Kinetics and thermodynamics of proton transfer to Cp*Ru(dppe)H: Via dihydrogen bonding and (η2-H2)-complex to the dihydride

Article abstract of DOI:10.1016/j.ica.2006.07.106

The interaction between Cp^*RuH(dppe) and a series of proton donors (HA) of increasing strength: CFH₂CH₂OH (MFE), CF₃CH₂OH (TFE), (CF₃)₂CHOH (HFIP), p-nitrophenol, CF₃COOH and HBF₄ has been investigated spectroscopically by variable-temperature IR, UV-Vis, and NMR spectroscopy in solvents of differing polarity (n-hexane, dichloromethane and their mixture). The low-temperature IR study shows the establishment of a hydrogen-bond which involves the hydride ligand as the proton accepting site. The basicity factor E_j for the hydride was found to be 1.39. All techniques indicate that an equilibrium exists between the dihydrogen-bonded complex and the cationic dihydrogen complex, [Cp^*Ru(η²-H₂)(dppe)]⁺, the formation of which is shown here for the first time. The proton transfer from HFIP is characterized by ΔH^{{ring operator}} = -8.1 ± 0.6 kcal mol^-1 and ΔS^{{ring operator}} = -17 ± 3 eu. The activation parameters for the subsequent irreversible isomerization leading to the classical dihydride complex, [Cp^*Ru(H)₂(dppe)]⁺, are ΔH^? = 20.9 ± 0.8 kcal mol^-1 and ΔS^? = 9 ± 3 eu as determined from ¹H NMR spectroscopy for protonation by HBF₄. Computational results at the DFT/B3PW91 level confirm the experimentally observed hydride basicity increase on descending the Group from iron to ruthenium and also the formation of the non-classical complex as an intermediate, prior to the thermodynamically favored dihydride.

Full text of DOI:10.1016/j.ica.2006.07.106

Inorganica Chimica Acta 360 (2007) 149–162
www.elsevier.com/locate/ica
Kinetics and thermodynamics of proton transfer to Cp*Ru(dppe)H:
Via dihydrogen bonding and (g2-H2)-complex to the dihydride
a,
a
b
Natalia V. Belkova *, Pavel A. Dub , Miguel Baya b,c, Jennifer Houghton
a
Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Vavilov Street 28, 119991 Moscow, Russia
´
b
` `
Laboratoire de Chimie de Coordination, UPR CNRS 8241 lie´e par convention a l’Universite Paul Sabatier et a l’Institut National
Polytechnique de Toulouse, 205 Route de Narbonne, 31077 Toulouse Cedex, France
c
`
Departament de Quı´mica, Edifici C.n, Universitat Autonoma de Barcelona, 08193 Bellaterra, Catalonia, Spain
Received 16 June 2006; received in revised form 31 July 2006; accepted 31 July 2006
Available online 18 August 2006
Inorganic Chemistry – The Next Generation.
Abstract
The interaction between Cp*RuH(dppe) and a series of proton donors (HA) of increasing strength: CFH2CH2OH (MFE),
CF3CH2OH (TFE), (CF3)2CHOH (HFIP), p-nitrophenol, CF3COOH and HBF4 has been investigated spectroscopically by variable-
temperature IR, UV–Vis, and NMR spectroscopy in solvents of differing polarity (n-hexane, dichloromethane and their mixture).
The low-temperature IR study shows the establishment of a hydrogen-bond which involves the hydride ligand as the proton accepting
site. The basicity factor Ej for the hydride was found to be 1.39. All techniques indicate that an equilibrium exists between the dihydro-
gen-bonded complex and the cationic dihydrogen complex, [Cp*Ru(g2-H2)(dppe)]+, the formation of which is shown here for the first
time. The proton transfer from HFIP is characterized by DHꢀ = ꢁ8.1 0.6 kcal molꢁ1 and DSꢀ = ꢁ17 3 eu. The activation parame-
ters for the subsequent irreversible isomerization leading to the classical dihydride complex, [Cp*Ru(H)2(dppe)]+, are DHꢀ = 20.9
0.8 kcal molꢁ1 and DSꢀ = 9 3 eu as determined from 1H NMR spectroscopy for protonation by HBF4. Computational results at
the DFT/B3PW91 level confirm the experimentally observed hydride basicity increase on descending the Group from iron to ruthenium
and also the formation of the non-classical complex as an intermediate, prior to the thermodynamically favored dihydride.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Dihydrogen bonding; Proton transfer; Hydride complexes; Ruthenium; DFT calculations
1. Introduction
called ‘‘kinetic protonation products’’) and further evolve
to thermodynamically favored products – classical polyhy-
drides [3]. From general considerations, the preponder-
ance of either reaction pathway will be determined by a
combination of both the nature of the metal center and
the electronic and steric properties of the ligands in its
coordination sphere. However, the current knowledge is
far from allowing to predict the reaction mechanism in
any given case. In particular, a change of central metal
atom down a Group may lead not only to an increase
of hydride ligand basicity and strengthening of dihydrogen
bonded complexes [4], but also to a change of proton
accepting site and to a direct interaction with the metal
(Mꢂ ꢂ ꢂHA hydrogen bonding). As an illustrative example,
Proton-transfer processes to and from transition metal
hydride complexes are the key steps in many stoichiome-
tric and catalytic chemical and biochemical processes [1].
It is nowadays well established that the interaction with
the hydride ligand leads to a dihydrogen complex via
dihydrogen bonded (MHꢂ ꢂ ꢂHA) adduct, whereas interac-
tion with the metal atom affords a Mꢂ ꢂ ꢂHA hydrogen
bond and then a polyhydride species [2]. Dihydrogen com-
plexes are sometimes unstable (therefore they are often
*
Corresponding author. Tel.: +7 495 1356448; fax: +7 495 1355085.
E-mail address: nataliabelk@ineos.ac.ru (N.V. Belkova).
0020-1693/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2006.07.106

150
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
hydrogen bond to the hydride ligand precedes the proton
transfer to Cp*MoH3(dppe) (Cp* = (g5-C5Me5), dppe =
1,2-bis(diphenylphosphino)ethane) [5], whereas the forma-
tion of an hydrogen bond to the metal atom is the first
step in protonation of tungsten analogue Cp*WH3(dppe)
[5a], the reaction product being classical tetrahydride
[Cp*MH4(dppe)]+ in both cases.
not been studied, the cationic [Cp*RuH2(dppe)]+ complex,
obtained by H2 addition to [Cp*Ru(dppe)]+, is known to be
a trans-dihydride [11].Therefore, one cannot exclude a pri-
ori the direct attack of a proton donor on the metal atom
and formation of Ruꢂ ꢂ ꢂHA hydrogen bond as the incipient
step of proton transfer in this case. Thus, in order to fur-
ther study the problem of metal influence on the nature
and properties of hydrogen bonded intermediates along
the proton transfer pathway to transition metal hydride
complexes, we have extended our investigations to the
ruthenium species Cp*RuH(dppe).
In this paper, we report the results of combined variable
temperature (190–290 K) spectroscopic and theoretical
studies of the interaction between Cp*RuH(dppe) and pro-
ton donors (HA) of increasing acid strength: CFH2CH2OH
(MFE), CF3CH2OH (TFE), (CF3)2CHOH (HFIP), p-nitro-
phenol, CF3COOH and HBF4 in solvents of different polar-
ities (n-hexane, dichloromethane and their mixture). We
have undertaken the equilibrium thermodynamics investi-
gation of the proton transfer step for the system [Cp*Ru(dp-
pe)H]/HFIP and a variable temperature kinetic study of the
subsequent process of the trans-[Cp*Ru(H)2(dppe)]+ dihy-
dride formation to gain information on the activation
parameters of this step.
Our previous works on the Cp*FeH(dppe) species [6,7]
have proved that, for this complex, the nucleophilic center
is the hydride site. The strength and concentration of the
proton donor determine the evolution of the dihydrogen-
bonded adduct to a dihydrogen species, which can further
evolve to the classical dihydride complex in an irreversible
process. We have found that the strength of the dihydro-
gen-bonding interaction correlates with the proton transfer
barrier. In the case of rather weak proton donors (like fluo-
rinated alcohols), the presence of a second molecule of acid
is required to trigger the proton transfer, through the
enhancement of the primary dihydrogen bond strength
and the stabilization of the transition state and of the final
product by homoconjugated anion formation. The activa-
tion parameters of the dihydrogen to dihydride isomeriza-
tion process in the reaction with strong acids (HBF4 and
TFA) have been determined, as well as the isotope effects
derived from the hydride H/D isotopic substitution [8].
The theoretical analysis of the potential energy surfaces
for the isomerization of the non-classical complex into
the classical one and comparison of theoretical and exper-
imental isotope effects have led to the final elucidation of
the mechanism (Scheme 1).
2. Experimental
All manipulations were performed under argon atmo-
sphere by standard Schlenk techniques. All solvents were
dried over appropriate drying agent (Na/benzophenone
for benzene or THF, CaH2 for dichloromethane, Na for
n-hexane) and freshly distilled under an argon atmosphere
prior to use. CD2Cl2 for NMR experiments (99.96 at.% D)
by Aldrich was degassed by three freeze–pump–thaw
cycles, and than purified by vacuum transfer at room tem-
perature. All Cp*Ru(dppe)H-solutions in CH2Cl2 were
The ruthenium analogue of the above iron hydride,
Cp*RuH(dppe), is known [9] and has been recently tested
together with other ruthenium hydride complexes
Cp0RuH(PP) (Cp0 = C5H5, C5H4Me, C5Me5; PP = chelat-
ing diphosphine) as an iminium cation hydrogenation cat-
alyst [10]. Although the Cp*RuH(dppe) protonation has
H-A
+ H-A
Fe
Fe
Fe
Ph2P
Ph2P
Ph2P
H
H
A
+
H
A
H
H
PPh2
PPh2
PPh2
H
+
A
+
A
Fe
Fe
PPh2
Ph2P
Fe
Ph2P
H
H
H
H
H
A
PPh2
H
H
Ph2P
PPh2
Scheme 1.

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
151
prepared and kept up at 190–200 K (iso-propanol/liquid
N2 baths) due to low stability of Cp*Ru(dppe)H in CH2Cl2
at room temperature (see below).
Standard Bruker software was used for the calculation of
the longitudinal relaxation time.
2.3. Electrochemical measurements
2.1. Preparation of Cp*Ru(dppe)H
Cyclic voltammograms were recorded with an EG&G
362 potentiostat connected to a Macintosh computer
through MacLab hardware/sofware. The electrochemical
cell was fitted with an Ag/AgCl reference electrode, a
1 mm diameter Pd-disk working electrode, and a plati-
num-wire counter electrode. [Bu4N]PF6 (ca. 0.1 M) was
used as the supporting electrolyte. All potentials are
reported relative to the ferrocene/ferrocenium couple. Fer-
rocene was added and measured as an internal standard at
the end of the experiment.
Metallic sodium (150 mg, 6.4 mmol) was added over dry
methanol (60 mL). When the reaction was completed, a
toluene solution (30 mL) of Cp*RuCl(dppe) (1.24 g,
1.85 mmol) [12] was added and the mixture was heated to
reflux for 5 h. The solvent was removed in vacuo, the resi-
due was extracted with toluene (20 mL) and filtered
through celite. The obtained solution was vacuum-dried
and the residue was washed with cold methanol (10 mL,
ꢁ40 ꢁC). The resulting solid was vacuum-dried and recrys-
tallized from n-hexane.
A light-yellow powder was
obtained. Yield: 0.85 g (72%).
2.4. Computational details
1H NMR (250 MHz, C6D6): d ꢁ13.43 (t, 2JH–P = 35 Hz,
1H, Ru–H), 1.84 (s, 15H, Cp*), 2.09–1.84 (m, 4 H, –CH2–
CH2–), 7.10–8.00 (m, 20H, Ph); T1 min = 787 ms at 233 K in
CD2Cl2. 31P{1H} NMR (101 MHz, C6D6): d 90.2 (s). IR
(nujol, CaF2): 1900 cmꢁ1 (mRuH).
Quantum mechanical calculations were performed with
the GAUSSIAN 03 package [13] at the DFT B3PW91 level
[14]. The phenyl groups of the dppe ligand were substituted
by hydrogens in the model complexes. Core electrons of the
Ru and P atoms were described using the effective core
pseudopotentials of Hay–Wadt [15] and valence electrons
were described with the standard LANL2DZ basis set
[13]. In the case of the P atoms, a set of d-type functions
was added [16]. All carbon atoms, the hydrogen atoms
not bonded to the metal, and the atoms of the TFE not
involved in a hydrogen bond have been described with a
6-31G basis set [17]. The hydrogen atom directly bonded
to the Ru center, together with hydrogen and oxygen
atoms of the proton donors involved in the hydrogen
bonding have been described with a 6-31G(d,p) basis set
[17]. Minima and transition states were characterized by
analytically computing the Hessian matrix. The basis set
superposition errors were calculated according to the coun-
terpoise method of Boys and Bernardi [18].
2.2. Spectroscopic studies
2.2.1. IR and UV–Vis investigations
The IR measurements were performed on the ‘‘Infralum
801’’ FT-IR and ‘‘Specord M82’’ spectrometers using CaF2
cells of 0.04–0.12 cm path length. UV measurements were
performed on ‘‘Specord M-40’’ spectrometer. All IR and
UV measurements were carried out by use of a home-mod-
ified cryostat (Carl Zeiss Jena) in the 190–290 K tempera-
ture range. The cryostat modification allows operating
under an inert atmosphere and transferring the reagents
(premixed either at low or room temperature) directly into
the cell which is pre-cooled to the required temperature.
The accuracy of the temperature adjustment was 1 K.
This setup was used both for the variable-temperature
equilibrium studies and for the kinetics investigations at
different temperatures.
3. Results
3.1. Synthesis, spectroscopic and electrochemical study of
Cp*Ru(dppe)H
2.2.2. NMR investigations
Samples of the hydride 1 in CD2Cl2 were prepared under
an argon atmosphere in 5 mm NMR tubes and cooled down
to 195 K immediately after dissolving. 58% HBF4 Æ Et2O or
TFE was added to the pre-cooled solution just before the
measurements.
The synthesis of Cp*RuH(dppe) (1) was originally
described in [9] where NaBH4 in EtOH was used to reduce
Cp*RuCl(dppe) (2) and had a reported yield of 41%. Thus
we explored other synthetic possibilities in order to
improve the yield. Reduction with LiAlH4 in THF at room
temperature for 2 h, the simplest method for preparing the
iron analogue, does not work in the ruthenium case, but 1
is obtained after a 2 h reflux, although with a poor yield
(42%). However, another well-known approach [19] –
reduction with sodium methoxide in a methanol/toluene
mixture – gave higher yields (up to 72%). The main spec-
troscopic features of 1 are a triplet at ꢁ13.43 ppm
1
The H and 31P{1H} data were collected with a Bruker
AV500 spectrometer operating at 500.3 and 202.5 MHz,
respectively. The temperature was calibrated using a meth-
anol chemical shift thermometer; the accuracy and stability
was 1 K. All samples were allowed to equilibrate at every
temperature for at least 3–5 min. The spectra were
calibrated with the residual solvent resonance (1H) and
with external 85% H3PO4 (31P). The conventional inver-
sion-recovery method (180-s-90) was used to determine
the variable-temperature longitudinal relaxation time T1.
1
(2JHP = 35 Hz) in the high field region of the H NMR
spectrum, and a singlet at 90.2 ppm in the 31P{1H} NMR

152
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
spectrum, measured at room temperature in a benzene-d6
solution. The position of the hydride resonance is solvent
and temperature dependent, shifting from ꢁ14.11 to
ꢁ14.20 ppm upon cooling the solution of 1 in dichloro-
methane from 253 to 193 K. In the IR spectrum (THF
methane [21]. Scanning to higher potentials allows a second
oxidation process that is completely irreversible and causes
partial loss in the reversibility of the first one. This suggests
that the 2-electron oxidation product [Cp*RuH(dppe)]2+ is
immediately consumed by a chemical process, which also
results in a concentration reduction for the 17-electron
species formed in the first observed process. The same phe-
nomenon was recently reported for the stepwise 2-electron
oxidation of Cp*MoH3(dppe) [22] and attributed to the pro-
ton transfer from the 16-electron product of double oxida-
tion to the starting material. For the oxidation of 1, this
corresponds to the process shown in Scheme 2 (S = solvent).
The proton transfer process leads to the consumption of one
equivalent of Cp*RuH(dppe) by a non-electrochemical pro-
cess for every two equivalents of the compound within the
diffusion layer, thus reducing the return wave of the first oxi-
dation process approximately by a factor of two. It is well
known that the acidity of transition metal hydride complexes
drastically increases upon oxidation [23]. Indeed, the stoichi-
ometric one-electron oxidation with ferrocenium yielded
[Cp*Ru(H)2(dppe)]+ as the only identified hydride species.
solution), it shows a mRuH band at 1916 cmꢁ1
.
Complex 1 is very soluble in benzene, toluene and THF,
and sparingly soluble in methanol. No significant H/D
exchange was observed at the hydride site when stirring a
suspension of 1 in methanol-d4 for 4 days, as observed by
2
1H, H and 31P{1H} NMR spectroscopy. As many other
hydride species, it is not stable in dichloromethane and
evolves to the chloride derivative 2 by H/Cl exchange with
the solvent. This is evidenced by the appearance of a quintu-
plet in the 1H NMR spectrum at 3.04 ppm (2JHD = 1.6 Hz),
assigned to CHD2Cl, and a singlet in the 31P{1H} spectrum
at 75.2 ppm, assigned to Cp*RuCl(dppe), when following
the evolution of 1 in CD2Cl2. At 300 K, the transformation
is of about 15% in 1 h, and almost quantitative (>90%) after
24 h, the pseudo-first order rate constant for this reaction
being 2.7 · 10ꢁ5 sꢁ1 at 300 K. At lower temperatures
(<273 K), 1 is perfectly stable in dichloromethane solutions
allowing to run variable temperature protonation studies.
Cyclic voltammetric measurements of 1 in THF solution
showed the presence of two oxidation processes at ꢁ0.38
and +0.09 V relative to the ferrocene/ferrocenium (Fc) cou-
ple (Fig. 1). The first process is reversible and corresponds to
the one electron oxidation of 1 to afford [Cp*RuH(dppe)]+.
The analogous CpRuH(PR3)2 complexes [(PR3)2 = (PPh3)2,
dppm, dppe, dppp] undergo one-electron oxidations at ꢁ0.1
to ꢁ0.3 V in THF or acetonitrile [20], whereas the related
(g5-C5H4R)RuH(PPh3) 2 complexes (R = H, t-Bu, CH2Ph,
CTol3) yield E1/2 = ꢁ0.36 V (independent of R) in dichloro-
3.2. Protonation by strong acids: characterization of
[Cp*Ru(dppe)H2]+
Complex [Cp*RuH2(dppe)]+ is known [11] to have clas-
sical nature while other members of the [Cp*RuH2(PP)]+
family (where PP is bidentate phosphine or 2PPh3) are
non-classical (g2-H2) complexes at low temperatures and
isomerize into trans-[Cp*Ru(H)2(PP)]+ upon warming
[24].The existence of the non-classical form was not
excluded [25], but was never confirmed.
Indeed, protonation of a dichloromethane-d2 solution of
1 at 273 K and above with 1 or 2 equiv. HBF4 immediately
affords the classical dihydride [Cp*Ru(H)2(dppe)]BF4 (3),
which is stable in dichloromethane for several days. No vis-
ible traces of any other species were observed. The two main
0.25
spectroscopic features of 3 are a triplet in the high field
0.15
2
region of the 1H NMR spectrum (ꢁ8.67 ppm, JH–P
=
27.5 Hz), and a singlet in the 31P{1H} NMR spectrum
located at 71.3 ppm. However, when addition of acid was
carried out at 193 K and the sample immediately monitored
by low temperature NMR (193 K), quantitative formation
0.05
þ
(>96%) of the non-classical ½CpꢃRuðg2-H2ÞðdppeÞꢄ BF4
ꢁ
-0.05
(4) species was observed. A broad signal (half-height width
w1/2 = 60 Hz) appeared at this temperature in the hydride
1
region of the H NMR spectrum, centered at ꢁ9.50 ppm
-0.15
in CD2Cl2. Its 31P{1H} NMR spectrum shows a singlet at
77.4 ppm. Warming above 243 K led to the irreversible
isomerization to the dihydride 3. The important spectro-
scopic data for 1, 3, and 4 are shown in Table 1.
-3.00
-2.00
-1.00
0.00
1.00
E, V vs Fc
Fig. 1. Cyclic voltammograms for the oxidation of Cp*RuH(dppe) in
THF. Red trace: ꢁ2.0 V / 0.5 V. Blue trace: ꢁ2.0 V/1.0 V. (Scan rate:
100 mV/s). (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)
Longitudinal relaxation times of the hydride signals have
been measured for the complexes 1, 3 and 4 (Table 1). The
[Cp*RuH(dppe)]2+ + Cp*RuH(dppe) + S → [Cp*Ru(dppe)(S)]+ + [Cp*Ru(H)2(dppe)]+
Scheme 2.

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
153
Table 1
Selected IR, 1H and P{1H} NMR spectroscopic characteristics for Cp*Ru(dppe)H and [Cp*Ru(dppe)H2]+BF4 in CD2Cl2
31
Compound
m
Ru–H/cmꢁ1 (e/Mꢁ1 cmꢁ1
)
dH/ppm (JHP/Hz)
T1 min/ms (T/K)
dP (ppm)
Cp*Ru(dppe)H (1)
1914 (255)a
ꢁ14.07 (32.5)b
ꢁ8.67 (27.5)b
ꢁ9.50 (broad)a
787 (233)
555 (213)
19.3 (233)
89.3
þ
½CpꢃRuðdppeÞðHÞ2ꢄ BF4 (3)ꢁ
2000 (90)b
71.3b
77.4a
ꢁ
c
½CpꢃRuðdppeÞðg2-H2Þꢄ þ BF4 (4)
a
At 200 K.
At 290 K.
Not observable (see text).
b
c
T1 minimum of 4 is found at about 233 K, and its value
(19.3 ms) proves the non-classical nature of the product.
Complex 3 shows a T1 minimum of about 555 ms at
213 K, coherent with its dihydride formulation. The mono-
hydride 1 shows a T1 minimum of 787 ms at about 233 K.
Addition of ca. 1 equiv. of trifluoroacetic acid (TFA) to
a CD2Cl2 solution of 1 at 193 K leads to the quantitative
formation of the non-classical species 4, as observed in
A
a
0.5
c
1
the H and 31P{1H} NMR spectra of this solution. An
analogous procedure using CF3COOD has allowed mea-
suring the H–D coupling constant for the monolabelled
dihydrogen ligand. No H/D scrambling has been H–D
observed at this temperature, in the experimental timescale
(approx. 30 min). The obtained value (23.8 Hz) corre-
b
0
2100
2000
1900
1800
ν, cm-1
˚
Fig. 2. IR spectra (mRuH region) of Cp*RuH(dppe) (0.025 M) and
Cp*RuH(dppe) (0.025 M) in the presence of 1 equiv. HBF4 in CH2Cl2.
Before the acid addition at 200 K (the spectrum of Cp*RuCl(dppe)
subtracted) (a); after addition of HBF4 at 200 K (b) and 250 K, time-
dependent spectra (c).
sponds to the H–D distance of 1.02 A (calculated accord-
ing to Eq. (1) [26]), which compares well with the value
˚
1.07 A obtained from the T1 data assuming slow rotation
of H2 (Eq. (2) [27]).
rHH ¼ ꢁ0:0167 JH–D þ 1:42;
ð1Þ
ð2Þ
plex where the phenyl groups of the dppe ligand are substi-
tuted by hydrogens. Minima corresponding to the two
expected isomers – dihydrogen complex and trans-dihy-
dride – were revealed. The optimized geometries are pre-
sented in Fig. 3 and selected geometrical parameters are
presented in Table 2. The energy difference between the
dihydride and dihydrogen isomers of [Cp*RuH2(dhpe)]+
model complex is 2.25 kcal molꢁ1 in favor of the nonclassi-
cal isomer, which is less than that of the iron analogue
(4.90 kcal molꢁ1 [8]). Theoretical consideration of the dif-
ferent models for the [Cp*FeH2(PP)]+ (PP = dppe or
dippe) [8] showed that the introduction of the substituents
on the phosphine ligand enhances the stability of the dihy-
1=6
rHH ¼ 5:815ðT1 min=mÞ
:
In a parallel experiment carried out in CH2Cl2, IR mon-
itoring in the Ru–H stretching region provided the spectral
changes shown in Fig. 2. The starting hydride complex 1 is
characterized by a relatively strong asymmetric mMH band at
1914 cmꢁ1 in CH2Cl2 solution, whose intensity grows up
with lowering the temperature (e = 255 and 227 L
molꢁ1 cmꢁ1 at 200 and 250 K, respectively). The asymmetry
of the band is due to its overlap with the overtones of the
aryl group vibrations and disappears after subtraction of
the spectrum of Cp*RuCl(dppe), see Fig. 2.
The HBF4 addition at 200 K causes the complete disap-
pearance of this band, see spectrum b. No new vibrations
that could be unambiguously assigned to m(H–H) and
m(M–H2) vibration modes for complex 4 appeared in the
IR spectrum. These bands may be very weak or hidden
under stronger CH vibrations or overtones. This spectrum
remains essentially unchanged over time (over 40 min) at
constant temperature, at temperatures below 230 K. At
higher temperatures, slow conversion to the classical dihy-
dride complex occurs, yielding eventually spectrum c. A
new band at 2000 cmꢁ1 is attributed to the cationic dihy-
dride complex. Such significant high-frequency shift
(Dm = +86 cmꢁ1) of the m(MH) band is typical for transition
metal protonation yielding a cationic classical product [28].
Theoretical calculations have been carried out at the
DFT B3PW91 level for the model [Cp*RuH2(dhpe)]+ com-
Fig. 3. Optimized geometries of the dihydrogen (left) and of the dihydride
(right) isomers of the model complex [Cp*RuH2(dhpe)]+.

154
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
Table 2
quency mOH(bonded) bands of hydrogen bonded OH groups
(Table 3). As expected, the interaction with the more
fluorinated alcohol–TFE is stronger than with MFE, as
˚
Selected optimized geometrical parameters (bond lengths in A and angles
in ꢁ) of the dihydrogen and dihydride isomers of the model complex
[Cp*RuH2(dhpe)]+
indicated by the greater band shift DmOH = mOH(free)
ꢁ
a
[Ru](g2-H2)a
[Ru](H)2
m
OH(bonded). The interaction enthalpies were obtained
Ru–CNTb
Ru–P
1.92
2.32
2.32
1.71
1.71
0.91
1.92
2.31
2.30
1.62
1.61
2.95
using Iogansen’s empirical correlation (Eq. (3)) [30,31]
and from the temperature dependence of H-bond forma-
tion constants (van’t Hoff method) (Table 3). Note that
the H-bond formation entropies DSꢀ are quite high in this
case, which explains the low formation constants K, e.g.
for the system 1/TFE in dichloromethane K200 K is only
10.4 Mꢁ1
Ru–H
H–H
\P–Ru–P
\H–Ru–H
a
82.4
30.7
87.0
131.9
[Ru]: [Cp*Ru(dhpe)]+.
CNT is the centroid of the Cp* ring.
18Dm
DHꢀ ¼
;
ð3Þ
ð4Þ
b
Dm þ 720
Ej ¼ DHij=DH11Pi:
dride form, reversing the stabilities in agreement with the
experimental observations that the dihydride is the thermo-
dynamic protonation product. Assuming a similar effect of
the model for ruthenium, the calculations indicate that the
relative stability of the dihydride isomer relative to the
dihydrogen complex is greater for ruthenium.
The combined study of IR changes in the mRuH region
and 1H NMR spectra showed that ruthenium bonded
hydride ligand is the coordination site, the dihydrogen
bonded (RuHꢂ ꢂ ꢂHOR) species being the intermediates of
proton transfer and non-classical complex formation
(Scheme 3).
Upon the addition of increasing amounts of TFE to the
solution of Cp*RuH(dppe) in n-hexane, whose low polarity
favors hydrogen bond formation but not proton transfer,
at room temperature, a gradual increase of the mRuH band
intensity was observed. The mRuH band integral intensity
was 1.3 times higher in the presence of 12 equiv. TFE is rel-
ative to free 1. Such changes are consistent with the hydro-
gen bonding to the hydride ligand [33] and are caused by
the overlap of the mRuH(1) and mRuH(1a) bands, of which
the latter has higher intensity. The observation of more dis-
tinct changes (separate bands) by temperature lowering in
this solvent was precluded by the low compound solubility.
However, use of a 2:1 v/v n-hexane-dichloromethane mix-
3.3. Interaction with weak proton donors: characterization of
hydrogen-bonded intermediate
To investigate the mechanism of the Cp*RuH(dppe)
protonation in more detail, we studied the interaction
of 1 with fluorinated alcohols CFH2CH2OH (MFE) and
(CF3)nCH3ꢁnOH (with n = 1 (TFE), 2 (HFIP)) according
to the well-established protocol [29], using combination of
IR and NMR spectroscopies. In the presence of excess
hydride, the IR spectra in the mOH region show a typical
picture of hydrogen bond formation. The intensity
decrease of the alcohol (MFE and TFE) mOH(free) bands
is accompanied by the appearance of new broad low fre-
Table 3
Parameters of the hydrogen-bonding between Cp*RuH(dppe) and MFE or TFE in CH2Cl2
ROH
Pia
mOH(free) (cmꢁ1
)
mOH(bonded) (cmꢁ1
)
Dm (cmꢁ1
)
DHꢀb (cmꢁ1
)
DHꢀc (cmꢁ1
)
DSꢀc (eu)
Ejd
MFE
TFE
a
0.74
0.89
3600
3590
3346
3254
254
336
ꢁ4.7
ꢁ5.7
ꢁ4.5
ꢁ5.7
ꢁ15.8
ꢁ23.8
1.38
1.39
Acidity factors of proton donors [32].
Calculated by Eq. (3), mean error 0.4 kcal molꢁ1
b
c
.
Calculated from the temperature dependence of H-bond formation constants.
d
Basicity factor as defined in Eq. (4), DHꢀ11 ¼ ꢁ4:6 kcal molꢁ1 for CH2Cl2 [30,31].
Scheme 3.

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
155
ture allowed to cool down to 250 K. No growth of the
1
A
20eqv
16eqv
12eqv
10eqv
8eqv
6eqv
4eqv
2eqv
RuH
dihydride mRuH band (at 2000 cmꢁ1) was observed under
these conditions, showing that the isomerization is slow
in comparison to the experiment timescale. In the presence
of the increasing amounts of TFE, the decrease of the mRuH
band of free hydride 1 at 1914 cmꢁ1 1 and the growth of a
new band at 1900 cmꢁ1 (DmRuH = ꢁ14 cmꢁ1), belonging to
the Ru–H stretching vibration in the dihydrogen-bonded
complex (mRuH bonded), was observed (Fig. 4).
0.8
0.6
0.4
0.2
0
Use of neat dichloromethane allowed working at lower
temperatures (down to 190 K) but the observation of dihy-
drogen bonded species was complicated by the proton
transfer process, which readily occurs in this solvent at
low temperatures. Thus, the addition of 2 equiv. of TFE
to a CH2Cl2 solution of 1 at 200 K causes a slight increase
of the mRuH band intensity due to the dihydrogen bond for-
mation. However, increasing the TFE excess up to 6 equiv.
2000
1950
1900
1850
ν, cm-1
Fig. 4. IR spectra (mRuH region) of Cp*RuH(dppe) (0.017 M) in the
presence of the increasing amounts of TFE. N-hexane–dichloromethane
2:1 v/v mixture, 250 K.
leads to a very small shift of the mRuH band (by ꢁ2 cmꢁ1
)
due to the increase of the mRuH (1a) contribution into the
overall band and to an intensity decrease because of partial
proton transfer with formation of the non-classical dihy-
dride, as was confirmed by NMR.
The 1H NMR spectra show an up-field shift of the
hydride resonance of 1 from ꢁ14.17 ppm (JHP = 35 Hz)
to ꢁ14.28 ppm (JHP = 30 Hz) in the presence of up to 2
equiv. TFE in CD2Cl2. The presence of a minor (1%)
amount of [Cp*Ru(g2-H2)(dppe)]+ (4) was shown by the
appearance of a signal at ꢁ9.49 ppm having a short T1
relaxation time (ca 30 ms). In the presence of 14 equiv. of
TFE, the hydride resonance of 1 shifted further to a stron-
ger field (dRu–H = ꢁ14.32 ppm, JHP = 30 Hz) (Fig. 5) and
the amount of non-classical complex present in the system
increased (14%). The signal of 4 disappeared when the tem-
perature increased above 230 K but reappeared upon cool-
ing back, evidencing the reversibility of the proton transfer
step (Scheme 3). Such substantial high-field shift of the
hydride ligand is in agreement with the formation of the
RuHꢂ ꢂ ꢂHA hydrogen bond [2].
Further confirmation of Hꢂ ꢂ ꢂH bond formation was
obtained by the measurement of the longitudinal relaxation
time (To1bs) of the hydride signal of 1 in the presence of
obs
1 min
TFE. As expected, the T
value decreases upon addition
of the proton donor [2,34]. This decrease is proportional to
the amount of alcohol added (Fig. 6), since the correspond-
ing hydride resonance belongs to the mixture of free and
hydrogen bonded hydride species and since the hydrogen
bonding equilibrium shifts to the right with the increase
of alcohol excess. It is also of significance that the temper-
ature corresponding to the T1 min increases slightly upon
H-bond formation from 230 to 240 K. This phenomenon
signals a decreased correlation time (sC), i.e. a slower tum-
bling motion, for the protonated species, in agreement with
its expected larger size. Thus, slightly higher temperatures
Fig. 5. 1H NMR spectra (hydride region, CD2Cl2, 200 K) of
Cp*RuH(dppe) (a) and Cp*RuH(dppe) in the presence of the increasing
amounts of TFE: 0.5 equiv. (b), 2 equiv. (c), 14 equiv. (d).
are needed to raise again sC to the conditions required
for the most efficient longitudinal relaxation (w20sC2 ꢅ 1).
No non-classical protonation product was present in the
solution at 240 K in the presence of 2 equiv. TFE. There-
fore, using the H-bond formation constant value obtained
by IR (0.96 Mꢁ1 at 240 K), it is possible to calculate the
amount of RuHꢂ ꢂ ꢂHOR species (x = [RuHꢂ ꢂ ꢂHOR]/
[RuH] = 0.11 at 2-fold TFE excess) and then the value of
obs
1 min
1
T
ðRuH ꢂ ꢂ ꢂ HORÞ ¼ 88 ms for the dihydrogen bonded
The partial proton transfer and formation of the non-classical cationic
hydride may contribute to the intensity loss at 1914 cmꢁ1
.
complex (Eq. (5)).

156
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
0.5
0
a
-0.5
-1
b
Fig. 7. Optimized geometries of the hydrogen bonded adducts at the
hydride and metal sites of model Cp*RuH(dhpe) complex with TFE.
-1.5
In agreement with such geometrical changes, the energy
changes associated with the hydrogen bond formation are
more negative for the hydride site than for the metal site,
DE = ꢁ10.46 and ꢁ6.72 kcal molꢁ1, respectively. These
values reduce to ꢁ5.83 and ꢁ1.64 kcal molꢁ1, respectively,
after basis set superposition error (BSSE) correction. Note
that the BSSE corrected energy for the hydride bonded
adduct is close to the experimentally determined hydrogen
bond formation enthalpy (Table 3). Recalculation of the
corresponding Cp*FeH(dhpe)/TFE complexes at the same
DFT B3PW91 level of theory gives BSSE corrected hydro-
gen bond formation energies DEBSSE = ꢁ0.42 and
ꢁ4.92 kcal molꢁ1 for Feꢂ ꢂ ꢂHOR and FeHꢂ ꢂ ꢂHOR com-
plexes, respectively. Thus, the calculations confirm the
preference of the proton donor interaction with the hydride
ligand and the increase of its basicity going from the iron to
ruthenium complex.
Concluding this section, formation of dihydrogen
bonded complexes between the ruthenium hydride 1 and
alcohols is the first step of the reversible low temperature
proton transfer reaction in this system, which precedes
the formation of the non-classical cationic hydride complex
4. In rather low polarity media like dichloromethane
(dielectric permittivity e is 14.63 at 200 K [35,36]), ionic
species can exist as contact ion pairs additionally stabilized
by hydrogen bonding between the g2-hydrogens of the cat-
ion and the counteranion oxygen atom. The formation of
such ion pairs has been shown for some non-classical
ruthenium hydrides [4a,37] as well as for the iron analogue
[Cp*Fe (g2-H2)(dppe)]+ [7].
c
-2
190
210
230
250
270
T, K
Fig. 6. Temperature dependence of longitudinal relaxation time (ln T1) for
the hydride signal of Cp*RuH(dppe) (a) in the presence of TFE (2 equiv. –
b, 14 equiv. – c) in CD2Cl2 solution.
obs
1min
obs
1=T
1=T
¼ xꢃ1=T
ðRuHꢂꢂꢂHORÞ þ ð1 ꢁ xÞꢃ1=T1minðRuHÞ;
ð5Þ
ðRuHꢂꢂꢂHORÞ ¼ 1=T1minðRuHÞ þ 1=T1minðHꢂꢂꢂHÞ:
1min
obs
1min
ð6Þ
In turn, Eq. (6) results in a T1 min(Hꢂ ꢂ ꢂH) time of 100 ms
for the hydride–proton dipole–dipole interaction. Applica-
tion of Eq. (2) to this system gives hydride–proton distance
˚
in dihydrogen bonded complex r(Hꢂ ꢂ ꢂH) of 1.38 A. The
˚
same Hꢂ ꢂ ꢂH distance (1.40 A) was obtained from the relax-
ation times measured in the presence of 14 equiv. TFE tak-
ing into account the amount of cationic complexes 3 (3%)
and 4 (18%) present in solution at 240 K.
Theoretical calculations have been carried out at the
DFT B3PW91 level for the model Cp*RuH(dhpe) complex
and its adducts with TFE in order to study the relative sta-
bility of hydrogen bonded complexes at the metal and
hydride sites. The optimised geometries of these two com-
plexes are presented in Fig. 7.
The geometrical changes that accompany hydrogen
bond formation are similar to those computed previously
at the DFT B3LYP level for CpRuH(CO)(PH3) as a model
for CpRuH(CO)(PCy3) [37] and for Cp*FeH(dhpe) as a
model for Cp*FeH(dppe) [7]. The hydrogen bond forma-
tion is reflected in the lengthening of the OH bond dis-
3.4. Interaction with TFA and p-nitrophenol – ion pairing in
[Cp*Ru(g2-H2)(dppe)]+Xꢁ
Since the conjugate bases of the fluorinated alcohols
used herein are colorless species and do not exhibit any typ-
ical IR absorptions that could signal their hydrogen bond-
ing with the product dihydrogen complex, we employed
p-nitrophenol and trifluoroacetic acid for this study. The
UV–Vis and IR bands of these proton donors have been
shown to be very sensitive to their protonation and hydro-
gen bonding status.
˚
tances (Dr(OH) = 0.019 and 0.022 A for the metal and
hydride bonded complexes, respectively) from their values
in the isolated proton donors. Note that Dr(OH) is greater
for the hydride site, signalling stronger bonding. The bind-
ing of the proton donors at the hydride site also produces a
˚
Ru–H bond lengthening by 0.011 A, whereas the Ru–H
bond does not change upon binding at the metal site
In the case of trifluoroacetic acid (TFA), IR spectros-
copy can be conveniently used due to sensitivity of the acid
˚
(Dr(Ru–H) = ꢁ0.001 A).

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
157
as
OCO
mCO band and the anion m
band to hydrogen bonding
A
[7,38,39]. The IR spectra in the carbonyl stretching region
were measured at various TFA concentrations (from
0.0075 to 0.075 M) and at different 1/TFA ratios (from
4:1 to 1:6). Interestingly, no CF3COOꢁ anion band was
observed in the spectrum for a 2:1 TFA/1 ratio; two bands
at 1734 and 1780 cmꢁ1 are assigned to the dihydrogen-
bonded complex 1a and to the acid dimer, respectively.
When the temperature was increased up to 240 K, the band
at 1734 cmꢁ1 decreased reversibly, showing the reversibility
of the hydrogen bond formation. Upon increasing the
Cp*RuH(dppe) amount, proton transfer occurred and the
major band observed under these conditions is at
1687 cmꢁ1 (Fig. 8). Note that the bands of the molecular
complexes are still present in the spectra at 1:1 TFA/1
ratio.
1
0.5
0
b
a
1850
1800
1750
1700
1650
1600
ν
, cm-1
Fig. 9. IR spectra in the mCO=mOasCO region of TFA (0.075 M) in the presence
of equimolar amounts of Cp*RuH(dppe) (a) and Cp*FeH(dppe) (b).
Comparison of the spectrum of the 1:1 Cp*RuH(dppe)/
TFA mixture with that of the CF3COOꢁPPhþ salt and with
4
that of the Cp*FeH(dppe)/TFA 1:1 mixture at the same con-
centration (Fig. 9) shows that the band at 1687 cmꢁ1 has
high frequency shoulder belonging to the hydrꢁogen bonded
ion pair [Cp*(dppe)M(g2- H2)]+ꢂ ꢂ ꢂ[OCOCF3] in the case
of iron, but is only slightly asymmetric in the case of ruthe-
nium. Thus, in these rather diluted solutions the major spe-
cies present is free CF3COOꢁ anion, the amount of
hydrogen bonded ion pairs being higher in the case of iron
than ruthenium.
When using a 6-fold excess of the acid, no bands attrib-
utable to the free anion or to the hydrogen-bonded com-
plex are visible, whereas a wide and low intensity band is
observed at 1620 cmꢁ1. This corresponds to the free homo-
conjugated ion, in which the [CF3COO]ꢁ anion is bonded
to two TFA molecules [40].
proportions as the free base in solution. In the presence
of excess acid, the only carbonyl species present in solution
are CF3COOH and [CF3COO(HOOCCF3)2]ꢁ.
The equilibrium resulting from the interaction between
Cp*RuH(dppe) and a weaker proton donor – p-nitrophe-
nol (PNP; Pi = 1.27) – was investigated by UV–Vis spec-
troscopy. Spectra were recorded for CH2Cl2 solutions of
PNP (0.001 M) in the presence of 1 at different ratios from
1:0.1 to 1:2, in the 200–230 K temperature range, where the
observed changes were fully reversible. The spectra show
wide overlapping bands of both the phenols in their various
forms and the hydride complexes (both free and dihydro-
gen bonded) (Fig. 10).
The absence of free phenolate is signaled by the absence
of a band at 430 nm. A band decomposition yields three
bands with maxima at 312, 342, and 384 nm. The first
two bands are assigned to free PNP and to the dihydrogen
Evidently, since the trifluoroacetate anion is a weak pro-
ton acceptor for hydrogen bonding, it is present in large
1687
1780
A
d
e
3
a
0.5
1734
b
2
1
0
c
d
e
0.0
1850
1800
1750
1700
c
1850
1800
1750
1700
1650
1600
ν, cm-1
Fig. 8. IR spectra in the mCO=maOsCO region of TFA (0.0075 M) (a) and TFA in the presence of Cp*RuH(dppe): 0.5 equiv. (b); 1 equiv. (c); 2 equiv. (d); 4
equiv. (e). CH2Cl2, 200 K.

158
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
0.4
A
c
b
1.0
0.3
0.2
0.1
0
e
a
0.5
0.0
d
250
300
350
400
450
500
0.2
0
0.6
0.8
0.4
λ, nm
RuH mole fraction
Fig. 10. UV–Vis spectra of PNP (0.001 M) (a), potassium p-nitrophen-
olate (0.001 M in the presence of [18]crown-6) (b), homoconjugated PNP
anion (band derived from the spectrum of a 1:1 mixture of (a) and (b)) (c),
Cp*RuH(dppe) (0.0005 M) (d) and PNP (0.001 M) in the presence of
Cp*RuH(dppe) (0.0005 M) (e). CH2Cl2, 200 K.
Fig. 11. Intensity changes at 380 nm vs the Cp*RuH(dppe) mole fraction
(mole fraction = CRuH/[CRuH + CPNP]).
The spectral changes are fully reversible in the 200–
230 K temperature range, showing that proton transfer is
reversible and no significant isomerization to the classical
dihydride complex occurs within this temperature range.
Upon the temperature decrease, the band of the free phenol
decreases and those of hydrogen-bonded phenol (342 nm)
and hydrogen bonded ion pair (384 nm) increase. Note that
neither the free phenolate band, expected at 420–430 nm,
nor the band of homoconjugated anion [ArOHOAr]ꢁ,
expected at ca. 400 nm, were observed, suggesting that
the hydrogen bonded ion pair does not essentially dissoci-
ate under these conditions.
bonded complex [Cp*(dppe)RuH]ꢂ ꢂ ꢂHOC6H4NO2. Note
that the 342 nm band is red shifted not only relative to free
PNP (Dk = 30 nm), but also slightly red-shifted with
respect to the related iron complex [Cp*(dppe)-
FeH]ꢂ ꢂ ꢂHOC6H4NO2 (340 nm) [7], in agreement with a
slightly stronger hydrogen bonding. The band at 384 nm
is attributed to a hydrogen-bonded phenolate ion, since
this is blue-shifted from the free phenolate band by
46 nm. The formation of the 1:2 ion pair, stabilized by a
hydrogen bond between the non-classical cation and
homoconjugated phenolate anion, [Cp*(dppe)M(g2-
H2)]+ꢂ ꢂ ꢂ[ArOHOAr]ꢁ, was shown previously for the iron
analogue [7]. The assignment of the absorption at 384 nm
to the 1:2 hydrogen-bonded ion pair in the present case
was confirmed by the titration experiment. Upon increas-
ing the amount of 1 at constant PNP concentration at
200 K, the bands at 342 nm and 384 nm grow in intensity
whereas the free phenol band at 312 nm decreases. The plot
of the intensity changes at 380 nm versus the
Cp*RuH(dppe) mole fraction (Fig. 11) gives a break point
for a mole fraction of (or near) 0.3, indicating a 1:2 binding
stoichiometry for the ionic species, [Cp*(dppe)Ru(g2-
H2)]+[ArOHOAr]ꢁ. As in the case of iron the 16 nm
blue-shift of this band relative to free [ArOHOAr]ꢁ 2 sug-
gests further hydrogen bonding of homoconjugated anion
with the cationic dihydrogen complex, [Cp*(dppe)Ru(g2-
H2)]+ꢂ ꢂ ꢂ[ArOHOAr]ꢁ. Note that the position of this band
is slightly different from that of the iron analogue, in agree-
ment with a higher basicity of 1 and thus a lower acidity of
[Cp*(dppe)Ru(g2-H2)]+.
3.5. Thermodynamics of proton transfer from HFIP
The UV/Vis spectral changes resulting from the interac-
tion between neutral hydride 1 and HBF4 are shown in
Fig. 12. The starting hydride complex has a relatively
strong and broad metal–ligand charge-transfer band [41]
with a maximum around 320 nm (e = 7200 L molꢁ1 cmꢁ1
A
0.5
a
b
c
0
250
300
350
400
450
500
λ, nm
2
Fig. 12. UV–Vis spectral changes observed for the protonation of
Cp*Ru(dppe)H (0.0025 M) by 1 equiv. HBF4 in CH2Cl2. Before addition
of HBF4 (T = 200 K) (a); after addition of HBF4 at 200 K (b) and 290 K
(c).
Variable-temperature UV–Vis spectra of the homoconjugated PNP
anion were obtained previously by some of us for the equimolar mixture of
PNP and potassium p-nitrophenolate in the presence of excess [18] crown-
6, see Ref. [7].

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
159
at 200 K, CH2Cl2), spectrum a. The dihydrogen complex
and the classical hydride, spectra b and c, correspondingly,
have much weaker and featureless absorptions (at 320 nm
e = 3100 and 970 L molꢁ1 cmꢁ1
, correspondingly, at
200 K in CH2Cl2).
Proton transfer from HFIP was studied by UV–Vis spec-
troscopy, which proved itself to be very useful for the quan-
titative measurements. The interaction of Cp*RuH(dppe)
with 5 equiv. HFIP in dichloromethane at low temperatures
(200–205 K) produces dihydrogen complex 4 (as was con-
Scheme 4.
1
firmed by H NMR spectroscopy) immediately and near
quantitatively, which transforms back to the starting
hydride complex upon temperature increase, see Fig. 13.
The spectrum obtained upon cooling back from 235 to
205 K, is identical to the original one at the same tempera-
ture. These UV–Vis spectral changes indicate, in agreement
with the IR and NMR data, absence of the isomerization
and the reversibility of the proton transfer equilibrium
(Scheme 3) within 200–235 K temperature range. These
enabled us to obtain the equilibrium constant K2 for the
proton transfer step assuming that all hydride 1 is in the
dihydrogen-bonded form 1a and that the equilibrium
involves a second alcohol molecule. The van’t Hoff plot
ꢁ
with the BF4 salt (Table 1), being 599 ms at 220 K. A pos-
sible interpretation of this difference is a stronger hydrogen
bonding between one or two hydride ligands of 3 with the
stronger conjugate base CF3CH2Oꢁ, thereby increasing the
average Hꢂ ꢂ ꢂH distances.
However, TFE gives only partial proton transfer even at
quite high concentrations (up to 20 equiv.), therefore, as
was shown for the iron analogue [7,8], the accurate determi-
nation of the rate constant for the isomerization of
[Cp*Ru(g2-H2)(dppe)]+ (4) obtained with these alcohols
at different temperatures would be thwarted by coupling
(Fig. 13) gives the enthalpy (DHꢀ = ꢁ8.1 0.6 kcal molꢁ1
)
between two key rate constants – k3 (Scheme 4) and k
ꢁ2
and the entropy (DSꢀ = ꢁ17 3 eu) of the proton transfer
(Scheme 3) leading from the dihydrogen complex back to
the starting monohydride. The isomerization rate constant
k3obs for HFIP at 250 K (2.5 equiv. HFIP give ca. 40% of
4) was estimated as 3 · 10ꢁ4 sꢁ1. The use of HBF4 Æ Et2O
allowed studying the process in Scheme 4 under clean first
order conditions. Thus, the dihydrogen complex 4 was gen-
erated quantitatively in situ by low temperature (195 K)
addition of the acid (58% HBF4 Æ Et2O) to the solution of
1 in dichloromethane. The isomerization reaction (Scheme
4) could be studied by IR measurements of the increase of
the dihydride complex 3 mRuH band at 2000 cmꢁ1 (represen-
tative set of spectral changes is shown in Fig. 2). However,
step.
3.6. Isomerization of [Cp*Ru(g2-H2)(dppe)]+ into
[Cp*Ru(H)2(dppe)]+
In the presence of more than 2 equiv. of TFE, the pro-
ton transfer occurs at 200 K in dichloromethane and subse-
quent isomerization of non-classical complex 3 into the
classical one 4 becomes evident above 240 K (Scheme 4).
Interestingly, the T1 min of the dihydride signal of 3
obtained under these conditions increases in comparison
11.5
10.5
9.5
0.75
A
a
0.5
0.25
0
2
K
c
b
8.5
0.0042
0.0044
0.0046
0.0048
250
300
350
400
450
500
1/T, K-1
λ, nm
Fig. 13. UV–Vis spectra of Cp*RuH(dppe) (0.0035 M) in the presence of 5 equiv. HFIP in CH2Cl2 (left) and the corresponding dependence of lnK2 vs
temperature (right). Free hydride at 200 K (a); in the presence of HFIP at 205 K (b) and 235 K (c). The intermediate spectra correspond to 230–205 K (5 K
steps).

160
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
4. Discussion
-0.5
-1.5
-2.5
-3.5
4.1. Proton accepting properties of the hydride ligand
243K
As pointed out in Section 1, the data on the metal influ-
ence on the hydrogen bonding site nature and the proton
accepting properties of the hydride ligand are still scarce.
Herein, we have found that the first step of the
Cp*RuH(dppe) interaction with proton donors is the dihy-
drogen bond formation. The preference of the proton
donor coordination to the hydride ligand is confirmed by
theoretical calculations. The basicity factor Ej of the ruthe-
nium bound hydride ligand (1.39) is only slightly higher
than that of the Cp*FeH(dppe) (Ej = 1.35 [6]). This is quite
surprising, since a much more pronounced basicity increase
down the Group was found for the series of PP3MH2
hydrides (PP3 = j4-P(CH2CH2PPh2)3), Ej values being
equal to 1.12 for M = Fe and 1.33 for M = Ru [4b]. The
possible explanation for this phenomenon could be in the
steric influence of the ancillary ligands, which apparently
impose high steric constraints in the case of
Cp*RuH(dppe). This is ꢀevident from the high hydrogen
bond entropy value DS = ꢁ23.8 eu for Cp*RuH(dppe)/
TFE complex, which is almost twice higher than that for
Cp*FeH(dppe)/TFE (ꢁ13.6 eu), whereas the enthalpy val-
ues differ by only 0.3 kcal molꢁ1 (DHꢀ = ꢁ5.7 and
ꢁ5.4 kcal molꢁ1 for RuH and FeH, respectively).
8.48 10-5 s-1
248K
1.82 10-4 s-1
250K
2.64 10-4 s-1
253K
260K
4.22 10-4 s-1
1.53 10-3 s-1
0
100
200
300
400
t, min
þ
ꢃ
Fig. 14. Logarithmic plots of the ½Cp Ruðg2-H2ÞðdppeÞꢄ BF4 resonance
decay. Temperatures and the corresponding rate constants are indicated
near to the lines.
ꢁ
-35
-36
-37
-38
-39
4.2. Proton transfer mechanism
Despite the steric encumbrances to hydrogen bond for-
mation, proton transfer easily occurs in this system yielding
the cationic non-classical [Cp*Ru(g2- H2)(dppe)]+Aꢁ com-
plex, whose formation is shown here for the first time. The
proton transfer from TFE, HFIP and PNP is not quantita-
tive under the proton donor excesses used and is reversible
within the temperature range 190–230 K, the equilibrium
(Scheme 3) shifting to the right with temperature decrease.
This signals a proton transfer exothermicity. In low polar-
ity media like dichloromethane, ionic species should exist as
contact ion pairs [43], which could additionally be stabi-
lized by hydrogen bonding between the cation and anion.
There is also a question about the anion composition,
which could be additionally bonded to its conjugated acid,
forming the so-called homoconjugate anion [AHA]ꢁ. In the
case of the quite strong acid TFA, the [AHA]ꢁ formation
occurs in the presence of high acid excess. The major spe-
cies in solution is the free CF3COOꢁ anion, the amount
of hydrogen bonded ion pairs being higher in the case of
iron than ruthenium. However, for the weaker proton
donor p-nitrophenol, formation of the hydrogen bonded
ion pair [Cp*(dppe)Ru(g2-H2)]+ꢂ ꢂ ꢂ[ArOHOAr]ꢁ was con-
firmed by means of UV–Vis spectroscopy. Thus, one can
argue that for even weaker proton donors such as TFE
and HFIP, having stronger conjugated bases, the non-clas-
sical complex does exist as the contact ion pair additionally
stabilized by hydrogen bonding between the cation and the
0.0037
0.0038
0.0039
0.004
0.0041
0.0042
1/T, K-1
Fig. 15. Eyring plot for the isomerization rate constant k3.
as in the case of the iron analogue [8], the NMR monitoring
of the process was much more convenient, providing the
concentrations of both non-classical (4) and classical (3)
species as well as directly probing for the formation of pos-
sible decomposition products during the measurements.3
Both the decay of the dihydrogen complex resonance and
the growth of the dihydride complex resonance were moni-
1
tored by H NMR spectroscopy. The non-classical reso-
nance decay gives an excellent fit to the first order rate
law at each temperature (Fig. 14), yielding the rate con-
stants k3 reported in Fig. 14. The Eyring analysis of the rate
constants k3 (Fig. 15) yields the activation parameters
DHꢀ = 20.9 0.8 kcal molꢁ1 and DSꢀ = 9 3 eu. The acti-
vation enthalpy and entropy values are close to those found
for the iron analogue (DHꢀ = 21.6 0.8 kcal molꢁ1
,
DSꢀ = 5 3 eu [8]) and similar to those previously reported
for related ruthenium derivatives [24d,42].
3
The sum of the integrals, after normalization relative to the solvent
peak used as an internal standard, was relatively constant throughout the
reaction, showing almost no decomposition.

N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
161
cess are similar to those established for the iron analogue,
Cp*FeH(dppe), as well as other transition metal hydride
complexes. In the course of this study, the formation of the
previously unknown non-classical complex, [Cp*Ru-
(g2-H2)(dppe)]+, was observed at low temperatures, and
was preceded by the dihydrogen bonding step. The thermo-
dynamic parameters for dihydrogen bond formation and
proton transfer are found to increase on descending the
Group from iron to ruthenium, whereas the activation
parameters for the subsequent [Cp*M(g2-H2)(dppe)]+ !
trans-[Cp*M(H)2(dppe)]+ isomerization are very similar.
Such a comparison has been made for the first time and will
be extended to the osmium analogue in our future work.
CF3
CF3
Ru H + HO CH
P
1
P
-6.7
kcal mol-1
1a
-8.1
kcal mol-1
Ru
H
HOR
P
P
4
Acknowledgments
+
H
H
Ru
The authors are very grateful to Profs. L.M. Epstein,
E.S. Shubina, R. Poli and A. Lledos, friends and tutors,
for their support of our research activities and for the help
in preparation of this manuscript. Thanks are expressed to
Dr. Y. Coppel (LCC, Toulouse) for the assistance in low
temperature NMR experiments. This work was supported
by the European Commission through the HYDROCHEM pro-
gram (contract HPRN-CT-2002-00176), a bilateral PICS
Grant (France–Russia), RFBR Grants (05-03-22001, 05-
03-32415) and the Division of Chemistry and Material Sci-
ences of RAS (Russia). M.B. thanks the Spanish Ministerio
de Educacio´n y Ciencia for a post-doctoral fellowship.
P
P
O
H OR
R
Fig. 16. Energy (DHꢀ) profile for protonation of Cp*RuH(dppe) by HFIP.
homoconjugate anion. Involvement of the second alcohol
molecule in the proton transfer to the hydride ligand
strengthens the primary hydrogen bond, making the activa-
tion barrier lower, and stabilizes the proton transfer prod-
uct through the formation of the hydrogen bonded ion pair
with the homoconjugate anion as was shown both experi-
mentally and theoretically for Cp*FeH(dppe) [6,7]. The
dihydrogen-bonding and [Cp*Ru(g2-H2)(dppe)]+ forma-
tion are immediate in the time-scale of the conventional
spectroscopies used. The equilibrium between these two
species shifts toward the dihydrogen complex when stron-
ger proton donors, higher alcohol/hydride ratios, or lower
temperatures are employed. Assuming the same scheme,
the thermodynamic parameters (DHꢀ and DSꢀ) of proton
transfer from HFIP to Cp*RuH(dppe) were calculated,
which appeared to be more negative than those for the
Cp*FeH(dppe) / HFIP system [7]. The hydrogen bond for-
mation enthalpy for this alcohol was obtained using Eq. (2)
(DHꢀ = ꢁ6.7 kcal molꢁ1) and thus the two-well energy pro-
file (Fig. 16) can be drawn for this system. The energy of the
transition state leading to complex 4 is drawn qualitatively
only. The free activation energy (12–16 kcal molꢁ1) can be
estimated on the basis of the NMR (lack of line broaden-
ing) and IR (lack of spectral evolution at the temperatures
used) results [2]. This is the first time that a comparison can
be made between two metals from the same Group in terms
of the enthalpy profile of a proton transfer reaction. The
data obtained show that both hydrogen bonding and pro-
ton transfer enthalpy increase down the Group.
References
[1] M. Peruzzini, R. Poli (Eds.), Recent Advances in Hydride Chemistry,
Elsevier, Amsterdam, 2001.
[2] N.V. Belkova, E.S. Shubina, L.M. Epstein, Acc. Chem. Res. 38
(2005) 624.
[3] (a) D.M. Heinekey, W.J. Oldham, Chem. Rev. 93 (1993) 913;
(b) G.J. Kubas, Metal Dihydrogen and r-Bond Complexes, Kluwer
Academic/Plenum Press, New York, 2001.
[4] (a) E.I. Gutsul, N.V. Belkova, G.M. Babakhina, L.M. Epstein, E.S.
Shubina, C. Bianchini, M. Peruzzini, F. Zanobini, Russ. Chem. Bull.,
Int. Ed. 52 (2003) 1204;
(b) E.I. Gutsul, N.V. Belkova, M.S. Sverdlov, L.M. Epstein, E.S.
Shubina, V.I. Bakhmutov, T.N. Gribanova, R.M. Minyaev, C.
Bianchini, M. Peruzzini, F. Zanobini, Chem. Eur. J. 9 (2003) 2219.
[5] (a) J. Andrieu, N.V. Belkova, M. Besora, E. Collange, L.M. Epstein,
A. Lledo´s, R. Poli, P.O. Revin, E.S. Shubina, E.V. Vorontsov, Russ.
Chem. Bull. 52 (2003) 2679;
(b) N.V. Belkova, P.O. Revin, M. Besora, M. Baya, L.M. Epstein,
´
A. Lledos, R. Poli, E.S. Shubina, E.V. Vorontsov, Eur. J. Inorg.
Chem. (2006) 2192.
[6] N.V. Belkova, P.O. Revin, L.M. Epstein, E.V. Vorontsov, V.I.
Bakhmutov, E.S. Shubina, E. Collange, R. Poli, J. Am. Chem. Soc.
125 (2003) 11106.
[7] N.V. Belkova, E. Collange, P. Dub, L.M. Epstein, D.A. Lemenovskii,
´
A. Lledos, O. Maresca, F. Maseras, R. Poli, P.O. Revin, E.S.
Shubina, E.V. Vorontsov, Chem. Eur. J. 11 (2005) 873.
[8] M. Baya, O. Maresca, R. Poli, Y. Coppel, F. Maseras, A. Lledo´s,
N.V. Belkova, P.A. Dub, L.M. Epstein, E.S. Shubina, Inorg. Chem.
(2006), in press.
[9] D.-H. Lee, J. Korean Chem. Soc. 36 (1992) 248.
[10] H. Guan, M. Iimura, M.P. Magee, J.R. Norton, G. Zhu, J. Am.
Chem. Soc. 127 (2005) 7805.
5. Conclusions
The combined experimental and theoretical study of the
interaction between Cp*RuH(dppe) and proton donors of
different strengths showed that general features of this pro-

162
N.V. Belkova et al. / Inorganica Chimica Acta 360 (2007) 149–162
[11] K. Kirchner, K. Mauthner, K. Mereiter, R. Schmid, J. Chem. Soc.,
Chem. Commun. (1993) 892.
[21] C. Rømming, K.-T. Smith, M. Tilset, Inorg. Chim. Acta 259 (1997)
281.
[12] F. Morandini, A. Dondana, I. Munari, G. Pilloni, G. Consiglio, A.
Sironi, M. Moret, Inorg. Chim. Acta 282 (1998) 163.
[13] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C.
Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B.
Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin,
R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma,
G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S.
Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G.
Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T.
Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challa-
combe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C.
Gonzalez, J.A. Pople, GAUSSIAN 03, Revision C.02, Gaussian, Inc.,
Wallingford, CT, 2004.
[14] (a) A.D. Becke, J. Chem. Phys. 98 (1993) 5648;
(b) J.P. Perdew, in: P. Ziesche, H. Eschrig (Eds.), Electronic Structure
of Solids, vol. 91, Akademie Verlag, Berlin, 1991, p. 11;
(c) J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1992) 13244.
[15] (a) W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284;
(b) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299.
[16] A. Ho¨llwarth, M. Bohme, S. Dapprich, A. Ehlers, A. Gobbi, V.
Jonas, K. Kohler, R. Stegmann, A. Veldkamp, G. Frenking, Chem.
Phys. Lett. 208 (1993) 237.
[22] R. Poli, M. Baya, R. Meunier-Prest, S. Raveau, New J. Chem. 30
(2006) 759.
[23] O.B. Ryan, M. Tilset, V.D. Parker, J. Am. Chem. Soc. 112 (1990)
2618.
[24] (a) G. Jia, C.-P. Lau, Coord. Chem. Rev. 190–192 (1999) 83;
(b) G. Jia, R.H. Morris, J. Am. Chem. Soc. 113 (1991) 875;
(c) G. Jia, A.J. Lough, R.H. Morris, Organometallics 11 (1992) 161;
(d) M.S. Chinn, D.M. Heinekey, J. Am. Chem. Soc. 112 (1990) 5166.
[25] G. Jia, W.S. Ng, H.S. Chu, W.-T. Wong, N.-T. Yu, I.D. Williams,
Organometallics 18 (1999) 3597.
[26] P.A. Maltby, M. Schlaf, M. Steinbeck, A.J. Lough, R.H. Morris,
W.T. Klooster, T.F. Koetzle, R.C. Srivastava, J. Am. Chem. Soc.
118 (1996) 5396.
[27] (a) K.A. Earl, G. Jia, P.A. Maltby, R.H. Morris, J. Am. Chem. Soc.
113 (1991) 3027;
(b) P.J. Desrosiers, L. Cai, Z. Lin, R. Richards, J. Halpern, J. Am.
Chem. Soc. 113 (1991) 4173.
[28] (a) L.M. Epstein, N.V. Belkova, E.I. Gutsul, E.S. Shubina, Polish J.
Chem. 77 (2003) 1371;
(b) E.S. Shubina, A.N. Krylov, N.V. Belkova, L.M. Epstein, A.P.
Borisov, V.D. Mahaev, J. Organomet. Chem. 493 (1995) 275;
(c) R.B. Girling, P. Grebenik, R.N. Perutz, Inorg. Chem. 25 (1986)
31;
(d) K. Abdur-Rashid, T.P. Fong, B. Greaves, D.G. Gusev, J.G.
Hinman, S.E. Landau, A.J. Lough, R.H. Morris, J. Am. Chem. Soc.
122 (2000) 9155.
[29] L.M. Epstein, N.V. Belkova, E.S. Shubina, in: M. Peruzzini, R. Poli
(Eds.), Recent Advances in Hydride Chemistry, Elsevier, Amsterdam,
2001, p. 391.
[17] (a) R. Ditchfield, W.J. Hehre, J.A. Pople, J. Chem. Phys. 54 (1971)
724;
[30] A.V. Iogansen, Spectrochim. Acta A 55 (1999) 1585.
[31] A.V. Iogansen, Theor. Exp. Khim. 7 (1971) 312.
[32] A.V. Iogansen, Theor. Exp. Khim. 7 (1971) 302.
[33] N.V. Belkova, E.S. Shubina, E.I. Gutsul, L.M. Epstein, I.L.
Eremenko, S.E. Nefedov, J. Organomet. Chem. 610 (2000) 58.
[34] V.I. Bakhmutov, Eur. J. Inorg. Chem. (2005) 245.
[35] S.O. Morgan, H.H. Lowry, J. Phys. Chem. 34 (1930) 2385.
[36] S. Sharif, G.S. Denisov, M.D. Toney, H.-H. Limbach, J. Am. Chem.
Soc. 128 (2006) 3375.
(b) W.J. Hehre, R. Ditchfield, J.A. Pople, J. Chem. Phys. 56 (1972)
2257;
(c) P.C. Hariharan, J.A. Pople, Mol. Phys. 27 (1974) 209;
(d) M.S. Gordon, Chem. Phys. Lett. 76 (1980) 163;
(e) P.C. Hariharan, J.A. Pople, Theo. Chim. Acta 28 (1973) 213;
(f) J.-P. Blaudeau, M.P. McGrath, L.A. Curtiss, L. Radom, J. Chem.
Phys. 107 (1997) 5016;
´
(g) M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, D.J. DeFrees,
J.A. Pople, M.S. Gordon, J. Chem. Phys. 77 (1982) 3654;
(h) R.C. Binning Jr., L.A. Curtiss, J. Comput. Chem. 11 (1990) 1206;
(i) V.A. Rassolov, J.A. Pople, M.A. Ratner, T.L. Windus, J. Chem.
Phys. 109 (1998) 1223;
[37] N.V. Belkova, M. Besora, L.M. Epstein, A. Lledos, F. Maseras, E.S.
Shubina, J. Am. Chem. Soc. 125 (2003) 7715.
[38] N.V. Belkova, A.V. Ionidis, L.M. Epstein, E.S. Shubina, S. Gruu-
endemann, N.S. Golubev, H.-H. Limbach, Eur J. Inorg. Chem.
(2001) 1753.
(j) V.A. Rassolov, M.A. Ratner, J.A. Pople, P.C. Redfern, L.A.
Curtiss, J. Comput. Chem. 22 (2001) 976.
[39] E.S. Shubina, N.V. Belkova, L.M. Epstein, J. Organomet. Chem.
536–537 (1997) 17.
[18] S.F. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553.
[19] (a) M.I. Bruce, M.G. Humphrey, A.G. Swincer, R.C. Wallis, Aust. J.
Chem. 37 (1984) 1747;
[40] J.Z. Dega-Szafran, M. Grundwald-Wyspianska, M. Szafran, Spec-
trochim. Acta A 47 (1991) 543.
[41] A.B.P. Lever, Inorganic Electronic Spectroscopy, second ed., Else-
vier, Amsterdam, 1984.
(b) H. Guan, M. Iimura, M.P. Magee, J.R. Norton, K.E. Janak,
Organometallics 22 (2003) 4084.
´
´
[42] I. de Los Rıos, M. Jimenez-Tenorio, J. Padilla, M.C. Puerta, P.
[20] K.-T. Smith, C. Rømming, M. Tilset, J. Am. Chem. Soc. 115 (1993)
8681.
Valerga, Organometallics 15 (1996) 4565.
[43] A. Macchioni, Chem. Rev. 105 (2005) 2039.