); 6.54 (t), 7.95 (d), 8.21 (d) (each 2H,
pz). ES/MS: 1010 (M+). IR: 1070 cm-1 (νOst14N); νOst15N,
1034 cm-1 (calcd 1033 cm-1). In the preparation of 7, the solution
was heated at 50 °C for 1 h to speed up the rate. UV/vis: 440
(310), 260 (8000), 230 (11000). Anal. Found: C, 26.03; H, 1.86;
N, 19.33. Calcd for C11H10BN7O4Os: C, 26.15; H, 1.99; N, 19.41.
TpOs(N)(ONO2)2 (8). Yield ) 42%. 1H NMR (CD2Cl2): 6.19 (t),
7.63 (d), 7.83 (d) (each 1H, pz); 6.54 (t), 7.95 (d), 8.21 (d) (each
2H, pz). ES/MS: 1010 (M+). IR: 1059 cm-1 (νOst14N); νOst
15N, 1024 cm-1 (calcd 1023 cm-1). UV/vis: 435 (330), 260 (10000),
230 (12000). Anal. Found: C, 19.79; H, 1.70; N, 23.20. Calcd for
C9H10BN9O6Os: C, 19.97; H, 1.86; N, 23.29.
recorded on Bruker AF-300, AM-499, and WM-500 spectrometers
at ambient temperatures and referenced to Me4Si or a residual
solvent peak: δ (multiplicity, coupling constant, number of protons,
assignment). All pyrazole JHH are 2 Hz. 15N NMR spectra were
recorded on a Bruker DMX-750 spectrometer at ambient temper-
15
ature and referenced using an external aqueous solution of 15NH4
-
NO3. IR spectra were obtained as KBr pellets using a Perkin-Elmer
1600 series FTIR spectrometer at 4 cm-1 resolution. Each complex
contains a number of stretches associated with the Tp ligands that
typically vary less than 3 cm-1: 982 (w), 3112 (m), and 1503 (m),
1075 (m), 657 (m), 615 (m), 1404 (s), 1312 (s), 1216, (s), 1118
(s), 768 (s), 715 (s), and 1046 (vs). Electronic absorption spectra
were acquired with a Hewlett-Packard 8453 diode array UV-visible
spectrophotometer in anhydrous CH2Cl2 and are reported as λ/nm
(ꢀ/M-1 cm-1).
Cyclic voltammograms were measured with a BAS model 100
potentiostat using a three-compartment cell with a Ag/AgNO3 (0.1
M in CH3CN) reference electrode, platinum disk working electrode,
and platinum wire auxiliary electrode. The electrolyte solution was
0.1 M Bu4NPF6 (recrystallized from ethanol) in CH3CN. The
solution in the working compartment was deoxygenated and kept
under an atmosphere of N2. Potentials are reported versus internal
Cp2Fe+/0. Electrospray ionization mass spectra (ESI/MS) were
obtained on a Applied Biosystems Mariner API-TOF mass spec-
trometer. Samples were introduced in a CH3CN/H2O mixture with
a nozzle potential set at 140 V. Elemental analyses were performed
by Atlantic Microlabs.
TpOs{N(Ph)BPh2}(OAc)2 (9). A solution of 2 (0.100 g, 0.205
mmol), BPh3 (49 mg, 0.205 mmol), and 20 mL of CH2Cl2 was
stirred for 10 min and then run down a column of silica gel with
70/30 CH2Cl2/hexane, yielding 9 as a dark orange solid (0.139 g,
0.178 mmol, 87%). 1H NMR (CD2Cl2): 5.92 (1H, t, pz), 6.00 (1H,
t, pz′), 6.39 (1H, t, pz′′), 6.07, 6.52 (each 1H, d, pz), 5.50, 7.14
(each 1H, d, pz′), 7.41, 7.53 (each 1H, d, pz′′); 4.92 (d, 7 Hz, 2H,
NPh0), 7.02 (t, 7 Hz, 2H, NPhm), 6.61 (t, 7 Hz, 1H, NPhp), (d, 7
Hz, 4H, BPh2ortho), (d, 7 Hz, 4H, BPh2meta), (d, 7 Hz, 2H, BPh2para);
3.28 (s, 3H, CH3), 2.05 (s, 3H, CH3). ES/MS: 780 (M+). UV/vis:
450 (5700), 365 (6100), 325 (8100), 270 (13700). Anal. Found:
TpOs(N)(OAc)2 (2). A 200 mL round-bottom flask was charged
with 1 (0.200 g, 0.410 mmol), AgOAc (0.480 g, 2.87 mmol), and
50 mL of acetonitrile. The flask was connected with a condenser
and heated with stirring in an 85 °C oil bath for 12 h. The solution
was cooled to room temperature, filtered to remove AgOAc and
AgCl, and reduced to 5 mL. Chromatography on silica gel with
CH2Cl2/acetone (95/5) as the eluant gave an orange band, which
610 Inorganic Chemistry, Vol. 42, No. 2, 2003

Properties of the Osmium Nitrido Group in TpOs(N)X2
C, 47.52; H, 4.76; N, 11.95. Calcd for C29H31B2N7O4Os‚C3H6O:
C, 47.36; H, 4.60; N, 12.08 (1 mol of acetone is observed in H
NMR spectrum was taken to confirm the formation of 9, the
complete depletion of 2, and the small excess of BPh3.
1
NMR spectra after recrystallization).
Crystal Structures of 3 and 8. Both crystals were grown by
slow evaporation of hexane/methylene chloride solutions. Crystals
were mounted on a glass fiber with epoxy. Data were collected on
a Nonius KappaCCD diffractometer. Data for 3 was collected at
20 °C by rotation about the φ axis in 0.8° increments over 129.9°.
8 was cooled to -143 °C, and then the data was collected by
rotation about the φ axis in 1° increments over 179.9°. The exposure
times per frame and the crystal-detector distances were 40 s and
30 mm for both 3 and 8. The merging R factors of 0.0933 (3) and
0.1701 (8) indicated that the data were not of good quality. Solutions
by direct methods produced a complete phasing model for 3 and 8
with the rest located by difference Fourier synthesis. All heavy
atoms were refined anisotropically by full-matrix least-squares. All
hydrogen atoms were located using a riding model.
Competition Studies. All manipulations were done in a glovebox
with anhydrous CD2Cl2. In a typical procedure, four 2 mL
volumetric flasks were charged with 3 (7 mg, 11 µmol,), 4 (8 mg,
11 µmol), BPh3 (11 mg, 45 µmol), and C6Me6 (40 mg, 0.247 mmol),
respectively. The volumetric flasks containing C6Me6 and BPh3
were each taken up with 2 mL of CD2Cl2. The C6Me6 solution (50
µL, 6 µmol) was added to the flasks with 3 and 4, and each was
taken up to 2 mL with CD2Cl2, giving solutions 3.1 mM in C6Me6
and 5.5 mM in 3 or 4. Three separate sealable NMR tubes were
charged with (i) 0.350 mL of the solution of 3, (ii) 0.350 mL of
the solution of 4, and (iii) 0.175 mL of each of the solutions of 3
and 4. 1H NMR spectra were taken, and the tubes were returned to
the glovebox. One equivalent of BPh3 (86 µL) was added to tubes
i and ii. BPh3 (0.5 equiv, 43 µL) was added to tube iii, tilting the
NMR tube sideways to make sure none of the BPh3 solution touched
the osmium solutions. The NMR tubes were then mixed vigorously
and quickly. Integration of the NMR spectra gave the product yields.
Kinetic Studies. All manipulations were done in a glovebox.
Two separate 25 mL volumetric flasks were charged with 2 (23
mg, 43 µmol) and BPh3 (11 mg, 45 µmol) and brought up to 25
mL with CH2Cl2 (1.7 mM 2, 1.8 mM BPh3). A quartz UV cell was
charged with the solution of 2 (1.5 mL, 0.0025 mmol) and covered
with a septum. The BPh3 solution (1.5 mL, 0.0027 mmol) was added
to the cuvette by gastight syringe, the solution was mixed, and
absorption spectra were taken every second for 5 min. The solution
was removed in vacuo and taken up in anhydrous CD2Cl2, and a
Acknowledgment. We are grateful to Dr. Nancy Doherty
(University of California at Irvine) for helpful discussion on
15N NMR. We thank Dr. Ross Lawrence for mass spectrom-
etry, and Dr. Joanna Long for 15N NMR experiments. We
are grateful to the National Science Foundation for financial
support. We thank Dr. Tom Crevier and Tara Conrad for
initial studies on TpOs(N)(OAc)2.
Supporting Information Available: Complete crystallographic
data (CIF) for complexes 3 and 8. This material is available via
the Internet at http://pubs.acs.org.
IC0260264
Inorganic Chemistry, Vol. 42, No. 2, 2003 611