Catalytic Hydrogenolysis and Isomerization of Light Alkanes over the Silica-Supported Titanium Hydride Complex (SiO)3TiH

Full text of DOI:10.1021/ja972488i

12408
J. Am. Chem. Soc. 1997, 119, 12408-12409
(TiH) at 1706, 1692, 1679, and 1647 cm-1 and two bands
associated with ν(SiH) at 2263 and 2196 cm-1. Methane and
ethane were formed in a 1:1 ratio (total carbon given off
corresponds to the release and hydrogenolysis of 3 equiv of
neopentane). Reaction of this complex with deuterium gas (1
atm, 10 mmol) showed rapid disappearance of the ν(TiH) bands
(within 15 min at room temperature) with no modification of
ν(SiH) bands: the expected ν(TiD) bands were not observed
as the silica is opaque in the region corresponding to these bands
(around 1200 cm-1). Subsequent reaction with hydrogen gas
completely restored the ν(TiH) bands. Further characterization
of this complex has been accomplished by chemical analyses7
which leads to the tentative formulation (tSiO)3TiH (2).
Catalytic Hydrogenolysis and Isomerization of Light
Alkanes over the Silica-Supported Titanium
Hydride Complex (tSiO)3TiH
Ce´cile Rosier, Gerald P. Niccolai,* and Jean-Marie Basset*
Laboratory of Surface Organometallic Chemistry
UMR 9986 CNRS-CPE Lyon
43, bouleVard du 11 noVembre 1918
F-69616 Villeurbanne Cedex, France
ReceiVed July 23, 1997
We have recently reported the hydrogenolysis of light alkanes
under very mild conditions of temperature and pressure cata-
lyzed by (tSiO)3ZrH1 and (tSiO)3HfH.2 This remarkable
reactivity can be attributed to the fact that the active species
are formally 8 e- and extremely coordinatively unsaturated. We
proposed, on the basis of the primary products observed, that
the hydrogenolysis proceeds by stepwise cleavage of carbon-
carbon bonds by â-alkyl elimination from surface metal-alkyl
intermediates. We now report that under identical conditions
the reactions of neopentane and isobutane with hydrogen over
(tSiO)3TiH cannot be simply described as hydrogenolysis but
that alkane skeletal isomerization also occurs.
The title complex, (tSiO)3TiH, is synthesized in two steps
from TiNp4 [Np ) CH2C(CH3)3] and silica (Degussa Aerosil,
200 m2/g). All reactions are performed in sealed tube reactors
pre-equipped with break-seal tubes for the introduction of
nongaseous reactants and CaF2 windows to allow aquisition of
infrared spectra, as described elsewhere.3
The formation of methane and ethane rather than neopentane
indicates that 2 catalyzes the hydrogenolysis of neopentane.
Similar observations were made regarding the formation and
reactivity of the analogous surface complexes (tSiO)3ZrH1 and
(tSiO)3HfH.2 What surprised us was the ratio of methane to
ethane at completion which was 1:1 in the case of Ti, but was
always observed in a strict 3:1 ratio for Zr and Hf. For these
latter cases, we have proposed a mechanism in which neopen-
tane undergoes C-H bond activation to form a metal-neopentyl
complex. The carbon-carbon bond cleavage step was one of
â-methyl elimination to form a metal-alkyl-olefin complex,8
which on further reaction with hydrogen produces methane and
isobutane (or a surface metal-isobutyl complex) (eq 1). The
A silica wafer was dehydroxylated under dynamic vacuum
for 15 h at 500 °C (silica500). The infrared spectrum indicates
the presence of surface silanols free of any hydrogen bonding.
Tetraneopentyl titanium4 is sublimed (60 °C, 10-4 mm Hg) onto
the disk at room temperature. The disk turns from white to
yellow, neopentane is given off, the ν(SiO-H) band (3747
cm-1) is consumed, and new bands corresponding to the
isobutane (or isobutyl fragment) is in turn converted by the same
mechanism to methane and propane and then propane to
methane and ethane. The carbon-carbon bond of ethane cannot
be broken by this mechanism as a surface metal-ethyl fragment
has no carbon-carbon bond â to the metal.
The fact that in the titanium case we observed a methane to
ethane ratio at completion of 1:1 indicates that there was some
skeletal rearrangement of the alkyl ligands occurring on the same
time scale as the overall hydrogenolysis reaction. We have
studied this by following the evolution of products in the
catalytic hydrogenolysis of neopentane and isobutane with 2 at
various temperatures and conversions (Figures 1 and 2).9
neopentyl ligand ν(CH) (2956, 2905, 2867, and 2800 cm-1
)
and δ(CH) (1466, 1393, 1365 cm-1) vibrations appear. When
the reaction is performed on a fully deuterated silica disk, 1
equiv of neopentane-d1 was evolved. It should be noted that
the surface complex is never isolated, per se, but rather the
formulations proposed would represent the vast majority of
titanium centers on the surface. Spectroscopy,5 other chemical
tests and elemental analyses6 are consistent with a surface
species in which a trineopentyl metal complex is linked to the
surface by one siloxy bridge, tSiOTiNp3 (1), as was the case
for the analogous zirconium and hafnium reactions.
(6) Elemental analyses were performed at the CNRS Central Analysis
Service, Solaize. For titanium analysis, the metal is dissolved in HF/HClO4/
HNO3, dried, taken up in HClO4/HNO3, and analyzed by AES. (a) 1H CP-
MAS NMR (300 MHz, Bruker): δ 0.46 (CH3), 1.73 (CH2). (b) 13C CP-
MAS NMR (75 MHz, Bruker): δ 36.7 (CH3), 37.2 (C(CH3)3). The CH2
was not observed. (c) Hydrolysis of the surface complex shows removal of
all alkyl ligands in the IR: quantification of neopentane given off and
elemental analysis of the solid yields a NpH/Ti ratio of 3. (d) Reaction
with O2 produces no gaseous product, and the elemental analysis of the
solid yields a C/Ti ratio of 14.8 (expected, 15). 1H CP-MAS NMR (300
MHz, Bruker): δ 1.10 (CH3), 2.03 (CH2). 13C CP-MAS NMR (75 MHz,
This alkyl complex (20 mg, 1 wt % Ti, 9.5 µmol) was treated
under hydrogen (1 atm, 10 mmol) at 150 °C for 1 h. The solid
turned from pale yellow to brownish yellow. In the infrared
spectrum, the bands associated with the neopentyl ligands
disappear with the formation of four bands associated with ν-
(1) (a) Lecuyer, C.; Quignard, F.; Choplin, A.; Olivier, D.; Basset, J. M.
Angew. Chem. 1991, 103, 1692-1694; Angew. Chem., Int. Ed. Engl. 1991,
30 (12) 1660-1661. (b) Quignard, F.; Lecuyer, C.; Choplin, A.; Olivier,
D.; Basset, J. M. J. Mol. Catal. 1992, 74, 353-63. (c) Corker, J.; Lefebvre,
F.; Lecuyer, C.; Dufaud, V.; Quignard, F.; Choplin, A.; Evans, J.; Basset,
J.-M. Science 1996, 271 (5251), 966-969.
t
Bruker): δ 25.1 (CH3), 33.9 (C(CH3)3), 89.1 (CH2). (e) Reaction with -
BuOH produced a solid, tSiOTi(OtBu)3 for which the elemental analysis
yields a C/Ti ratio of 13 (expected, 12). 13C CP-MAS NMR (75 MHz,
Bruker): δ 29.6 (C(CH3)3), and no peaks at δ 36-38. The quaternary carbon
of the tBu was not observed. (f) Reaction with hydrogen at 150 °C produces
methane and ethane in quantities consistent with three neopentyl ligands
(vide infra).
(2) d’Ornelas, L.; Reyes, S.; Quignard, F.; Choplin, A.; Basset, J. M.
Chem. Lett. 1993, 1931-1934.
(3) Quignard, F.; Lecuyer, C.; Bougault, C.; Lefebvre, F.; Choplin, A.;
Olivier, D.; Basset, J. M. Inorg. Chem. 1992, 31, 928-30.
(7) (a) Reaction with OdC(tBu)2 produces a white solid (presumably
ketone insertion into Ti-H bond). Microanalysis revealed C/Ti ratio of 7
(expected, 9 for the monohydride). (b) Reaction with CH3I produced 0.7
equiv of methane per Ti (expected, 1.0). (c) Reaction with tBuOH converts
hydrides to tert-butoxides. Microanalysis revealed C/Ti ratio of 5 (expected,
4).
(4) The complex is synthesized as decribed in: Mowat, W.; Wilkinson,
G. J. Chem. Soc., Dalton Trans. 1973, 1120-1124.
(5) Infrared spectra were obtained using a Nicolet Magna-IR 550
spectrometer with a Mid-IR (4000-400 cm-1) detector at 4 cm-1 resolution.
Solid state CP-MAS NMR were obtained on a Bruker DSX 300 spectrom-
(8) The direct observation of the â-alkyl elimination has been reported:
Horton, A. D. Organometallics 1996, 15, 2675-2677. This mechanism has
been implicated in a number of reactions. See refererences in the above
article.
1
eter (90°, 3.8 ms H, contact time 10 ms, recyle delay 2 s). Mass spectra
were obtained on a HP-GCD GC-MS system at 120 °C (gas chromatog-
raphy, Chromopack KCl/Al2O3, 50 m × 250 µm, isothermal at 80 °C).
S0002-7863(97)02488-8 CCC: $14.00 © 1997 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 119, No. 50, 1997 12409
Figure 1. Evolution of products and selectivity for the reaction of
neopentane with hydrogen in the presence of 2 at 150 °C.
Figure 2. Evolution of products and selectivity for the reaction of
isobutane with hydrogen in the presence of 2 at 150 °C.
The skeletal alkane isomerization is not difficult to explain.
Following an initial step of C-H bond activation the surface
alkyl fragment can undergo â-methyl elimination. For titanium,
the intermediate metal-alkyl-olefin complex persists long
enough (relative to olefin rotation) to undergo reinsertion of
the rotated olefin into the metal-carbon bond. This isomeri-
sation step is illustrated for the neopentyl fragment in eq 2.
For neopentane at low conversion, methane, isobutane, and
n-butane are primary products. Although isopentane was not
observed, the presence and importance of n-butane are a clear
indication of carbon skeletal rearrangement prior to the forma-
tion of products. Furthermore, the fact that propane and ethane
are initially formed in a 1:1 ratio can also be considered as the
signature of the isomerization of surface neopentyl fragments
to isopentyl fragments prior to product formation.
A similar result is obtained in the reaction of isobutane with
hydrogen at 150 °C (Figure 2). Ethane is clearly a primary
product, the implication being that the initially formed surface
isobutyl fragment has rearranged to an n-butyl fragment.10 In
this case, the C4 isomerization product n-butane is observed in
significant quantities and the selectivity for this product increases
at very low conversion, suggesting that it is indeed a primary
product (although a minor one).
These reactions were repeated at 50 and 100 °C (see
Supporting Information). For the hydrogenolysis of neopentane,
this effect was clearly present at these temperatures. For
isobutane, the process of isomerization is not clearly indicated.
Thus the remarkable activity of these surface complexes with
respect to carbon-carbon bonds is thus more interesting given
the possibility of light hydrocarbon isomerization at very mild
temperatures.
(9) In a typical reaction, a reactor is charged with 2 (isolated from
reactants by a break-seal, ∼25 mg, 1-1.7 wt % Ti, ∼10 µmol Ti-H; weight
of wafer and wt % Ti varies from one experiment to another), the alkane
(50 mm Hg, 60-80 mol/mol of Ti) and hydrogen (1 atm, 15.2 mol/mol of
alkane). The reactor was heated to the reaction temperature, and the break-
seal was broken (t ) 0). Gases were sampled by opening the reactor to a
closed known volume equipped with a septum. The reactor was then
reisolated, the known volume was filled to atmospheric pressure with air,
and the gases were removed by syringe for quantification by gas chroma-
tography (Chrompack KCl/Al2O3 column, 50 m × 250 µm, FID detector).
(10) We have previously demonstrated that the hydrogenolysis of
n-butane catalyzed by (tSiO)3HfH2 leads to the formation of primarily
ethane by a mechanism of initial primary C-H bond activation to produce
surface n-butyl and subsequent â-ethyl elimination to form C2 fragments.
Methane and propane are formed in a minor pathway by first-step secondary
C-H bond activation to give surface Hf-CH(CH3)CH2CH3 which then
undergoes â-methyl elimination eventually leading to methane and propane.
Acknowledgment. The authors thank Dr. Fre´deric Lefebvre for the
CP-MAS NMR spectroscopy and analysis.
Supporting Information Available: Infrared spectra (4000-1300
cm-1) corresponding to the syntheses of 1 and 2; NMR spectra cited
in footnotes; graphs of the evolution of gaseous products in the reaction
of neopentane and isobutane at 50 and 100 °C (equivalents of product
vs time, selectivity vs neopentane conversion) (10 pages). See any
current masthead page for ordering and Internet access instructions.
JA972488I