1H- and 2H-T1 relaxation behavior of the rhodium dihydrogen complex [(triphos)Rh(η2-H2)H2]+

Article abstract of DOI:10.1021/ic990667q

Protonation of the classical trihydride [(triphos)RhH₃] (2) at 210 K in either THF or CH₂Cl₂ by either HBF₄·OMe₂ or CF₃SO₂OH gives the nonclassical η²-H₂ complex [(triphos)Rh(η²-H₂)H₂]⁺ (1) [triphos = MeC(CH₂PPh₂)₃]. Complex 1 is thermally unstable and highly fluxional in solution. In THF above 230 K, 1 transforms into the solvento dihydride complex [(triphos)Rh(η¹-THF-d₈)H₂]⁺ (5) that, at room temperature, quickly converts to the stable dimer trans-[{(triphos)RhH}₂(μ-H)₂]²⁺ (trans-6). In CH₂Cl₂, 1 is stable up to 240 K. Above this temperature, the η²-H₂ complex begins to convert into a mixture of trans- and cis-6, which, in turn, transform into the bridging-chloride dimers trans- and cis-[{(triphos)RhH}₂(μ-Cl)₂]²⁺ at room temperature. Complex 1 contains a fast-spinning H₂ ligand with a T₁(min) of 38.9 ms in CD₂Cl₂ (220 K, 400 MHz). An NMR analysis of the bis-deuterated isotopomer [(triphos)RhH₂D₂]⁺ (1-d₂) did not provide a J(HD) value. At 190 K, the perdeuterated isotopomers [(triphos)RhD₃] (2-d₃) and 1-d₄ show T₁(min) values of 16.5 and 32.6 ms (76.753 MHz), respectively, for the rapidly exchanging deuterides. An analogous 2-fold elongation of T₁(min) is also observed on going from [(triphos)IrD₃] to [(triphos)Ir(η²-D₂)D₂]⁺. A rationale for the elongation of T₁(min) in nonclassical polyhydrides is proposed on the basis of both the results obtained and recent literature reports.

Full text of DOI:10.1021/ic990667q

Inorg. Chem. 2000, 39, 1655-1660
1655
2
1H- and H-T1 Relaxation Behavior of the Rhodium Dihydrogen Complex
[(Triphos)Rh(η2-H2)H2]+
Vladimir I. Bakhmutov,*,† Claudio Bianchini,*,‡ Maurizio Peruzzini,*,‡ Francesco Vizza,‡ and
Evgeny V. Vorontsov†
Institute of Organo-Element Compounds of the Russian Academy of Sciences, 28 Vavilova,
Moscow 117813, Russia, and Istituto per lo Studio della Stereochimica ed Energetica dei
Composti di Coordinazione, ISSECC-CNR, Via Jacopo Nardi, 39, 50132 Firenze, Italy
ReceiVed June 9, 1999
Protonation of the classical trihydride [(triphos)RhH3] (2) at 210 K in either THF or CH2Cl2 by either HBF4‚OMe2
or CF3SO2OH gives the nonclassical η2-H2 complex [(triphos)Rh(η2-H2)H2]+ (1) [triphos ) MeC(CH2PPh2)3].
Complex 1 is thermally unstable and highly fluxional in solution. In THF above 230 K, 1 transforms into the
solvento dihydride complex [(triphos)Rh(η1-THF-d8)H2]+ (5) that, at room temperature, quickly converts to the
stable dimer trans-[{(triphos)RhH}2(µ-H)2]2+ (trans-6). In CH2Cl2, 1 is stable up to 240 K. Above this temperature,
the η2-H2 complex begins to convert into a mixture of trans- and cis-6, which, in turn, transform into the bridging-
chloride dimers trans- and cis-[{(triphos)RhH}2(µ-Cl)2]2+ at room temperature. Complex 1 contains a fast-spinning
H2 ligand with a T1min of 38.9 ms in CD2Cl2 (220 K, 400 MHz). An NMR analysis of the bis-deuterated isotopomer
[(triphos)RhH2D2]+ (1-d2) did not provide a J(HD) value. At 190 K, the perdeuterated isotopomers [(triphos)-
RhD3] (2-d3) and 1-d4 show T1min values of 16.5 and 32.6 ms (76.753 MHz), respectively, for the rapidly exchanging
deuterides. An analogous 2-fold elongation of T1min is also observed on going from [(triphos)IrD3] to [(triphos)-
Ir(η2-D2)D2]+. A rationale for the elongation of T1min in nonclassical polyhydrides is proposed on the basis of
both the results obtained and recent literature reports.
Introduction
The scarcity of rhodium dihydrogen complexes in the
literature is highly intriguing, as rhodium is the essential
Kubas et al. first recognized in 1984 the existence of the η2-
H2 ligand coordinated in the complexes M(η2-H2)(CO)3(PiPr3)2
(M ) Mo, W).1 Since then, more than 200 dihydrogen
complexes have been prepared with metal ions from groups 4,
5, 6, 7, 8, 9, and 10.2,3 However, only two authentic rhodium-
(III) derivatives are known, namely, the isolated complex
[TpMe2RhH2(η2-H2)] and the unstable derivative [TpRhH2(η2-
H2)] described by Venanzi et al. and by Heinekey et al.,
respectively [TpMe2 ) hydrotris(3,5-dimethylpyrazolyl)borate;
Tp ) hydridotris(pyrazolyl)borate].4-6 A third, highly fluxional,
Rh(I) complex, [(PP3)RhH2]+,7 was postulated to contain a
dihydrogen ligand in solution, but this formulation was later
questioned on the basis of the absence of detectable J(HD)
coupling in the deuterium-hydride isotopomer [PP3 ) P(CH2-
CH2PPh2)3].8,9
ingredient in most homogeneous and heterogeneous hydrogena-
tion catalysts. Moreover, several hydrogenation reactions have
been postulated to involve the intermediacy of elusive η2-H2
Rh species.10 Motivated by this evidence, we are presently trying
to develop new spectroscopic techniques and theoretical models
for the correct evaluation of H-H bonding interactions in
polyhydrido metal complexes, especially in rhodium compounds,
which are often thermally unstable and highly fluxional in
solution. Indeed, low stability and/or high fluxionality may
thwart, if not prevent, the measurement of either T1min or J(HD)
in polyhydride metal complexes.
In this work, we describe the synthesis and characterization
of the new complex [(triphos)Rh(η2-H2)H2]+ (1), obtained by
protonation of the trihydride [(triphos)RhH3] (2) [triphos )
MeC(CH2PPh2)3] with different protic acids at low temperature
in either CD2Cl2 or THF-d8. To further support the nonclassical
dihydrogen structure of the rhodium complex, we have carried
† Institute of Organo-Element Compounds of the Russian Academy of
Sciences.
‡ Istituto per lo Studio della Stereochimica ed Energetica dei Composti
di Coordinazione.
1
2
out a detailed analysis of the H- and H-relaxation properties
of 1, of its perdeuterated isotopomer [(triphos)Rh(η2-D2)D2]+
(1-d4), and of the authentic nonclassical iridium analogues
[(triphos)Ir(η2-H2)H2]+ (3) and [(triphos)Ir(η2-D2)D2]+ (3-d4).
(1) Kubas, G. J.; Ryan, R. R.; Swanson, B. I.; Vergamini, P. J.;
Wasserman, H. J. Am. Chem. Soc. 1984, 106, 451.
(2) Heinekey, D. M.; Oldham, W. J., Jr. Chem. ReV. 1993, 93, 913.
(3) Jessop, P. G.; Morris, R. H. Coord. Chem. ReV. 1992, 121, 155.
(4) Bucher, U. E.; Lengweiler, D.; Philipsborn, W.; Venanzi, L. M. Angew.
Chem., Int. Ed. Engl. 1990, 29, 548.
Experimental Section
(5) Nanz, D.; Philipsborn, W.; Bucher, U. E.; Venanzi, L. M. Magn. Reson.
Chem. 1991, 29, 538.
(6) Oldham, W. J., Jr.; Hinkle, A. S.; Heinekey, D. M. J. Am. Chem. Soc.
1997, 119, 11028.
(7) Bianchini, C.; Mealli, C.; Peruzzini, M.; Zanobini, F. J. Am. Chem.
Soc. 1987, 109, 5548.
(8) Heinekey, D. M.; van Roon, M. J. Am. Chem. Soc. 1996, 118, 12134.
(9) Bianchini, C.; Elsevier, C. J.; Ernsting, J. M.; Peruzzini, M.; Zanobini,
F. Inorg. Chem. 1995, 34, 84.
The NMR studies were carried out in standard 5-mm NMR tubes
using CD2Cl2 or THF-d8 as the solvent. All solvents were dried using
conventional procedures and were then distilled under inert atmosphere
prior to use. 1H- and 31P{1H}-NMR spectra were recorded on either a
Bruker AMX 400 or an AVANCE DRX 500 spectrometer operating
(10) Esteruelas, M. A.; Oro, L. A. Chem. ReV. 1998, 98, 577.
10.1021/ic990667q CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/18/2000

1656 Inorganic Chemistry, Vol. 39, No. 8, 2000
Bakhmutov et al.
at 400.13 and 500.13 MHz (1H) and at 162.04 and 202.52 MHz (31P),
respectively. Peak positions are relative to tetramethylsilane and were
calibrated against the residual solvent resonance (1H) or against external
its safety apparatus, in a dry ice/acetone bath, and ∼20 µL of HBF4‚
OMe2 was added via syringe. Immediately after the introduction of
the acid, the tube was pressurized with hydrogen to ∼30 bar and
carefully shaken. The tube was then inserted into the spectrometer,
which was precooled at 210 K. 1H- and 31P{1H}-NMR spectra,
immediately recorded at this temperature, showed the formation of 1
and 5 in an approximately 1:1 ratio.
When the temperature was increased, the concentration of 5 increased
steadily at the expense of that of 1 until above 240 K, at which point
only 5 was visible in solution, together with traces of the tetrahydride
dimer trans-[{(triphos)RhH}2(µ-H)2]2+ (trans-6). At higher temperature,
5 rapidly transformed into trans-6 (∼90%), together with other unknown
products.16
2
85% H3PO4 with downfield values taken as positive (31P). The H-
NMR data were collected with either the Bruker AMX 400 spectrometer
operating at 61.40 MHz or the Bruker DRX 500 instrument operating
at 76.77 MHz. The conventional inversion-recovery method (180°-
τ-90°) was used to determine the variable-temperature longitudinal-
relaxation time T1. The calculation of the relaxation times was made
using the nonlinear three-parameter fitting routine of the spectrometers.
In each experiment, the waiting period was 5 times longer than the
expected relaxation time, and 16-20 variable delays were employed.
The duration of the pulses was controlled at each temperature. The
errors in determinations of T1 were lower than 5%. The high-pressure
NMR (HPNMR) experiments were performed in 10-mm sapphire tubes
(Saphikon Inc.) assembled with Ti-alloy pressure heads constructed at
ISSECC-CNR.11 The HPNMR spectra were recorded using a standard
10-mm probe tuned to 31P and 1H nuclei on a Bruker AC 200
spectrometer.
Caution: All manipulations inVolVing high pressures are potentially
hazardous. Safety precautions must be taken at all stages of NMR
studies inVolVing high-pressure tubes.12
The hydride [(triphos)RhH3] (2) was prepared as described in the
literature,13 whereas [(triphos)IrH3] (4) was synthesized via a procedure
that was improved with respect to that reported in the literature.14 The
trideuterides [(triphos)RhD3] (2-d3) and [(triphos)IrD3] (4-d3) were
prepared in 90% isotopic purity (1H NMR) following the procedure
reported for the protonated analogue by simply using deuterated solvents
and reagents.
Preparation of [(Triphos)IrH3] (4). A THF solution (20 mL) of
[IrCl(COE)2]215 (500 mg, 0.74 mmol) (COE ) cyclo-octene) was treated
under nitrogen with 1 equiv of triphos. After gentle heating to 40 °C,
a 10-fold excess of solid LiAlH4 was added in small portions with
stirring to give a slurry that was refluxed for 4 h. The resulting solution
was cooled to room temperature, and the excess of LiAlH4 was
hydrolyzed with a THF/H2O mixture (6 mL, 1:1, v/v). The solution
was filtered through Celite and concentrated under vacuum to ∼10
mL before adding a 1:1 mixture of EtOH/n-hexane (10 mL). Slow
concentration under a flow of nitrogen gave 4 as off-white microcrystals
(65% yield). The product was authenticated by comparison with an
authentic specimen prepared using the procedure reported by Venanzi
et al.14
In Situ Protonation of [(Triphos)RhH3] with HBF4 in CD2Cl2
under Nitrogen. Replacing THF-d8 with CD2Cl2 in the in situ
protonation of 2 at 210 K gave 1 in high yield (>90%), independent
of either the acid (CF3COOH, CF3SO2OH, or HBF4‚OMe2) or the
gaseous atmosphere (N2 or H2) employed. After standing at 240 K for
15 min, 1 converted to a 1:1.2 mixture of cis- and trans-[{(triphos)-
RhH}2(µ-H)2]2+ (cis-6 and trans-6). The formation of a 1:4 mixture of
the known µ-chloride dimers cis- and trans-[{(triphos)RhH}2(µ-Cl)2]2+
(cis-7 and trans-7)16 took place at room temperature within 1 h.
NMR data for 1 in CD2Cl2: 31P{1H} NMR δ 23.20 [3P, d, J(PRh)
96.6 Hz]; 1H NMR δ -5.84 [4H, br qd, J(HP) 31 Hz, J(HRh) 15 Hz].
NMR data for cis-6 in CD2Cl2: 31P{1H} NMR δ 17.15 [6P, d, J(PRh)
1
101.1 Hz]; H NMR δ -10.53 [4H, qut, J(HP) 23.1 Hz, J(HRh) 4.8
Hz].
NMR data for trans-6 in CD2Cl2: 31P{1H} NMR δ 25.23 [6P, d,
J(PRh) 95.9 Hz]; 1H NMR δ -7.86 [4H, qut, J(HP) 22.8 Hz, J(HRh)
14.0 Hz].
Results and Discussion
Synthesis and Characterization of [(Triphos)Rh(η2-H2)-
H2]+. The synthesis and characterization of the octahedral Ir(III)
nonclassical tetrahydride [(triphos)Ir(η2-H2)H2]+ (3) has recently
been described.17 This complex was obtained either by the solid-
state hydrogenation of [(triphos)Ir(η2-C2H4)H2]BPh4 at H2
pressures higher than 5 bar or by the in situ protonation of the
trihydride [(triphos)IrH3] (4) with a slight excess of HBF4‚OMe2
in CD2Cl2 at low temperature. The latter procedure was
employed to prepare a sample of 3 in THF-d8 at 210 K that
was studied, for comparative purposes, with the same NMR
apparatus and in the same experimental conditions as those
employed for the characterization of the new rhodium complex
[(triphos)Rh(η2-H2)H2]+ (1).
Because of the presence of an excess of acid, 3 exhibits a
single resonance for the four hydrogen atoms of the Ir(η2-H2)H2
moiety, even at 210 K.17 In these experimental conditions, the
hydride signal shows a quite short T1 time of 77 ms (500 MHz),
much shorter than the relaxation time measured for 4 (560 ms
at 210 K in THF-d8). Above 230 K, the resonances of the free
acid (δ 9.3) and the hydride ligands broaden, suggesting the
occurrence of H+/IrH4 exchange on the NMR time scale. At
higher temperature, 3 starts to decompose, yielding, first, the
solvate adduct [(triphos)Ir(η1-THF-d8)H2]+ (7) and, then, the
classical tetrahydride trans-[{(triphos)IrH}2(µ-H)2]2+ (8).17,18
The η2-H2 rhodium complex 1 was similarly generated in
solution by reacting the classical trihydride 2 in THF-d8 at 210
K with 3 equiv of either HBF4‚OMe2 or HOSO2CF3 (Scheme
1).
In Situ Protonation of [(Triphos)RhH3] with HBF4‚OMe2 in
THF-d8 (NMR Experiment). A 5-mm NMR tube was charged under
nitrogen with 30 mg of 2 (0.041 mmol) and 1 mL of THF-d8. Into this
solution, cooled to 210 K with a liquid nitrogen/acetone bath, was added
by syringe an excess of HBF4‚OMe2 (20 µL, 0.16 mmol). The solution
immediately turned pale yellow. The tube was inserted into the
1
spectrometer, which was precooled at 210 K. H- and 31P{1H}-NMR
spectra, immediately recorded at this temperature, revealed the forma-
tion of two compounds: the molecular hydrogen complex [(triphos)-
Rh(η2-H2)H2]+ (1) and the THF-d8 adduct [(triphos)Rh(η1-THF-d8)H2]+
(5) in a 1:4 ratio.
Repeating the protonation reaction under 1 bar of H2 did not
significantly change the product composition.
NMR data for 1 in THF-d8: 31P{1H} NMR δ 21.28 [3P, d, J(PRh)
96.7 Hz]; 1H NMR δ -5.68 [4H, br qd, J(HP) 30 Hz, J(HRh) 15 Hz].
NMR data for 5 in THF-d8: 31P{1H} NMR δ 53.18 [1P, dt, J(PRh)
138.9 Hz, J(PP) 22.3 Hz], δ 6.67 [2P, dd, J(PRh) 79.4 Hz, J(PP) 22.3
1
Hz]; H NMR δ -6.53 [2H, dt, J(HPtrans) 164 Hz, J(HPcis) ≈ J(HRh)
12.0 Hz].
In Situ Protonation of [(Triphos)RhH3] with HBF4 in THF-d8
under a High Pressure of H2. A 10-mm sapphire HPNMR tube was
charged with 30 mg of 2 (0.041 mmol) and 2 mL of THF-d8 that had
been degassed under nitrogen. The tube was immersed, together with
(14) Janser, P.; Venanzi, L. M.; Bachechi, F. J. Organomet. Chem. 1985,
296, 229.
(15) Herde, J. L.; Lambert, J. C.; Senoff, C. V. Inorg. Synth. 1974, 15, 18.
(16) Bianchini, C.; Meli, A.; Laschi, F.; Ramirez, J. A.; Zanello, P.; Vacca,
A. Inorg. Chem. 1988, 27, 4429.
(17) Bianchini, C.; Moneti, S.; Peruzzini, M.; Vizza, F. Inorg. Chem. 1997,
36, 5818.
(18) Bianchini, C.; Caulton, K. G.; Folting, K.; Meli, A.; Peruzzini, M.;
Vizza, F. J. Am. Chem. Soc. 1992, 114, 7290.
(11) CNR (Bianchini, C.; Meli, A.; Traversi, A.). Italian Patent FI A000025,
1997.
(12) Elsevier, C. J. J. Mol. Catal. 1994, 92, 285.
(13) Ott, J.; Venanzi, L. M.; Ghilardi, C. A.; Midollini, S.; Orlandini, A.
J. Organomet. Chem. 1985, 291, 89.

Relaxation Behavior of [(Triphos)Rh(η2-H2)H2]+
Inorganic Chemistry, Vol. 39, No. 8, 2000 1657
Scheme 1
1
Figure 1. Hydride region of the H-NMR spectrum (220 K, THF-d8,
500 MHz) after protonation of 2 with HBF4‚OMe2 at low temperature.
above 240 K. At room temperature, however, the THF adduct
was unstable and slowly converted to the red-orange tetrahydride
dimer trans-[{(triphos)RhH}2(µ-H)2]2+ (trans-6).16 The forma-
tion of trans-6, which is scarcely soluble in THF and which
may separate as red-orange crystals of the BF4 salt depending
on the concentration, is generally accompanied by the formation
of other small impurities which, altogether, may sum up to
∼10%. A much more selective method for the synthesis of 1
in solution was provided by the reaction of 2 with a 3-fold
excess of CF3COOH in CD2Cl2 [RhH4: δ -5.84, J(HP) 31 Hz,
J(RhH) 15 Hz]. As in THF, 1 is highly fluxional in CD2Cl2,
where the 31P{1H}-NMR spectrum consists of a temperature-
invariant narrow doublet [δ 23.2, J(RhP) 96.6 Hz] down to 180
K. Above 240 K in CD2Cl2, 1 quickly loses H2 to form a mixture
of cis- and trans-6 that irreversibly transformed into the double-
bridging chloride dimers cis- and trans-7 via H/Cl exchange
(Scheme 1).16
The NMR parameters of 1 are quite similar to those of 3. In
the hydride region, the 1H-NMR spectrum at 220 K contains a
broad signal at -5.68 ppm that appears as a quartet of doublets
with J(HP) ) 30 Hz and J(HRh) ) 15 Hz (Figure 1) in the
spectrum obtained with resolution enhancement via Gaussian
multiplication of the FID. Information about the 1H-31P
coupling was provided by 1H{31P}-NMR experiments. The
signal in these spectra corresponds to four hydrogen atoms (1H-
NMR integration with respect to the aliphatic resonances of the
triphos ligands) and exhibits a very short T1 time of 65.5 ms
(500 MHz). Like the iridium analogue, 1 is thus very fluxional
in solution, as shown also by the 31P{1H} spectrum, which
consists of a single narrow doublet at 21.28 ppm with J(PRh)
) 96.7 Hz. The experimental relaxation time of the resonance
due to the four exchanging hydrides is close to the value found
for the Ir derivative 3 and is much shorter than the value
determined for the parent trihydride 2 (577 ms) under the same
experimental conditions.
1H-Relaxation Study of [(Triphos)Rh(η2-H2)(H)2]+. The
proton-relaxation data recorded at 500 MHz for the hydride
ligands of 2 and 1 in THF-d8 give a T1min value of 311 ms at
260 K for the classical trihydride 2 and a value decreasing from
65.5 ms at 220 K to 46.0 ms at 235 K for 1. These data allow
one to suggest a nonclassical structure for 1. Unfortunately, a
direct T1min measurement for 1 in THF-d8 was not possible,
because this complex is stable only at temperatures below 240
K. However, a T1min value of 38.9 ms was observed at 220 K
(400 MHz) in CD2Cl2. This value, scaled up to 500 MHz,
corresponds to a T1min of 48.6 ms. The absence of mixing of
the T1 times for the free acid and the hydride signal of 1 in the
temperature range investigated eliminates the possibility of any
H/H exchange between the RhH4 moiety and the free acid and,
therefore, indicates that the T1min value measured in CD2Cl2 is
absolutely correct.
Complex 1 was not the sole product obtained upon protona-
tion of 2 at 210 K. In fact, a compound that we assign as the
THF adduct [(triphos)Rh(η1-THF-d8)H2]+ (5) was the major
product (80%), because of competitive coordination of the
solvent to the rhodium center (Scheme 1). In comparable
reaction conditions, the Ir derivative [(triphos)Ir(η1-THF)H2]-
BPh4 has also been isolated.18
A standard treatment of the 1H-relaxation data in CD2Cl2 for
1 yields a rH-H bond distance of either 0.91 or 1.14 Å, assuming
the presence of either a fast-spinning or a slow-spinning
dihydrogen ligand, respectively.19,20 Inelastic neutron scattering
studies recently performed on the complex [TpMe2RhH2(η2-H2)]
are consistent with a very low rotational barrier (Ea ) 0.56 kcal/
mol) and a short rH-H distance of 0.94 Å.21 It has been proposed
that cis-hydride/dihydrogen interactions can effectively lower
the barrier to H2 rotation.22,23 On the basis of these arguments,
it is most likely that the dihydrogen ligand in 1 is spinning very
rapidly.24
The ratio of 1 to 5 did not change when the protonation
reaction of 2 was performed under 1 bar of H2. However, when
the protonation was performed in a sapphire NMR tube at a H2
pressure of 30 bar, 1 and 5 were formed in a 1:1 ratio.
The proton-NMR spectrum of the THF adduct 5 shows a
hydride resonance at -6.53 ppm, which appears as a slightly
perturbed doublet of triplets with a very long T1 of 690 ms at
210 K (500 MHz). The 31P{1H}-NMR spectrum consists of an
AM2 spin system with δ 53.18 [1P, dt, J(PRh) 138.9 Hz, J(PP)
22.3 Hz] and δ 6.67 [2P, dd, J(PRh) 79.4 Hz, J(PP) 22.3 Hz].
The formation of 5 from 1 is accompanied by the appearance
1H-NMR spin-lattice relaxation experiments are recognized
as an important tool for elucidating the solution-phase structure
of transition metal polyhydride complexes.2,19,22 In particular,
notwithstanding some criticism associated with the evaluation
(19) Bakhmutov, V. I.; Vorontsov, E. V. ReV. Inorg. Chem. 1998, 3, 183.
(20) Desrosiers, P. J.; Cai, L.; Lin, Z.; Richards, R.; Halpern, J. J. Am.
Chem. Soc. 1991, 113, 4173.
(21) Eckert, J.; Albinati, A.; Bucher, U. E.; Venanzi, L. M. Inorg. Chem.
1996, 35, 1292.
1
of the typical H resonance at 4.53 ppm for free H2.
The stability of 1 did not remarkably increase under 30 bar
of H2, as its irreversible transformation into 5 occurred already

1658 Inorganic Chemistry, Vol. 39, No. 8, 2000
Bakhmutov et al.
of this parameter,25 the measurement of T1 is particularly
important when the complexes are thermally unstable and may
exist only in solution and only at low temperature.26 A more
reliable diagnostic tool for assessing the nonclassical structure
of a metal polyhydrido complex is provided by the measurement
of the J(HD) coupling constant (g10 Hz) in the monodeuterated
isotopomer.2,27,28 In an attempt to synthesize 1-d1 for this kind
of measurement, the trihydride 2 was protonated with CF3COOD
in CD2Cl2 at 210 K. Irrespective of the amount of acid employed
(1-3 equiv), this reaction gave a largely predominant isoto-
pomer that we assigned as [(triphos)RhH2D2] (1-d2) on the basis
1
of the relative integral intensities in the H-NMR spectrum.
Moreover, deuterium incorporation into 1 was evident by the
1
appearance of free H2 and HD. The hydride region in the H-
NMR spectrum of 1-d2, recorded with 31P-decoupling at 230 K
and 400 MHz, contains a broad doublet with J(RhH) ) 15 Hz
(∆ν ) ∼8-10 Hz) with no appreciable J(HD).
2
Complexes 1 and 1-d2 are highly fluxional, showing an
averaged hydride resonance even at low temperature. This
remarkable fluxionality should effectively reduce the J(HD)
value in 1-d2 because of the different locations of deuterium in
the complex. Indeed, unresolved J(HD) splitting in low-
temperature NMR spectra has already been reported for other
partially deuterated nonclassical polyhydrides, such as RuH2-
(H2)(CO)(PPri3)2.29 It has been proposed that this phenomenon
may be due to deuterium relaxation resulting in scalar relaxation
Figure 2. Plot of the H-T1 data collected at 76.75 MHz in CH2Cl2
for the OD resonance of the acid (9) and the RhD4 resonance of 1-d4
(3) and of 2-d3 (O).
disappeared in the 2H-NMR spectrum [δ -8.60, J(RhD) 3 Hz,
2J(DP) + 2 × J(DP′) ) 20 Hz]. Formed in its place was a
2
broad resonance at -5.70 ppm that could be ascribed to the
four freely exchanging deuterides of 1-d4. Similar spectral
changes were observed upon low-temperature protonation of
4-d3 with an excess of CF3COOD in CH2Cl2: a single deuteride
resonance was observed at -8.20 ppm for 3-d4 even at 170 K.
Because the 2H spin-lattice relaxation is dominated by
quadrupolar interactions,30 the deuterium quadrupole coupling
constant (DQCC) is a relevant parameter for the characterization
of polyhydride metal complexes (eq 1).31-33
of the second type for the proton coupled to deuterium.30
A
broad, featureless hydride resonance has also been observed for
partially deuterated [TpRhH2(η2-H2)].4,5
In the case of 1-d2, a simple calculation using the equation
1/T2 ) 8/3π2J2(H-D)T1(D),29-32 where T1(D) was measured for
[(triphos)RhD4]+ (1-d4) as 0.0385 s at 220 K (see below), shows
that the expected line width of the hydride resonance (∆ν )
1/πT2) in 1-d2 might be ∼8 Hz for a J(HD) value of ∼5 Hz. It
is therefore likely that both of these effects are responsible for
the nondetection of the J(HD) constant in 1-d2.
1/T1 ) 3/50π2(2I + 3)[I2(2I - 1)]-1(e2qZZQ/h)2(1 + η2/3)
[τmol/(1 + ωD τmol2) + 4τmol/(1 + 4ωD τmol2)] (1)
2
2
To provide further support for the nonclassical structure of
1, the relaxation behaviors of the perdeuterated isotopomers 2-d3
and 1-d4 have been studied by variable-temperature 2H-T1
measurements. For comparative purposes, these relaxation
experiments were carried out for the Ir derivatives 4-d3 and 3-d4,
the latter containing an intact D2 ligand.17
Complex 1-d4 was readily prepared by protonation of 2-d3
with a 2-fold excess of CF3SO2OD in CH2Cl2 at 195 K. Upon
acid addition, the deuteride resonance due to 2-d3 completely
τmol ) τ0 exp(Ea/RT)
η ) |qXX - qYY|/qZZ
In eq 1, I is the spin of deuterium, η the asymmetry parameter
of the electric field gradient on D, τmol the correlation time of
molecular reorientations with the activation energy Ea, and the
term e2qZZQ/h the DQCC on the nucleus D.32
When the T1 time reaches a minimum, the DQCC value can
be calculated by means of eq 2.
(22) Eisenstein, O.; Jackson, S. A. Inorg. Chem. 1990, 29, 3910.
(23) Bianchini, C.; Masi, D.; Peruzzini, M.; Casarin, M.; Maccato, C.; Rizzi,
G. A. Inorg. Chem. 1997, 36, 1061.
1/2
DQCC ) 1.2201(1 + η2/3)-1/2(ν/T1min
)
(2)
(24) Ab initio calculations carried out during the preparation of this article
have shown that the H-H distances in the model compounds
[HC(CH2PH2)3M(η2-H2)H2]+ are 0.7983 Å for M ) Rh and 0.8349
Å for M ) Ir: Maseras, F. Universitat Autonoma de Barcelona (Spain),
private communication.
(25) See, for example: Gusev, D. G.; Vymenits, A. B.; Bakhmutov, V. I.
Inorg. Chem. 1991, 30, 3116. See also ref 17.
(26) See, for example: Hamilton, D. G.; Crabtree, R. H. J. Am. Chem.
Soc. 1988, 110, 4126.
(27) Maltby, P. A.; Schlaf, M.; Steinbeck, M.; Lough, A. J.; Morris, R.
H.; Klooster, W. T.; Koetzle, T. F.; Srivastava, R. C. J. Am. Chem.
Soc. 1996, 118, 5396.
(28) Hush, N. S. J. Am. Chem. Soc. 1997, 119, 1717.
(29) Gusev, D. G.; Vymenits, A. B.; Bakhmutov, V. I. Inorg. Chem. 1992,
31, 2.
(30) Abragam, A. The Principles of Nuclear Magnetism; Oxford University
Press: New York, 1971.
(31) Nietlispach, D.; Bakhmutov, V. I.; Berke, H. J. Am. Chem. Soc. 1993,
115, 9191 and references therein.
where DQCC, ν, and T1min are measured in kilohertz, megahertz,
and second, respectively.31-33 According to recent molecular
orbital calculations,32 the asymmetry parameter η can be taken
as 0 and 0.62 for classical hydrides and dihydrogen complexes,
respectively.
Figure 2 compares the 2H-T1 data for the RhD4 resonance in
1-d4 at 76.753 MHz at different temperatures with the OD signal
of the acid and with the deuteride resonance of 2-d3. Complexes
2-d3 and 1-d4 show T1min values of 16.5 and 32.6 ms, respec-
tively, at ∼190 K. Given that the η parameter is approximately
0 in terminal deuterides,34-36 eq 2 gives a very reasonable
DQCC value of 83.2 kHz for 2-d3.32 The T1min time for the
classical ligands in 4-d3 is also quite short and is measured as
(32) Bakhmutov, V. I.; Bianchini, C.; Maseras, F.; Lledos, A.; Peruzzini,
M.; Vorontsov, E. V. Chem. Eur. J. 1999, 5, 3318.
(33) Facey, G. A.; Fong, T. P.; Gusev, D.; Macdonald, P. M.; Morris, R.
H.; Schlaf, M.; Xu, W. Can. J. Chem. 1999, 77, 1899.

Relaxation Behavior of [(Triphos)Rh(η2-H2)H2]+
Inorganic Chemistry, Vol. 39, No. 8, 2000 1659
10.1 ms at 190 K (61.402 MHz), corresponding to a DQCC
value of 95.0 kHz.
[RuD(D2)(dppe)2]+, for which the minimum T1 value of the D2
ligand is 12 times greater than that of the terminal deuteride
ligand (dppe ) diphenylphosphinoethane).33
It follows from the 2H-relaxation data that the value of T1min
for the η2-D2 complex 1-d4 is significantly longer than that for
the classical trihydride 2-d3 (32.6 vs 16.5 ms, 76.753 MHz).
This T1min elongation in the nonclassical η2-D2 complex 1-d4
might be associated with the occurrence of slow D+/D2 exchange
between the D2 ligand and the free acid, which actually exhibits
a quite long T1 time. This type of exchange on the NMR time
scale has been detected by 1H-NMR spectroscopy for the iridium
obs
A perusal of eq 4 shows that T1min ) 2T1min(D) when
T1min(D2) f ∞. According to our 2H-relaxation measurements,
the ratio T1min(1-d4)/T1min(2-d3) is 1.98, which is close to the
maximum elongation of the relaxation time on going from a
classical hydride complex to a very fluxional nonclassical
tetrahydride. In conclusion, 1-d4 can be formulated as a
dihydrogen complex that contains a D2 ligand undergoing a fast
rotation with the angle R close to the magic angle.
1
derivative 3 in THF-d8 at temperatures g230 K,17 but the H-
For comparative purposes, a 2H-relaxation study has also been
performed for the iridium η2-D2 complex 3-d4 prepared by
protonation of the trihydride 4-d3 with CF3COOD in CH2Cl2.17
Interestingly, for the averaged resonance of 3-d4, a value of 20.1
ms (61.402 MHz) was experimentally measured at 170 K when
no D+/(D2) exchange occurs as the minimum value from the
temperature dependence of T1. Like the Rh analogues, the
NMR spectra of the rhodium derivative 1 in either CD2Cl2 or
THF-d8 did not reveal this process in the temperature range
investigated. On the other hand, the absence of D+/D2 exchange
in CH2Cl2 solutions of 1-d4 was evident also from an analysis
2
of the H-NMR data. Indeed, the standard calculations of the
ln(T1) vs 1/T relaxation curves using eq 1 (solid lines in Figure
2) provide the same Ea value of 2.7 kcal/mol for the molecular
reorientations of 2-d3 and 1-d4, whereas the activation energy
calculated for the acid is significantly lower (1.8 kcal/mol). All
of the curves in Figure 2 are thus independent from the
relaxation viewpoint, showing the absence of D+/D2 exchange
on the T1 NMR time scale.
transformation of the classical trihydride 4-d3 into the nonclas-
obs
sical tetrahydride 3-d4 causes a 2-fold increase in T1min
.
On the basis of the preceding results, one may conclude that
the dihydrogen ligands in both 3-d4 and 1-d4 are spinning very
rapidly and that the direction of the main axis of the electric
field gradient shows an R value close to the magic angle. A
similar R value (52.5° or 49.7°) for the D2 ligand in [Os(D2)-
Cl2(CO)(PPri3)2] has recently been inferred from variable-
2
A recent analysis of the H-T1min behavior for a series of
dideuterium complexes has shown that a significant elongation
of T1 takes place when the D2 ligand undergoes a fast
intramolecular rotation (or fast libration) around the axis
2
temperature solid-state H-NMR measurements.33 These data
perpendicular to the D-D bond with the correlation time τrot
somewhat contrast with recent molecular orbital calculations
on several Os, W, and Ru dihydrogen complexes, according to
which the main axis of the electric-field gradient lies along the
D-D vector and R takes values between 72° and 84°.32
Variable-temperature 1H-NMR spectra of solid IrH2(H2)(Cl)-
(PPri3)2 have revealed a surprising hydride scrambling (eq 5)
with an energy barrier lower than 3 kcal/mol.39
32,33
, τmol
.
For example, when the angle R between the rotation axis and
the direction of the main axis of the electric field gradient is
equal to 90° and, thus, the main axis is aligned with the D-D
bond, a fast rotation has been found to cause a 4-fold elongation
IrH2(H*2)(Cl)(PPri3)2 a IrH*2(H2)(Cl)(PPri3)2
(5)
of T1min.
37 In general, the DQCC value for a spinning D2 ligand
can be calculated with eq 3, showing that a maximal elongation
of T1min should be expected when R is close to the magic angle
(∼54°) and T1min f ∞.
A standard treatment of the relaxation data in terms of eq 1
has shown that the molecular reorientations in the rhodium
complexes 1, 1-d4, 2, and 2-d3 in THF or CH2Cl2 require
activation energies (Eactmol) ranging between 2.8 and 3.6 kcal/
mol, which is consistent with the variable-temperature 1H- and
2H-NMR spectra of 1 and 1-d4. Accordingly, one may expect
DQCC )
2.4402(1 + η2/3)-1/2[ν/T1min(3 cos2 R - 1)2]1/2 (3)
exch
mol
that Eact
< Eact
and hence that τmol > τexch. The 2H
Complex 1-d4 undergoes a fast deuteride/dideuterium ex-
change, and therefore T1min for its D2 ligand [T1min(D2)] can be
relaxation is caused by reorientations of the main axis of the
electric-field gradient at D, generally as a result of molecular
motions with the correlation time τmol. This is absolutely valid
for the classical deuteride ligands in 1-d4 or 3-d4, whereas
additional reorientations of the gradient, attributable to rapid
intramolecular rotation with τrot < τmol, are apparently operative
for the D2 ligands. This rotation leads to the elongation of the
relaxation for processes in which the rotation and the hydride/
obs
calculated as 1.34 s ((5%) from T1min (0.0326 s) using eq
4,38 where the value of T1min(D) for the deuteride ligands in
2-d3 is 0.0165 s.
1/T1minobs ) 0.5/T1min(D) + 0.5/T1min(D2)
(4)
The T1min time thus undergoes approximately an 80-fold increase
on going from the classical deuteride ligand to the intact D2
ligand. A related effect has been observed for the complex
dihydrogen exchange are independent. However, when τrot
∼
τexch, the two processes may couple with each other. In this
eventuality, during τmol, the main axis of the electric-field
gradient on the classical D ligands in 1-d4 and 3-d4 might also
undergo additional reorientations as an effect of motion transfer.
Intuitively, this effect might contribute to the elongation of
T1minobs in nonclassical D2 complexes. In the absence of further
experimental data, however, the effect of motion transfer
remains an intriguing working hypothesis.
(34) Butler, L. G.; Keitler, E. A. J. Coord. Chem. 1994, 32, 121 and
references therein.
(35) Kim, A. J.; Fronczek, F. R.; Butler, L. G.; Chen, S.; Keitler, E. A. J.
Am. Chem. Soc. 1991, 113, 9090.
(36) Guo, H.; Jarrett, W. L.; Butler, L. G. Inorg. Chem. 1987, 26, 3001.
(37) Notice that a similar behavior has been also observed for 1H-T1min of
a fast-spinning H2 ligand: Morris, R. H.; Witterbort, R. J. Magn.
Reson. Chem. 1997, 35, 243.
(38) Kaplan, J. I.; Fraenkel, G. NMR of Chemically Exchanging Systems;
Academic Press: New York, 1980; p 78.
(39) Wisniewski, L. L.; Mediati, M.; Jensen, C. M.; Zilm, K. W. J. Am.
Chem. Soc. 1993, 115, 7533.

1660 Inorganic Chemistry, Vol. 39, No. 8, 2000
Bakhmutov et al.
Conclusions
experimental data obtained in this work, it is confirmed that
the 2H-T1min time elongates significantly in going from classical
to nonclassical structures of polydeuterido metal complexes.
In addition to reporting a new example of a rhodium η2-H2
complex, this work describes the 2H variable-temperature spin-
lattice relaxation behavior of two structurally related dihydrogen
complexes of rhodium and iridium in solution. It has been shown
that deuterium-relaxation studies are extremely useful for the
characterization of polyhydride metal complexes, especially
when the complexes are thermally unstable and/or highly
fluxional in solution and the measurement of their T1min values
and/or J(HD) parameters is prevented. On the basis of the
Acknowledgment. Thanks are given to NATO (Contract
“2H-Relaxation as a Tool To Study Hydrogen Transfer Reac-
tions”, Grant HTECH-CRG 972135) and the CNR/RAS coop-
eration bilateral agreement for financial support. Thanks are also
given to Prof. R. H. Morris for disclosing some unpublished
results to us.
IC990667Q