Bridging the gap between homogeneous and heterogeneous catalysis: Ortho/para H2 conversion, hydrogen isotope scrambling, and hydrogenation of olefins by Ir(CO)Cl(PPh3)2

Article abstract of DOI:10.1021/ja0475961

Some transition metal complexes are known to catalyze ortho/para hydrogen conversion, hydrogen isotope scrambling, and hydrogenation reactions in liquid solution. Using the example of Vaska's complex, we present here evidence by NMR that the solvent is no

Full text of DOI:10.1021/ja0475961

Published on Web 06/16/2004
Bridging the Gap between Homogeneous and Heterogeneous Catalysis:
Ortho/Para H2 Conversion, Hydrogen Isotope Scrambling, and Hydrogenation
of Olefins by Ir(CO)Cl(PPh3)2
Jochen Matthes,†,‡ Tal Pery,† Stephan Gru¨ndemann,† Gerd Buntkowsky,† Sylviane Sabo-Etienne,‡
Bruno Chaudret,‡ and Hans-Heinrich Limbach*,†
Institute of Chemistry, Freie UniVersita¨t Berlin, Takustrasse 3, D-14195 Berlin, Germany, and Laboratoire de
Chimie de Coordination du CNRS 205, Route de Narbonne, F-31077 Toulouse-Cedex 4, France
Received April 26, 2004; E-mail: limbach@chemie.fu-berlin.de
The application of concepts of homogeneous catalysis to
heterogeneous catalysis and vice versa constitutes an area of current
interest. Using the example of Vaska’s catalyst Ir(CO)Cl(PPh3)2
(1),1 we show that compounds that catalyze the chemical ortho-
para nuclear spin conversion (SC) of dihydrogen (eq 1),2 the
associated isotope scrambling (IS) reaction of dihydrogen isoto-
pologs (eq 2),3 and the hydrogenation of unsaturated compounds
in organic solvents4 (eq 3) can do so also in the solid state. Our
findings shed new light on the mechanisms of these reactions.
in which both SC and IS could be detected at the same time. In
experiment I, a saturated solution of 1 in C6D6 was frozen and kept
at 77 K in the dark, in order to prevent photochemical reactions,
for 5 days in an NMR tube equipped with a Teflon needle valve
under a gaseous 1:1 mixture of H2 and D2 (each 400 mbar, 298
K). The addition of dihydrogen to 1 was then followed by 1H NMR
at 298 K for about 1 h. For comparison, experiment II was
performed in a similar way but without solvent, i.e., with poly-
crystalline 1. After 31 days at 77 K, the resulting gas mixture was
transferred by vacuum methods into an NMR tube containing in
the lower part a degassed, frozen solution of 1 in C6D6. The analysis
was then done as in experiment I.
Hv-Hv + Hv-Hv a Hv-HV + Hv-HV (chemical SC) (1)
H2 + D2 a 2HD (IS)
H-H + R2CdCR2 a R2CH-CHR2 (hydrogenation) (3)
Hv-Hv a Hv-HV (magnetic SC)
(4)
(2)
Initially, we wanted to study the magnetic SC (eq 4) of single
dihydrogen pairs induced by binding to a transition metal. This
binding increases the H-H distance and reduces the energy dif-
ference ∆E between the ortho and para states.5,6 Eventually, we
expected magnetic SC to take place in the absence of an external
magnetic field, induced by dipolar or scalar nuclear magnetic inter-
actions.7 However, magnetic SC in dihydrogen complexes seems
to be masked by chemical SC and IS. For example, polycrystalline
W(CO)3(PiPr3)2(H2) catalyzes IS,5b and the solid HD complex
disproportionates into a statistical mixture of the H2, HD, and D2
isotopologs. Therefore, our attention turned to transition metal com-
plexes that are able to form dihydride complexes containing two
inequivalent scalar coupled hydrogen sites, where ∆E is very small.
The 1H NMR signals of these sites exhibit just after the incorpora-
tion of p-H2 a para-hydrogen-induced polarization (PHIP), a phe-
nomenon that has been described by Eisenberg et al.8 and Bowers
et al.9 PHIP has been used for the elucidation of the mechanisms
of hydrogenation reactions catalyzed by transition metal catalysts
in liquid solution.10 In fact, PHIP is the result of a magnetic SC,
where the nuclear interaction involved is the frequency shift differ-
ence ∆ν between the two hydrogen sites, induced by an external
magnetic field. As has been shown by some of us,11 ∆ν has to be
on the order of ∆E or smaller for magnetic SC to take place.
On the other hand, in the experiments performed by Eisenberg
et al.,8 which led to the discovery of PHIP, p-H2 had been formed
in the absence of a magnetic field in frozen solutions of transition
metal dihydride precursors in contact with dihydrogen at 77 K. After
warming up to room temperature, the p-H2 was detected via the
PHIP phenomenon. As the mechanism of the low-temperature SC
was unexplored to date, we performed the following experiments
1
Figure 1. 500 MHz H NMR spectra at 298 K of 1 dissolved in C6D6 in
the presence of dihydrogen mixtures. (a) Experiment I: 400 mbar H2 and
400 mbar D2 were stored over a frozen solution of 1 in C6D6 for 5 days at
77 K prior to the NMR experiment. (b) Experiment II: 400 mbar H2 and
400 mbar D2 were stored over polycrystralline 1 for 31 days at 77 K and
transferred into an NMR tube containing a frozen solution of 1 in C6D6.
The spectra were taken after thawing.
The oxidative addition of H2 to 1 leading to IrH2(CO) Cl(PPh3)2
(2) does not occur at 77 K but starts at room temperature and is
complete after about 1 h. The partial room-temperature H NMR
1
spectra of the samples 2 min after thawing are depicted in Figure
1. The two hydride sites of 2 give rise to triplets of doublets (JHP
) 12 Hz; JHH ) 4.5 Hz) at -6.6 and -17.35 ppm; moreover, they
exhibit the well-known PHIP pattern with alternating enhanced
positive and negative signal components as was described for this
system by Duckett et al.12 Around 4.45 ppm, we observe not only
o-H2 but also a signal for HD giving rise to the well-known triplet
(JHD ) 43 Hz). These observations show that in the frozen solution
and in the polycrystalline material, not only p-H2 but also HD is
formed. Control experiments performed without the catalyst showed
no p-H2 enrichment and no formation of HD. Although a quantita-
tive determination of the relative rates of both processes was beyond
the scope of this study, we can now conclude that a substantial
part of p-H2 is generated in the solid state by chemical SC, rather
† Freie Universita¨t Berlin.
‡ Laboratoire de Chimie de Coordination.
9
8366
J. AM. CHEM. SOC. 2004, 126, 8366-8367
10.1021/ja0475961 CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
therefore, take place either in the bulk solid or only in the crystal
surfaces. It is unlikely that a purely magnetic spin conversion can
take place in 3.7
The details of the reaction pathways are not known, but they
could involve species similar to those shown in Chart 1c. It is,
however, natural to assume that 3 is also the active species of the
hydrogenation of ethene. Evidence for a similar mechanism that
excludes a direct binding of the substrate to the metal center was
provided by Noyori et al.13a and by Morris et al.13b for the case of
hydrogenation of ketones by Ru complexes. Currently, we are
exploring the various mechanisms in more detail, both theoretically
and experimentally.
We conclude that exploring both the liquid- and solid-state
activity of a catalyst may help to bridge the gap between the
homogeneous and heterogeneous catalysis and to provide additional
information about the reaction mechanisms.
1
Figure 2. 500 MHz gas-phase H NMR spectrum of H2 and ethylene in
an NMR tube that was (a) empty and (b) filled with Ir(Cl)CO(PPh3)2
crystalline powder. Measurements were taken at 298 K after storage for x
days at 298 K in the dark.
Acknowledgment. This work has been supported by the DFG,
the CNRS, the Fonds der Chemischen Industrie, and the (DFH/
UFA) Robert Bosch Stiftung. We thank Dr. Bo¨ttcher, FU-Berlin,
for the optical microscope experiments.
Chart 1
Supporting Information Available: Experimental details and
additional NMR spectra (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Tadros, M. E.; Vaska, L. J. Colloid Interface Sci. 1982, 85, 389.
(2) For studies of SC of diamagnetic transition metal complexes in solution,
see: (a) Calvin, M. Trans. Faraday Soc. 1938, 34, 1181. (b) Wilmarth,
W. K.; Barsh, M. K. J. Am. Chem. Soc. 1953, 75, 2237. (c) Brown, J. M.;
Canning, L. R.; Downs, A. J.; Forster, A. M. J. Organomet. Chem. 1983,
225, 103.
(3) For studies of IS of diamagnetic transition metal complexes in solution,
see: (a) Eberhardt, G. G.; Vaska, L. J. Catal. 1967, 8, 183. (b) Ando, N.;
Maruya, K.; Mizoroki, T.; Ozaki, A. J. Catal. 1972, 20, 299. (c) Kubas,
G. J.; Ryan, R. R.; Wroblewski, D. J. Am. Chem. Soc. 1986, 108, 1339.
(d) Upmacis, R. K.; Poliakoff, M.; Turner, J. J. J. Am. Chem. Soc. 1986,
108, 3654. (e) Kubas, G. J.; Unkefer, C. J.; Swanson, B. I.; Fukushima,
E. J. Am. Chem. Soc. 1986, 108, 7000. (f) Aubart, M. A.; Pignolet, L. H.
J. Am. Chem. Soc. 1992, 114, 7901. (g) Aubart, M. A.; Chandler, B. D.;
Gould, R. A. T.; Krogstad, D. A.; Schoondergang, M. F. J.; Pignolet, L.
H. Inorg. Chem. 1994, 33, 3724. (h) Collman, J. P.; Fish, H. T.; Wagen-
knecht, P. S.; Tyvoll, D. A.; Chng, L. L.; Eberspacher, T. A.; Braumann,
J. I.; Bacon, J. W.; Pignolet, L. H. Inorg. Chem. 1996, 35, 6746.
(4) Strohmeier, W.; Steigerwald, H.; Lukacs, M. J. Organomet. Chem. 1978,
144, 135 and references therein.
(5) (a) Kubas, G. J.; Ryan, R. R.; Swanson, B. I.; Vergamini, P. J.;
Wassermann, H. J. Am. Chem. Soc. 1984, 106, 451. (b) Kubas, G. J. Acc.
Chem. Res. 1988, 21, 120. (c) Kubas, G. J. Metal Dihydrogen and σ-Bond
Complexes; Fackler, J. P., Ed.; Kluwer Academic/Plenum Publishers: New
York, 2001.
(6) (a) Limbach, H. H.; Scherer, G.; Maurer, M.; Chaudret, B. Angew. Chem.
1992, 104, 1414; Angew. Chem., Int. Ed. Engl. 1992, 31, 1369. (b)
Limbach, H. H.; Ulrich, S.; Buntkowsky, G.; Gru¨ndemann, S.; Sabo-
Etienne, S.; Chaudret, B.; Kubas, G. J.; Eckert, J. J. Am. Chem. Soc. 1998,
120, 7929.
(7) Preliminary model calculations show that a dipolar magnetic SC is
operative in the solid state and a scalar magnetic SC in the liquid state
when ∆E is reduced to the order of the magnetic interactions involved.
For that purpose, the H-H distances need to be increased to more than
about 1.3 Å.
(8) (a) Eisenschmid, T. C.; Kirss, R. U.; Deutsch, P. P.; Hommeltoft, S. I.;
Eisenberg, R.; Bargon, J.; Lawler, R. G.; Balch, A. L. J. Am. Chem. Soc.
1987, 109, 8089. (b) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc.
1987, 109, 5541. (c) Eisenberg, R. Acc. Chem. Res. 1991, 24, 110.
(9) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc. 1987, 109, 5541.
(10) (a) Bargon, J.; Kandels, J.; Woelk, K. Z. Phys. Chem. 1993, 180, 65. (b)
Natterer J.; Bargon, J. Prog. Nucl. Magn. Reson. Spectrosc. 1997, 31,
293. (c) Duckett, S. B.; Sleigh, C. J. Prog. Nucl. Magn. Reson. Spectrosc.
1999, 34, 71.
than by magnetic SC. The latter dominates, however, in the liquid
state in the presence of a magnetic field.
This result was surprising for us because magnetic SC requires
a strong interaction of only one dihydrogen molecule with a
transition metal catalyst, whereas for chemical SC, at least two
dihydrogen molecules are required. By contrast, in the case of Kubas
dihydrogen complex5b only one additional dihydrogen molecule is
required for IS to occur. It was also surprising that dihydrogen could
penetrate easily the frozen solid solutions and access the catalyst.
This problem incited us to study the nature of the frozen solutions
of 1 in benzene by optical microscopy at 258 K. We observed a
microscopic phase separation into irregular formed particles of 1
and solid benzene and pores with typical in the 20 µm range. Thus,
it is conceivable that dihydrogen can penetrate the frozen mixture
via the pores and interact with small particles of 1, in a similar
way as with the polycrystalline samples. In the latter case, however,
the particles are larger and the surface area is smaller; hence, the
conversion efficiency is smaller.
Encouraged by these results, we checked whether the crystallites
of 1 are also active catalysts for the hydrogenation of larger
molecules such as alkenes (eq 3). For this purpose we stored a 1:2
mixture of ethene and H2 (300 mbar C2H4 and 600 mbar H2, 298
K) in (a) an empty NMR tube and (b) in an NMR tube filled with
polycrystalline 1. The resulting gas mixtures were analyzed by gas-
1
phase H NMR (Figure 2). Without catalyst, the gas mixture did
not react within 7 days; only a sharp peak of ethene is detected at
5.15 ppm and a broad one for o-H2 at 4.45 ppm. However, when
the mixture is stored over polycrystalline 1, a sharp ethane peak
appears at 0.83 ppm. The ethene peak is gone after 7 days.
All results are summarized in Chart 1. In the liquid phase, 1
reacts with dihydrogen to 2, presumably via a (stretched)5 dihy-
drogen complex 3, which also probably catalyzes the reactions of
eq 1-3. In the solid state, the reaction of 1 to 2 is inhibited because
it would require a major ligand reorientation. On the other hand,
the formation of 3 may not require a ligand orientation and may,
(11) Buntkowsky, G.; Bargon, J.; Limbach, H. H. J. Am. Chem. Soc. 1996,
118, 8677.
(12) (a) Messerle, B. A.;Sleigh, C. J.; Partridge, M. G.; Duckett, S. B. J. Chem.
Soc., Dalton Trans. 1999, 1429. (b) Hasnip, S. K.; Duckett, S. B.; Sleigh,
C. J.; Taylor, D. R.; Barlow, G. K.; Taylor, M. J. Chem. Commun. 1999,
1717.
(13) (a) Yamakawa, M.; Ito, H.; Noyori, R. J. Am. Chem. Soc. 2000, 122,
1466. (b) Abdur-Rashid, K.; Lough, A. J.; Morris, R. H. Organometallics
2000, 19, 2655.
JA0475961
9
J. AM. CHEM. SOC. VOL. 126, NO. 27, 2004 8367