Reactivity of the Hydrido/Nitrosyl Radical MHCl(NO)(CO)(P iPr3)2, M = Ru, Os

Article abstract of DOI:10.1021/ic0349407

The reaction of equimolar NO with the 16 electron molecule RuHCl(CO)L ₂ (L = PⁱPr₃) proceeds, via a radical adduct RuHCl(CO)(NO) L₂, onward to form RuCl(NO)(CO)L₂ (X-ray structure determination) and RuHCl(HNO)(CO)L₂, in a 1:1 mole ratio. The HNO ligand, bound by N and trans to hydride, is rapidly degraded by excess NO. The osmium complex behaves analogously, but the adduct has a higher formation constant, permitting determination of its IR spectrum; both MHCl(CO)(NO)L₂ radicals are characterized by EPR spectroscopy, and DFT calculations on the Ru system show it to have a half-bent Ru-N-O unit with the spin density mainly on nitrogen. DFT (PBE) energies rule out certain possible mechanistic steps for forming the two products. A survey of the literature leads to the hypothesis that NO should generally be considered as a (neutral) Lewis base (2-electron donor) when it binds to a 16 electron complex which is resistant to oxidation or reduction, and that the resulting N-centered radical has a M-N-O angle of ～140°, which distinguishes it from NO^- (bent at <140°) and from NO ⁺ (>170°).

Full text of DOI:10.1021/ic0349407

Inorg. Chem. 2004, 43, 351−360
i
Reactivity of the Hydrido/Nitrosyl Radical MHCl(NO)(CO)(P Pr3)2, M )
Ru, Os
Alexei V. Marchenko, Andrei N. Vedernikov, David F. Dye, Maren Pink, Jeffrey M. Zaleski,* and
Kenneth G. Caulton*
Indiana UniVersity, Department of Chemistry, Bloomington, Indiana 47405
Received August 6, 2003
i
The reaction of equimolar NO with the 16 electron molecule RuHCl(CO)L2 (L ) PPr3) proceeds, via a radical
adduct RuHCl(CO)(NO) L2, onward to form RuCl(NO)(CO)L2 (X-ray structure determination) and RuHCl(HNO)-
(CO)L2, in a 1:1 mole ratio. The HNO ligand, bound by N and trans to hydride, is rapidly degraded by excess NO.
The osmium complex behaves analogously, but the adduct has a higher formation constant, permitting determination
of its IR spectrum; both MHCl(CO)(NO)L2 radicals are characterized by EPR spectroscopy, and DFT calculations
on the Ru system show it to have a “half-bent” Ru−N−O unit with the spin density mainly on nitrogen. DFT (PBE)
energies rule out certain possible mechanistic steps for forming the two products. A survey of the literature leads
to the hypothesis that NO should generally be considered as a (neutral) Lewis base (2-electron donor) when it
binds to a 16 electron complex which is resistant to oxidation or reduction, and that the resulting N-centered radical
+
has a M−N−O angle of ∼140°, which distinguishes it from NO- (bent at <140°) and from NO (>170°).
Introduction
complexes HMLn (generally 18 valence electrons) nearly
always have an even electron count, any adduct HM(NO)Lq
cannot have a 16 or 18 valence electron count. In the rare
case of the hydride-alternative pure σ ligand CH3, W(CH3)6
reacts with excess NO to give9 (η2-MeN(NO)O)2WMe4 so
here an even electron count is achieved by adding two NO
per methyl modified.
We study here one new aspect of this subject in that we
bind one NO to a 16 electron, Ru(II) complex, but the reagent
complex is a monohydride. The hydride ligand is always a
reactive functionality, and here we find that it is destined to
react by hydrogen atom transfer, since that redox chemistry
is a way to move from radical intermediates to even-electron
products, a path that conventional wisdom suggests should
be thermodynamically favorable. This is found to offer a
new synthesis of coordinated nitroxyl, HNO. The reason that
this is an unprecedented route to an HNO ligand is that, in
the past, the NO ligand has been installed on a metal first,
and then, some form of H, either H+ 10 or H-,11,12 is furnished
in a bimolecular reaction. We also contrast the NO reaction
chemistry of RuHCl(CO)L2 with that of its osmium analogue,
which lengthens the lifetime of one intermediate to permit
its infrared spectral detection.
The ways in which a formally 19 valence electron
molecule responds to minimize the instability associated with
this electron count are revealing. They reinforce the signifi-
cance of the 18 electron rule, but they show unusual bonding
and/or redox alternatives (isomers) and thus they can inform
us about the reactivity consequences of a 19 valence electron
count. The reactivity patterns of radical organometallic
species are currently underdeveloped.1-6
For binary metal carbonyls, NO reacts to replace an even
number of electrons by virtue of two NO replacing three
conventional Lewis bases, often CO.7,8 The reactions of NO
with transition metal hydride complexes are not simple
substitutions, and invariably lead to the “disappearance” of
the hydrides, either to identifiable H2 product (when the
reagent complex is a dihydride), or (for monohydrides) to
an unspecified fate.7,8 The latter occurs most often when a
monohydride complex is involved. Since monohydride
* To whom correspondence should be addressed. E-mail: caulton@
indiana.edu (K.G.C.); zaleski@indiana.edu (J.M.Z.).
(1) Tyler, D. R. Prog. Inorg. Chem. 1988, 36, 125.
(2) Tyler, D. R.; Mao, F. Coord. Chem. ReV. 1990, 97, 119.
(3) Tyler, D. R. Acc. Chem. Res. 1991, 24(11), 325.
(4) Schut, D. M.; Keana, K. J.; Tyler, D. R.; Rieger, P. H. J. Am. Chem.
Soc. 1995, 117(35), 8939.
(5) Braden, D. A.; Tyler, D. R. J. Am. Chem. Soc. 1998, 120 (5), 942.
(6) Braden, D. A.; Tyler, D. R. Organometallics 1998, 17 (18), 4060.
(7) Hayton, T. W.; Legzdins, P.; Sharp, W. B. Chem. ReV. 2002, 102,
935.
(9) Shortland, A. J.; Wilkinson, G. J. Chem. Soc., Dalton Trans. 1973,
872.
(10) Southern, J. S.; Green, M. T.; Hillhouse, G. L.; Guzei, I. A.; Rheingold,
A. L. Inorg. Chem. 2001, 40 (23), 6039.
(11) Sellmann, D.; Blum, N.; Heinemann, F. W.; Hess, B. A. Chem.s
Eur. J. 2001, 7, 1874.
(12) Sellman, D.; Sutter, J. Z. Anorg. Allg. Chem. 2003, 629, 893.
(8) Richter-Addo, G. R.; Legzdins, P. Oxford University Press: New York,
1992.
10.1021/ic0349407 CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/10/2003
Inorganic Chemistry, Vol. 43, No. 1, 2004 351

Marchenko et al.
Scheme 1
Figure 1. (a) ORTEP drawing (50% probability) of Ru(NO)Cl(CO)(Pi-
Pr3)2 omitting alkyl carbons and attached hydrogen. (b) DFT-optimized
geometry of RuHCl(NO)(CO)(PiPr3)2 with iPr carbons and hydrogens
omitted.
Table 1. Selected Bond Distances and Angles for
RuCl(NO)(CO)(PiPr3)2
Ru(1)
Ru(1)
Ru(1)
Ru(1)
C(11)
N(1)
C(11)
N(1)
P(1)
Cl(1)
O(2)
O(1)
1.7814(19)
1.8566(11)
2.4117(2)
2.4905(5)
1.149(2)
perature is an effective way to clarify this point. These experi-
ments, in d8-toluene beginning at -70°, show that, even at
low temperature, the mole fraction conversion of RuHCl-
(CO)L2 by substoichiometric NO (2:1 Ru to NO ratio) is
only small, leading to slight broadening of the signals of
RuHCl(CO)L2, but no NMR signals detectable for the 1:1
adduct (hence paramagnetic) RuHCl(NO)(CO)L2. Spectral
monitoring, as the temperature is raised in 10° increments,
shows growth of the equimolar diamagnetic products shown
in Scheme 1.
While 3 is the product identified in the room temperature
reaction in the presence of excess NO, 2 is new, and because
it contains two hydrogens derived from RuHCl(CO)L2, it
establishes material balance in the reaction. The most
interesting spectral features of 2 are a hydride multiplet
(triplet of doublets) and a resonance at 20.9 ppm character-
istic of an HNO ligand.10 At -20 °C, the coordinated HNO
proton signal is a doublet, due to coupling with the hydride
trans to itself. Compound 2 persists at 23 °C for more than
12 h provided the NO concentration is kept very low. When
RuHCl(HNO)(CO)L2 does slowly transform, due to even
trace free NO, NMR and IR evidence shows HONO (+6.2
ppm, but variable, due to hydrogen bonding; 3642 cm-1)
and N2O (2210 cm-1).
1.1386(18)
C(11)
C(11)
N(1)
P(1)a
C(11)
N(1)
P(1)a
P(1)
O(2)
C(1)
C(4)
C(7)
O(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
C(11)
P(1)
N(1)
98.92(7)
89.32(7)
P(1)
P(1)
100.448(5)
159.105(11)
160.27(7)
100.796(12)
89.715(13)
86.391(13)
172.9(2)
111.76(3)
114.35(3)
112.31(3)
138.75(11)
P(1)
Cl(1)
Cl(1)
Cl(1)
Cl(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
Ru(1)
P(1)
P(1)
N(1)
The results reported here fall into two distinct categories:
(1) What is the nature of bonding and charge distribution
when the odd electron species NO forms a 1:1 adduct with
16 valence electron species MHCl(CO)(PiPr3)2? (2) How
does the hydride play a role in the subsequent reaction
chemistry of that adduct?13
Results
The reaction of RuHCl(CO)L2 with NO is one whose
outcome depends critically on the presence or absence of
free NO, and thus on the efficiency of mixing and on overall
stoichiometry. An effective way to avoid excess NO, which
we show below would take the primary product on to
secondary products, is to employ a stoichiometry of less than
one NO per Ru. In this way, and beginning at -70 °C in
The nitroxyl group is sensitive to conversion, by excess
free NO, as already established (eq 1).14,15 Thus, when
RuHCl(CO)L2 is combined with excess NO even at -198
°C in d8-toluene, then slowly thawed and mixed, already at
-60 °C there is essentially complete conversion to RuCl-
(NO)(CO)L2 as the only metal complex product.
d8-toluene, one can see, by recording 31P and H NMR
spectra, the production of essentially equimolar Ru(NO)Cl-
(CO)L2 and RuHCl(HNO)(CO)L2.
1
HNO + 2NO f HONO + N2O
(1)
If, in contrast, RuHCl(CO)L2 (L ) PiPr3) is reacted with
NO without exceptionally careful avoidance of excess NO,
the product is Ru(NO)Cl(CO)L2 alone. This square pyramidal
(Figure 1a and Table 1) molecule has a bent NO in the apical
coordination site and is thus isoelectronic with the hydride
reagent. Apparently this structure is more stable than a satur-
ated alternative with linear coordination of the nitrosyl.
Despite the structural similarity to the hydride reagent, this
product alone fails to account for the fate of the hydride li-
gand. As stated above, monitoring the reaction at lower tem-
Additional evidence for the formation of the described pro-
ducts is obtained by using IR spectroscopy. When RuHCl-
(CO)(PiPr3)2 is reacted with NO in C6D6, the reaction mixture
shows two sets of new bands, at 1911 and 1582 cm-1, at-
tributed to ν(CO) and ν(NO) of Ru(NO)(CO)Cl(PiPr3)2 which
are in good agreement with the previously reported data for
16
the analogous complex Ru(NO)(CO)Cl(PtBu2Me)2 and at
1943 and 1392 cm-1, for RuH(HNO)(CO)Cl(PiPr3)2. The in-
(14) Bunte, S. W.; Rice, B. M.; Chabalowski, C. F. J. Phys. Chem. 1997,
101, 9430.
(15) Lee, T. J.; Dateo, C. E. J. Chem. Phys. 1995, 103, 9110.
(13) Marchenko, A. V.; Vedernikov, A. N.; Dye, D. F.; Pink, M.; Zaleski,
J. M.; Caulton, K. G. Inorg. Chem. 2002, 41, 4087.
352 Inorganic Chemistry, Vol. 43, No. 1, 2004

Hydrido/Nitrosyl Radical MHCl(NO)(CO)(PiPr3)2
Figure 2. Observed (s) and calculated (- - -) X-band EPR spectra of RuHCl(NO)(CO)(PiPr3)2 (a) and its RuD analogue (b) in toluene at 77 K.
crease in ν(CO) for 3 in comparison to RuHCl(CO)(PiPr3)2,
1910 cm-1, is due to a weaker donor power of the nitrosyl
ligand in comparison to the hydride, and the higher ν(CO)
and lower ν(NO) for the nitroxyl complex can be explained
by the substantial back-donation from the metal to the ni-
troxyl ligand, and participation of RuH(dN+(H)sO-)(CO)Cl-
(PiPr3)2 resonance structure. Similar evidence for the π-ac-
cepting ability of the nitroxyl ligand and IR data has been
reported recently.10
bends, in which case it is a 17 electron complex (B). Nineteen
The X-ray structural study shows RuCl(NO)(CO)L2 (Fig-
ure 1a) to have a square pyramidal structure with an apical
bent nitrosyl. It is therefore unsaturated and Ru(II) rather
than the linear RuNO 18 electron alternative, which would
be Ru(0). All cis apical/basal angles are 99° ( 1°, and trans
basal/basal angles are 160° ( 1° The similarity of all Ru-
P-C angles rules out any agostic interaction trans to the
nitrosyl ligand. The NO group bends in the Cl-Ru-CO
plane, although the disorder (crystallographic C2 symmetry
along Ru1-N1) prevents knowing whether the nitrosyl O
bends toward CO or toward Cl. The Ru-NO distance is
longer (by 0.075 Å or 25σ) than the Ru-CO distance,
consistent with a lower Ru-N bond order.
If 1 is indeed a good H-atom acceptor (to form 2), it might
be possible to intercept it with a H-atom donor, and thus
divert the product from 3 (a hydrogen-loss product) toward
2. Reaction of RuHCl(CO)L2 with NO (1:0.66 mole ratio)
at 20 °C in C6D6, in the presence of 20 equiv of 1,4-cyclo-
hexadiene as a H atom donor, shows, within 5 min of com-
bining the reagents, the formation of equimolar RuHCl(H-
NO)(CO)L2 and RuCl(NO)(CO)L2, together with 25% un-
reacted RuHCl(CO)L2 (confirming the intended deficiency
of NO); the yield of 2 is not enhanced. Under these con-
ditions, it is thus impossible to trap any metal containing
radical. The product distributions are the same even when
the reaction is run in neat 1,4-cyclohexadiene. No products
of any transformation (e.g., polymerization) of the diene were
observed, thus indicating the highly selective reactivity of
free radicals involved in this metal hydride chemistry. Further
trapping experiments with the weaker Sn-H bond of (n-
Bu)3SnH fail because this molecule reacts rapidly with
RuHCl(CO)L2.
electron complexes have been intensely studied among
carbonyl complexes,1-6 and the outcome is that the popula-
tion of an M-ligand antibonding orbital is avoided by
bending one MCO unit, and with radical character then
residing primarily on that carbonyl. That carbonyl is then a
one electron donor to the metal, so a 19 valence electron
configuration at the metal is avoided. In effect, a metalla-
formyl radical is created. Another way to synthesize such a
species is by binding CO to an odd electron complex, e.g.,
RhII (porphyrin) + CO.17 This additional option, in com-
parison to A and B above, permits the metal to achieve an
18 electron configuration, as in C. To the extent that the 18
electron rule truly semiquantitatively predicts reaction ener-
gies and “stability,” C may be the best representation of
reality. We next review additional spectroscopic evidence.
Under low temperature EPR conditions (77 K), the equi-
librium position is shifted enough toward adduct that the
(broad) EPR signal of free NO is not detected. The EPR
1
On the surface, compound 1 might be considered as an
analogue of 18 electron RuHCl(CO)2L2, but with one more
electron, and one proton added to the carbon nucleus. It is
thus a 19 valence electron complex (A) unless the nitrosyl
spectrum of 1 (Figure 2a) reveals an S ) /2 spin system
with near axial symmetry (gx ≈ gy > gz) and simulated spin
Hamiltonian parameters gx ) 2.006, gy ) 1.993, and gz )
1.910 which were obtained using Monte Carlo methods.18
Inorganic Chemistry, Vol. 43, No. 1, 2004 353

Marchenko et al.
a
Strong nitrogen (I ) 1, NO) hyperfine coupling (AxN ) 34.5
G) splits gx into a 3-line pattern. Comparable superhyperfine
coupling of the hydride ligand trans to the nitrogen further
splits the three line pattern of gx (AxH ) 36.5 G) as well as
gy (AyH ) 35.1) generating multiple overlapping signals
which are responsible for the four line pattern observed at g
≈ 2. Additional minor coupling of the electron spin to NO
and resolved coupling to hydride are also observed on the
remaining axes (AyN ) 3.9 G, AzN ) 1.7 G, AzH ) 35.8 G).
The EPR spectrum of the deuteride analogue (Figure 2b)
confirms these assignments as the newly refined g-values
are conserved (gx ) 2.001, gy ) 1.994, and gz ) 1.910) and
only strong coupling to NO is observed (AxN ) 34.3 G, AyN
) 4.9 G, AzN ) 4.5 G, AxD ) 2.1 G, AyD ) 2.1 G, AzD ) 1.9
G) since the ratio γD / γH ) 0.15 makes coupling to
deuterium unresolved. The large nitrogen hyperfine coupling
constants, low g-values relative to Ru(III) EPR signals,19 and
strong similarities of the spin Hamiltonian parameters to well
characterized Ru-NO‚ species20-22 confirm that the unpaired
electron resides primarily on the NO ligand.
Scheme 2
a ∆G°298 in kcal/mol; L ) PMe3.
The observed reactivity of RuHCl(CO)L2 toward NO
appears to be relatively insensitive to the electron donating
ability of the phosphine ligand L. When L ) PiPr3 in RuHCl-
(CO)L2 is replaced by a significantly less electron-donating
phosphine PiPr2(3,5-(CF3)2C6H3), and this complex23 is
combined with NO under the conditions of slight deficiency
of NO, the NMR indicates the presence of RuCl(NO)(CO)-
(PiPr2(3,5-(CF3)2C6H3))2 and RuHCl(HNO)(CO)(PiPr2(3,5-
(CF3)2C6H3))2 in 1:1 ratio within 10 min after mixing.
Mechanistic Insights from DFT Calculations. The key
event in the mechanism in Scheme 1 is H atom transfer from
Ru to N. This is necessary to form a nonhydride product,
and to eliminate odd-electron species. The rest are “merely”
steps forming Ru-N bonds. DFT (PBE) reaction energies,
calculated on a model with PMe3 ligands, permit exclusion
of certain homolytic elementary processes as being too
endergonic to account for the observed rates.24,25
There are two different metal hydrides that are candidates
for serving as the hydrogen atom donor. The reagent hydride
complex is not a viable source for H-atom transfer (eq 2)
because ∆H°298, its bond dissociation enthalpy, is too large.
The DFT energies (kcal/mol) of the two homolytic dissocia-
tions show the 19 electron radical (eq 3) to be the
thermodynamically preferred H atom donor; this ranking
certainly rests on the greater stability of RuCl(NO)(CO)L2
(eq 3), an observed product, than that of the unknown 15
valence electron radical RuCl(CO)L2 (eq 2).
The two candidates to abstract a hydrogen atom are NO
itself (eq 4) and RuHCl(NO)(CO)L2 (eq 5), and the N-H
bond formation energies of these two are similar, although
H transfer to free NO is slightly less preferred at this level
of calculation. Even at high computational levels,26,27 the
(16) Ogasawara, M.; Huang, D.; Streib, W. E.; Huffman, J. C.; Gallego-
Planas, N.; Maseras, F.; Eisenstein, O.; Caulton, K. G. J. Am. Chem.
Soc. 1997, 119, 8642.
(17) Wayland, B. B.; Sherry, A. E.; Bunn, A. G. Inorg. Chem. 1981, 20,
3093.
(18) Neese, F. QCPE Bull. 1995, 5, 15.
(19) Munshi, P.; Samanta, R.; Lahiri, G. K. J. Organomet. Chem. 1999,
586, 176.
(20) Sellmann, D.; Gottschalk-Gaudig, T.; Haussinger, D.; Heinemann, F.
W.; Hess, B. A. Chem.s Eur. J. 2001, 7, 2099.
N-H bond in free HNO was found to be very weak (43.6
kcal/mol27).
On the basis of these preliminaries, the two mechanisms
in Scheme 2 (which shows standard free energies) can be
evaluated.24,25 Mechanism B is “direct” in that H-atom
transfer between two identical metal complexes produces
both observed products in a single step (i.e., a single bond
rupture). Mechanism A has more steps, and ends with free
HNO binding to available RuHCl(CO)(PMe3)2 to form the
second observed product. Alternatively, the HNO could serve
(21) McGarvey, B. R.; Ferro, A. A.; Tfouni, E.; Bezerra, C. W. B.; Bagatin,
I.; Franco, D. W. Inorg. Chem. 2000, 39, 3577.
(22) Wanner, M.; Scheiring, T.; Kaim, W.; Slep, L. D.; Baraldo, L. M.;
Olabe, J. A.; Zalis, S.; Baerends, E. J. Inorg. Chem. 2001, 40, 5704.
(23) Marchenko, A. V.; Huffman, J. C.; Valerga, P.; Jimenez Tenorio, M.;
Puerta, M. C.; Caulton, K. G. Inorg. Chem. 2001, 40, 6444.
(24) Koch, W.; Holthausen, M. C. A Chemist’s Guide to Density Functional
Theory; Wiley-VCH: Weinheim, 2001.
354 Inorganic Chemistry, Vol. 43, No. 1, 2004

Hydrido/Nitrosyl Radical MHCl(NO)(CO)(PiPr3)2
Table 2. Calculated (DFT) Vibrational Frequencies (cm-1
)
as an H-atom shuttle by transferring its H• (eq 6) to RuHCl-
(NO)(CO)(PMe3)2 ( ∆H°298 ) -10 kcal/mol).
that, in general, a ∼140° M-N-O should be taken as
possible evidence for nitrogen-centered radical Lewis base
behavior of neutral NO. There is good empirical evidence28
that this is the case for NO bound to a variety of unsaturated
d6 complexes (even when that yields only a 16 valence
electron species).
The Ru/N Bonds. By the criterion of RusN bond
dissociation enthalpy (BDE) of either NO (eq 7a) or HNO
(eq 7b) toward RuHCl(CO)(PMe3)2, there is a difference
between these two, but NO forms a satisfactory bond to
RuHCl(CO)(PMe3)2 despite the radical character of the
product.
However, the transition state for H-atom abstraction from
RuHCl(NO)(CO)L2 will certainly have less steric repulsion
when it is done by NO (A, Scheme 2) than by another
molecule of RuHCl(NO)(CO)L2 (B, Scheme 2), thus favoring
path A.
The structure of 1 was calculated (Figure 1b) by geometry
optimization of the DFT (PBE) energy of the full RuHCl-
(CO)(NO)(PiPr3)2 species. While it is certainly an octahedral
structure around Ru, the interesting features are the long Rus
N bond (2.015 Å, long even compared to the RusN distance,
1.857(1) Å in Ru(NO)(Cl)(CO)L2), and the bend at N. The
N/O distance, 1.187 Å, is longer than in free NO. Since the
spin density on Ru is only 0.04 e, and the NO carries the
other 0.87 e, the ground state of this adduct is truly a ligand-
centered radical. A Lewis structure consistent with all of the
above is D, which shows that adduct formation involves a
16 electron Ru achieving an 18 (not 19) electron configu-
ration by localizing the extra electrons on oxygen (note the
two lone pairs on oxygen). As a result, this N/O bond order
For comparison, the BDE of ammonia on the analogous
species (eq 7c) is 7 kcal/mol. The low value for binding
ammonia shows that, confronted with the trans effect of a
hydride ligand, this pure σ donor forms only a very weak
bond to RuHCl(CO)L2.
Trends in Vibrational Frequencies. The calculated CO
vibrational frequency stands as a useful gauge of π-basicity
at the metal, and thus of concepts such as metal oxidation
state. The three species involved in our reactivity studies
are assembled in Table 2 in a way to illustrate their being
related by formal hydrogen atom transfers. The H‚ in step a
is added to the metal, and the rise in νCO confirms that this
is a net oxidation of the metal. The H‚ added in step b goes
to nitrogen, and the constancy of νCO in this step indicates
that there is no significant redox change at the metal; more
generally, the π-basicities of Ru in the presence of radical
NO and of HNO are similar. The increase in the nitrosyl
stretching frequency in step a is in agreement with a change
NO- f NO‚ as Ru is oxidized, since the NO bond order
increases from free diatomic NO- to NO‚. In step b,
hydrogenation of coordinated NO‚ reduces the NO bond
order to ca. 2.
approximates 2, and the unpaired electron is localized heavily
on N (although this is really derived from an N/O π* orbital,
so there is some radical character on oxygen also). Simple
reasoning on degree of bonding versus occupancy of the lone
“pair” orbital on N suggests that the Ru-N-O angle will
be larger than the 120° of RNO when the nitrogen orbital is
only half occupied, hence the 143° angle here. We suggest
The Osmium Analogue. The reaction of NO with OsHCl-
(CO)L2 (L ) PiPr3) in benzene shows general similarities to
the Ru analogue, but with quantitative differences. Under
only 80 mm NO (0.5 mol/mol Os) in benzene at 23 °C, there
is essentially complete (stoichiometrically limited) formation
of OsHCl(NO)(CO)L2, which causes complete loss of the
(25) Cramer, C. J. Essentials of Computational Chemistry: Theories and
Models; J. Wiley: West Sussex, England, 2002.
(26) Tschumper, G. S.; Schaefer, H. F., III. J. Chem. Phys. 1997, 107,
2529.
(27) Mordaunt, D. H.; Flothmann, H.; Stumpf, M.; Keller, H.-M.; Beck,
C.; Schinke, R.; Yamashita, K. J. Chem. Phys. 1997, 107, 6603.
(28) Scheidt, W. R.; Ellison, M. K. Acc. Chem. Res. 1999, 32, 350.
Inorganic Chemistry, Vol. 43, No. 1, 2004 355

Marchenko et al.
the Ru analogue, but with a greater separation of the gz from
the gx and gy values. The superhyperfine parameters are
analogous to those for the Ru analogue, and as for Ru, any
coupling to P was unresolvable. Noteworthy is the significant
coupling to hydrogen, which has only a σ orbital. This is
useful experimental evidence that the SOMO cannot have
π*(NO) character (i.e., perpendicular to the MNO plane),
since that does not have the symmetry to mix with the
hydrogen orbital. The half filled N (sp2) orbital does have
symmetry to mix with a metal orbital and the H (1s) orbital,
to give the observed AH coupling. The calculated spin
densities for the model MHCl(NO)(CO)(PMe3)2 help account
for the observed superhyperfine coupling to the hydride
ligand. Thus, although the spin densities are mainly (Ru/Os
values given) on N (0.56/0.55), O (0.31/0.32), and M (0.04/
0.04), there is nonzero hydride participation in the SOMO
such that the hydride populations are 0.09 and 0.08. While
there is some unpaired spin on the metals, it is less than that
on the hydride, and the radical character at hydride is a
reminder that hydrogen atom transfer becomes an attractive
mode of reactivity. This is also why the RusH BDE (eq 3)
is so low. It must be recognized that the nonzero AH value
to hydride is direct evidence that the Ru-N-O unit is
nonlinear, since only then does the H 1s orbital mix with
nitrogen orbitals in the SOMO.
Figure 3. Observed (s) and calculated (- - -) X-band EPR spectra of
OsHCl(NO)(CO)(PiPr3)2 in toluene at 77 K: gx ) 1.998, gy ) 1.974, gz )
1.812, AxN ) 35.9 G, AyN ) 5.1 G, AzN ) 2.5 G, AxH ) 32.0 G, AyH ) 30.7
G, AzH ) 32.7 G.
1H NMR signals in the normal position of those of OsHCl-
(CO)L2, as well as complete loss of 31P {1H} NMR signals.
This higher equilibrium concentration of NO adduct is
consistent with the generally higher adduct formation
constant for Os versus Ru for nonradical Lewis bases. There
is also a weak and broad 1H NMR peak at 4.0 ppm attributed
to HONO. Upon addition of 1 atm (i.e., excess) NO to this
solution, 1H and 31P{1H} NMR spectra taken within 15 min
show OsHCl(HNO)(CO)L2 and OsCl(NO)(CO)L2 29 in a 1:4
mole ratio as the major products. The 4.0 ppm HONO peak
is now stronger and broader. The peak due to coordinated
nitroxyl is far downfield, 20.9 ppm, and is a doublet with
coupling identical to the coupling in the triplet of doublets
pattern due to hydride at -8.38 ppm; this identical doublet
coupling proves that these protons are both in the same
molecule and that there is only one of each type. Thus, a
nitroxyl ligand is also produced in the osmium case, and it
is subject to degradation by excess NO, but much more
slowly than in the case of ruthenium.
In this osmium system, the reaction of H atom abstrac-
tion is slower, and the initially formed radical OsH(NO‚)-
(CO)Cl(PiPr3)2 is present in higher mole fraction, even with
substoichiometric NO, at -60 °C. The radical complex is
therefore detectable by IR spectroscopy with ν(CO) at 1903
cm-1 and ν(NO) at 1559 cm-1. The increase in ν(CO) from
OsHCl(CO)(PiPr3)2 (1888 cm-1) to its NO adduct is ex-
plained by back-donation from the Os center to the NO
ligand; this also explains the longer life of this radical in
the case of Os since it is more reducing than Ru. Upon
addition of a larger amount of NO to the reaction mixture,
the formation of OsCl(NO)(CO) (PiPr3)2 and OsH(HNO)-
(CO)Cl(PiPr3)2 is confirmed by the appearance of ν(CO)
and ν(NO) of these products at 1892 and 1552 cm-1 for
the nitrosyl complex, and at 1929 and 1388 cm-1 for the
nitroxyl complex. These values are lower than for the
analogous Ru products due to the higher extent of the back-
donation from more reducing Os to CO, NO, and (HNO)
ligands.
Discussion
The collective results presented here show that MHCl-
(CO)L2 (M ) Ru, Os) is neither readily oxidizable nor
reducible by NO, thus frustrating one NO molecule from
behaving in its “typical” way upon encountering a metal (eq
8a,b).
NO f NO+ (linear) + e-
or e- + NO f NO- (bent)
(8a)
(8b)
The dominant reactivity of unsaturated MHCl(CO)L2 is
to find a conventional Lewis base, and it is in this way that
NO is forced to act, despite thus creating a diatomic ligand-
centered radical. Indeed, it has been concluded recently,30
for d6 metals, “NO behaves as a normal nucleophile in ligand
substitution reactions...even though it is a radical.”
A recent report shows how the reactivity shown in eq 8b
can lead to porphyrin ligand oxidation, which is then manifest
as C-C bond formation to a second porphyrin com-
plex.31
The hydride chemical shift of RuHCl(HNO)(CO)L2 is
sufficiently far downfield from that of RuHCl(CO)L2 that it
indicates a “strong” ligand trans to itself. The nitroxyl proton
chemical shift is also very far downfield at 20 °C, suggesting
some strong influence by the metal (e.g., anistropic deshield-
ing). Taken together, and arguing as we have earlier about
The EPR spectrum of OsHCl(NO)(CO)L2 in a frozen glass
at 77 K (Figure 3) is generally similar in shape to that of
(30) Wolak, M.; van Eldik, R. Coord. Chem. ReV. 2002, 230, 263-282.
(31) Rath, S. P.; Koerner, R.; Olmstead, M. M.; Balch, A. L. J. Am. Chem.
Soc. 2003, 125, 11798.
(29) Renkema, K. B.; Caulton, K. G. New J. Chem. 1999, 23, 1027.
356 Inorganic Chemistry, Vol. 43, No. 1, 2004

Hydrido/Nitrosyl Radical MHCl(NO)(CO)(PiPr3)2
carbene complexes32 and about vinyl complexes,33,34 par-
ticipation by resonance form E is suggested. Note that the
νNO value for the osmium nitrosyl complex, 1388 cm-1, is
close to the value for N/O single bonds, consistent with Os
being a stronger π acid than Ru.
where intramolecular redox transfer (eq 10) is not favored,
and NO binds as a “simple” Lewis base, but with a nonlinear
Cu-N-O geometry.
CuI-(NO•) N CuII-(NO-)
(10)
An earlier example of addition of NO to a 16 electron
molecule involves RhCl(porphyrin).37
It has been previously established from spectroscopic
2+ 38
+ 39,40
studies that Ru(NO)(NH3)5
,
Ru(NO)Cl(bipy)2 ,
and
CpRh(NO)(PR3)41 are all bent nitrosyls with considerable
spin density on N. In the case of CpCr(NO)2(PR3), the radical
character appears to be delocalized over two nitrosyls, and
no significant CrNO bending is evident in the crystal
structure.42 In the case of M(NO)2(porphyrin), where M )
Fe or Ru, both nitrosyls bend.17,28 Only in the case of Ru-
(NO)(NH3)52+ was free radical reactivity described, and the
product of reaction with added R• was stated to be
coordinated RNO.38
The net conversion of two 19 electron radical hydrides is
in some ways analogous to that of two ethyl radicals to give
an alkane and an olefin (eq 9). Both are net H-atom transfers,
or disproportionation.
The units (tetradentate)(FeNO2+) have been further char-
acterized43 as having low spin d6 FeII with coordinated NO•
radical, further complicated by coexistence of a second spin
state; such spin-equilibrium behavior is most frequent for
3d metals, but not for the 4d and 5d metals studied here.
Comparison to Other Ligand-Centered Radicals. We
point out that F, G, and H are analogous to the MHCl(CO)-
(NO)L2 species reported here in being the product of bonding
a radical (PhS• or I•) to a 16 electron fragment.
H3CCH2 + H3CCH2 f H3CCH3 + H2CdCH2 (9a)
2RuHCl(NO)(CO)L2 f RuHCl(HNO)(CO)L2 +
RuCl(NO)(CO)L2 (9b)
The (probable) mechanistic difference is that eq 9b, for
steric reasons, does not occur by a step bimolecular in Ru,
but instead requires free NO as a sterically compact H atom
“shuttle,” and the weak HsRu bond in RuHCl(NO)(CO)L2
facilitates this shuttle role. The weak HsN bond in HNO
leads to its final destruction by excess NO.
i
i
CpMn(CO) (SAr)
W(SPh)(CO)3(P Pr3)2 WI(CO)3(P Pr3)2
2
F
G
H
Comparison to Other Nitrosyls. A closely analogous
situation was reported for the d10, 16 electron reagent
TpR,R′CuI, incorporating various tris-pyrazolylborate ligands.35,36
These bind one NO, incompletely at 1 atm and 25 °C, to
give the adducts TpR,R′Cu(NO) which spectroscopic and
crystallographic studies reveal to have η3-Tp and a bent
( Cu-NO ) 163.4°) nitrosyl. EPR studies show large (∼30
G) hyperfine coupling to N (and also to copper), and g values
significantly different from those of TpCuIIN3. These “elec-
tron excess” metal complexes (19 valence electrons if three
electrons of NO are counted) were concluded to be best
represented as CuI (NO•) oxidation levels. Clearly all of this
is an unsaturated d10 + NO analogue of our unsaturated d6
+ NO chemistry. Moreover, the single crystal structure of a
TpR,R′CuNO species shows large anisotropic thermal el-
lipsoids for the nitrosyl oxygen, indicative perhaps of
disorder, and so the reported Cu-N-O may be in error.
Since the 163.4° angle was fixed in the quantum calculations,
geometry optimization by more modern methods may show
an even smaller angle, as found here in RuHCl(NO)(CO)-
L2. In sum, this seems to be another d electron configuration
While the metal oxidation state of each of these is
represented in the literature44-46 assuming anionic SPh and
I (thus MnII and WI), each may to varying extents involve
ligand-centered radicals. For SAr especially, the bonding
could be represented as I, which is to say that MnI and WO,
in the ligand environment, are not oxidized by the radicals
I or ArS.
Indeed, the EPR g values of species F, 1.977 f 2.068,
are close enough together to suggest major light atom
participation in the SOMO, but certainly with some metal
character (manganese hyperfine structure is resolved). The
similarity here is that both the NO adducts reported here
and F-H involve the interaction of a d6 metal fragment with
a radical which also has lone pairs on the coordinating atom.
(37) Wayland, B. B.; Newman, A. R. Inorg. Chem. 1981, 20, 3093.
(38) Armor, J. N.; Hoffman, M. Z. Inorg. Chem. 1975, 4, 444.
(39) Callahan, R. W.; Meyer, T. J. Inorg. Chem. 1977, 16, 574.
(40) Callahan, R. W.; Brown, G. M.; Meyer, T. J. J. Am. Chem. Soc. 1975,
97, 894.
(41) Geiger, W. E.; Rieger, P. H.; Tulyathan, B.; Rausch, M. D. J. Am.
Chem. Soc. 1984, 106, 7000.
(32) Huang, D.; Spivak, G. J.; Caulton, K. G. New J. Chem. 1998, 22 (10),
1023.
(33) Marchenko, A. V.; Gerard, H.; Eisenstein, O.; Caulton, K. G. New J.
Chem. 2001, 25 (10), 1244.
(34) Caulton, K. G. J. Org. Chem. 2001, 617, 56.
(35) Schneider, J. L.; Carrier, S. M.; Ruggiero, C. E.; Young, V. G., Jr.;
Tolman, W. B. J. Am. Chem. Soc. 1998, 120, 11408.
(36) Ruggiero, C. E.; Carrier, S. M.; Antholine, W. E.; Whittaker, J. W.;
Cramer, C. J.; Tolman, W. B. J. Am. Chem. Soc. 1993, 115, 11285.
(42) Yu, Y. S.; Jacobson, R. A.; Angelici, R. J. Inorg. Chem. 1982, 21,
3106.
(43) Li, M.; Bonnet, D.; Eckhard, B.; Neese, F.; Weyhermueller, T.; Blum,
N.; Sellmann, D.; Wieghardt, K. Inorg. Chem. 2002, 41, 3444-3456.
Inorganic Chemistry, Vol. 43, No. 1, 2004 357

Marchenko et al.
In each case, a potential (but weak) oxidant encounters a
metal/ligand fragment which is only weakly reducing.
favorable because, following H atom abstraction from Ru
or Os, the nitrosyl ligand can undergo intramolecular electron
transfer from M to neutral radical NO, to give bent NO-, in
M(NO)Cl(CO)L2. That is, the thermodynamics of H atom
transfer become especially favorable because of the stability
of the metal complex product. (4) The hydride hydrogen is
fated to become part of light atom products because HNO
reacts further, when additional NO is present, according to
known metal-free reactions.
A similar conclusion holds for the binding of the phenoxy
group to a number of transition metals.47-51 For chromium,
manganese, iron, cobalt, and zinc, oxidation of the MOPh
group beyond a certain level occurs primarily at the ligand,
yielding a metal-coordinated phenoxy radical. The same is
true for the unit CuOPh2+, which is established to be CuII
and a coordinated phenoxy radical, a situation which has
been implicated in the function of galactose oxidase.52,53
Another class of potentially oxidizing radicals is the
nitroxides, R2NO. Since these are quite “stable” radicals, the
rarer case is when they are themselves reduced to their
monoanion,54 and the more common cases are those (gener-
ally involving 3d metal ions) where they persist as ligand-
localized radicals upon coordination.55,56
Experimental Section
General Considerations. All manipulations were carried out
using standard Schlenk and glovebox techniques under prepurified
argon. All solvents were dried and distilled over appropriate agents
and stored in airtight solvent bulbs with Teflon closures under argon
prior to use. Nitrogen monoxide (Aldrich) and deuterium gas
(Cambridge Isotope Laboratories) were used as received. RuHCl-
57
(CO)(PiPr3)2 was prepared according to a previously reported
1
2
procedures. H (referenced to residual solvent impurity), H, and
31P (referenced to 85% H3PO4) NMR spectra were collected on
Varian Gemini 2000 (300 MHz 1H, 121 MHz 31P) and Varian Inova
1
2
(400 MHz H, 73 MHz H) spectrometers. Infrared spectra were
recorded on Nicolet 510P FT-IR and React-IR spectrometers. EPR
spectra were obtained on a Bruker 300ESP spectrometer operating
at X-band (∼9.5 GHz): microwave power, 20 mW; modulation
amplitude, 5.0 G, modulation frequency, 100 kHz; receiver gain,
1.25 × 104. All EPR spectra were observed in toluene as frozen
glasses at 77 K. The values of unresolved coupling (A < 5.2 G)
are derived from the fitting program and are thus less accurate.
Computations. Geometry optimization, frequency analysis,
calculations of energy, and Mulliken spin population in this work
have been performed using density functional theory (DFT)
method,58 specifically functional PBE,59 implemented in an original
program package “Priroda”.60,61 In PBE calculations, relativistic
Stevens-Basch-Krauss (SBK) effective core potentials (ECPs)62-64
optimized for DFT calculations have been used. The basis set was
311-split for main group elements with one additional polarization
p-function for hydrogen, and an additional two polarization
d-functions for elements of higher periods. Full geometry optimiza-
tion was performed without constraints on symmetry. For all species
under investigation, frequency analysis has been carried out. All
minima have been checked for the absence of imaginary frequen-
cies. Calculated IR frequencies were not scaled.
Conclusions
This work shows the following: (1) The radical adducts
MHCl(NO)(CO)L2 (M ) Ru,Os) are persistent. This adduct
makes full utilization of the metal valence orbitals (i.e. 18
electron configuration), and the radical character is best borne
by the NO ligand, especially at nitrogen. Thus, the radical
character being borne by light atoms is analogous to
nitroxide, R2NO. (2) The adduct next confronts the problem
that the two reactive functionalitites, coordinated H and NO,
are mutually trans in a nonfluxional species. This leaves the
adduct metastable, at least with respect to intramolecular
reactions. Bimolecular reactions thus become viable. (3) The
radical appears to react further only with additional radical
NO, and does so by H-atom transfer. This is especially
(44) Lang, R. F.; Ju, T. D.; Kiss, G.; Hoff, C. D.; Bryan, J. C.; Kubas, G.
J. Inorg. Chim. Acta 1997, 259, 317.
(45) Lau, P.; Braunwarth, H.; Huttner, G.; Guenauer, D.; Evertz, K.; Imhof,
W.; Emmerich, C.; Zsolnai, L. Organometallics 1991, 10, 3861.
(46) Milukov, V. A.; Sinyashin, O. G.; Ginzburg, A. G.; Kondratenko, M.
A.; Loim, N. M.; Gubskaya, V. P.; Musin, R. Z.; Morozov, V. I.;
Batyeva, E. S. J. Org. Chem. 1995, 493, 221.
Reaction of RuHCl(CO)(PiPr3)2 with Excess NO. A solution
of 13 mg (0.027 mmol) of RuHCl(CO)(PiPr3)2 in 0.5 mL of C6D6
was placed in an NMR tube fitted with a Teflon stopcock. The
solution was freeze-pump-thaw-degassed 3 times in liquid N2,
the headspace evacuated, and 1 atm NO (0.1 mmol) was introduced
into the tube. The color of the solution became orange from yellow-
(47) Bill, E.; Mueller, J.; Weyhermueller, T.; Wieghardt, K. Inorg. Chem.
1999, 38, 5795-5802.
(48) Chaudhuri, P.; Hess, M.; Mueller, J.; Hildenbrand, K.; Bill, E.;
Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 1999, 121,
9599-9610.
(49) Sokolowski, A.; Mueller, J.; Weyhermueller, T.; Schnepf, R.; Hilde-
brandt, P.; Hildenbrand, K.; Bothe, E.; Wieghardt, K. J. Am. Chem.
Soc. 1997, 119, 8889-8900.
1
orange. After stirring for 10 min at 23 °C, H and 31P{1H} NMR
showed complete conversion of RuHCl(CO)(PiPr3)2 to Ru(NO)Cl-
(50) Sokolowski, A.; Adam, B.; Weyhermueller, T.; Kikuchi, A.; Hilden-
brand, K.; Schnepf, R.; Hildebrandt, P.; Bill, E.; Wieghardt, K. Inorg.
Chem. 1997, 36, 3702-3710.
(51) Sokolowski, A.; Bothe, E.; Bill, E.; Weyhermueller, T.; Wieghardt,
K. Chem. Commun. 1996, 14, 1671-1672.
(57) Esteruelas, M. A.; Werner, H. Organomet. Chem 1986, 303 (2), 221.
(58) Parr, R. G.; Yang, W. Density-functional theory of atoms and
molecules; Oxford University Press: Oxford, 1989.
(59) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett 1996, 77,
3865.
(60) Laikov, D. N. Chem. Phys. Lett 1997, 281, 151.
(61) Ustynyuk, Y. A.; Ustynyuk, L. Y.; Laikov, D. N.; Lunin, V. V. J.
Organomet. Chem. 2000, 597, 182.
(52) Jazdzewski, B. A.; Tolman, W. B. Coord. Chem. ReV. 2000, 200-
202, 633-685.
(53) Halfen, J. A.; Young, V. G., Jr.; Tolman, W. B. Angew. Chem., Int.
Ed. 1996, 35, 1687-1690.
(54) Evans, W. J.; Perotti, J. M.; Doedens, R. J.; Ziller, J. W. Chem.
Commun. 2001, 2326.
(55) Kahn, M. L.; Sutter, J.-P.; Golhen, S.; Guionneau, P.; Ouahab, L.;
Kahn, O.; Chasseau, D. J. Am. Chem. Soc. 2000, 122, 3413.
(56) Ishii, K.; Fujisawa, J.-I.; Adachi, A.; Yamauchi, S.; Kobayashi, N. J.
Am. Chem. Soc. 1998, 120, 3152.
(62) Cundari, T. R.; Stevens, W. J. J. Chem. Phys. 1993, 98, 5555.
(63) Stevens, W. J.; Basch, H.; Krauss, M. J. Chem. Phys. 1984, 81, 6026.
(64) Stevens, W. J.; Basch, H.; Krauss, M.; Jasien, P. Can. J. Chem. 1992,
70, 612.
358 Inorganic Chemistry, Vol. 43, No. 1, 2004

Hydrido/Nitrosyl Radical MHCl(NO)(CO)(PiPr3)2
(CO)(PiPr3)2. 1H NMR (C6D6, 20 °C), ppm: 1.12 (br m, 36H,
PCH3), 2.64 (br m, 6H, PCH). 31P{1H} NMR (C6D6, 20 °C): 49.6.
IR (C6D6), cm-1: ν 1911 (CO), 1582 (NO). The crystalline sample
of Ru(NO)Cl(CO)(PiPr3)2 for X-ray structure determination was
obtained by slow evaporation of a toluene solution.
solution cooled by an ethyl acetate/liquid N2 bath. The resulting
EPR spectra were independent of the mixing technique. An identical
procedure was followed with RuDCl(CO)(PiPr3)2. One of the EPR
experiments showed, at earliest observation times, the presence of
a radical different from RuH(NO)(CO)Cl(PiPr3)2. This first product
was determined to be a Ru(III) product of oxidation of RuHCl-
(CO)(PiPr3)2 by adventitious oxygen. This assignment was con-
firmed by observation of the identical EPR spectrum when a
solution of RuHCl(CO)(PiPr3)2 was contacted with air. In this case,
the orange solution turns green, and the EPR spectrum resembles
that reported in the literature for the Ru(III) product of a reaction
of a Ru(II) complex with air.65
Preparation of RuDCl(CO)(PiPr3)2. A 0.1 g (0.2 mmol) portion
of RuHCl(CO)(PiPr3)2 was dissolved in 10 mL of benzene and
placed in a 100 mL flask with a Teflon-coated stirbar. The solution
was frozen in liquid N2, the headspace evacuated, and 1 atm D2
added to headspace. After stirring for 12 h at +50 °C, the solvent
was removed to give RuDCl(CO)(PiPr3)2. Yield: 0.1 g, 100%. 2H
NMR for RuDCl(CO)(PiPr3)2 (C6D6, 20 °C), ppm: -24.4 (br, Ru-
D).
Reaction of RuDCl(CO)(PiPr3)2 with a Deficiency of NO. A
solution of 16 mg (0.033 mmol) of RuHCl(CO)(PiPr3)2 in 0.9 mL
of C7H8 was placed in an EPR tube closed by a rubber septum.
The solution was freeze-pump-thaw-degassed in liquid N2, the
headspace was evacuated, and 150 mm NO (0.02 mmol) was
introduced into the tube. Then the solution was thawed, stirred
briefly, frozen in liquid N2, and inserted into a precooled EPR probe.
In an alternative mixing technique, NO was bubbled through the
solution cooled by an ethyl acetate/liquid N2 bath. The resulting
EPR spectra were independent of the mixing technique. If O2
inadvertently contacts Ru(H or D)Cl(CO(PiPr3)2 in solution, an EPR
active species is produced; it shows signals at higher g value, which
allows detection of the artifact.
Variable-Temperature Reaction of RuHCl(CO)(PiPr3)2 with
a Deficiency of NO. A solution of 16 mg (0.033 mmol) RuHCl-
(CO)(PiPr3)2 in 0.5 mL C7D8 was placed in an NMR tube fitted
with Teflon stopcock. The solution was freeze-pump-thaw-
degassed 3 times in liquid N2, the headspace was evacuated over
the frozen solution, and 150 mm NO (0.02 mmol) was introduced
into the tube. Then the solution was thawed, stirred briefly, and
1
inserted into a precooled NMR probe. H and 31P{1H} NMR at
-60 °C showed 8% conversion of RuHCl(CO)(PiPr3)2 to Ru(NO)-
Cl(CO)(PiPr3)2 and 2% conversion to RuH(HNO)Cl(CO)(PiPr3)2.
Data follow for -60 °C. 1H NMR (C7D8), ppm: 1.20 (br m, PCH3),
2.43 (br m, PCH), -24.35 (br s, RuHCl(CO)(PiPr3)2), -7.65 (br,
RuH(HNO)Cl(CO)(PiPr3)2), 20.95 (br, RuHNO). 31P{1H} NMR
(C7D8), ppm: 48.5 (8%, s, Ru(NO)Cl(CO)(PiPr3)2), 57.1 (90%, br
s, RuHCl(CO)(PiPr3)2), 61.5 (2%, br s, RuH(HNO)Cl(CO)(PiPr3)2).
Data follow for -40 °C. 1H NMR (C7D8), ppm: 1.25 (br m, PCH3),
2.50 (br m, PCH), -24.22 (br s, RuHCl(CO)(PiPr3)2), -7.70 (td,
JPH ) 23.0 Hz, JHH ) 9.8 Hz) RuH(HNO)Cl(CO)(PiPr3)2), 20.95
(d, JHH ) 9.8 Hz, RuHNO) 31P{1H} NMR (C7D8), ppm: 49.0 (14%,
s, Ru(NO)Cl(CO)(PiPr3)2), 57.6 (84%, br s, RuHCl(CO)(PiPr3)2),
61.5 (2%, br s, RuH(HNO)Cl(CO)(PiPr3)2). Upon warming to +20
°C, the 1H and 31P{1H} peaks for the starting material are extremely
broad and the only products that are present in the reaction mixture
are Ru(NO)Cl(CO)(PiPr3)2 and RuH(HNO)Cl(CO)(PiPr3)2 in 1.5:1
ratio. IR of RuH(HNO)Cl(CO)(PiPr3)2 (C6D6), cm-1: ν 1959 (CO),
1528 (NO). After 11 h at +23 °C, 31P{1H} NMR shows Ru(NO)-
Cl(CO)(PiPr3)2, RuHCl(CO)(PiPr3)2, and RuH(HNO)Cl(CO)(PiPr3)2
in 4:1.4:1 ratio.
Reaction of RuHCl(CO)(PiPr3)2 with a Deficiency of NO in
the Presence of 1,4-Cyclohexadiene. A solution of 14.5 mg (0.030
mmol) of RuHCl(CO)(PiPr3)2 and 55 µL (0.6 mmol) of 1,4-
cyclohexadiene in 0.5 mL of C6D6 was placed in an NMR tube
fitted with Teflon stopcock. The solution was frozen in liquid N2,
the headspace was evacuated, and 200 mm NO (0.02 mmol) was
introduced into the tube. After stirring for 5 min at 23 °C, 1H and
31P{1H} NMR showed complete conversion of RuHCl(CO)(PiPr3)2
to a 1:1 mixture of Ru(NO)Cl(CO)(PiPr3)2 and RuH(HNO)Cl(CO)-
(PiPr3)2. The same result is observed when RuHCl(CO)(PiPr3)2 and
NO are mixed in the same ratio in neat 1,4-cyclohexadiene.
Variable-Temperature Reaction of RuHCl(CO)(PiPr3)2 with
Excess NO. A solution of 16 mg (0.033 mmol) of RuHCl(CO)-
(PiPr3)2 in 0.5 mL of C7D8 was placed in an NMR tube fitted with
Teflon stopcock. The solution was frozen in liquid N2, the headspace
was evacuated, and 760 mm NO (0.1 mmol) was introduced into
the tube. Then the solution was thawed, stirred briefly, and inserted
Reaction of RuHCl(CO)(PiPr2(3,5-(CF3)2C6H3))2 with NO. A
solution of 18.6 mg (0.025 mmol) of RuHCl(CO)(PiPr2(3,5-
(CF3)2C6H3))2 in 0.5 mL of C6D6 was placed in an NMR tube fitted
with a Teflon stopcock. The solution was freeze-pump-thaw-
degassed 3 times in liquid N2, the headspace above the frozen
solution was evacuated, and 150 mm atm NO (0.020 mmol) was
introduced into the tube. The color of the solution turned red from
1
yellow. After stirring for 10 min at 23 °C, H and 31P{1H} NMR
showed complete conversion of RuHCl(CO)(PiPr2(3,5-CF3)2C6H3))2
to a mixture of RuHCl(HNO)(CO)(PiPr2(3,5-CF3)2C6H3))2 and
1
RuCl(NO)(CO)(PiPr2(3,5-(CF3)2C6H3))2 in a 1:1 ratio. H NMR
(C6D6, 20 °C), ppm: 7.96 (td, JHP ) 21.9 Hz, JHH ) 9.6 Hz,
RuH(HNO)), 0.5-1.15 (overlapping m, PCH3), 2.26, 2.49, 2.55,
3.01 (m, PCH), 7.70 (s, p-3,5-(CF3)2C6H3), 8.08, 8.17 (o-3,5-
(CF3)2C6H3), 21.23 (d, JHH ) 9.6 Hz, Ru(HNO)). 31P{1H} NMR
(C6D6, 20 °C): 49.9 (s, RuCl(NO)(CO)(PiPr2(3,5-(CF3)2C6H3))2),
63.3 (s, RuHCl(HNO)(CO)(PiPr2(3,5-(CF3)2C6H3))2).
Crystal Structure Determination of RuCl(NO)(CO)(PiPr3)2.
An orange crystal (approximate dimensions 0.20 × 0.20 × 0.15
mm3) was placed onto the tip of a 0.1 mm diameter glass capillary
and mounted on a SMART6000 (Bruker) at 120(2) K. A preliminary
set of cell constants was calculated from reflections harvested from
three sets of 20 frames. These initial sets of frames were oriented
such that orthogonal wedges of reciprocal space were surveyed.
This produced initial orientation matrices determined from 341
reflections. The data collection was carried out using Mo KR
radiation (graphite monochromator) with a frame time of 7 s and
1
into a precooled NMR probe. H and 31P{1H} NMR at -60 °C
showed conversion of RuHCl(CO)(PiPr3)2 to Ru(NO)Cl(CO)(Pi-
Pr3)2 and RuH(HNO)Cl(CO)(PiPr3)2 in a 31:1 product ratio. In
addition to the signals corresponding to these two compounds, 1H
NMR showed a broad peak at + 6.2 ppm attributed to HONO.
EPR Monitoring of the Reaction of RuHCl(CO)(PiPr3)2. EPR
tubes were charged with solution of 16 mg (0.033 mmol) of RuHCl-
(CO)(PiPr3)2 in 0.9 mL of C7H8 and closed by a rubber septum.
The solution was freeze-pump-thaw-degassed in liquid N2, the
headspace was evacuated, and 150-450 mm NO (0.02-0.06 mmol)
was introduced into the tube. The solution was then thawed, stirred
briefly, frozen in liquid N2, and inserted into a precooled EPR probe.
In an alternative mixing technique, NO was bubbled through the
(65) James, B. R.; Mikkelsen, S. R.; Leung, T. W.; Williams, G. M.; Wong,
R. Inorg. Chim. Acta 1984, 85, 209.
Inorganic Chemistry, Vol. 43, No. 1, 2004 359

Marchenko et al.
a detector distance of 5 cm. A randomly oriented region of
reciprocal space was surveyed to the extent of 1.2 spheres and to
a resolution of 0.51 Å. Five major sections of frames were collected
with 0.30° steps in ω at 5 different φ settings and a detector position
of -43° in 2θ. An additional set of 50 frames was collected in
order to model decay. The intensity data were corrected for
absorption.66 Final cell constants were calculated from the xyz
centroids of 8152 strong reflections from the actual data collection
after integration (SAINT).67 The space group C2/c was determined
on the basis of systematic absences and intensity statistics. The
structure was solved using SHELXS-97 and refined using SHELXL-
disappearance of the hydride signal of the starting material at -32.5
ppm, as well as a significant decrease in intensity for the alkyl
signals of the phosphine ligands. A new broad signal was observed
at 4.00 ppm and attributed to HONO. After the additional 760 mm
NO (0.1 mmol) was introduced, the color of the reaction mixture
turned orange. After 10 min stirring at 23 °C, 31P{1H} NMR showed
the presence of OsCl(NO)(CO)(PiPr3)2 (25.3 ppm) and OsHCl-
(HNO)(CO)(PiPr3)2 (34.2 ppm) in 4.4:1 ratio. 1H NMR (C6D6), ppm:
-8.37 (td, JPH ) 25.1 Hz, JHH ) 8.8 Hz, OsH(HNO)Cl(CO)(Pi-
Pr3)2), 0.9-1.2 (m, PCH3), 2.60 (m, PCH), 3.98 (br, HONO), 21.90
(d, JHH ) 8.8 Hz, OsHNO). After 18 h of stirring at 23 °C, 31P-
{1H} NMR indicated that the OsCl(NO)(CO)(PiPr3)2/OsHCl(HNO)-
(CO)(PiPr3)2 ratio changed to 22:1.
97.68 A Patterson solution was calculated which provided the
.
position of the heavy atoms. Full-matrix least-squares/difference
Fourier cycles were performed which located the remaining atoms.
All non-hydrogen atoms were refined with anisotropic displacement
parameters. All hydrogen atoms were placed in ideal positions and
refined as riding atoms with individual isotropic displacement
parameters. The ruthenium and nitrogen atom are located on a 2-fold
axis, and the oxygen atom of the nitrosyl group is disordered over
two positions (50:50). Likewise, the chlorine atom and the carbonyl
moiety are disordered with each other (50:50). Ruthenium is located
0.45 Å above the basal plane (least squares calculated from Cl1,
Cl1a, P1, P1a). The residual electron density is located in the
vicinity of heavy atoms and on bonds.
ReactIR-Monitored Reaction of OsHCl(CO)(PiPr3)2 with NO.
A solution of 50 mg (0.075 mmol) of OsHCl(CO)(PiPr3)2 in 3 mL
of C6H6 was placed in a ReactIR flask. A 1.2 mL portion of NO
(0.05 mmol) at 1 atm was slowly bubbled through the solution using
a needle and a gastight syringe. The IR difference spectra (relative
to OsHCl(CO)(PiPr3)2 solution) taken during 2 h showed a gradual
decrease for ν(CO) of starting material (1888 cm-1) and the growth
of two new peaks at 1903 and 1559 cm-1, corresponding to ν(CO)
and ν(NO) of OsH(NO)Cl(CO)(PiPr3)2. After an additional 1.2 mL
of NO was bubbled through the reaction mixture, the IR difference
spectrum showed the growth of new peaks at 1929 and 1388 cm-1
,
NMR Tube Reaction of OsHCl(CO)(PiPr3)2 with NO. A
solution of 14 mg (0.021 mmol) of OsHCl(CO)(PiPr3)2 in 0.5 mL
of C6D6 was placed in an NMR tube fitted with Teflon stopcock.
The solution was freeze-pump-thaw-degassed 3 times in liquid
N2, the headspace above the frozen solution was evacuated, and
80 mm NO (0.011 mmol) was introduced into the tube. After stirring
for 10 min at 23 °C, 31P{1H} NMR showed the absence of signals
for starting material and any products; 1H NMR showed the
corresponding to ν(CO) and ν(NO) of OsH(HNO)Cl(CO)(PiPr3)2,
and at 1892 and 1552 cm-1, corresponding to ν(CO) and ν(NO) of
Os(NO)Cl(CO)(PiPr3)2.
Acknowledgment. This work was supported by the
Department of Energy. We also thank Prof. P. A. Evans for
use of his React-IR equipment.
Supporting Information Available: Full crystallographic data
(CIF). This material is available free of charge via the Internet at
http://pubs.acs.org.
(66) Blessing, R. Acta Crystallogr. 1995, 33.
(67) SAINT 6.1; Bruker Analytical X-ray Systems: Madison, WI.
(68) SHELXTL-Plus V5.10; Bruker Analytical X-ray Systems: Madison,
WI.
IC0349407
360 Inorganic Chemistry, Vol. 43, No. 1, 2004