The effect of ligand basicity on the unconventional hydrogen-bond in H(μ-H)Os3(CO)10L (L=amine) derivatives

Article abstract of DOI:10.1016/S0020-1693(02)00863-0

The reaction of amine ligands with the unsaturated 46 e^- cluster Os₃(μ-H₂)(CO)₁₀ yields a series of complexes of formula Os₃H(μ-H)(CO)₁₀(amine). In these derivatives the presence of an unconventional hydrogen-bond M-H?H-N is detected by the evaluation of proton spin-lattice relaxation times. In order to elucidate the relationship between pK_b and H?H length, the M-H?H-N distances were quantified for amines of known pK_b. In practice, this was achieved by comparing the T₁ values of the terminal hydride resonances of the protonated isotopomer with the T₁ values of the corresponding derivative containing selectively deuterated amine. A nice correlation between the M-H?H-N distances and the pK_b of the amine ligands was found. DFT calculations were carried out for the cluster on Os₃H(μ-H)(CO)₁₀(Benzylamine) to gain a fuller understanding of the possible coordination geometries of an amine ligand on the surface of a tri-osmium cluster.

Full text of DOI:10.1016/S0020-1693(02)00863-0

Inorganica Chimica Acta 334 (2002) 448Á454
/
www.elsevier.com/locate/ica
The effect of ligand basicity on the unconventional hydrogen-bond in
amine) derivatives
H(m-H)Os3(CO)10L (Lꢀ
/
Silvio Aime a, Fabrizio Bertone a, Roberto Gobetto a,*, Luciano Milone a,
Andrea Russo a, Marc J. Stchedroff a, Marco Milanesio b
a Dipartimento di Chimica Inorganica, Chimica Fisica, Chimica dei Materiali, Universita` degli Studi di Torino, via Pietro Giuria 7, 10125 Turin, Italy
b Dipartimento di Scienze e Tecnologie Avanzate, Universita` del Piemonte Orientale, Corso T. Borsalino 54, 15100 Alessandria, Italy
Received 12 December 2001; accepted 26 February 2002
Abstract
The reaction of amine ligands with the unsaturated 46 eꢁ cluster Os3(m-H2)(CO)10 yields a series of complexes of formula Os3H(m-
H)(CO)10(amine). In these derivatives the presence of an unconventional hydrogen-bond MÃ
/
HÁ Á ÁHÃ
/
N is detected by the evaluation
N
of proton spin-lattice relaxation times. In order to elucidate the relationship between pKb and HÁ Á ÁH length, the MÃ
/
HÁ Á ÁHÃ
/
distances were quantified for amines of known pKb. In practice, this was achieved by comparing the T1 values of the terminal
hydride resonances of the protonated isotopomer with the T1 values of the corresponding derivative containing selectively
deuterated amine. A nice correlation between the MÃ
calculations were carried out for the cluster on Os3H(m-H)(CO)10(Benzylamine) to gain a fuller understanding of the possible
coordination geometries of an amine ligand on the surface of a tri-osmium cluster. # 2002 Elsevier Science B.V. All rights reserved.
/
HÁ Á ÁHÃ
/
N distances and the pKb of the amine ligands was found. DFT
Keywords: Triosmium clusters; Dihydrogen-bond; Spin-lattice relaxation; NMR; DFT; Amine; 187Os NMR
1. Introduction
approaching Ndꢁ Ã
ability strengthens the HÁ Á ÁH interaction and (c) the
influence of trans-substituents on the dꢁ charge of the
/
Hdꢂ moiety, as a greater polariz-
Since it was first reported [1], the occurrence of weak
electrostatic interactions between metal hydrides and
hydrogens bonded to an electronegative main group
atom (N or O) have become widely acknowledged [2].
Weak HÁ Á ÁH interactions may even be responsible for
the stabilization of transient species, such as catalytic
intermediates. Investigations of a number of metal
complexes by IR spectroscopy [3] have established that
/
hydride ligand [5]. The description view of the hydride
as a soft and polarizable species would be corroborated
by observation of a relationship between the experimen-
tally determined HÁ Á ÁH distance and the ligand basicity.
H(m-H)Os3(CO)10(XÃ
/
H) adducts were chosen for this
study as they are easily formed by the reaction of the
cluster (m-H2)Os3(CO)10 with XÃH containing bases.
/
Previously we have developed a methodology for
evaluating the HÁ Á ÁH distances by T1 measurements,
and have structurally characterized several imine [6] or
the strength of these bonds is in the range 3Á
molꢁ1. Theoretical studies support the view that MÃ
HÁ Á ÁHÃX interactions (where XꢀO or N) [4] involve
/
7 kcal
/
/
/
amine [7] systems containing the MÃ
/
HÁ Á ÁHÃ
/
N interac-
polarization of the HÁ Á ÁH interaction leading to a
formally polarized MÃ
/
HdꢁÁ Á ÁHdꢂ Ã
/
X interaction. It
tions. DFT calculations have now been performed on
Os3H(m-H)(CO)10(Benzylamine) cluster (shown in Fig. 1
with the labeling scheme) to gain a fuller understanding
of the possible coordination geometries of an amine
ligand on the surface of a tri-Osmium cluster. Particular
attention was paid to the possibility of the amine group
has been suggested that there are three key factors that
affect such an interaction: (a) a geometric arrangement
which favors the close approach of the two interacting
groups; (b) the polarizability of the Mdꢂ Ã
/
Hdꢁ upon the
to rotate around the Os2Ã
/
Os1Ã
/
NÃ/C torsion angle.
Transition metal chemical shifts are a very sensitive
indication of chemical environment [8]. With this in
* Corresponding author. Tel.: ꢂ
6707 855.
/
39-011-6707 523; fax: ꢂ39-011-
/
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 0 8 6 3 - 0

S. Aime et al. / Inorganica Chimica Acta 334 (2002) 448Á
/
454
449
2.3. NMR methods
Non selective inversion recovery was used to obtain
1H T1 values [12]. Samples for T1 measurements were
degassed using standard freezeÁ
/
pumpÁthaw techniques.
/
Spectra were acquired at 90, 270 and 400 MHz 1H.
Similarly, selectively deuterated samples were prepared
and inversion recovery was used to obtain 1H T1 values.
From the differences in the observed T1 values, caused
by deuteration, the effects of the HÁ Á ÁH weak interaction
were evaluated. Temperature calibrations were carried
Fig. 1. The model compound for the theoretical calculations with the
adopted labeling scheme indicated, showing the degrees of freedom
(black arrows), explored by the conformational search.
1
out with a standard methanol H thermometer. Errors
in the reported T1 values were estimated to be in the 9
/
2% range.
mind, we have also attempted, with only partial success,
to relate changes in the amine basicity with the observed
187Os chemical shifts.
2.4. 187Os chemical shift determination
For the gradient selected HMQC and HMBC experi-
ments, the gradient ratios were calculated according to
the g of the individual nuclei using the spectrometer
2. Experimental
software [13]. For 1HÁ187
Os HMQC experiments,
/
gradient ratios of 50:21:30 were used.
2.1. Materials
2.5. 2D Acquisition and processing strategy
Routine NMR spectra were acquired on a JEOL
1
EX400 spectrometer operating at 400 MHz H, a JEOL
Initially, a spectrum was acquired with a spectral
width of 100 kHz in the indirect dimension to ensure
that the signal was not folded: ultimately this was
reduced to 20 kHz and the number of increments
increased to 512. The data were then transformed into
a 2 by 4K data matrix with a SINE function. In all cases
a magnitude calculation was performed in F1. Chemical
shifts were referenced using the J convention, with JOs
taken as 2.303164 [13].
1
GSX270 operating at 270 MHz H, or a JEOL EX90
spectrometer operating at 90 MHz 1H. Gradient selected
spectra were acquired using
equipped with an XYZ 10A gradients amplifier (z-
a Bruker DMX600
gradient strength 0.05 T mꢁ1) and a 5 mm BBO
1
probehead (inner coil X, outer coil H) equipped with
self-shielding z-gradients. Os3(CO)12 and Os3(m-
H)2(CO)10 were prepared as previously reported [9,10].
Os3(m-H)2(CO)10 was recrystallized before use. 13CO
(99% 13C) gas was used as supplied by Isotec, Os3(CO)12
was enriched as previously reported by gaseous ex-
change in octane solution at elevated temperature [11].
Benzylamine, 4-aminopyridine, p-anisidine, aniline, 4-
chloroaniline, 3-fluoroaniline, 3-nitroaniline, 2,4-di-
chloroaniline, 4-nitroaniline were used as purchased
from Aldrich. Selectively deuterated amines were pre-
pared by the addition of 20 ml of D2O to 4 mg of the
ligand in a Younge’s valve NMR tube. The mixture was
shaken and allowed to stand for 1 min and then 0.7 cm3
of CD2Cl2 was added.
2.6. Theoretical calculations
DFT SCF-MO calculations [14] were performed on
Os3H(m-H)(CO)10(NPhH2) (see Fig. 1), employing the
BLYP [15] and B3-LYP [16] methods, as implemented in
the GAUSSIAN 98 [17] program. The initial geometry of 1
was obtained by graphical manipulation (with the
MOLDRAW [18] program) of the crystal structure of the
related compound H(m-H)Os3(CO)10(HNÄ
/CPh2) [19],
after which an exhaustive conformational search was
performed, employing the MOLDRAW [17] and SPARTAN
[20] programs. Geometrical optimizations were carried
out at BLYP level [21], using Stuttgart RSC ECP
pseudo-potentials and a related basis set [22] for the
Os atoms and different basis sets, up to the 6-31G(d,p)
[23], for the remaining atoms. Since full geometry
2.2. The in situ synthesis of H(m-H)Os3(CO)10(L)
In a typical reaction, 10 mg of Os3(m-H)2CO)10
(1.17ꢃ
10ꢁ5 mol) was dissolved in 0.7 cm3 of CD2Cl2
/
and transferred into a Younge’s valve NMR tube. A 2
molar equiv. of base was then added. In the case of the
more reactive bases (1 and 2) the solution immediately
changed from purple to yellow, in the other cases there
was little immediate observable color change.
optimizations gave unreasonably long OsÃ
the OsÃOs distances were constrained to the values
observed in the crystal structure of the related com-
pound Os3H(m-H)(CO)10(HNÄCPh2) [19]. Single point
energy calculations, using the BLYP/6-31G(d,p) opti-
/
Os bonds,
/
/

450
S. Aime et al. / Inorganica Chimica Acta 334 (2002) 448Á454
/
mized geometries, were carried out at the DFT level by
adopting the hybrid B3-LYP [15] functional, employing
the Stuttgart RSC ECP for the Os atoms and different
minimum, the actual tc value can easily be evaluated
(tcꢀ
calculated from the dipolar relaxation terms as pre-
/
0.62 per v0). Thus the HÁ Á ÁH distances may be
basis sets, up to 6-311ꢂ
atoms. Only the results derived for the highest level of
calculation [B3LYP/6-311ꢂ G(2d,2p)//BLYP/6-
/
ꢂ
/
G(2d,2p), for the remaining
viously reported [19].
However, there are possible errors in such a basic
procedure as this would have neglected the differences in
/
ꢂ
/
31G(d,p)] will be discussed here.
the OsÃ
/
H dipolar interaction between the two hydrides
and also the possibility that the NÃ
/H proton gives a
dipolar contribution to the bridging hydride. Our
proven methodology consists of determining the differ-
ences in the relaxation rates of the two isotopomers H,H
and H,D which then allows us to quantify the dipolar
contribution in the relaxation of a given hydride and
then to evaluate the HÁ Á ÁH distances.
3. Results and discussion
The violet 46 eꢁ, formally unsaturated [24], cluster
Os3(m-H)2(CO)10 readily reacts, in CD2Cl2 solution at
273 K, with amine ligands (L) to form the corresponding
Whilst there is no great difference in the results
obtained by the compensated and uncompensated
methods, we believe that the selective deuteration
method is likely to be more accurate. However, in
most cases the difference is of a similar magnitude to the
H(m-H)Os3(CO)10L derivative (Lꢀ
/
benzylamine 1, p-
anisidine 2, aniline 3, 4-chloroaniline 4, 3-fluoroaniline
5) while 3-nitroaniline, 2,4-dichloroaniline and 4-nitroa-
niline do not react at all (Scheme 1).
1
The low temperature limiting H spectrum (213 K,
400 MHz) of the adducts shows two major resonances at
experimental error. The HÃ/H distances evaluated by
approximately ꢁ
/
10.4 and ꢁ15.5 ppm, which are
/
selective deuteration method are reported in Table 1.
Whilst there was not 100% formation of the cluster
ligand adduct, except in the case of 1, we observed that
the stronger the basicity of the ligand the higher the
percentage of adduct formed. More interestingly the
assigned to the terminal and bridging hydride ligands,
respectively, the precise shift varies according to the
ligand used (see Table 1).
Some of us have previously demonstrated that long-
itudinal relaxation time measurements (T1) provide a
viable method for the determination of the HÁ Á ÁH
distances in polyhydridic cluster systems [6,7,25]. As
we wished to evaluate any correlation between the
interaction strength and the ligand basicity, for a
homologous series, we were constrained to use ligands
of known basicity. Measurements of all the adducts
stronger the base the longer the interproton OsÃ
/
HÁ Á ÁHÃ
/
N distance in the resulting adduct (see Table 1).
As all the nitrogen bases used in this work are primary
amines, an open question is whether only one or both
the hydrogen atoms of the NH2 moiety interact with the
terminal hydride. This has been tackled by an extended
conformational search on compound 1 (see Fig. 1) using
DFT to find its possible stable conformations. As
carbonyl and hydride ligands have no conformational
flexibility, they are therefore kept fixed in the conforma-
tional search. On the other hand, the degrees of freedom
of the amine ligand were fully assessed by moving
around the three torsion angles indicated by the arrows
in Fig. 1. From this search process three conformers, c1,
c2 and c3 shown in Fig. 2, were found. The relative
stability and main geometrical features of these con-
formers are summarized in Table 2. These three
formed (1Á
terminal hydride had a shorter relaxation time than its
bridging counterpart. The contribution from 189Os (Iꢀ
/5) reveals that for any given adduct the
/
3/2) cannot be responsible for this as the bridging
hydride would be expected to relax faster (as it has
32% probability of being bound to 189Os) than the
terminal hydride (which had a 16% probability of being
bound to 189Os). Moreover, the T1 measurements at
three different magnetic field strengths allow us to rule
out any preferential contribution, from the chemical
shift anisotropy term, to the terminal hydride. One may,
therefore, conclude that the difference in observed T1’s
between bridging and terminal hydride comes mainly
from the dipolar contribution of the amine proton.
The relaxation rates of the terminal and bridging
hydrides, were then measured as function of the
temperature in order to calculate the minimum T1
values for each species. At the temperature of T1
conformers differ mainly in the value of the Os1Ã
/
Os2ÃNÃC torsion angle, which result in different
/
/
intra-molecular contact patterns (shown as dotted lines
in Fig. 2). Both c1 and c2 conformers form a single di-
hydrogen bond (DHB), involving H4 and H3, respec-
tively. Whilst the geometry of conformer c1 was that
expected from the crystal structure (see for example the
related compound H(m-H)Os3(CO)10(HNÄ/CPh2) [19],
conformer c2 displays a quite different arrangement, in
which the amine group points toward the axial carbonyl
on Os3, but without showing particularly short contacts
with this ligand (see Fig. 2(b)). Conformer c3 is
Scheme 1. Reaction scheme leading to Os3H(m-H)(CO)10NH2R1.
characterized by the presence of a NÃ
/
HÁ Á ÁOC interac-

Table 1
NMR data for the Os3H(m-H)(CO)10L adducts
Ligand
dH1 (ppm)
dH2 (ppm)
Kb1
%
rHÃH 10ꢁ10
m
dOs2 (ppm) d 187Os (N)
dOs1 (ppm)
JOs2ÃH1 (Hz)
JOs2ÃH2 (Hz)
JOs1ÃH2 (Hz)
Benzylamine 1
p-anisidine 2
Aniline 3
ꢁ10.47
ꢁ10.42
ꢁ10.38
ꢁ10.47
ꢁ10.47
ꢁ16.22
ꢁ15.54
ꢁ15.56
ꢁ15.53
ꢁ15.52
2.14ꢃ10ꢁ5
2.19ꢃ10ꢁ9
4.27ꢃ10ꢁ10
1.41ꢃ10ꢁ10
3.16ꢃ10ꢁ11
2.92ꢃ10ꢁ12
1.12ꢃ10ꢁ12
1.00ꢃ10ꢁ13
100.0 1.99
95.8 1.95
ꢁ13288
ꢁ13740
ꢁ13430
ꢁ13480
ꢁ13780
ꢁ12291
ꢁ12568
ꢁ12267
ꢁ12311
ꢁ12615
37.4
37.9
38.7
37.1
39.9
33.5
35.5
34.6
35.0
34.0
38.0
37.6
36.1
37.1
37.8
89.7 1.93
32.3 1.93
18.9 1.90
4-chloroaniline 4
3-fluoroaniline 5
3-nitroaniline 6
2,4-dichloroaniline 7
4-nitroaniline 8
0
0
0
For atoms labeling see Fig. 1. rHÃH were measured by selective deuteration T1 methods: ꢁ70 8C in CD2Cl2 at 400 MHz. % refers to the percentage of the product with respect to the starting
amount of Os(m-H)2(CO)10
.

452
S. Aime et al. / Inorganica Chimica Acta 334 (2002) 448Á454
/
˚
Fig. 2. Stable energy minima for compound 1 (HÁ Á ÁH distance in A).
amine hydrogen atoms are involved in a DHB interac-
tion (conformation c4 in Fig. 3) is not a minimum
energy conformation on the BLYP/6-31G(d,p) potential
surface. It is worth noting that geometrical optimization
conformation c4 (unstable) leads to conformer c3
(stable), which is a minimum on the BLYP/6-31G(d,p)
potential surface. Therefore, it could be suggested that
H1Á Á ÁOC contact in conformer c3 is more stabilizing (or
less repulsive) than H1Á Á ÁCO contact in conformation
c4. However, a better steric arrangement of the ligands
may equally well be the driving force for the stabiliza-
tion of conformer c3 with respect to c4.
Table 2
Geometrical parameters, energetic features and Mulliken charges of
the conformations for compound 1 obtained after the conformational
ꢁ1
˚
search (angles in 8, distances in A, relative stability in kcal mol
)
Conformer
NMR data
c1
c2
c3
Geometrical parameter
Os2ÃOs1ÃNÃC
H1Á Á ÁH3
162
239
3.10
204
1.93
2.48
2.14
ꢂ0.8
1.99
H1Á Á ÁH4
2.87
0.34
1.73
0.0
a
DE
Mulliken charges
˚
The HÁ Á ÁH distance of 1.99 A found for 1 by NMR
H1
H2
H3
H4
ꢁ0.58
ꢁ0.41
0.20
ꢁ0.59
ꢁ0.44
0.24
ꢁ0.55
ꢁ0.39
0.15
data in CD2Cl2 solution should be an average of the
three conformers depicted in Fig. 2, with a rapid
exchange between them. This dynamic exchange in
solution is able to explain the differences between the
observed NMR distance in solution, and that from
theoretical calculations, as already demonstrated for the
0.22
0.23
0.25
a
Based on a B3LYP/6-311ꢂꢂG(2d,2p)//BLYP/6-31G(d,p) calcu-
lations (see supplementary material for further details).
related Os3H(m-H)(CO)10(HNÄ
/
CPh2) compound [19].
tion, besides the OsÃ
/
HÁ Á ÁHÃ
/
N bond, as shown in Fig.
The other amine derivatives studied by NMR may
probably be described by a similar dynamic model,
since the small changes on the amine ligand should not
be able to alter significantly the relative stability of
conformations c1, c2 and c3. It is worth noting that the
smaller benzylamine ligand in compound 1 may undergo
larger molecular motions with respect to the bulkier
2(c).
DFT calculations indicate that the three conformers
in Fig. 2 have similar stability [23]. Whilst the hypothe-
tical ‘bi-dentate’ DHB conformation, in which both
benzophenoneimine in Os3H(m-H)(CO)10(HNÄ
/
CPh2),
where movements are limited to small vibration around
the conformation observed in the X-ray structure.
3.1. 187Os NMR spectroscopy
We were interested to see if trends in the ligand
basicity were reflected in the metal chemical shifts. If we
were able to measure the shifts of all the three metals it
would, at least in principle, determine the actual effects
of the steric and electronic parameters. Previously the
chemical shifts of a series of related compounds Os3(m-
Fig. 3. The unstable c4 conformation, with the ‘bi-dentate’ DHB
shown as dotted lines and the probably unfavorable contact as dashed
arrow.
H)2(CO)8m-L2 (L2ꢀ
/
dppm, dppe, dppp and BINAP)

S. Aime et al. / Inorganica Chimica Acta 334 (2002) 448Á
/
454
453
and Os3(m-H)2(CO)10 have been measured and found to
be indicative of the metal-ligand ring size, as were the
metal phosphine couplings and the hydride couplings
[26].
5. Supplementary material
All the details about relative stability at the different
level of theory are given in table S1, available as
supplementary material.
In the H(m-H)Os3(CO)10L derivatives there are of
course three different osmium environments. Unfortu-
nately the coupling from the bridging hydride to the
remote osmium site proved to be too small to make such
a measurement even by HMBC; for example, in the case
of Os3(m-H)2(CO)10 the coupling is only 1.4 Hz and,
given that the hydride is likely forced further out of
plane in H(m-H)Os3(CO)10L derivatives, the coupling
will be even smaller. For the osmium atom bearing the
bridging and the terminal hydrides a reasonable correla-
tion was found between the basicity and chemical shift
for adducts 1, 3, 4 and 5. We observed a high field shift
of the osmium atom when the basicity of the
amine ligand decreases. However, it is evident that other
factors are involved in the case of 2. Disappointingly
no discernible trends have been found for
the metal bearing the N donor and the bridging
hydride.
Acknowledgements
We gratefully acknowledge the Italian MURST for a
grant (COFIN 2000). We acknowledge INSTM for
providing privileged access to the CINECA supercom-
puting facility. M.M. thanks the ‘Fondazione G. Done-
gani’ for a grant and D. Viterbo of the University of
Piemonte Orientale for useful suggestions. The EU also
provided funding for the TMR project, contract number
ERBFMRXCT960091, for M.J. Stchedroff.
References
[1] (a) J.C. Lee, E. Peris, A.L. Rheingold, R.H. Crabtree, J. Am.
Chem. Soc. 116 (1994) 11014;
(b) J.C. Lee, A.L. Rheingold, B. Muler, P.S. Pregosin, R.H.
Crabtree, J. Chem. Soc., Chem. Commun. (1994) 1021;
(c) S. Park, R. Ramachandra, A.J. Lough, R.H. Morris, J. Chem.
Soc., Chem. Commun. (1994) 2201.
4. Conclusions
[2] R.H. Crabtree, P.E.M. Siegbahn, O. Eisenstein, A.L. Rheingold,
T.F. Koetzle, Acc. Chem. Res. 29 (1996) 348 (and references
therein).
In this study we were able to estimate by T1
measurements the intramolecular hydrogen-bond dis-
[3] L.M. Epstein, E.S. Shubina, J. Phys. Chem. 102 (1998) 359.
[4] D. Braga, F. Grepioni, E. Tedesco, M.J. Calhorda, E.M. Lopes,
New J. Chem. (1999) 219.
tances between metal hydrides and NÃ
in a series of H(m-H)Os3(CO)10(L) (Lꢀ
tives.
/
H bonds present
/amine) deriva-
[5] E. Peris, J.C. Lee, J.R. Rambo, O. Einsentein, R.H. Crabtree, J.
Am. Chem. Soc. 117 (1995) 3485.
Moreover, we demonstrated that there is a clear
relationship between the basicity of the ligand, the
magnitude of the weak interaction and the interprotonic
HÁ Á ÁH distance.
[6] (a) S. Aime, M. Ferriz, R. Gobetto, E. Valls, Organometallics 18
(1999) 2030;
(b) S. Aime, M. Ferriz, R. Gobetto, E. Valls, Organometallics 19
(2000) 707.
[7] S. Aime, R. Gobetto, E. Valls, Organometallics 16 (1997) 5140.
[8] W. Leitner, M. Buehl, R. Fornika, C. Six, W. Baumann, E.
Dinjus, M. Kessler, C. Krueger, A. Rufinska, Organometallics 18
(1999) 1196.
The large freedom of motion of the mono-aromatic
amines allows the presence of three possible conformers,
with different hydrogen-bond patterns. However, ex-
perimentally a unique stable conformer was found for
[9] C.W. Bradford, R.S. Nyholm, Chem. Commun. (1967) 384.
[10] H. Kaesz, Chem. Brit. (1973) 344.
the related H(m-H)Os3(CO)10(HNÄ
/
CPh2) compound
[11] See for example: (a) S. Aime, W. Dastru, R. Gobetto, J. Krause,
L. Violano, Inorg. Chim. Acta 235 (1995) 357;
(b) S. Aime, W. Dastru, R. Gobetto, A.J. Arce, Organometallics
13 (1994) 3737;
[19], where the bulkier bi-aromatic HNÄ
/CPh2 ligand
causes greater rigidity. It has been shown for this
compound that Mulliken charges of opposite sign are
present in the hydride and the iminic proton, confirming
an electrostatic contribution to the DHB [19]. The
results obtained in this work qualitatively agree with
this view. Thus a more electron depleted aminic hydro-
gen is expected when the basicity of the amine ligand
decreases thus yielding a stronger DHB interaction. The
correlation between the DHB distances and the amine
basicity indicates that a simple polarizable model could
explain the observed interaction between the hydride
and the soft amine ligand.
(c) S. Aime, R. Bertoncello, V. Busetti, R. Gobetto, G. Granozzi,
D. Osella, Inorg. Chem. 25 (1986) 4004.
[12] (a) A. Abragham, The Principles of Nuclear Magnetic Resonance,
Oxford University Press, Oxford, 1961;
(b) A.E. Derome, Modern NMR Techniques for Chemistry
Research, Pergamon, Oxford, 1987;
(c) T.D.W. Claridge, High-Resolution NMR Techniques in
Organic Chemistry, Tetrahedron Organic Chemistry Series, vol.
19, Pergamon, Oxford, 1999.
[13] S. Berger, R. Braun, K.O. Kowinski, 150 and More NMR
Experiments, Wiley VCH, Weinheim, 1999.
[14] Various platforms, among which the SGI O3K supercomputer of
CINECA (www.cineca.it) were employed.

454
S. Aime et al. / Inorganica Chimica Acta 334 (2002) 448Á454
/
[15] Becke88 exchange functional: A.D. Becke, Phys. Rev. A 38 (1988)
3098 and LYP correlation functional: C. Lee, W. Yang, R.G.
Parr, Phys. Rev. B 37 (1988) 785.
[20] SPARTAN 5.0.2, Wavefunction Inc., 18401 Von Karman Ave., Ste.
370 Irvine, CA 92612 USA.
[21] BLYP, a pure DFT functional, was chosen to fully exploit the
calculation speedup, obtainable using the fast multipole
method (FMM algorithm, see J.M. Milliam, G.E. Scuseria,
J. Chem. Phys. 106 (1997) 5569), as implemented in Gaussian
98.
[16] A.D. Becke J. Chem. Phys. 98 (1993), p. 5648.
[17] Gaussian 98, Revision A.7, M.J. Frisch, G.W. Trucks, H.B.
Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G.
Zakrzewski, J.A. Montgomery, Jr., R.E. Stratmann, J.C. Burant,
S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V.
Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T.
Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonza-
lez, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W.
Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S.
Replogle, J.A. Pople (1998) Gaussian, Inc., Pittsburgh, PA.
[18] P. Ugliengo, D. Viterbo, G. Chiari, Z. Kristallogr. 207 (1993) 9
[22] D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss,
Theor. Chim. Acta 77 (1990) 123. STUTTGART ECPs can be
retrieved in the following web site: http://www.theochem.uni-
stuttgart.de/.
[23] W.J. Hehre, L. Radom, P.V. Schleyer, J. Pople, Ab Initio
Molecular Orbital Theory, first ed. (Chapter 4), Wiley, New
York, 1986.
[24] M.P. Brown, P.A. Dolby, M.M. Harding, A.J. Matthews, A.K.
Smith, J. Chem. Soc., Dalton Trans. 11 (1993) 1671.
[25] S. Aime, M. Cisero, R. Gobetto, D. Osella, A.J. Arce, Inorg.
Chem. 30 (1991) 1614.
[26] (a) A.E. Aliev, A.J. Deeming, M.J. Stchedroff, J.W. Steed,
Organometallics, submitted for publication;
(web site:
draw.html).
http://www.ch.unito.it/ifm/fisica/moldraw/mol-
(b) M.J. Stchedroff, S. Aime, R. Gobetto, L. Salassa, E.
Nordlander, Magnetic Reson. Chem. 40 (2002) 107.
[19] S. Aime, E. Diana, R. Gobetto, M. Milanesio, E. Valls, D.
Viterbo, Organometallics 21 (2002) 50.