In-depth NMR and IR study of the proton transfer equilibrium between [{(MeC(CH2PPh2)3}Ru(CO)H2] and hexafluoroisopropanol

Article abstract of DOI:10.1139/v00-196

The (carbonyl)dihydride complex [(triphos)Ru(CO)H₂] (2) has been synthesized by reaction of the ruthenate [(triphos)RuH₃]K (triphos = MeC(CH₂PPh₂)₃) with ethanol saturated with CO. A single crystal X-

Full text of DOI:10.1139/v00-196

479
In-depth NMR and IR study of the proton transfer
equilibrium between [{(MeC(CH2PPh2)3}Ru(CO)H2]
and hexafluoroisopropanol
Vladimir I. Bakhmutov, Ekaterina V. Bakhmutova, Natalia V. Belkova,
Claudio Bianchini, Lina M. Epstein, Dante Masi, Maurizio Peruzzini,
Elena S. Shubina, Evgenii V. Vorontsov, and Fabrizio Zanobini
Abstract: The (carbonyl)dihydride complex [(triphos)Ru(CO)H2] (2) has been synthesized by reaction of the ruthenate
[(triphos)RuH3]K (triphos = MeC(CH2PPh2)3) with ethanol saturated with CO. A single crystal X-ray analysis and IR
and NMR experiments have shown that 2 adopts in both the solid state and solution an octahedral coordination geome-
try with a facial triphos ligand, two cis terminal hydrides, and a terminal carbonyl. The reaction of hexafluoro-2-
propanol (HFIP) with 2 has been studied in CH2Cl2 solution by IR and NMR spectroscopy. The proton donor interacts
with a terminal hydride of 2 forming a rather strong hydrogen bond. The resulting H-bonded adduct
[{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a) has fully been characterized by in situ NMR and IR techniques. Com-
pound 2a is in equilibrium with the nonclassical η2-H2 complex [(triphos)Ru(CO)H(H2)]+ (2b), which can independ-
ently be prepared by protonation of 2 with a strong protic acid at low temperature. Unequivocal characterization of the
obs
dihydrogen complex (2b) has been achieved by a multifaceted spectroscopic investigation (T 1min = 0.005 s (200 MHz),
JH,D ≈ 30 Hz, DQCC = 78.3 kHz). A combined IR and NMR study of the proton transfer reaction involving 2 and
HFIP in CH2Cl2 to give, first, the H-bonded adduct (2a) and, then, the dihydrogen complex (2b) has demonstrated that
all these species are in equilibrium in the temperature range from 190 to 260 K. The thermodynamic parameters for
the formation of 2a have independently been determined by NMR and IR methods, while those for the formation of 2b
have been obtained by IR spectroscopy. An energetic profile for the reaction sequence 2 → 2a → 2b is proposed and
discussed.
Key words: hydrides, hydrogen bonding, ruthenium, IR spectroscopy, NMR spectroscopy. 489
Résumé : On a synthétisé le complexe [(triphos)Ru(CO)H2] (2) (triphos = MeC(CH2PPh2)3), un (carbonyl)dihydrure, en
faisant réagir le ruthénate [(triphos)RuH3]K avec de l’éthanol saturé de CO. La diffraction des rayons X sur un cristal
unique et des expériences d’IR et de RMN ont permis de montrer que le composé 2, tant à l’état solide qu’en solution,
adopte une géométrie de coordination octaédrique comportant un ligand triphos facial, deux hydrures terminaux cis et
un carbonyle terminal. Faisant appel à la spectroscopie IR et RMN, on a étudié la réaction du composé 2 avec
l’hexafluoropropan-2-ol (HFIP) en solution dans le CH2Cl2. Le donneur de proton interagit avec un hydrure terminal
du composé 2 en formant une liaison hydrogène relativement forte. Faisant appel in situ d’IR et de RMN, on a fait un
caractérisation complète de l’adduit à liaison hydrogène qui en résulte [{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a).
Le composé 2a est en équilibre avec le complexe non classique η2-H2, [(triphos)Ru(CO)(H2)]+ (2b) que l’on peut pré-
parer de façon indépendante en procédant à la protonation du composé 2 à l’aide d’un acide protique fort, à basse tem-
pérature. On a effectué la caractérisation non ambiguë du complexe dihydrogéné, 2b, en faisant appel à une étude
obs
spectroscopique à multiples facettes (T 1min = 0,005 s (200 MHz), JH,D ≈ 30 Hz, DQCC = 78,3 kHz). Faisant appel à
une combinaison de techniques d’IR et de RMN, on a étudié la réaction de transfert de proton du composé 2 avec
l’HFIP, dans le CH2Cl2, qui donne premièrement l’adduit à liaison hydrogène 2a et ensuite le complexe dihydrogéné
2b; ces études ont démontré que toutes ces espèces sont en équilibre sur toute la plage de températures allant de
Received August 30, 2000. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on May 23, 2001.
Dedicated to Professor Brian James on the occasion of his 65th birthday in recognition of his outstanding work in organometallic
chemistry and homogeneous catalysis.
V.I. Bakhmutov,1,2 E.V. Bakhmutova,1 N.V. Belkova, L.M. Epstein,1,2 E.S. Shubina, and E.V. Vorontsov. A.N. Nesmeyanov
Institute of Organoelement Compounds, Vavilov str. 28, V-334, Moscow, Russia, 117813.
C. Bianchini,2 D. Masi, M. Peruzzini, F. Zanobini. Istituto per lo Studio della Stereochimica ed Energetica dei Composti di
Coordinazione, ISSECC-CNR, Via Jacopo Nardi, 39, 50132, Firenze, Italy.
1Present Address: Centro de Investigación y de Estudios Avanzados del I.P.N., Apdo postal 14–740, 07000, México D.F., México.
2Corresponding authors (fax: 39-055-247-8366; e-mail: bianchini@fi.cnr.it; vladimir@mail.cinvestav.mx; epst@ineos.ac.ru;
peruz@fi.cnr.it).
Can. J. Chem. 79: 479–489 (2001)
© 2001 NRC Canada
DOI: 10.1139/cjc-79-5/6-479

480
Can. J. Chem. Vol. 79, 2001
190–260 K. Faisant appel à l’IR et à la RMN, on a déterminé de façon indépendante les paramètres thermodynamiques
pour la formation du composé 2a alors que ceux pour la formation de 2b ont été obtenus par spectroscopie IR. On
propose et on discute d’un profil énergétique pour la séquence des réactions 2 → 2a → 2b.
Mots clés : hydrures, liaison hydrogène, ruthénium, spectroscopie IR, spectroscopie RMN.
[Traduit par la Rédaction] Bakhmutov et al.
[(triphos)RuH(η2-BH4)] (3) was prepared as described in the
literature (3).
Introduction
The occurrence of intra- and intermolecular hydrogen
bonds between transition-metal hydride complexes and
Brønsted acids has recently been established for a variety of
metals and proton donors (1). It is proposed that such M-
H···H-X interactions in solution may lead to the formation of
dihydrogen (η2-H2) complexes; however, well-documented
examples of proton transfers involving the formation of
η2-H2 metal complexes via M-H···H-X intermediates are still
very scarce (2). This scarcity of data is essentially due to the
NMR studies were carried out in standard 5-mm-NMR
tubes in either CD2Cl2 or THF-d8. The deuterated solvents
were dried by using conventional procedures and were
1
freshly distilled under inert atmosphere prior to use. H and
2H NMR spectra were recorded on Bruker AC200, WP-200
or AMX-400 spectrometers and were calibrated with respect
to tetramethylsilane. 31P NMR spectra were recorded on the
Bruker AC200 instrument operating at 81.01 MHz. Chemi-
cal shifts were measured relative to external 85% H3PO4
with downfield values taken as positive. 13C NMR spectra
were recorded on Varian VXR 300 or Bruker AC200 spec-
trometers operating at 75.42 or 50.32 MHz, respectively.
Peak positions are relative to tetramethylsilane and were cal-
ibrated against the residual solvent resonance. The conven-
tional inversion-recovery method (180-τ-90) was used to
determine the variable-temperature longitudinal-relaxation
time (T1). The relaxation times were calculated using the
nonlinear three-parameter fitting routine of the spectrome-
ters. In each experiment, the waiting period was five times
larger than the expected relaxation time and 16–20 variable
delays were employed. The duration of the pulses were con-
trolled at every temperature. The errors in T1 determinations
were lower than 5% (this was checked with different sam-
ples). The proton donor-hydride reactions were carried out
in cold (195–250 K) CD2Cl2 solutions in 5-mm-NMR tubes
and all the NMR experiments were run starting at low tem-
perature with precooled probe-heads.
M(H )+
+
low temperature at which the M-H···H-X X
X– equilibrium becomes spectroscopically visible (2b, 2d, f).
Nonetheless, the energetic profile associated with the M-H +
M(H )+ + RO– sequence relative to
ROH X M-H···HOR X
2
the reaction of the rhenium(I) complex [(triphos)Re(CO)2H]
(triphos = MeC(CH2PPh2)3) (1) with perfluoro-tert-butanol
has recently been estimated using a combined NMR–IR ap-
proach (2e). In particular, the H-bonded adduct
[(triphos)Re(CO)2H···HOR] (1a) was unambiguously inter-
cepted by in situ IR and NMR experiments before degrada-
tion to the thermodynamically stable dihydrogen complex
[(triphos)Re(CO)2(η2-H2)]+ (1b) (2b). The energy associated
with each step of the conversion of 1a to 1b was determined,
and a neat energy profile for the whole proton transfer
process was obtained (2e). Prior to the rhenium study, ther-
modynamic parameters relative to the (M-H + HOR X
M-H···HOR) → M-(η2-H2) step had been reported exclusively
for the reaction of the dihydride trans-[RuH2(dppm)2] (dppm =
bis-diphenylphosphinomethane) with phenol in toluene (2a).
This paper is aimed at expanding our knowledge on
proton-transfer reactions involving transition metal hydrides,
as well as showing the reliability and wide applicability of
Infrared spectra were recorded as Nujol™ mulls on a
PerkinElmer 1600 series FT-IR spectrometer between KBr
plates or in CH2Cl2 solutions on a Specord M82 spectrome-
ter using 0.01–0.2 cm CaF2 cells. The latter measurements
were performed under a dry argon atmosphere. Low-
temperature IR measurements were carried out in CH2Cl2
solutions using a Carl Zeiss Jena cryostat. The accuracy of
the experimental temperature was ±0.5 K. The cells width
was 0.04–0.12 cm.
the combined NMR–IR approach to study M-H + ROH X
M-H···HOR
M(H )+ + RO– solution equilibria. To this
X
purpose, the new di2hydride complex [(triphos)Ru(CO)H2]
(2) has been synthesized, and the energetic profile associated
with the formation of the hydrogen-bonded adduct
[{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a) and of the
dihydrogen complex [(triphos)Ru(CO)H(H2)]+ (2b) has been
determined by means of low-temperature NMR and IR ex-
periments.
Elemental analyses (C, H) were performed at ISSECC
CNR using a Carlo Erba model 1106 elemental analyzer.
Preparation of [(triphos)Ru(CO)H2] (2)
A vigorously stirred suspension of [(triphos)RuH(η2-
BH4)] (3) (600 mg, 0.81 mmol) in THF (30 mL) was treated
with solid KO-t-Bu (270 mg, 2.41 mmol) added in small
portions. After 30 min stirring, the starting pale yellow com-
plex dissolved to yield a colorless solution which was lay-
ered with absolute ethanol (20 mL) saturated with carbon
monoxide (CO). After 15 min, the CO flow was stopped and
a brisk stream of nitrogen was passed throughout the solu-
tion until ivory colored crystals of 2 separated. Yield:
415 mg (78%). Slow crystallization of the mother liquor in
the air gave a second crop of 2 (80 mg, 15%). Ivory-colored
truncated prisms of 2 suitable for an X-ray crystal analysis
Experimental
General procedures
Tetrahydrofuran (THF) was purified by distillation under
nitrogen over LiAlH4. All other reagents and chemicals were
reagent grade and, unless otherwise stated, were used as re-
ceived by commercial suppliers. All reactions and manipula-
tions were routinely performed under a dry nitrogen
atmosphere by using standard Schlenk-tube techniques. The
solid complexes were collected on sintered glass-frits and
washed with ethanol and light petroleum ether (bp 40–60°C)
before being dried with a stream of nitrogen. The complex
© 2001 NRC Canada

Bakhmutov et al.
481
were obtained by slow evaporation of the solvent from a
diluted CH2Cl2–EtOH solution (2:1 v/v). IR (KBr, Nujol™)
(cm–1): 1920 (CO), 1884 (RuH), 1859; (CH2Cl2 solution)
of 2a in 2b occurred in almost quantitative yield. Complex
2b was also obtained by addition 10–12 equiv of HFIP to a
dichloromethane-d2 solution of 2 (NMR tube experiment
carried out at –80°C).
1934 (CO), 1871 (RuH). 31P{H} NMR (22°C, CD2Cl2,
2
81.01 MHz) δ: 34.71 (AM2 spin system, t, JP ,P
=
A
M
2
1
32.9 Hz, PA), 27.07 (d, JP
= 32.9 Hz, PM). H NMR
Preparation of [(triphos)Ru(CO)2H](BPh4) (5)
M ,PA
(22°C, CD2Cl2, 200.13 MHz) δ: 7.8–6.9 (m, 30H, aromatic
A solution of 2 (250 mg, 0.33 mmol) in THF (15 mL) was
saturated with CO and then 1 equiv of MeOTf (40 µL,
0.36 mmol) was added via syringe under stirring. After
15 min, CO was replaced by nitrogen, and NaBPh4 (300 mg,
0.88 mmol) in ethanol (15 mL) was added. Slow concentra-
tion of the resulting colorless solution gave off-white crys-
tals of 5 in ca. 95% yield. The IR and NMR characteristics
of 5 are fully coincident with those reported in the literature
for the same compound prepared by a different route (4).
4
protons), 2.4–2.1 (m, 6H, CH2(triphos)), –1.48 (q, JH,P
=
=
=
2.9 Hz, 3H, CH3(triphos)), –7.32 (AA′XX′Y, 2JH
A ,HA ′
2
2
6.0 Hz, 2JH
= 18.2 Hz, JH
= 66.1 Hz, JH
A,PX ′
A,PY
A,PX
18.0 Hz, 2JP ,P = 38.0 Hz, 2H, RuH). 13C{H} NMR (22°C,
X
X′
2
CD2Cl2, 50.32 MHz) δ: 209.98 (dt, JC,P = 77.5 Hz,
trans
2JC,P = 7.8 Hz, CO), 142, 138 (m, aromatic ipso carbons),
cis
134–128 (m, aromatic meta, ortho, and para carbons), 39.36
3
2
(q, JC,P = 9.4 Hz, CH3), 39.11 (q, JC,P = 7.0 Hz, CCH3),
35.71 (td, J = JC,P + JC,P′
= 12.5 Hz, JC,P = 4.7 Hz,
equat
equat
CH2Pequat), 33.98 (dt, JC,P = 20.3 Hz, JC,P axial = 6.2 Hz,
X-ray diffraction study of [(triphos)Ru(CO)H2] (2)
An ivory-colored crystal of 2 with the dimensions 0.22 ×
0.12 × 0.17 mm was chosen for an X-ray analysis. A sum-
mary of crystal and intensity data for the compound is pre-
sented in Table 1. Experimental data were recorded at room
temp on a PHILIPS PW1100-FEBO diffractometer using a
graphite-monochromated Cu Kα radiation. A set of 25 care-
fully centered reflections in the range 10° ≤ θ ≤ 15° was used
for determining the lattice constants. As a general procedure,
the intensity of three standard reflections was measured peri-
odically every 2 h for orientation and intensity control. This
procedure did not reveal any decay of intensities. The data
were corrected for Lorentz and polarization effects. Atomic
scattering factors were those tabulated by Cromer and Waber
(5) with anomalous dispersion corrections taken from ref 6.
An empirical absorption correction was applied via Ψ scan
with transmission factors in the range 1.8794–1.0042. The
computational work was performed with a Pentium-II™ per-
sonal computer using the programs SHELXL 93 (7) and
ZORTEP (8). Final atomic coordinates with equivalent iso-
tropic thermal parameters of all atoms and structure factors
are available as supplementary material.
The structure was solved via the heavy atom technique us-
ing the program SIR 92 (9) and all the non-hydrogen atoms
were found through a series of F0 Fourier maps. Refinement
was done by full-matrix least-squares calculations, initially
with isotropic thermal parameters, and then, in the last least-
squares cycle, with anisotropic thermal parameters for ruthenium,
phosphorus, and carbon and oxygen atoms of the carbonyl
ligand. All of the phenyl rings were treated as rigid bodies
with D6h symmetry and C—C distances fixed at 1.39 Å. Hy-
drogen atoms of the triphos ligand were introduced in calculated
positions, but not refined. At an advanced stage of the refine-
ment, two hydrogen atoms were located in the Fourier map
and successfully refined as ruthenium coordinated hydrides.
axial
CH2Paxial). Anal. calcd. for C42H41OP3Ru: Cequ6at6.8, H 5.5;
found C 66.4, H 5.6.
Preparation of [(triphos)Ru(CO)D2] (2-d2)
The perdeuterated isotopomer [(triphos)Ru(CO)D2] (2b-
d2) was prepared as described above using [(triphos)RuD(η2-
BD4)] (3-d4) in the place of 3 and C2H5OD instead of etha-
1
nol. IR (KBr) (cm–1): 1349 (m) (RuD). H NMR analysis
confirmed an isotopic purity higher than 95%.
Preparation of [(triphos)Ru(CO)H(H2)]+ (2b)
A 5-mm screw-cap NMR tube was charged with 2
(40 mg, 5.3 × 10–2 mmol) and degassed dichloromethane-d2
(0.8 mL). Into this solution, cooled to –78°C with a dry
ice–acetone bath, was added via syringe an excess of
HBF4·OMe2 (20 µL, 0.16 mmol). The tube was immediately
inserted into the spectrometer precooled at –80°C and a
31P{H} NMR spectrum, immediately recorded at this tem-
perature, revealed the quantitative transformation of 2 into
2b. 31P{H} NMR (–80°C, CD2Cl2, 81.01 MHz) δ: 20.19
(AM2 spin system, t, 2JP
35.3 Hz, PM). H NMR (–80°C, CD2Cl2, 200.13 MHz) δ:
7.7–7.0 (m, 30H, aromatic protons), 2.7–2.1 (m, 6H,
CH2(triphos)), 1.58 (br, 3H, CH3(triphos)), –5.39 (br, 3H, RuH).
= 35.3 Hz, PA), 22.90 (d, 2JP ,P
=
M ,PA
A
M
1
Monitoring the reaction between 2 and HFIP
Interception of the hydrogen-adduct
[{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a)
Under similar experimental conditions, two equiv of HFIP
(11.2 µL, 0.11 mmol) were added via syringe to a solution
of 2 prepared as described above. Low temperature NMR
spectra showed the complete conversion of 2 into the hydrogen-
adduct 2a. 31P{H} NMR (–80°C, CD2Cl2, 81.01 MHz) δ:
2
30.95 (AM2 spin system, t, JP
= 34.5 Hz, PA), 24.47 (d,
M,PA
2JP ,P = 34.5 Hz, PM). 1H NMR (–80°C, CD2Cl2, 200.13 MHz)
A
M
Results and discussion
δ: 8.1–6.8 (m, 30H, aromatic protons), 2.4–2.0 (m, 6H,
CH2(triphos)), 1.43 (br, 3H, CH3(triphos)), –7.79 (br, 2H, RuH).
Synthesis and characterization of the dihydride
complex [(triphos)Ru(CO)H2] (2)
Treatment of the known (tetrahydroborate)hydride com-
plex [(triphos)RuH(η2-BH4)] (3) with a slight excess of KO-
t-Bu in anhydrous THF at room temp gave a yellow-orange
solution of the trihydride [(triphos)RuH3]K (4) as the only
Ru-containing product (Scheme 1) (10). Although 4 can be
Transformation of the hydrogen-adduct into
[(triphos)Ru(CO)H(H2)] (2b)
Addition of a large excess of HFIP (12 equiv, 67 µL) to
the above solution at –80°C immediately transformed the
hydrogen-adduct 2a into the molecular hydrogen complex
2b. 31P NMR spectroscopy showed that the transformation
© 2001 NRC Canada

482
Can. J. Chem. Vol. 79, 2001
Table 1. Crystal data and structure refinement for 2.
Empirical formula
Formula weight
Temperature (K)
Wavelength (Å)
Crystal system
C42H41OP3Ru
755.78
293(2)
1.54180
Orthorombic
P21/b
Space group
Unit cell dimensions
a = 16.886(6) Å
α = 90.00(2)°
b = 20.845(9) Å
β = 90.00(2)°
c = 10.2770(10) Å
γ = 90.00(2)°
3617(2)
Volume (Å3)
Z
4
Density (calcd.) (Mg m–3)
Absorption coefficient (mm–1)
F(000)
1.388
5.003
1560
Crystal size (mm)
Theta range for data collection (°)
Index ranges
Reflections collected
Independent reflections
Refinement method
Data/restraints/parameters
GoF on F2
0.22 × 0.12 × 0.17
4.24–50.02
–2 ≤ h ≤ 16; 0 ≤ k ≤ 20; –2 ≤ l ≤ 10
1832
1831 (R(int) = 0.0374)
Full-matrix least-squares on F2
1831/1/158
1.074
Final R indices [I > 2σ (I)]
R indices (all data)
Absolute structure parameter
Largest diff. peak and hole (e Å–3)
R1 = 0.0796, wR2 = 0.2039
R1 = 0.0836, wR2 = 0.2107
–0.03(3)
1.286 and –1.400
isolated as potassium salt in the presence of a 18-crown-6
ether, its difficult storage, due to remarkable sensitivity to
water and protic solvents, makes freshly prepared THF solu-
tions excellent vehicles to study the chemistry of the
ruthenate 4 and, in particular, to generate the 16e– fragment
[(triphos)Ru(H)2]. As a matter of fact, a THF solution of 4
reacts with protic solvents such as water or alcohols to give
quantitatively the tetrahydrido complex [{(triphos)RuH}2(µ-
H)2] and H2 via spontaneous dimerization of the
coordinatively unsaturated fragment [(triphos)RuH2].3 In the
presence of σ-donor ligands, this 16e– system can be inter-
cepted to form stable octahedral complexes of the formula
[(triphos)Ru(L)H2] (L = nitriles, phosphines, arsines, CO,
etc.).3 With carbon monoxide, the new carbonyl dihydride
complex [(triphos)Ru(CO)H2] (2) was readily obtained by
simply adding ethanol saturated with CO to a THF solution
of 4. Once formed, 2 is stable in THF and can be isolated in
the solid state. Compound 2 is air-stable in both the solid
state and solution (THF, halocarbons, acetone) and has been
characterized by multinuclear (1H, 31P, and 13C) NMR and
IR spectroscopy, and elemental and X-ray diffraction analy-
ses (vide infra).
(fac) triphos and three other positions occupied by two hydrides
and one carbon monoxide ligand (11). A narrow doublet of
triplets is contained in the 13C NMR spectrum at δ 209.98
for the CO carbon atom, which agrees with the proposed
structural formulation for 2.
Complex 2 is stereochemically rigid on the NMR time
scale; neither the 31P nor the 1H NMR spectrum show signif-
icant changes over the temperature window of dichloro-
methane. In keeping with a classical dihydride nature, a
variable-temperature analysis of the spin-lattice relaxation
time (T1) of the isochronous hydride ligands gave a quite
long minimum value of 0.178 s at 200 K using a 200 MHz
instrument (12). The solid state IR spectrum contains a
ν(CO) band at 1920 cm–1 and two ν(RuH) bands at 1884 and
1859 cm–1. The observation of a medium intensity absorp-
tion at ca. 1350 cm–1 in the spectrum of the deuterated
isotopomer 2b-d2 confirmed the correctness of our ν(RuH)
assignment (kH/D = 1.39). The IR spectrum in CH2Cl2 solu-
tion shows ν(CO) at 1934 cm–1 and ν(RuH) at 1871 cm–1.
This latter band is a combination of both νs(RuH) and
νas(RuH), which are well-resolved in the solid state (νs(RuH)
1884 cm–1 and νas(RuH) 1859 cm–1).
The AM2 pattern featured in the 31P NMR spectrum of 2
and the typical AA′XX′Y (X, X′, Y = 31P) pattern of the
high-field hydride resonance (δ = –7.32) are consistent with
the presence of two cis-disposed hydride ligands in an octa-
hedral coordination with three positions taken by a facial
Well-shaped crystals of 2 were grown from a dilute di-
chloromethane–ethanol solution. Figure 1 shows a ORTEP
view of the complex molecule, while crystallographic data
and selected bond distances and angles are presented in Ta-
bles 1 and 2, respectively. The results of a single crystal
3 C. Bianchini and M. Peruzzini, unpublished results.
© 2001 NRC Canada

Bakhmutov et al.
483
Scheme. 1.
bonyl. The observed distortions from a regular octahedron
are typical of triphos complexes containing hydride ligands
(13). The three Ru—P distances are very similar to each
other (average Ru—P = 2.34 Å), and reflect comparable
trans influence exerted by the hydride and the carbonyl lig-
ands. The Ru—CO distance of 1.84(2) Å is typical for ruthe-
nium(II) carbonyl complexes (14). The two hydride ligands
were located in the difference Fourier maps and refined as
isotropic atoms together with the rest of the non-hydrogen
atoms of the structure, giving Ru—H(1) and Ru—H(2) sepa-
rations of 1.46(18) and 1.50(11) Å, respectively.
Fig. 1. ORTEP drawing of the molecular structure of 2. Only the
ipso carbons of the six triphos phenyl rings are shown.
Synthesis and spectroscopic characterization of the
dihydrogen complex [(triphos)Ru(CO)(H)(H2)]+ (2b)
Treatment of
2 with strong protic acids (HBF4,
HOSO2CF3) at room temp gave a mixture of products that
denied first-order analysis, thus indicating that the molecular
hydrogen complex [(triphos)Ru(CO)(H)(H2)]+ (2b), which
actually forms upon protonation of 2 (vide infra), is unstable
at room temp. This behavior agrees with the results by
Michos
et
al.
(15)
for
the
protonation
of
[{PhP(CH2CH2PPh2)2}RuH2L] (L = CO, P(OCH2)3CEt,
PMe2Ph), which did not afford isolable molecular hydrogen
complexes, although the spectroscopic data were provided.
However, when the protonation was repeated in the presence
of carbon monoxide, the known octahedral hydrido
dicarbonyl complex [(triphos)Ru(CO)2H]+ was quantitatively
produced and isolated as the tetraphenylborate salt
[(triphos)Ru(CO)2H](BPh4) (5). Complex 5 has previously
been synthesized by oxidation of the ruthenium(0) complex
X-ray diffraction analysis are consistent with the solution
structure determined by NMR spectroscopy.
The ruthenium atom is octahedrally coordinated by a fac
triphos ligand, by two cis hydrides, and by a terminal car-
© 2001 NRC Canada

484
Can. J. Chem. Vol. 79, 2001
Table 2. Selected bond lengths and angles for 2.
Fig. 2. 31P NMR spectra (81.01 MHz) illustrating the stepwise
transformation of 2 (4.63 × 10–5 M solution in CD2Cl2) (top
trace), into 2a (middle trace) upon addition of HFIP (2 equiv)
and, eventually, into 2b (bottom trace) upon addition of further
10 equiv of HFIP. All the spectra were taken at 193 K.
Bond lengths (Å)
Ru(1)—C(6)
Ru(1)—P(1)
Ru(1)—P(2)
Ru(1)—P(3)
Ru(1)—H(1)
Ru(1)—H(2)
O(1)—C(6)
P(1)—C(1)
P(2)—C(2)
P(3)—C(3)
C(1)—C(4)
C(2)—C(4)
C(3)—C(4)
C(4)—C(5)
1.84(2)
2.334(5)
2.340(5)
2.344(5)
1.46(18)
1.50(11)
1.20(3)
1.873(19)
1.881(19)
1.856(18)
1.56(2)
1.50(3)
1.57(2)
1.58(3)
Bond angles (°)
107.4(7)
101.9(7)
89.54(17)
162.1(8)
87.17(18)
88.19(17)
81(8)
110(7)
158(7)
84(8)
65(4)
172(4)
96(4)
99(4)
66(8)
C(6)-Ru(1)-P(1)
C(6)-Ru(1)-P(2)
P(1)-Ru(1)-P(2)
C(6)-Ru(1)-P(3)
P(1)-Ru(1)-P(3)
P(2)-Ru(1)-P(3)
C(6)-Ru(1)-H(1)
P(1)-Ru(1)-H(1)
P(2)-Ru(1)-H(1)
P(3)-Ru(1)-H(1)
C(6)-Ru(1)-H(2)
P(1)-Ru(1)-H(2)
P(2)-Ru(1)-H(2)
P(3)-Ru(1)-H(2)
H(1)-Ru(1)-H(2)
O(1)-C(6)-Ru(1)
176(2)
[(triphos)Ru(CO)2] with HCl or MeCOCl (4). To intercept
the molecular hydrogen complex 2b, the protonation of 2
was then carried out in CH2Cl2 solution at low-temperature
(200 K) using either a stoichiometric amount of strong acid
or a large excess (10–12 equiv) of a medium-strength proton
donor such as the activated alcohol hexafluoroisopropanol
(HFIP). Indeed, the addition of 12 equiv of HFIP to a
CD2Cl2 solution of 2 in a 5 mm NMR tube cooled to 195 K
generated a new broad hydride resonance centered at ca.
–5.4 ppm (3H) featured by a very short T1min value of
0.005 s at 190 K (200 MHz), which is typical for a
nonclassical dihydrogen structure (16, 17). The appearance
of the η2-H2 resonance in the proton spectrum of 2b was
paralleled by that of a well-resolved AMM′ pattern with
δ P 20.19 and δ P 22.90 ppm (JP ,P = 35.3 Hz) in the 31P
NMR spectrum. The PA donor atom, trans to the CO ligand,
undergoes a significant high-field shift (∆δ ≈ 14.5 ppm) in
line with the formation of a cationic carbonyl complex. Fig-
ure 2 collects the NMR spectra illustrating the stepwise
transformation of 2 into 2b via the H-bonded adduct
[{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a). Indeed,
this latter compound is formed in very low concentration
when a large excess of HFIP is employed (Fig. 2c), while it
is the only product visible on the 31P NMR timescale when 2
is reacted with two equiv of alcohol (vide infra).
36
34
32
30
28
26
24
22
20
18
16
(ppm)
The dihydrogen complex 2b is highly fluxional in CD2Cl2
solution as a consequence of a fast hydride-dihydrogen ex-
change occurring even at the lowest temperature investigated
(180 K). Compound 2b is quite stable in solution below 250
K but slowly eliminates H2 above this temperature as shown
by the appearance of the typical H2 resonance at 4.56 ppm in
the proton NMR spectrum. The decomposition of 2b is not
chemoselective; in fact, a single, well-defined product was
obtained only in the presence of external ligands, like carbon
monoxide, which are capable of trapping the 16e– fragment
[(triphos)Ru(CO)H]+.
A
M
A
M
The nonclassical nature of 2b was confirmed by deute-
rium labeling experiments. A H{31P} NMR spectrum, re-
1
corded after addition of excess CF3COOD to a solution of 2
in CD2Cl2 at 200 K, exhibits a poorly resolved resonance at
–5.48 ppm, which, upon increasing the temperature to
250 K, transforms into a nonbinomial quintet with a line
separation of ca. 10 Hz. Accordingly, this resonance can
© 2001 NRC Canada

Bakhmutov et al.
485
1
Fig. 3. Hydride region of the H{31P} NMR spectrum of the 2-
d2 isotopomer (250 K, 400 MHz) prepared by low-temperature
protonation of 2 with an excess of CF3COOD in CD2Cl2.
Since the relaxation properties of the deuterium nucleus
are a powerful tool to investigate the dynamic behavior of
dihydrogen ligands (16b, 21), 2H-T1 spin-lattice measure-
ments were carried out on complexes 2b-d2 and 2b-d3 in
CH2Cl2. The perdeuterated isotopomer 2b-d3 has been prepared
by low-temperature protonation of 2b-d2 with CF3COOD.
2
The H NMR spectrum of 2b-d3 contains a broad resonance
at ca. –5.4 ppm that can safely be assigned to the three
2
freely exchanging deuterides. Like 2b, the H resonance of
2b-d3 remains unchanged even at 180 K due to fast
deuteride–deuterium exchange.
2
The H-T1 relaxation experiments on the parent classical
deuteride 2b-d2 give a T1min (2Ru,D2) value of 0.0149 s at
61.402 MHz. By applying eq. [3] and introducing therein
η = 0 (21), a deuterium quadrupole coupling constant
(DQCC) (22) as large as 78.3 kHz can be calculated for 2b-d2
1/2
[3]
DQCC = 1.2201(1 + η2/3)–1/2 (ν/T1min
)
where DQCC, ν, and T1min are given in kHz, MHz, and s, re-
spectively.
The DQCC value for 2-d2 is very close to those recently
determined for classical deuteride ligands in the related com-
plexes [(dppe)2RuD(D2)]BPh4 (dppe = PPh2CH2CH2PPh2)
safely be assigned to the bis-deuterated isotopomer
[(triphos)Ru(CO)HD2]+ (2b-d2) (Fig. 3). In the 1H NMR
spectrum, the typical 1:1:1 triplet of free HD is also ob-
served at 4.51 ppm. In agreement with the formation of a
deuterated isotopomer of 2b, the deuteration results in a
high-field isotopic shift of the “HD2” signal of 120 ppb.
Similar isotopic shifts are commonly observed for HD metal
complexes (17) and may be due to the IPR effect. A JH,D
constant of ca. 30 Hz can be calculated for 2b assuming a
very fast H–H2 exchange (17). Introducing this value in the
empirical correlation, r(HH) = (1.42 – 0.0167)JH,D Å, recently
proposed by Morris and co-workers (18) for dihydrogen
complexes, a rather elongated H—H distance of 0.92 Å may
be figured out for the nonclassical ligand in 2b.
(DQCC
= 79 kHz) (20b) and [(PP3)RuD2] (PP3 =
P(CH2CH2PPh2)3) (DQCC = 73–76 kHz) (20a). The average
2H resonance in 2b-d3 exhibits a T1min (2bRu,D3) time of
0.0174 s. Taking into account a fast D–D2 exchange (see
eq. [1]), it is then possible to calculate a T1min value of
0.0190
s for the nonclassical D2 ligand in 2b-d3.
Nonclassical transition metal hydride complexes have been
2
reported to display longer H-T1min times as compared to
classical polyhydrides due to motions of the dihydrogen lig-
ands (20). The elongation observed on going from 2b-d2 to
2b-d3, however, is very weak, which may be due to a slow-
2
An independent estimation of the H—H separation can
also be obtained from the relaxation data. The T1min mea-
rotating dihydrogen ligand. Thus, the evaluation of H-T1min
(or DQCC) is important because it provides an additional
support for the genuine nonclassical nature of 2b.
The IR spectrum of the dihydrogen complex 2b, obtained
by treatment of 2 with HBF4 at 200 K in dichloromethane
solution, shows a ν(CO) band at 2023 cm–1, which is shifted
at higher energy as compared to the ν(CO) band of the initial
complex. Such high-frequency shifts are commonly ob-
served for transformations of hydride(carbonyl) complexes
into cationic nonclassical molecular hydrogen-carbonyl de-
rivatives (2b, f).
sured for both the hydride signals in 2 (T1min (2Ru,H2) =
obs
0.178 s) and 2b (T
(2bRu,H3) = 0.005 s) allows one to
1min
calculate the T1min for the dihydrogen ligand in 2b by means
of eq. [1], and hence the T1min (Ru(H2)) applying eq. [2].
obs
[1]
[2]
1/T
(2bRu,H3) = (1/3)/T1min (2Ru,H2
+ (2/3)/T
)
1min
obs
(2bRu(H2)
)
1min
obs
1/T
(2bRu(H2)) = 1/T1min (2Ru,H2
)
1min
Synthesis and spectroscopic characterization of the H-
bonded adduct [{(triphos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a)
Upon addition of 2 (0.07 M) to a CH2Cl2 solution of
trifluoroethanol (TFE), HFIP, MeOH, or monofluoroethanol
(MFE) at 298 K, significant changes were observed in the
νOH region of the corresponding IR spectrum. To avoid self-
association, proton-donor concentrations in the range 0.01–
0.03 M were used, so as to work with an excess of 2. Irre-
spective of the proton donor, the changes affecting the OH
stretching region (Table 3) were indicative of a hydrogen
bonding interaction between the alcohol and the hydride lig-
ands in the ruthenium complex.4 Indeed, the intensity of the
+ 1/T1min (2bRu(H2)
)
As a result, a T1min (2bRu(H2)) value of 0.0035 s, exclusively
occasioned by hydride–hydride dipole–dipole interactions in
the dihydrogen ligand, is obtained, which ultimately gives an
H—H distance ranging from 0.74 to 0.94 Å depending on
the spinning model adopted (fast or slow rotation) for the
motion of the dihydrogen ligand (19a). As the separation of
0.94 Å calculated using the low-spinning model is very close
to the distance determined from the JH,D coupling constant
in 2b-d2, it is conceivable that 2b contains, as usually, a
rather “immobile” H2 ligand (19b).
4 A hydrogen-bonding interaction between the activated alcohol and the oxygen atom of the carbonyl ligand was ruled out in view of the
high-frequency shift encompassed by νCO upon addition of HFIP (1d, 2b).
© 2001 NRC Canada

486
Can. J. Chem. Vol. 79, 2001
Table 3. Spectral characteristics of the H-bonded adducts
[{(triphos)Ru(CO)(H)H}···{HOR}] and enthalpy values (–∆H)
relative to their formation in CH2Cl2.
Fig. 4. Temperature-dependence of the equilibrium constant
(K2–2a) for the formation of 2a from 2 and HFIP in CH2Cl2 (ob-
tained from IR data).
bonded
8
7
6
5
4
3
ν OH
∆νOH
–∆H
(cm–1)
(cm–1)
(kcal mol–1)
Ej
ROH
CH3OH
MFE
TFE
3415
3355
3254
3160
212
254
346
420
4.1
4.7
5.8
6.6
1.41
1.38
1.40
1.37
HFIP
band due to the free OH groups decreased, while a new
broad and intense band was formed at lower frequency. As
free
previously observed (1d, 2f), the frequency shifts (∆ν = ν
OH
– ν OboHnded) increase with the proton-donor ability of the alco-
hol in the order MeOH < MFE < TFE < HFIP.
The Iogansen correlation shown in eq. [4] may be used to
estimate the hydrogen-bonding enthalpy values at 298 K
from the experimental frequency shifts listed in Table 3 (1d,
2b).
3.50
4.00
4.50
103/T
5.00
5.50
[4]
–∆H° = 18 ∆ν/(720 + ∆ν)
On decreasing the temperature, the band shift (∆νCO) gradu-
ally moved to high-frequency reaching a maximum of
+15 cm–1 at 190 K for a 10-fold excess of alcohol, while a
band shift to low-frequency was displayed by the Ru—H
stretching of 2. Indeed, after addition of HFIP (2 to 3 equiv)
and cooling down the solution, the νRuH band at 1871 cm–1
moved to low-frequency by 20 cm–1, which is consistent
with a hydrogen bonding interaction between the alcohol
and a hydride ligand of 2 (1d, 2f). This behavior led us to as-
sign the two bands at 1949 (νCO) and 1851 (νRuH) cm–1 to the
H-bonded complex 2a. In contrast, when both the parent hy-
dride 2 and the H-bonded adduct 2a were contemporane-
ously present in solution, a unique νCO and νRuH band was
observed in a position intermediate between the two limits
and depending on the molar fraction of 2a (p(2a)). This lat-
ter can readily be calculated from eq. [5]
Fig. 5. Variable-temperature IR spectra (νCO range) of complex 2
(0.006 M) in CH2Cl2 at 190 K (curve 1) and in the presence of
a 10-fold excess of HFIP in the temperature range between 230
and 190 K (curves 2–6).
νCO(2 + 2a) − νCΟ(2)
[5]
p(2a) =
νCO(2a) − νCO(2)
Using the p(2a) data obtained in this way, the formation
constants K2 – 2a = [2a]/[2][HFIP] have been determined in
the temperature range from 200 to 260 K, and hence the
enthalpy and entropy values associated with the formation of
2a have been calculated by the van’t Hoff method. The cor-
responding temperature dependence of ln K2 – 2a vs. 1/T
2 might explain why this dihydride complex is easily
protonated to give the dihydrogen complex 2b even with rel-
atively weak proton donors such as HFIP and PFTB.
shown in Fig. 4 gave a –∆H2a value of 6.7 kcal mol–1 in
°
fairly good agreement with the value calculated from eq. [4].
°
A –∆ S2a value of 19.6 eu was similarly determined from
the plot.
Variable-temperature IR and NMR studies of the
proton transfer reactions between 2 and HFIP
Following the method reported by Iogansen (23), the ba-
sicity factor (Ej) of the hydride ligand in 2 was calculated as
1.39 ± 0.02 (1d, 2b). It is worth noticing that this parameter,
which characterizes the absolute proton accepting ability of
the hydride ligand, irrespective of both proton donor partner
and solvent, is significantly greater than those reported for
rhenium hydride 1 (Ej = 0.97) and other metal hydride com-
plexes (2b) as well as organic bases such as Et2O (Ej = 1.0)
and DMSO (Ej = 1.27) (23). The relatively high Ej value of
The reaction of 2 with HFIP (10–12 equiv) was followed
by IR spectroscopy in CH2Cl2 starting from 200 K. Selected
variable-temperature spectra in the νCO region are reported
in Fig. 5. The spectrum at 200 K showed the presence of the
dihydrogen complex 2b, with νCO at 2023 cm–1, and of the
H-bonded adduct 2a with νCO at 1949 cm–1. When the tem-
perature was increased to 260 K, the νCO band of 2b de-
creased in intensity while that of 2a increased.
© 2001 NRC Canada

Bakhmutov et al.
487
Fig. 6. Temperature-dependence of the chemical shift of the hy-
dride resonance of 2 in CD2Cl2 (ٗ), and in the presence of a
twofold excess of HFIP (᭺).
Table 4. Temperature-dependence of the equilibrium constants
between 2, 2a, and 2b in the presence of HFIP determined from
the variable-temperature H NMR and IR spectra (range 200–240 K).
1
K2 – 2a (mol–1)a
K2a – 2b
K2 – 2b (mol–1)b
b
T (K)
200
210
215
220
225
230
240
4.6 × 103
1.8 × 103
1.0 × 103
8.3 × 102
5.2 × 102
3.7 × 102
2.6 × 102
2.9 × 103
2.7 × 100
3.4 × 103
8.6 × 10–1
4.9 × 101
1.6 × 101
2.9 × 10–1
1.6 × 10–1
aDetermined from NMR data.
bDetermined from IR data.
Concomitantly, an enhancement of the νRuH band due to 2a
at ca. 1855–1860 cm–1 was observed.
Since the equilibrium between 2a and 2b is reversible in
the temperature range from 200 to 260 K, the IR spectra
allowed us to determine the equilibrium constants K2a – 2b
=
[2b]/[2a] and the overall constant K2 – 2b = [2b]/[2][HFIP]
for the proton transfer from HFIP to a terminal hydride of 2
to give the dihydrogen complex 2b via the intermediate H-
bonded complex 2a.
The equilibrium concentrations [2b] were obtained di-
rectly from the IR experiments. The equilibrium concentra-
tions [2], [2a], [2b], and [HFIP] are related to each other and
to the total concentration of hydride (c2) and proton donor
(cHFIP) by the mass-balance eqs. [6a] and [6b].
for this solution. The hydride chemical shift of 2 is prac-
tically independent from the temperature in the absence
of the proton donor (curve 1), while it reaches a plateau
at –7.87 ppm below 200 K (curve 2). Therefore, the reso-
nance at –7.87 ppm can be assigned to the H-bonded adduct
2a. Obviously, the observed hydride resonance is averaged
between two nonequivalent positions and, therefore, the
chemical shift of the hydride involved in hydrogen bonding
can be calculated as –8.29 ppm by eq. [8].
[6a] c2 = [2] + [2a] + [2b]
[6b] cHFIP = [HFIP] + [2a] + [2b]
[8]
δ(RuH)obs = 0.5 δ(RuHfree) + 0.5 δ(RuH···HOR)
Substituting for [2] and [HFIP] from eqs. [6a] and [6b] into
equation for K2 – 2a and rearranging, eq. [7] was obtained
whose solution gave [2a].
From curve 2 in Fig. 6, the molar fractions of 2 (pfree) and
2a (pbonded) were obtained at different temperatures (eq. [9])
and the corresponding equilibrium constants for the forma-
tion of the H-bonded complex could be calculated (Table 4).
Finally, the –∆H° (7.1 kcal mol–1) and –∆S° (19.0 eu) values
relative to the formation of 2a were obtained from a stan-
dard ln K vs. 1/T plot.
1
[7]
[2a]2 + [2a] 2[2b] − c2 − cHFIP
−
K 2 −2a
+ {c2cHFIP + [2b]2 − [2b](c2 + cHFIP)} = 0
Once the concentrations of each species participating in the
proton transfer equilibrium were known, the calculation of
the equilibrium constants K2a – 2b and K2 – 2b in the tempera-
ture range from 200 to 230 K was straightforward (Table 4).
The temperature dependencies of the equilibrium con-
stants with inverse temperature (1/T) are linear and provide
[9]
δ(RuH)obs = pfree δ(RuHfree
)
+ pbonded δ(RuHbonded
)
It is worth noticing that the thermodynamic parameters cal-
culated using variable-temperature NMR and IR experiments
are in good agreement with each other.
°
the following thermodynamic parameters: –∆H2a −2b
=
–1
°
Positive enthalpy and entropy changes (∆H° = 17 kcal
mol–1, ∆S° = 75.8 eu) obtained by NMR methods have been
reported for the proton transfer from phenol to the hydride
ligand in [RuH2(dppm)2] (2a), while the thermodynamic pa-
rameters for the proton transfer to [RuH2(dmpe)2] (dmpe =
bis(dimethylphosphino)ethane) in ethanol (the H-bonded
complex was not observed, however) exhibit a negative sign
and are only slightly smaller than ours (–∆H°= 6.0 kcal mol–1,
–∆S° = 19.9 eu) (24).
8.7 kcal mol , –∆ S2a −2b = 42 eu for the formation of 2b
–1
°
°
from 2a and –∆H2 −2b = 16.1 kcal mol , –∆ S2 −2b = 65 eu
for the overall proton transfer. Since the formation of 2a has
been found to involve a –∆H2 −2a value of 6.7 kcal mol–1, it
°
is apparent that the formation of 2a plays the major role in
the proton transfer reaction.
The proton transfer reaction for the system (2 + HFIP) has
also been studied by NMR spectroscopy. The low-
temperature addition (200 K) of a twofold excess of HFIP to
a CH2Cl2 solution of 2 results in a high-field shift of the hy-
dride resonance to –7.86 ppm. Figure 6 illustrates the tem-
perature dependence of the hydride chemical shift observed
It is also worth noticing that the H-bonded complex (2a)
is thermodynamically more stable than corresponding rhe-
nium complex (1a) (–∆H° = 4.8 kcal mol–1) (2b), which
© 2001 NRC Canada

488
Can. J. Chem. Vol. 79, 2001
Scheme 2.
the IR data; and (iv) the energy level of the transition state
for the proton transfer 2a → 2b has a purely qualitative
character.
2
Notwithstanding the qualitative character of the profile,
the proposed energy levels are quite reasonable. Indeed, it is
well known that the formation of H-bonded adducts is a dif-
fusion-controlled process with no energy barrier (25). The
observation of separated signals for the hydride ligands in 2a
and 2b, with no broadening of the resonances at low-
temperature, is consistent with a lifetime longer than 10 s for
this reaction (26). Accordingly, an upper limit of 10–1 s–1
can be envisaged for the corresponding rate constant. The
bottom limit of rate constant may roughly be estimated con-
sidering that our IR technique does not allow one to monitor
band-intensity changes for processes faster than 1 × 10–2 s–1
(2e). Since no intensity change with time is observed, it may
be assumed that the lifetime of the proton transfer is shorter
than 1 to 2 min, which provides a lower limit of 1 × 10–2 s–1
for the rate constant. It may be thus concluded that the trans-
formation of 2a into 2b has an activation barrier –∆G‡ (200
K) of ca. 12 to 13 kcal mol–1, which is in nice agreement
with the barrier previously calculated for the related rhenium
system [(triphos)Re(CO)2H] with PFTB (2e) as well as other
literature data. In particular, theoretical studies of proton
transfer to Re- and Ru-hydrides have recently shown that the
proton-transfer barriers depend on both the metal complex
basicity and the strength of the acid and may vary from 10
to 0.7 kcal mol–1 (27). On the other hand, rather low rates of
proton transfer and dihydrogen complex generation (varying
from 1 × 106 to 1 × 10–1 s–1 with corresponding activation
barriers of ca. 6.5–12 kcal mol–1) have been reported for the
reactions of Fe- and Ru-hydrido complexes with strong acids
(28).
6.7
2a
8.7
2b
1.81
0.93
rH·H , Å
reflects the lower basicity of the hydride ligand in 1 (see the
Ej factor).
The variable-temperature spin-lattice relaxation data for a
CH2Cl2 solution of 2 in the presence of a twofold excess of
HFIP showed that the relaxation time of the observed
hydride resonance reaches a minimum at 200 K (T
0.119 s at 200 MHz). As it is evident from Fig. 5, under
these conditions the equilibrium is completely shifted to-
wards the formation of 2a. Using eq. [10], the observed
value was employed to calculate T
(0.0894 s) for the H-bonded hydride ligand in 2a. Finally,
obs
=
1min
obs
obs
T
(RuH···HFIP)
1min
1min
obs
the combination of T
(RuH···HFIP) and of T1 (0.178 s)
1min
measured for 2 (see above) provided a T1min (H···H) value of
0.183 s for 2a, which is exclusively due to hydride–proton
dipole–dipole interactions (12).
Conclusions
The (carbonyl)dihydride complex [(triphos)Ru(CO)H2] (2)
has been found to react with hexafluoro-2-propanol to
give the new dihydrogen complex [(triphos)Ru(CO)H(H2)]+
obs
obs
[10] 1/T
= 0.5/0.178 + 0.5/T
(RuH···H)
1min
1min
Using the procedure described by Halpern and co-workers
(12), the T1min (H···H) relaxation time has been employed to
calculate a hydride—proton distance of 1.81 Å in 2a, which
is quite comparable to the value reported for the rhenium
system (1 + PFTB) (1.83 Å) (2b). This H—H distance lies in
the expected range (1.7–2.0 Å) for transition metal hydride
complexes with either intra- (1a–c) or intermolecular hydro-
gen bonds (1d, e).
(2b)
through
the
H-bonded
adduct
[{(tri-
phos)Ru(CO)(H)H}···{HOCH(CF3)2}] (2a). All these com-
pounds have been characterized by in situ NMR and IR
techniques. Variable-temperature IR and NMR studies in
CH2Cl2 solution have shown that 2, 2a, and 2b are in equi-
librium in the temperature range from 190 to 260 K. An en-
ergetic profile describing the formation of the dihydrogen
complex via the H-bonded adduct has been proposed on the
basis of thermodynamic parameters determined experimen-
tally. The proposed energy profile suggests that, at least in
thermodynamic terms, the formation of the H-bonded adduct
contributes substantially to the overall proton transfer lead-
ing to the dihydrogen complex.
We have already shown that 2 (0.008 M) is converted to
2b in the presence of a 12-fold excess of HFIP in CD2Cl2.
The dihydrogen complex is clearly observed in the H NMR
1
spectrum at 183 K (see above), which also contains an aver-
aged hydride resonance for 2 and 2a. Consistent with the IR
data, it has been found that increasing the temperature leads
to a completely reversible redistribution of integral intensi-
ties of both signals, thus indicating that all these species are
in equilibrium.
An energetic profile for the overall hydrogen-transfer re-
action is illustrated in Scheme 2. It is worth noticing that:
(i) the hydride—proton distance represents the reaction co-
ordinate (determined by NMR spectroscopy); (ii) the energy
of 2a is obtained by IR and NMR data; (iii) the depth of the
second well corresponds to the enthalpy value obtained from
Acknowledgments
Thanks are due to NATO (contract “2H-relaxation as a
tool to study hydrogen transfer reactions”, grant HTECH-
CRG 972 135), CNR (Italy) – RAS (Russian Federation)
agreement, and RFBR (project No 99–03–33270) for financial
support. Support from COST Chemistry Action D17/WG003
is also acknowledged.
© 2001 NRC Canada

Bakhmutov et al.
489
12. P.J. Desrosiers, L. Cai, Z. Lin, R. Richards, and J. Halpern. J.
Am. Chem. Soc. 113, 4173 (1991).
References
13. J. Ott, L.M. Venanzi, C.A. Ghilardi, S. Midollini, and A.
Orlandini. J. Organomet. Chem. 291, 89 (1985).
14. A.G. Orpen, L. Brammer, F.H. Allen, O. Kennard, D.G. Wat-
son, and R. Taylor. J. Chem. Soc. Dalton Trans. S1 (1989).
15. D. Michos, X.-L. Luo, and R.H. Crabtree. Inorg. Chem. 31,
4245 (1992).
16. (a) P.G. Jessop and R.H. Morris. Coord. Chem. Rev. 121, 155
(1992); (b) V.I. Bakhmutov and E.V. Vorontsov. Rev. Inorg.
Chem. 18, 183 (1998).
17. W.J. Oldham, A.S. Hinkle, and D.M. Heinekey. J. Am. Chem.
Soc. 119, 11028 (1997).
1. (a) A.J. Lough, S.Park, R. Ramashandrar, and R.H. Morris. J.
Am. Chem. Soc. 116, 8356 (1994); (b) J.C. Lee, E. Peris, A.
Rheinhold, and R.H. Crabtree. J. Am. Chem. Soc. 116, 11014
(1994); (c) E. Peris, J.C. Lee, J.R. Rambo, O. Eisenstein, and
R.H. Crabtree. J. Am. Chem. Soc. 117, 3485 (1995); (d) E.S.
Shubina, N.V. Belkova, A.N. Krylov, E.V. Vorontsov, L.M. Ep-
stein, D.G. Gusev, M. Niedermann, and H. Berke. J. Am.
Chem. Soc. 118, 1105 (1996); (e) N.V. Belkova, E.S. Shubina,
A.V. Ionidis, L.M. Epstein, H. Jacobsen, A. Messmer, and H.
Berke. Inorg. Chem. 36, 1522 (1997).
2. (a) J.A. Ayllon, C. Gervaux, S. Sabo-Etienne, and B.
Chaudret. Organometallics, 16, 2000 (1997); (b) E.S. Shubina,
N.V. Belkova, E.V. Bakhmutova, E.V. Vorontsov, V.I. Bakhmutov,
A.V. Ionidis, C. Bianchini, L. Marvelli, M. Peruzzini,
and L.M. Epstein. Inorg. Chim. Acta, 280, 302 (1998); (c) Y.
Guari, J.A. Aylonn, S. Sabo-Etienne, and B. Chaudret. Inorg.
Chem. 37, 640 (1998); (d) S. Grundemann, S. Ulrich, H.H.
Limbach, N.S. Golubev, G.S. Denisov, L.M. Epstein, S. Sabo-
Etienne, and B. Chaudret. Inorg. Chem. 38, 2550 (1999);
(e) N.V. Belkova, E.V. Bakhmutova, E.S. Shubina, C.
Bianchini, M. Peruzzini V.I. Bakhmutov and L.M. Epstein.
Eur. J. Inorg. Chem. 2163 (2000); (f) N.V. Belkova, E.S.
Shubina, E.I. Gutsul, L.M. Epstein, I.L. Eremenko, and S.E.
Nefedov. J. Organomet. Chem. (2000) in press.
18. P.A. Maltby, M. Schlaf, M. Steinbeck, A.J. Lough, R.H. Mor-
ris, W.T. Klooster, T.F. Koetzle, and R.C. Srivastava. J. Am.
Chem. Soc. 118, 5396 (1996).
19. (a) R.H. Morris and R.J. Witterbort. Magn. Reson. Chem. 35,
243 (1997); (b) D.G. Gusev, R.L. Kuhlman, K.B. Renkema, O.
Eisenstein, and K.G. Caulton. Inorg. Chem. 35, 6775 (1996).
20. (a) V.I. Bakhmutov, C. Bianchini, F. Maseras, A. Lledos, M.
Peruzzini, and E.V. Vorontsov. Chem. Eur. J. 5, 3318 (1999);
(b) G.A. Facey, T.P. Fong, D. Gusev, P.M. Macdonald, R.H.
Morris, M. Schlaf, and W. Xu. Can. J. Chem. 77, 1899 (1999);
(c) V.I. Bakhmutov, C. Bianchini, M. Peruzzini, F. Vizza, and
E.V. Vorontsov. Inorg. Chem. 39, 1655 (2000).
21. D. Nietlispach, V.I. Bakhmutov, and H. Berke. J. Am. Chem.
Soc. 115, 9191 (1993).
3. L.F. Rhodes and L.M. Venanzi. Inorg. Chem. 26, 2692 (1987).
4. S.I. Hommeltoft and M.C. Baird. Organometallics, 5, 190 (1986).
5. D. Cromer and J. Waber. Acta Crystallogr. 18, 104 (1965).
6. A.J.C. Wilson (Editor). International tables for X-ray crystal-
lography. Vol. C. Kluwer Academic Press, Dordrecht. 1992.
p. 219
7. G.M. Sheldrick. SHELXL. 93 Program for crystal structure re-
finement. University of Göttingen, Germany. 1993.
8. C.K. Johnson. Report ORNL-5138. Oak Ridge National Labo-
ratory, Oak Ridge, TN. 1976; as modified by L. Zsolnai and H.
Pritzkow. University of Heidelberg, Germany. 1994.
9. A. Altomare, G. Cascarano, G. Giacovazzo, and A. Gualandi.
J. Appl. Crystallogr. 26, 343 (1993).
10. C. Bianchini, A. Meli, S. Moneti, and F. Vizza.
Organometallics, 17, 2636 (1998).
11. (a) P. Barbaro, C. Bianchini, A. Meli, M. Peruzzini, and F.
Vizza. Organometallics, 10, 2227 (1991); (b) C. Bianchini, S.
Moneti, M. Peruzzini, and F. Vizza. Inorg. Chem. 36, 5818
(1997); (c) C. Bianchini, A. Meli, M. Peruzzini, F. Vizza, P.
Frediani, and J.A. Ramirez. Organometallics, 9, 226 (1990).
22. (a) H. Guo, W.L. Jarret, and L.G. Butler. Inorg. Chem. 26,
3001 (1987); (b) L.G. Butler and E.A. Keiter, J. Coord. Chem.
32, 121 (1992); (c) A.J. Kim, F.R. Fronczek, L.G. Butler, S.
Chen, and E.A. Keiter. J. Am. Chem. Soc. 113, 9090 (1991).
23. A.V. Iogansen. Theor. Experim. Khim. 7, 302 (1971).
24. L.D. Field, T.W. Hambley, and B.C.K. Yau. Inorg. Chem. 33,
2009, (1994).
25. S. Scheiner. Hydrogen bonding: a theoretical perspective. Ox-
ford University Press, Oxford, N.Y. 1997.
26. H. Gunter. NMR spectroscopy. John Wiley
Chichester, N.Y. 1980.
& Sons,
27. (a) G. Orlova and S. Scheiner. J. Phys. Chem. A, 102, 4813
(1998); (b) G. Orlova, S. Scheiner, and T. Kar. J. Phys. Chem.
A, 103, 514 (1999).
28. (a) M.G. Basallote, J. Duran, M.J. Fernandez-Trujillo, M.A.
Manez, and J.R. Torre. J. Chem. Soc. Dalton Trans. 7445
(1998); (b) M. G.Basallote, J. Duran, M.J. Fernandez-Trujillo,
and M.A. Manez. J. Chem. Soc. Dalton Trans. 2205 (1998).
© 2001 NRC Canada