Change in reaction pathway induced by deuteration: Thermal decomposition of neopentyl groups on Pt(111) surfaces

Full text of DOI:10.1021/ja962606m

J. Am. Chem. Soc. 1997, 119, 1169-1170
Change in Reaction Pathway Induced by
1169
Deuteration: Thermal Decomposition of Neopentyl
Groups on Pt(111) Surfaces
Ton V. W. Janssens, Gaolong Jin, and Francisco Zaera*
Department of Chemistry
UniVersity of California, RiVerside, California 92521
ReceiVed July 29, 1996
Previous surface science studies on the chemistry of alkyl
groups on various metal surfaces have revealed that their thermal
decomposition often involves an initial â-H elimination step,
in a fashion similar to that of alkyl ligands in organometallic
compounds, to produce the corresponding alkene and surface
hydrogen.1 Since in most cases this â-H elimination reaction
is dominant, much less is known about other hydrocarbon
reactions such as R- or γ-H eliminations or C-C bond scission
steps. The thermal decomposition of adsorbed neopentyl groups
[(CH3)3CCH2-], which do not have hydrogen atoms at the â
position, can provide more information about those less common
reactions. In this paper we present temperature-programmed
desorption (TPD) data for nondeuterio- [(CH3)3CCH2-], R-deu-
terio- [(CH3)3CCD2-], γ-deuterio- [(CD3)3CCH2-], and per-
deuterio- [(CD3)3CCD2-] neopentyl groups adsorbed on Pt(111)
surfaces. It was found that in the decomposition of both R-
and γ-deuterated neopentyl iodides a hydrogen, not deuterium,
is eliminated first. This indicates that while γ-H elimination
takes place in the former case, R-H elimination occurs in the
latter system instead. It is suggested that such a change in
reaction path is due to the kinetic isotope effect associated with
the deuterium substitution, which makes the reactions involving
deuterium atoms slower. This explanation implies that the rates
of the R- and γ-H elimination reactions are probably of the same
order of magnitude, since their difference must be less than the
changes caused by deuterium substitution.
The TPD experiments were performed in an ultra-high-
vacuum chamber pumped to a base pressure below 10-10 Torr
and equipped with an ion sputtering gun for cleaning of the
sample, a UTI-100C mass spectrometer for temperature-
programmed desorption experiments, and a Mattson Sirius 100
FT-IR spectrometer for reflection-absorption infrared spectro-
scopy (RAIRS).2,3 The mass spectrometer was controlled by a
personal computer, in a setup that allowed for the simultaneous
data collection of up to 15 masses. A moveable cone was
mounted on the top of the mass spectrometer to minimize the
signals from both the background gases and the desorption from
the sample holder. The (isotopically labeled) neopentyl groups
were produced by thermal decomposition of the respective
neopentyl iodides on the Pt surface; RAIRS spectra show
definite changes associated with the breaking of the C-I bond
upon increasing the adsorption temperature of the neopentyl
iodide from 115 to 170 K.4 The nondeuterioneopentyl iodide
was purchased from Aldrich (98% purity) and used as supplied,
the R-deuterioneopentyl iodide was synthesized by reduction
of trimethylacetyl chloride (Aldrich, 99%) with LiAlD4 (Aldrich,
98 D-atom %) to R-deuterioneopentyl alcohol5 and subsequent
iodine substitution,6 and the γ- and perdeuterioneopentyl iodides
were prepared from acetone-d6 (Aldrich, 99.5 D-atom %), via
pinacolone-d12, γ-deuteriotrimethyl acetic acid,7,8 and either γ-
Figure 1. Left: Temperature-programmed desorption from 3.0 L of
nondeuterioneopentyl iodide adsorbed on Pt(111) showing the desorp-
tion of the main products, namely neopentane (57 amu) and hydrogen
(2 amu). Left, inset: Mass spectrum for normal neopentane. Right:
Neopentane temperature-programmed desorption from 3.0 L of per-
deuterioneopentyl iodide on Pt(111). The high intensity of the 66-amu
signal shows that the neopentane produced in this case is mainly C5D12.
or perdeuterioneopentyl alcohol.5 The purity of the vapors was
checked by mass spectrometry in situ in the vacuum chamber.
Doses are reported in langmuirs (1 L t 10-6 Torr‚s).
Figure 1 displays typical TPD spectra for the decomposition
of neopentyl groups on Pt(111). The left panel of that figure
shows the hydrogen (2 amu) and neopentane (57 amu) TPD
traces obtained after a 3.0 L dose of nondeuterioneopentyl iodide
on the surface; those were the only two desorption products
observed in this case (molecular desorption occurs at 185 K,
but only after exposures above 5.0 L). The hydrogen desorption
that occurs between 400 and 600 K is typical for decomposition
of surface hydrocarbon fragments, and the trace for the 57 amu
is accompanied by similar signals for 43, 41, 39, and 29 amu
(not shown), with relative intensities matching those for
neopentane (Figure 1, inset). It is important to notice here that
the intensity of the parent molecular peak of neopentane (72
amu) is more than 100 times weaker than that of the terbutyl
fragment (57 amu),9,10 and could therefore not be resolved in
our TPD experiments.
The right panel of Figure 1 shows the desorption traces for
65 and 66 amu obtained in TPD experiments with 3.0 L of
perdeuterioneopentyl iodide: the 66 amu signal is clearly the
most intense, which means that perdeuterioneopentane, C5D12,
is preferentially formed in this case. This neopentane must form
via the reductive elimination of surface neopentyl groups with
the deuterium atoms that originate from dehydrogenation of
other neopentyl fragments, which means that the first step in
the decomposition of neopentyl groups can be identified by
monitoring the deuterium content of the neopentane formed in
the TPD experiments with partially deuterated neopentyl iodides.
In addition, the low intensity of the 65-amu trace in Figure 1
indicates that the amount of neopentane produced by reaction
with background hydrogen is relatively small. The two peaks
observed in this desorption trace point to isotope effects in the
decomposition and hydrogenation reactions of the neopentyl
moieties.
(7) Adams, R.; Adams, E. W. In Organic Syntheses, 2nd ed.; Gilman,
H., Blatt, A. H., Eds.; J. Wiley: New York, 1941; Vol. I, p 459. Hill, G.
A.; Flosdorf, E. W. Ibid., p 462; Sandborn, L. T.; Bousquet, E. W. Ibid., p
526.
(8) Vogel, A. I. In A Textbook of Practical Organic Chemistry, 3rd ed.;
J. Wiley & Sons: New York, 1962; pp 349-351.
(9) Handbook of Data on Organic Compounds, 3rd ed.; Lide, D. R.,
Milne, G. W. A., Eds.; CRC Press, Inc.: Boca Raton, 1994; Vol. V.
(10) Atlas of Mass Spectral Data; Stenhagen, E., Abrahamsson, S.,
McLafferty, F. W., Eds.; J. Wiley & Sons: New York, 1969; Vol. 1.
* Corresponding author.
(1) Zaera, F. Chem. ReV. 1995, 95, 2651.
(2) Hoffmann, H.; Griffiths, P. R.; Zaera, F. Surf. Sci. 1992, 262, 141.
(3) Janssens, T. V. W.; Zaera, F. Surf. Sci. 1995, 344, 77.
(4) Janssens, T. V. W.; Zaera, F. In preparation.
(5) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc. 1947, 69, 2548.
(6) Rydon, H. N. In Organic Synthesis; Benson, R. E., Ed.; J. Wiley &
Sons: New York, 1971; Vol. 51; pp 44-45.
S0002-7863(96)02606-6 CCC: $14.00 © 1997 American Chemical Society

1170 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Communications to the Editor
steps must dominate regardless of the position of the hydrogen
atoms in the molecule. Based on rate ratios from the data in
Figure 2, the isotope effect for the hydride elimination steps
was estimated to be between five and ten (factors larger than
ten have in fact been reported for the dissociation of the
neopentane ligand in (Cp*Rh(CO)(CH3)4C to neopentyl and
hydride ligands at temperatures below 200 K).11 The same
calculations also indicated that the rate constant for R-H
elimination is about five times larger than that for γ-H
elimination, but since the latter is statistically more favorable
than the former (there are nine γ and only two R hydrogens in
neopentyl moieties), the two rates end up being about the same
(within 15% of each other).
Such a change in decomposition pathway induced by deu-
teration was not observed in the decomposition of dineopen-
tylbis(triethylphosphine) platinum(II), where γ-H elimination
was preferred in both the R- and γ-deuterated neopentyl groups
instead.12 A different behavior was also observed on Ni(100)
surfaces, where R-H elimination, which is the preferred reac-
tion,13 is followed by a subsequent C-C bond scission reaction
to isobutene;14 this latter reaction was not observed on the
Pt(111) surface.
3,3-Dimethylplatinacyclobutane is an example of the type of
metallacycles that may result from the R,γ-coordination of
hydrocarbon moieties believed to occur during the catalytic
hydrogenolysis and isomerization of alkanes.15 The data
presented in this paper strongly suggest the formation of a
R,γ-diadsorbed species on Pt(111) in the case of neopentyl
moieties on Pt(111), though this reaction seems to compete with
the production of neopentylidene species. Our data are also
consistent with the relatively high selectivity toward the
formation of neopentane-d1 to -d3 in H/D exchange processes
over supported Pt catalysts, which can be achieved via a
neopentylidene intermediate, compared to the production of
neopentane-d4 to -d9, which involves an R,γ-diadsorbed inter-
mediate.15,16
Figure 2. Neopentane temperature-programmed desorption from 3.0
L of R-deuterio- (left) and γ-deuterio- (right) neopentyl iodide adsorbed
on Pt(111). The neopentane is in both cases formed via a hydrogen,
not deuterium, transfer.
Figure 2 summarizes the most important neopentane desorp-
tion traces obtained in TPD experiments with 3.0 L of R- (left)
and γ- (right) deuterioneopentyl iodides. In the case of
R-deuterioneopentyl iodide the most intense signal is that for
59 amu, indicating the predominant desorption of neopentane-
d2. This means that the formation of this neopentane mainly
involves hydrogen, not deuterium atoms, implying that the first
step in the decomposition of R-deuterioneopentyl groups is a
γ-H elimination step (which is presumably accompanied by the
formation of a 3,3-dimethylplatinacyclobutane species on the
surface). The low intensity seen for the 60-amu trace may be
due to a small amount of deuteration, which would imply that
either some R-D elimination takes place as well or some
previous H/D exchange occurs within the surface species. The
most intense peak in TPD experiments with the γ-deuterioneo-
pentyl iodide, on the other hand, is that for the 63 amu, indicative
of the formation of neopentane-d9. Quite remarkably, a
hydrogen atom is transferred in this case too, which means that,
in contrast to the case of the R-deuterioneopentyl species, the
first step in the γ-deuterioneopentyl decomposition is an R-H
elimination reaction (probably leading to the formation of
surface neopentylidene groups). The intensity of the 64-amu
peak can again be caused by either γ-D elimination or H/D
exchange.
Acknowledgment. Financial support for this research was provided
by a grant from the National Science Foundation (CHE-9530191).
JA962606M
(11) Bengali, A. A.; Schultz, R. H.; Moore, C. B.; Bergman, R. G. J.
Am. Chem. Soc. 1994, 116, 9585.
(12) Foley, P.; DiCosimo, R.; Whitesides, G. M. J. Am. Chem. Soc. 1980,
102, 6713.
(13) Zaera, F.; Tjandra, S. J. Am. Chem. Soc. 1996, 118, 12738.
(14) Zaera, F.; Tjandra, S. J. Am. Chem. Soc. 1993, 115, 5851.
(15) Oliver, J. A.; Kemball, C.; Brown, R.; Jamieson, E. W. J. Chem.
Soc., Faraday Trans. 1 1985, 81, 1871.
(16) Brown, R.; Kemball, C.; Oliver, J. A.; Sedler, I. H. J. Chem. Res.
1985, 3201.
A possible explanation for the different reaction paths for R-
and γ-deuterioneopentyl groups is that the kinetic isotope effect
associated with deuterium substitution controls the selectivity
of the reaction: since reactions involving deuterium atoms are
expected to be slower than those with hydrogen, H-elimination