The thermal chemistry of neopentyl iodide on Ni(100) surfaces: Selectivity between α-C-H and γ-C-H and between C-H and C-C bond-scission steps in chemisorbed neopentyl moieties

Article abstract of DOI:10.1021/ja9628056

The thermal chemistry of neopentyl iodide on Ni(100) single-crystal surfaces was characterized under vacuum by using temperature-programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS). The first decomposition step, which takes place around 150 K, is the scission of the C-I bond, the same as in other chemisorbed alkyl halides. Owing to the absence of β hydrogens, however, no easy decomposition pathway is available for the resulting neopentyl surface species. Neopentane is produced via neopentyl reductive elimination with surface hydrogen, and desorbs in two stages around 140 and 180 K. The yield for this pathway is approximately 45% of the initial neopentyl iodide at saturation (which is approximately 0.2 ML) on the clean nickel, but reaches a value close to 100% if enough hydrogen (or deuterium) is predosed on the surface. The other major carbon-containing product from neopentyl iodide activation is isobutene, which desorbs around 400 K. Isotope labeling experiments demonstrated that the C-C bond that breaks in that reaction is the one between the α and β carbons, and highlighted the fact that the kinetics of the overall reaction displays strong isotope effects upon deuterium substitution at either the α or γ positions. In addition, the hydrogen TPD traces indicated that one of the two hydrogens from the α carbon of the neopentyl group is removed at low temperatures (below 300 K), suggesting that the precursor to isobutene formation is a neopentylidene intermediate.

Full text of DOI:10.1021/ja9628056

12738
J. Am. Chem. Soc. 1996, 118, 12738-12746
The Thermal Chemistry of Neopentyl Iodide on Ni(100)
Surfaces: Selectivity between R-C-H and γ-C-H and between
C-H and C-C Bond-Scission Steps in Chemisorbed Neopentyl
Moieties
Francisco Zaera* and Sariwan Tjandra
Contribution from the Department of Chemistry, UniVersity of California,
RiVerside, California 92521
X
ReceiVed August 12, 1996
Abstract: The thermal chemistry of neopentyl iodide on Ni(100) single-crystal surfaces was characterized under
vacuum by using temperature-programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS). The
first decomposition step, which takes place around 150 K, is the scission of the C-I bond, the same as in other
chemisorbed alkyl halides. Owing to the absence of â hydrogens, however, no easy decomposition pathway is
available for the resulting neopentyl surface species. Neopentane is produced via neopentyl reductive elimination
with surface hydrogen, and desorbs in two stages around 140 and 180 K. The yield for this pathway is approximately
4
5% of the initial neopentyl iodide at saturation (which is approximately 0.2 ML) on the clean nickel, but reaches
a value close to 100% if enough hydrogen (or deuterium) is predosed on the surface. The other major carbon-
containing product from neopentyl iodide activation is isobutene, which desorbs around 400 K. Isotope labeling
experiments demonstrated that the C-C bond that breaks in that reaction is the one between the R and â carbons,
and highlighted the fact that the kinetics of the overall reaction displays strong isotope effects upon deuterium
substitution at either the R or γ positions. In addition, the hydrogen TPD traces indicated that one of the two
hydrogens from the R carbon of the neopentyl group is removed at low temperatures (below 300 K), suggesting that
the precursor to isobutene formation is a neopentylidene intermediate.
1
. Introduction
ature-programmed desorption (TPD) and X-ray photoelectron
spectroscopy (XPS) studies on the chemistry of neopentyl iodide
on Ni(100) surfaces. Since this is one of the simplest molecules
with no hydrogens at the â position, it is ideal for testing the
selectivity between R- and γ-hydride elimination steps. It was
found that removal of an R-hydrogen occurs selectively below
Previous studies have shown that alkyl iodides are ideal
precursors for the preparation of alkyl groups cleanly on metal
surfaces.1
-4
Indeed, the initial thermal activation of alkyl
iodides on most transition metals leads to the scission of their
C-I bond and to the generation of alkyl fragments and iodine
atoms on the surface. The resulting alkyl fragments can then
be used to study the reactions involved in hydrocarbon conver-
sion processes such as Fisher-Tropsch synthesis and alkane
activation.5 Our recent studies on Ni(100) have shown that
small alkyl fragments undergo both hydrogenation and dehy-
drogenation reactions to form alkanes and alkenes, respec-
3
00 K and yields neopentylidene moieties on the surface, and
that the subsequent activation of that intermediate produces gas-
phase isobutene. The C-C bond scission involved in the latter
reaction occurs exclusively at the R-â position, and displays
strong kinetic isotope effects upon substitution of hydrogens
for deuteriums at either the R or γ positions. The results
reported here are discussed in terms of a possible mechanism
for the overall neopentyl conversion.
,6
7
-10
tively;
the alkene is produced via a â-hydride elimination
step, while the alkane is generated via the reductive elimination
of alkyl species with hydrogen atoms originating from either
adsorption of background gases or the â-hydride elimination
step. The latter reaction is favored by the presence of
coadsorbed surface hydrogen.1
2
. Experimental Details
All experiments were performed in an ultrahigh vacuum (UHV)
7,14
apparatus described in detail elsewhere
The chamber is evacuated
1-13
-10
by a turbomolecular pump to a base pressure below 1 × 10 Torr,
and is equipped with instrumentation for temperature-programmed
desorption (TPD), X-ray photoelectron (XPS), secondary-ion mass
As part of our continuing studies on the chemistry of alkyl
iodides on metal surfaces, here we report results from temper-
(
SIMS), Auger electron (AES), and ion scattering (ISS) spectroscopies.
X
Abstract published in AdVance ACS Abstracts, December 1, 1996.
TPD spectra were obtained by using a heating rate of about 10 K/s,
and by recording the mass-spectrometric signals of up to 15 different
masses simultaneously in a single experiment with an interfaced
computer. The TPD signal intensities are reported in absolute
desorption rates, in monolayer/s, after calibrating the original signals
from the mass spectrometer as follows: (1) the raw data, which
represent desorption fluxes, were first converted into values proportional
to partial pressures by correcting for the differences in desorption
temperature (by multiplying them by the square root of the temperature);
(2) the signal for hydrogen was then calibrated by comparison with
the TPD from a hydrogen-saturated Ni(100), which was estimated to
be about 1 ML (defined as one adsorbate per nickel surface atom) of
(
(
(
(
1) Zaera, F. Acc. Chem. Res. 1992, 25, 260.
2) Zaera, F. J. Mol. Catal. 1994, 86, 221.
3) Zaera, F. Chem. ReV. 1995, 95, 2651.
4) Zhou, X.-L.; Zhu, X.-L.; White, J. M. Acc. Chem. Res. 1990, 23,
3
27.
(
5) Vannice, M. A. Catal. ReV.-Sci. Eng. 1976, 14, 153.
(
6) Hubert, A. Catalysis in C1 Chemistry, Keim, W., Eds.; (Reidel
Publishing Co.: Dordretcht, 1983.
(7) Tjandra, S.; Zaera, F. Langmuir 1992, 8, 2090.
(8) Tjandra, S.; Zaera, F. Surf. Sci. 1993, 289, 255.
(9) Tjandra, S.; Zaera, F. Langmuir 1994, 10, 2640.
(10) Tjandra, S.; Zaera, F. J. Am. Chem. Soc. 1995, 117, 9749.
(11) Tjandra, S.; Zaera, F. J. Catal. 1993, 144, 361.
(12) Tjandra, S.; Zaera, F. J. Catal. 1994, 147, 598.
(13) Tjandra, S.; Zaera, F. Surf. Sci. 1995, 322, 140.
(14) Zaera, F. Surf. Sci. 1989, 219, 453.
S0002-7863(96)02805-3 CCC: $12.00 © 1996 American Chemical Society

Thermal Chemistry of Neopentyl Iodide
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12739
atomic hydrogen;15,16 and (3) the other signals were referenced to that
of hydrogen by measuring relative sensitivities when leaking equal
pressures of the gases of interest (hydrogen, methane, isobutene,
neopentane, and neopentyl iodide). XPS spectra were taken by using
an aluminum-anode X-ray source and a hemispherical electron energy
analyzer of 50 mm radius set at a constant pass energy of 50 eV, which
corresponds to a resolution of about 1.2 eV full width at half-maximum
(fwhm). The energy scale was calibrated by using binding energy
values for the Pt 4f7/2 and Cu 2p3/2 signals of 70.9 and 932.4 eV,
1
4
respectively.
The nickel single crystal was cut and polished in the (100) orientation
by using standard procedures, and mounted in a manipulator capable
of cooling to liquid nitrogen temperatures and resistively heating to up
to 1300 K. The temperature of the sample was measured with a
chromel-alumel thermocouple spot-welded to the side of the crystal.
Cleaning of the crystal was done by oxygen treatment and argon ion
sputtering cycles followed by annealing to about 1000 K until no
impurities were detected by either AES or XPS. The nonisotopically
Figure 1. Hydrogen (2 amu, a), isobutene (56 amu, b), neopentane
(
57 amu, c), and neopentyl iodide (198 amu, d) temperature-
programmed desorption (TPD) spectra from neopentyl iodide adsorbed
on Ni(100) at 90 K as a function of initial exposure. The scale bar
provided in panel b for the desorption rates applies to all the panels in
this figure. A heating rate of about 10 K/s was used in these
experiments.
labeled neopentyl iodide (C
was subjected to several freeze-pump-thaw cycles before introduction
into the vacuum chamber. The neopentyl-R-d iodide was synthesized
by reduction of trimethyl acetyl chloride (Aldrich, 99%) with lithium
5
H11I) was purchased from Aldrich, and
2
17
aluminum deuteride (Aldrich, 98% D) and subsequent reaction of the
resulting neopentyl-R-d alcohol with phosphorus triiodide (Aldrich,
9%) in carbon disulfide (Aldrich, 99.9%).1 The neopentyl-γ-d
was made via the reaction of neopentyl-γ-d alcohol with triphenyl
2
temperatures. Increasing the initial coverage of neopentyl iodide
on the surface leads to a shift of the hydrogen desorption to
higher temperatures, and to a split into three features (Figure
8
9
9
iodide
was in
9
1
9
phosphite (Aldrich, 97%) and methyl iodide. The alcohol-d
9
1
a); after a 6.0 L dose the three peaks are centered at
turn prepared from acetone-d (Aldrich, 99.5%), via pinacol-d12 and
6
2
0
approximately 330, 405, and 490 K, and correspond to roughly
20, 70, and 10% of the total area, respectively.
the corresponding pinacolone; the oxidation of that pinacolone to
pivalic-d acid was accomplished by using a solution of sodium
9
2
1
hydroxide and bromine (Aldrich, 99.99%), and the reduction of the
pivalic acid to the alcohol by adding lithium aluminum hydride (Aldrich,
Isobutene desorption starts after initial exposures of 4.0 L or
more, and the TPD peak, initially centered around 395 K, grows
and shifts slightly (to approximately 400 K) with increasing
coverage (Figure 1b). A small amount of neopentane desorption
does occur at low coverages, as shown by the broad double
peak with maxima around 150 and 180 K seen in the trace for
the 2.0 L exposure (Figure 1c), and that signal grows, sharpens,
and shifts to lower temperatures (140 K) by 4.0 L. Even higher
neopentyl iodide exposures result in the appearance of three
features in the neopentane TPD, one sharp one that goes from
1
7
9
5%) to the solution. The purity of all compounds was checked
initially by NMR and periodically in situ in our vacuum chamber by
mass spectrometry. Surface exposures are reported in langmuirs (1 L
-
6
)
1 × 10 Torr‚s), not corrected for differences in ion gauge
sensitivities.
3. Results
The thermal chemistry of neopentyl iodide on Ni(100) was
first studied by temperature-programmed desorption (TPD). The
identification of the desorbing products was achieved by
following the evolution of a large number of molecular
fragments during several TPD experiments, including the 2-4,
1
35 to 110 K as the dose is increased from 6.0 to 15.0 L, a
broader peak centered around 175 K that grows later in intensity,
and a third small feature around 380 K that appears as a bump
in the high-temperature tail of the second peak. The low
temperature peak may perhaps be associated either with the
16-20, 28, 39-44, 55-60, 70-73, 112, 127, 142, 198, and
200 amu’s. Only four major products were seen to desorb from
ejection of neopentyl radicals from neopentyl iodide decomposi-
this system, namely, hydrogen, isobutene, neopentane, and
neopentyl iodide. A small amount (less than 0.05 ML) of
methane was occasionally detected around 200 K, but this was
not reproducible, and most likely originated from either impuri-
ties or surface defects.
22,23
tion
or with a direct reaction between neopentyl iodide and
11
chemisorbed hydrogen, but the other features are most likely
associated with the reductive elimination of neopentyl surface
moieties with hydrogen (see Discussion). Finally, molecular
desorption is seen only after monolayer saturation, after an
exposure slightly below 10.0 L, and appears as a single sharp
peak about 178 K.
The TPD traces for hydrogen (2 amu, panel a), isobutene
(56 amu after deconvolution of the low-temperature contribution
from neopentane, panel b), neopentane (57 amu, panel c), and
neopentyl iodide (198 amu, panel d) are shown as a function
of initial neopentyl iodide exposure in Figure 1. Hydrogen,
which is the only desorbing product at low coverages (after
exposures below 2.0 L), appears as one main peak around 360
K and a high temperature tail about 450 K. The temperature
of the main feature is approximately the same as for hydrogen
desorption from H2 adsorbed on clean Ni(100),15 suggesting that
the neopentyl iodide molecules decompose completely at low
The product yields calculated from the areas of the TPD traces
shown in Figure 1 are displayed as a function of initial neopentyl
iodide dose in Figure 2. The yield for molecular hydrogen
desorption is highest at low coverages, but never amounts to
more than 0.25 ML (where 1 ML is again defined as one
adsorbate moiety per nickel surface atom). Mass balance
arguments indicate that this desorption corresponds to the total
decomposition of approximately 0.04 ML of neopentyl iodide
for the 2.0 and 4.0 L cases, and to only 0.024 and 0.008 ML
for the 6.0 and 10.0 L exposures, respectively. The nonzero
value (about 0.04 ML) shown for hydrogen after a 0.0 L
neopentyl iodide dose was obtained by following the exact same
protocol as in the other measurements (except for the neopentyl
(
15) Christmann, K.; Schober, O.; Ertl, G.; Neumann, M. J. Chem. Phys.
1
974, 60, 4528.
(
(
(
(
(
(
16) Zhu, X.-Y.; White, J. M. J. Phys. Chem. 1988, 92, 3970.
17) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc. 1947, 69, 2548.
18) Hudson, H. R. J. Chem. Soc. 1968, 664.
19) Rydon, H. N. Org. Syn. 1971, 51, 44.
20) Adams, R.; Adams, E. W. Org. Syn. 1941, 1, 459.
21) Puntambeker, S. V.; Zoellner, E. A. Org. Syn. 1941, 1, 524.
(22) Lin, J.-L.; Bent, B. E. J. Am. Chem. Soc. 1993, 115, 2849.
(23) Zaera, F.; Tjandra, S. J. Phys. Chem. 1994, 98, 3044.

12740 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Zaera and Tjandra
Figure 4. Left: C 1s XPS from 6.0 L of neopentyl iodide adsorbed
on Ni(100) as a function of annealing temperature. Right: Difference
spectra calculated from the raw data shown in the left panel to highlight
the changes that occur on the surface upon heating. Inset: Calculated
carbon surface coverages as a function of annealing temperature. The
peak shifts seen below 160 K are associated with the conversion of
neopentyl iodide to neopentyl surface species, while the reduction in
carbon surface concentration reflects the depletion of the surface species
due to the desorption of neopentane.
Figure 2. Desorption yields from TPD experiments with neopentyl
iodide adsorbed on Ni(100) surfaces as a function of initial exposure.
the iodide at 90 K and annealing to 180 K to desorb any
molecular neopentyl iodide in order to calibrate the surface
coverages. It was estimated that saturation is reached after
exposures of about 7.0 L, and corresponds to approximately
1
0
0
.20 ML of iodine, in reasonable agreement with the results
obtained by TPD.
The corresponding C 1s XPS annealing data are shown in
Figure 4. The 6.0 L of molecular neopentyl iodide initially
adsorbed at 90 K yields a peak around 284.9 eV and a possible
small shoulder in the low binding energy side. Annealing of
that sample at different increasing temperatures induces several
changes, as better illustrated by the difference spectra displayed
in the right panel of Figure 4. Heating to temperatures as low
as 100 K results in a shift in intensity from the main original
peak to the shoulder around 283.2 eV. No significant changes
are then observed up to 120 K, but at 140 K the peak intensity
is reduced by about 15%, and further intensity shifting from
the high to the low binding energy features occurs at 160 K.
Between 160 and 180 K there is a significant loss in peak
intensity, about another 25% of the initial signal (see inset in
the right panel), and by 220 K only one small peak below 284.0
eV is left in the spectra. The intensity loss in going from 90 to
Figure 3. I 3d X-ray photoelectron spectra (XPS) from 6.0 L of
neopentyl iodide adsorbed on Ni(100) as a function of annealing
temperature. The shift in the I 3d5/2 and 3d3/2 peaks around 150 K is
associated with the dissociation of the adsorbed iodide molecules into
neopentyl surface species and atomic iodine.
2
20 K is most likely associated with the desorption of
dosing), and represents an upper limit for the amount of
adsorption from the background that takes place in these
experiments (see also the isotope labeling experiments discussed
later). Neopentane and isobutene desorption start above about
neopentane, and amounts to a total of about 50% of the initial
signal, in good agreement with the TPD data (which gave a
value of 45 ( 5%).
The mechanism for neopentane formation was probed further
by TPD experiments with coadsorbed deuterium. The left panel
of Figure 5 displays typical TPD raw data for the case of a
1.0 and 3.0 L, respectively, and their yields are approximately
equal and grow approximately linearly with exposure afterwards,
reaching values of up to about 45% of the total neopentyl iodide
conversion each at high coverages.
Ni(100) surface sequentially dosed with 2.0 L of D and 6.0 L
of normal (nondeuterated) neopentyl iodide at 90 K. Two major
2
The TPD data presented above were complemented with both
iodine and carbon XPS studies. Figure 3 shows the I 3d XPS
traces obtained after adsorbing 6.0 L of neopentyl iodide on
Ni(100) at 90 K and then annealing to the indicated tempera-
tures. Two peaks are initially seen around 620.2 and 631.7 eV,
corresponding to the 3d5/2 and 3d3/2 photoelectron signals from
iodine in neopentyl iodide, respectively.8 Those features do
not change in shape or area as the surface is heated, but do
shift by about 0.4 eV toward lower binding energies around
species desorb in this case, namely, hydrogen (of all isotopic
compositions) and neopentane. Hydrogen desorption is identi-
fied by the traces for 2, 3, and 4 amu, which correspond to H2,
HD, and D2, respectively; the preadsorbed deuterium desorbs
mainly around 265 K, although high temperature shoulders can
be seen at 320 and 340 in the D2 and HD traces, respectively,
while normal hydrogen from neopentyl iodide decomposition
evolves predominantly about 385 K. With respect to the
neopentanes, their production is manifested by the large peaks
at 145 and 235 K in the remaining traces (41 and 56-59 amu).
In order to analyze these data further, it is necessary to point
,14
1
50 K, a transition associated with the scission of the C-I
8,14
bond.
Additional spectra were taken after dosing 15.0 L of

Thermal Chemistry of Neopentyl Iodide
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12741
Figure 6. Hydrogen (H
normal ((CH CCH I, a), γ-d
CCD I, c) neopentyl iodides adsorbed on clean Ni(100) at 90 K. The
low-temperature dehydrogenation of neopentyl surface groups involves
only one hydrogen atom from the R position, as indicated by the 330
K peak seen for deuterium in panel c (and the absence of any D signal
2
2
, HD, and D
2
) TPD spectra from 6.0 L of
3
)
3
2
9
3
((CD )
3
CCH I, b), and R-d ((CH
2
2
3 3
) -
Figure 5. Left: 2, 3, 4, 59, 56, 41, 57, and 58 amu TPD raw traces
from 6.0 L of neopentyl iodide adsorbed on a Ni(100) surface predosed
with 2.0 L of deuterium. Right: Corresponding total H, total D,
2
neopentane-d
0 1
, neopentane-d and isobutene calculated TPD data. The
formation of isobutene is inhibited almost completely by the coadsorbed
deuterium, and no neopentane isotopomers other than d and d are
0 1
produced in this system either.
below 400 K in panel b), and presumably produces neopentylidene
moieties on the nickel substrate.
D2 was explored next (data not shown). After exposures above
out that the cracking pattern of neopentane in the mass
spectrometer shows almost no signal for the molecular peak;
the main contribution to the mass spectra of those species comes
1
.0 L of the iodide, the total yields for both hydrogen and
deuterium remain approximately constant (at about 0.12 and
0.30 ML, respectively), but the peak shapes change significantly
upon increasing hydrocarbon coverage, with the bulk of the
signal intensity shifting to the low-temperature features. Since
this effect is more pronounced in the case of deuterium, which
is preadsorbed on the surface and therefore does not originate
from hydrocarbon decomposition, it must be related to changes
in adsorption energy with coadsorption, as reported previously
24
from the formation of a tert-butyl cation in the ionizing region.
Therefore, while the spectrum of normal neopentane displays
one prominent peak at 57 amu (and a small one at 58 amu 0.044
times the size of that at 57 amu due to the normal abundance
1
3
of C in the hydrocarbon molecules), that of neopentane-d1
has two peaks at 57 and 58 amu with 1:3 relative intensities (if
isotope effects are neglected), and neopentane-d2 has peaks at
1
1
for other systems. In terms of the formation of neopentane,
the main effect induced by the coadsorbed deuterium is to
increase its yield at the expense of neopentyl decomposition.
This is particularly noticeable for the low-temperature peak,
which is most likely associated with reactions involving the
direct conversion of the neopentyl iodide to neopentane2 In
addition, increasing coverages of neopentyl iodide on the
deuterium-covered nickel surface lead to a broadening of the
neopentane TPD peaks toward higher temperatures, from below
5
7, 58, and 59 amu with relative intensities going from 1:0:3
for pure 2,2-dimethylpropane-1,1-d2 to 0:2:2 for pure 2,2-
dimethylpropane-1,3-d2. With this in mind, we can proceed
with the interpretation of the TPD in Figure 5. First of all, the
low-temperature peaks seen in the traces for 56 and 59 amu
3,25
can be accounted for by the cracking and 13C contributions from
neopentane-d0 and neopentane-d1, so no neopentane-d2 forms
in this system. This means that no H-D scrambling occurs on
the surface, and that only normal and monodeuterated neopen-
tanes are detected, from which the production of the d1
isotopomer accounts for about 80% of the total. Also, results
from other isotope-labeling experiments (not shown here)
indicate that the d0 production is due to incorporation of normal
hydrogen from background adsorption into neopentyl surface
groups. It is interesting to point to two more observations
related to neopentane desorption: (1) the relative high-to-low
temperature yield ratio (between the 235 and 145 K contribu-
tions to the TPD traces) is higher for neopentane-d1 than for
neopentane-d0 (that is, more monodeuterated neopentane is
produced in relative terms at high temperatures); and (2) a small
amount of neopentane-d0 (but no neopentane-d1) is produced
around 385 K, as in the case of normal neopentyl iodide on
clean Ni(100). Finally, a small amount of isobutene is also
produced even in the presence of surface deuterium, note the
peaks about 400 K in the traces for 41 and 56 amu, but this
2
00 K after a 0.5 L dose to around 235 K after 6.0 L. The
relative production of neopentane-d1 over neopentane-d0 is more
important in this latter feature (as compared to that at the low
temperature), either because of kinetic isotope effects or because
of a difference in reaction mechanisms (reactions with neopentyl
iodide vs neopentyl surface species). Interestingly enough, the
small desorption peak around 380 K is only seen in the
neopentane-d0 traces, most likely because the hydrogen atom
required for neopentyl hydrogenation at those high temperatures
originates from the decomposition of other surface moieties,
not from hydrogen (deuterium) preadsorption; note that this
reaction occurs at the onset of the formation of isobutene.
The results from TPD experiments with partially-deuterated
neopentyl iodides are shown in Figures 6, 7, and 9; the
desorption traces for hydrogen (in all its isotopomers, H2, HD
and D2) are shown in Figure 6, and those for neopentane and
isobutene in Figures 7 and 9, respectively. Three different
isotopically substituted neopentyl iodides were used here,
namely, the normal, nondeuterated compound ((CH3)3CCH2I),
and the R-d2 ((CH3)3CCD2I) and γ-d9 ((CD3)3CCH2I) deuterium-
substituted isotopomers. Regarding the desorption of hydrogen,
the most interesting observation from these experiments comes
from comparing the low-temperature (330 K) peaks in the traces
(normal) isobutene amounts to less than 2% of the total initial
neopentyl iodide adsorbed on the surface, and is only observed
after alkyl iodide exposures above 5.0 L. The processed data
are shown in the right panel of Figure 5.
The evolution of the TPD traces as the initial neopentyl iodide
exposure is increased in the experiments with 2.0 L of predosed
(24) Registry of Mass Spectral Data; Stenhagen, E., Abrahamson, S.,
McLafferty, F., Eds.; Wiley-Interscience: New York, 1974.
(25) Tjandra, S.; Zaera, F. J. Am. Chem. Soc. 1992, 114, 10645.

12742 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Zaera and Tjandra
scission of any of the γ-C-H bonds must require temperatures
much higher than those leading to neopentane desorption. In
contrast, a significant amount of neopentane-d3 is produced in
the case of neopentyl-R-d2 iodide, by incorporation of a
deuterium atom which can only come from R-deuterium
elimination from surface neopentyl groups. That deuterium is
the one that desorbs in the low-temperature hydrogen TPD
reported above.
The first dehydrogenation step mentioned in the last paragraph
appears to be irreversible. This hypothesis was tested by
performing experiments where a clean Ni(100) surface was first
dosed with neopentyl iodide at 90 K, then annealed at 250 K,
and finally exposed to D2 again at 90 K. The idea here was to
try to induce the deuteration of the surface species that forms
after thermal activation of the neopentyl iodide past the
temperature of the first dehydrogenation step. The results from
three of those TPD runs are shown in Figure 8. Both the
neopentyl iodide and the deuterium exposures used were varied
in an attempt to balance two opposite trends, namely, the
difficulty in hydrogenating neopentyl groups at low coverages,
and the blocking of sites for deuterium adsorption after high
neopentyl iodide exposures. Some deuterium adsorption does
take place even after a 3.0 L dose of neopentyl iodide on the
surface, as indicated by the observable signal in the D2 and
Figure 7. Neopentane TPD spectra from 6.0 L of normal ((CH
CCH I, a), R-d ((CH CCD I, b), and γ-d ((CD CCH I, c) neopentyl
iodides adsorbed on clean Ni(100) at 90 K. The significant production
of neopentane-d iodide seen in panel b implies
3 3
) -
2
2
3
)
3
2
9
)
3 3
2
3
from neopentyl-R-d
2
the incorporation of deuterium atoms into the neopentyl surface groups
at temperatures as low as 150 K, and therefore points to an early R-H
(R-D) elimination step from those moieties.
(more noticeable) HD TPD traces in the right panel of Figure
8
, but that deuterium uptake is much lower than when the
adsorption order is reversed (compare the corresponding traces
in Figures 5 and 8), so lower alkyl iodide exposures were used
as well (Figure 8, left panels). The main observation from these
experiments is that, once heated to 250 K, the neopentyl groups
dehydrogenate irreversibly on the surface, because no deuterated
neopentane can be produced afterwards; either 2.0 or 3.0 L doses
of neopentyl iodide are sufficient to induce significant hydro-
genation if the adsorption of both the alkyl iodide and the
deuterium is done at low temperatures (data not shown). Also
noteworthy is the fact that some normal (nondeuterated)
neopentane is indeed made around 280 K, most likely by
hydrogenation of the few neopentyl moieties that may remain
undissociated at that temperature; when the annealing temper-
ature before deuterium dosing was increased to 280 K, no
neopentane formation was observed at all (results not shown).
Finally, a small amount of isobutene desorbs around 400 K,
especially in the 3.0 L neopentyl case, another indication of
the irreversible dehydrogenation that leads to surface decom-
position.
2 2 0 1
Figure 8. H , HD, D , isobutene, neopentane-d , and neopentane-d
TPD spectra from Ni(100) surfaces first dosed with neopentyl iodide
at 90 K, then annealed at 250 K to activate the C-I bond, and finally
exposed to deuterium at 90 K. Different doses were used in these
experiments: 0.75 L of neopentyl iodide and 2.0 L of deuterium (a),
2
.0 and 2.0 L of the iodide and D
2
, respectively (b), and 3.0 and 3.0 L
(c). The absence of any significant neopentane desorption in all cases
points to the irreversible nature of the first R-H elimination step from
neopentyl surface species.
for the γ- and R-deuterated compounds shown in the two right
panels of Figure 6; while most of the hydrogen produced at
that temperature comes out as H2 in the first case, it contains a
significant amount of deuterium in the second. In fact, about
half of the total deuterium that originates from neopentyl-R-d2
decomposition does desorb below 390 K, indicating that an
R-hydrogen (deuterium) elimination step from those surface
moieties takes place at temperatures much below those needed
for either isobutene production or total dehydrogenation.
The fact that at least one R-H elimination step occurs at low
temperatures, probably below 150 K, is corroborated by the
neopentane TPD data displayed in Figure 7. The formation of
neopentane from neopentyl iodide requires the incorporation
of an additional hydrogen atom, which originates either from
surface decomposition of some of the adsorbed hydrocarbons
or from adsorption from the background. The fact that some
normal H is added to the neopentyl moieties in all three cases
studied here implies that the latter source of hydrogen can never
be ruled out, but the absence of any significant production of
neopentane-d10 from neopentyl-γ-d9 iodide indicates that no
γ-hydrogens are involved in the hydrogenation process; the
Additional information on the mechanism for isobutene
formation can be extracted from the isobutene TPD data from
the experiments with the isotopically-labeled compounds shown
in Figure 9. Notice in particular the high selectivity for normal
and perdeuterio isobutene formation from the neopentyl-R-d2
and -γ-d9 iodides, respectively. This result points to the
specificity of the carbon-carbon bond-breaking step at the R-â
position, and therefore rules out the â-methyl elimination
26
mechanism proposed in an earlier report, because in that case
d2 and d6 olefins (respectively) should have been produced
instead (see Discussion). There is also a clear shift in the
temperature maxima of the isobutene TPD peak with deuterium
substitution, an effect that is more noticeable in the case of the
neopentyl-R-d2 case. In order to minimize the problems
associated with comparing data from different TPD experiments,
the TPD for both the normal and d2-substituted compounds were
repeated approximately a dozen times in different occasions over
a period of more than a year; the isobutene TPD peaks were
found to reach maxima at 387 ( 8 and 415 ( 15 K (95%
(26) Zaera, F.; Tjandra, S. J. Am. Chem. Soc. 1993, 115, 5851.

Thermal Chemistry of Neopentyl Iodide
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12743
Figure 9. Isobutane TPD spectra from 6.0 L of normal ((CH
3
)
3
CCH
2
I, a), R-d
2
((CH
3
)
3
CCD
2
9 3 3 2
I, b), γ-d ((CD ) CCH I, c), and a 1:1 mixture of
normal and γ-d
R-d and -γ-d iodides, respectively, indicating that the only C-C bond that breaks in this reaction is that between the R and â positions. Significant
kinetic isotope effects upon deuterium substitutions at either R or γ positions are also evidenced by the shifts in the corresponding TPD peaks.
9
neopentyl iodides adsorbed on clean Ni(100) at 90 K. Normal and perdeuterio isobutenes are produced exclusively from neopentyl-
2
9
confidence), respectively. This large difference is clearly outside
the statistical fluctuations of the experiments.
4. Discussion
As already mentioned in the Introduction, alkyl halides (the
iodides in particular) have been used extensively in recent years
as precursors for the preparation of alkyl surface groups on metal
A kinetic isotope effect was also detected between the
neopentyl-d0 (normal) and neopentyl-γ-d9 iodides, but of a
smaller magnitude than in the case of the d2 compound; here
the isobutene peak maximum shifts only from 387 to 406 K.
Fortunately, the products in this case have different isotopic
composition (C4H8 vs C4D8), and can therefore be separated
easily by mass spectroscopic methods. In order to eliminate
any errors from running two separate experiments, a single TPD
determination was made for the kinetics of formation of both
isobutene products by using a 1:1 (CH3)3CCH2I + (CD3)3CCH2I
mixture, the result of which is shown in the last panel of Figure
1
surfaces. The weak C-I bond in those compounds can be
easily activated, and breaks below 200 K on Ni(100) and other
10,14,28-30
surfaces.
The same low-temperature C-I bond scission
was seen here for the case of neopentyl iodide on Ni(100): the
XPS spectra shown in Figure 3 display the shifts in I 3d binding
energies (from 620.2 to 619.8 eV when heating above 150 K
in the case of the I 3d5/2 peak) typically associated with the
detachment of the iodine atom from the molecule into the
14
surface, and the data presented in Figure 4 indicate that the C
9. Only two isotopomers of the olefin were detected, the normal
1s XPS signal changes in that temperature range as well.
Activation of the C-I bond in adsorbed neopentyl iodide most
likely leads to the formation of neopentyl groups on the surface.
Spectroscopic evidence has been provided in the past for the
and perdeuterio isobutenes, and those peaked at significantly
different temperatures, 378 and 391 K, respectively. This clearly
demonstrates that deuteration at the γ position also induces a
change in the kinetics for the production of isobutene.
1,9,29,31-34
isolation of other alkyl groups on similar systems,
and,
Finally, the possibility of an early C-C bond-breaking step,
at temperatures below those where isobutene desorption is
observed, was also probed. The TPD from 3.0 L of isobutene
adsorbed on clean Ni(100) (not shown) indicate a high degree
of decomposition, as seen by a large signal for H2 desorption
between 250 and about 500 K; only about 5% of the original
olefin desorbs molecularly, in a sharp peak at 138 K and a
broader and smaller feature around 230 K. Preadsorption of
methyl iodide on the surface does not significantly alter the
chemistry of adsorbed isobutene, but the TPD in that case
displays an increase in molecular isobutene desorption (still at
although no direct evidence for the formation of chemisorbed
neopentyl groups was acquired in our case, some of the TPD
results reported above support the hypothesis that such groups
do form on the Ni(100) surface as well. In particular, alkyl
groups have been shown to undergo a facile reductive elimina-
tion step with coadsorbed hydrogen (or deuterium) to yield the
1
1-13,35,36
corresponding alkane,
and that was observed here:
almost half of the initial neopentyl iodide molecules adsorbed
on clean Ni(100) produce neopentane at high coverages, and
almost quantitative conversion via that reaction is obtained on
hydrogen (or deuterium) predosed surfaces. Furthermore, the
reaction of neopentyl iodide with surface deuterium produces
almost exclusively neopentane-d1, strongly suggesting that it is
a surface neopentyl group that is being reduced in this process.
It should be noticed that some of the neopentane detected in
the TPD experiments appears at quite low temperatures, below
1
38 and 230 K) and the formation of a small amount of methane
1
1
about 220 K (from hydrogenation of methyl groups ). One
last experiment was designed to better simulate the surface that
may result from neopentyl thermal activation in which the
preadsorbed methyl iodide was annealed to 160 K before
7
,27
adsorption of the isobutene in order to break the C-I bond.
(
28) Tjandra, S.; Zaera, F. J. Vac. Sci. Technol. 1992, A10, 404.
The results in this case were quite similar to those where the
adsorption was all carried out at low temperature, only that a
little less methane was produced. No high-temperature isobutene
desorption was detected in any of these cases, and extensive
dehydrogenation was always observed below 400 K (over 95%
of the hydrogen produced desorbs below that temperature).
(29) Lin, J.-L.; Bent, B. E.; J. Phys. Chem. 1992, 96, 8529.
(
30) Zaera, F.; Hoffmann, H. J. Phys. Chem. 1991, 95, 6297.
(31) Jenks, C. J.; Bent, B. E.; Bernstein, N.; Zaera, F. J. Am. Chem.
Soc. 1993, 115, 308.
(32) Zaera, F.; Bernstein, N. J. Am. Chem. Soc. 1994, 116, 4881.
(
(
33) Zhou, X.-L.; White, J. M. Surf. Sci. 1988, 194, 438.
34) Lloyd, K. G.; Roop, B.; Campion, A.; White, J. M. Surf. Sci. 1989,
214, 227.
(
35) Zaera, F. J. Phys. Chem. 1990, 94, 8350.
(27) Tjandra, S.; Zaera, F. Langmuir 1993, 9, 880.
(36) Zaera, F. Surf. Sci. 1992, 262, 335.

1
1
2744 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Zaera and Tjandra
40 K, and could therefore originate from a reaction involving
or dehydrogenate to surface carbon and hydrogen (as indicated
by the TPD experiments with coadsorbed methyl iodide and
isobutene mentioned in the Results). All this leads to the
conclusion that the production of isobutene seen in the TPD
data from neopentyl iodide on Ni(100) occurs via a concerted
rearrangement on neopentylidene moieties.
23
the original iodide. Nevertheless, the peak about 180 K, which
accounts for more than half of the total neopentane, is clearly
associated with neopentyl surface species.
The interesting and unique observation from the studies
reported in this paper relates to the subsequent thermal
decomposition of the neopentyl surface moieties on the Ni(100)
surface. Most alkyl groups dehydrogenate readily via a â-hy-
dride elimination step to yield the corresponding olefin, but
since neopentyl groups have no hydrogens at the â position,
they can not follow such a pathway. The only hydrocarbon
product other than neopentane detected in the TPD experiments
with neopentyl iodide on Ni(100) is isobutene, which desorbs
A few additional points can be made concerning the conver-
sion of neopentylidene to isobutene based on the results
presented here. First of all, the isobutene TPD experiments with
R- and γ-labeled compounds shown in Figure 9 illustrate the
high regioselectivity of the C-C bond-breaking step: normal
(d ) and perdeuterio (d ) isobutenes are produced exclusively
1
,3
0
9
from neopentyl-R-d and neopentyl-γ-d iodides, respectively,
2
9
2
6
around 400 K. The rest of this discussion will deal with the
mechanism that leads to the formation of that olefin.
indicating that the bond that breaks is that between the R and
â carbons. In addition, deuterium substitutions at either the R
or γ positions induce noticeable normal kinetic isotope effects.
The magnitude of these rate changes upon deuteration can be
estimated by comparing reaction rates at a fixed temperature in
the TPD data, and comes out to be on the order of factors of
5-7 and 2-3 for the R-d2 and γ-d9 cases, respectively.
Alternatively, the activation energy for isobutene formation,
The first point that needs to be made is that neopentyl surface
species dehydrogenate at relatively low temperatures, perhaps
below 150 K, via an R-hydride elimination step. Both the
regioselectivity and the stoichiometry of this step were proven
by the hydrogen TPD data from partially-labeled neopentyl
iodides shown in Figure 6; the traces from experiments with
40
(CD3)3CCH2I show practically no deuterium desorption below
estimated by using either Redhead’s equation or a leading edge
41
3
80 K, indicating that no γ-C-H (C-D) bonds break before
analysis, is about 18 ( 2 kcal/mol for normal neopentyl iodide
10
-1
that temperature, and the intensity of the deuterium desorption
signal from (CH3)3CCD2I decomposition is split approximately
evenly between the two peaks below and above 400 K, which
means that one of the two deuterium atoms in that molecule is
released early on during the heating of the sample. The point
at which this R-H elimination takes place can not be established
accurately with the TPD data from this study, but several pieces
of information do point to a low temperature for this process.
In particular, the TPD results from (CH3)3CCD2I displayed in
Figure 7 indicate a high yield for the formation of neopentane-
d3; that molecule is produced via the reductive elimination of
the neopentyl-R-d2 groups with surface deuterium, and the only
possible source for the latter is the R-H (R-D) elimination step
mentioned above. Since neopentane production via hydrogena-
tion of surface neopentyl groups starts around 150 K, that must
be an upper limit for the temperature of the dehydrogenation
reaction.
(the preexponential factor being approximately 10 s ), and
approximately 1.5 and 1.0 kcal/mol higher for the d and d9
2
compounds, respectively. These changes in energy barrier with
deuterium substitution are typical of primary isotope effects and
can not be easily accounted by changes in the mass of the
leaving group (which would lead to a maximum change in rate
of about 5% for the γ-substituted system). In conclusion, the
data presented here strongly suggest that the formation of
isobutene from neopentylidene on Ni(100) surfaces involves a
concerted step where all C-C and R- and γ-C-H bond scissions
occur simultaneously (or at least are all involved in the rate-
limiting step). Alternatively, it is possible (although less likely)
for neopentylidene to undergo two slow and competitive R- and
γ-H elimination steps in parallel. The resulting neopentylidyne
and R,R,γ-tricoordinated metallacycle intermediates would then
have to produce isobutene via a series of fast decomposition
reactions.
R-Hydride elimination from neopentyl moieties presumably
produces surface neopentylidene ((CH3)3CCHdNix) species. It
appears as if such intermediates, once formed on Ni(100), cannot
be easily hydrogenated back to either neopentyl or neopentane,
and are stable up to close to 400 K. All attempts to either
hydrogenate or deuterate this species by post-dosing deuterium
after neopentyl iodide activation failed (see Results). In
addition, no other reactions were seen to take place on these
surfaces before the production of isobutene. That no more
dehydrogenation steps occur below 400 K is proven by the fact
that the only hydrogen desorption seen below that temperature
in the TPD is that associated with the R-H elimination reaction;
all other hydrogen atoms desorb above 400 K, implying that
their detection is limited by the surface reaction(s) that produces
them. With respect to C-C bond-breaking steps, they must
also take place at high temperatures, because (1) carbon-carbon
Even though it is difficult to envision how the concerted
conversion of neopentylidene to isobutene may take place on
this Ni(100) surface (a proposal for this is shown schematically
in Scheme 1), several of the mechanisms seen in other systems
can be ruled out here. First of all, the hypothesis of a â-methyl
26
elimination pathway postulated in our earlier studies, which
42
was based on analogies with organometallic systems, can now
be discarded because of the results shown in Figure 9, which
unequivocally prove the high selectivity for the cleavage of the
C-C bond at the R-â position (no loss of methyl groups is
involved in this step). The most common reaction followed
by alkyl groups with no â hydrogens on late transition metals
43
is, in fact, a γ-hydride elimination, either involving a metal-
44
hydride intermediate, or via a direct abstraction by electrophilic
attack of another metal-carbon bond in a four-center transition
45
state; the resulting metallacyclobutane can then in principle
bond-scission steps usually display high activation energies,
undergo a ring opening to the observed olefin (and a surface
above 15 kcal/mol;3
,37-39
(2) if tert-butyl groups were to be
made on the surface, they would undergo â-hydride elimination
and produce isobutene below 180 K;10 and (3) if isobutene were
to be produced at any point above 140 K during the TPD
experiments, it would either immediately desorb molecularly
(
40) Redhead, P. A. Vacuum 1962, 12, 203.
(41) Habenschaden, E.; K u¨ ppers, J. Surf. Sci. 1984, 138, L147.
(
(
42) Watson, P. L.; Parshall, G. W. Acc. Chem. Res. 1985, 18, 51.
43) Collman, J. P.; Hegedus, L. S.; Norton, J. R. Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, CA, 1987.
(
(
(
37) Goodman, D. W. Surf. Sci. 1982, 123, L679.
38) Zaera, F.; Somorjai, G. A. J. Phys. Chem. 1985, 89, 3211.
39) Rodriguez, J. A.; Goodman, D. W. Surf. Sci. Rep. 1991, 14, 1.
(44) Foley, P.; DiCosimo, R. Whitesides, G. M. J. Am. Chem. Soc. 1980,
102, 6713.
(45) Fendrick, C. M.; Marks, T. J. J. Am. Chem. Soc. 1986, 108, 425.

Thermal Chemistry of Neopentyl Iodide
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12745
Scheme 1
carbene), as in metathesis processes.46,47 Both the formation
of a 3,3-dimethyl nickelacyclobutane from the corresponding
dineopentyl-metal complex and its subsequent decomposition
to isobutene have in fact been reported recently.4 There are
nevertheless several reasons why this mechanism is not con-
sistent with our observations for the neopentyl/Ni(100) sys-
tem: (1) no γ-hydrogens are activated below 400 K, which
means that if a metallacycle were to form on the nickel surface,
it would have to be produced at that high temperature, and would
have to decompose immediately afterwards; (2) because an early
R-H elimination is seen in the hydrogen TPD of deuterium-
labeled compounds, an alkylidene, not alkyl, species is believed
to be the starting moiety in the high-temperature conversion of
neopentyl to isobutene; (3) if isobutene were to be produced
via the ring opening of a 3,3-dimethyl nickelacyclobutane
intermediate, the probabilities for R-â and â-γ C-C bond
scissions would be approximately equal, an expectation incon-
sistent with the high selectivity seen here toward the activation
of the CR-Câ bond; and (4) a γ-hydride elimination rate-limiting
step can not account for the large kinetic isotope effect upon
deuterium substitution at the R position of the neopentyl moiety
reported above.
nation to µ-vinylidene has been observed instead,57 but such a
reaction does not lead to the R-â C-C bond scission reported
here, and is not even available on neopentyl groups (which do
not have â hydrogens). Two more reactions are available to
8
5
8
coordinated carbenes, namely, direct dimerization and 1,2-
5
9
shift to the corresponding alkenes, but both those steps are
more common with Fischer-type carbenes, and do not explain
the isobutene formation seen here either.
Given the importance of alkyls as intermediates in the
catalytic conversion of alkanes (in reactions such as hydrogena-
tion-dehydrogenation, re-forming, and partial and total oxida-
tion), extensive work has been done on those systems as
6
0-62
well.
The catalytic H-D exchange of neopentane on nickel
films has been shown to be limited mostly to one methyl group,
and to yield neopentane-d and -d as the major products.
63
1
3
A
mechanism was proposed for this reaction where the initial
surface neopentyl intermediate that forms after neopentane
activation follows one of two competing pathways, a reductive
elimination with surface deuterium to neopentane-d , or a fast
1
interconversion with an R,R-diadsorbed species (neopentylidene)
followed by a slower hydrogenation to neopentane-d . These
3
ideas are entirely consistent with the results of our work.
A related mechanism to the one discussed in the previous
The catalytic H-D exchange on neopentane is usually
accompanied by slower hydrogenolysis and isomerization
processes. Three basic mechanisms have been proposed to
explain these reactions: (1) a direct isomerization via an R,R,γ-
paragraph involves an initial direct carbene extrusion from the
original neopentyl moiety4
3,49
and a subsequent â-hydride
elimination from the resulting tert-butyl surface species to
isobutene. The preferential C-C bond cleavage at the R
position via this type of mechanism was proven for the case of
6
4
tricoordinated bridged species; (2) the intermediate isomer-
ization of that tricoordinated metallacycle to a monocoordinated
5
0
65
a nickelacyclohexane complex by using deuterium labeling.
This idea, however, can be discarded in our case as well, because
1) again, the starting reactant on the nickel surface is neo-
cyclopentyl species; and (3) a methyl transfer via a cyclo-
pentane-like intermediate, the same as in the bond-shift mech-
6
6,67
(
anism in carbonium ions.
It is important to point out,
pentylidene, not neopentyl; and (2) since the â-hydride elimina-
tion that must follow the extrusion is fast, this mechanism would
not induce any kinetic isotope effects upon substitution at the
γ position.
however, that these ideas have been put forward mainly to
explain isomerization reactions (from neopentane to 2-meth-
ylbutane in this case), and such processes are seen almost
exclusively on platinum-based catalysts. No isomerization is
observed in the case of nickel, only the production of isobutane
(as well as smaller amounts of methane, ethane and propane).63
The bottom line is that none of the mechanisms enumerated
above is supported by the evidence obtained from our experi-
ments. Finally, it is interesting to note that on Pt(111) surfaces
The alternative initial R-hydride elimination reported here has
indeed been seen in organometallic compounds, especially of
early transition metals.5
1-53
However, the resulting alkylidenes
most often dehydrogenate further to alkylidynes,5 and those
4
are quite stable, even though they occasionally dimerize to
5
5,56
produce alkynes.
In some clusters alkylidyne dehydroge-
(
56) Kim, H. P.; Angelici, R. J. AdV. Organometal. Chem. 1987, 27, 51.
(
46) Lee, J. B.; Ott, K. C.; Grubbs, R. H. J. Am. Chem. Soc. 1982, 104,
491.
47) Dragutan, V.; Balaban, A. T.; Dimonie, M. Olefin Metathesis and
Ring Opening Polymerization of Cycloolefins; Wiley: New York, 1986.
48) Miyashita, A.; Ohyoshi, M.; Shitara, H.; Nohira, H. J. Organometal.
Chem. 1988, 338, 103.
49) Stenstrøm, Y.; Klauck, G.; Koziol, A.; Palenik, G. J.; Jones, W. M.
Organometallics 1986, 5, 2155.
(57) Rashidi, M.; Puddephatt, R. J. Organometallics 1988, 7, 1636.
(58) Merrifield, J. H.; Lin, G.-Y.; Kiel, W. A.; Gladysz, J. A. J. Am.
Chem. Soc. 1983, 105, 5811.
7
(
(59) Fischer, E. O.; Schubert, U.; Fischer, H. Pure Appl. Chem. 1878,
50, 587.
(
(60) Bond, G. C. Catalysis by Metals; Academic Press: London, 1962.
(61) Kemball, C. AdV. Catal. Relat. Subj. 1959, 11, 223.
(62) Zaera, F.; Somorjai, G. A. In Hydrogen Effects in Catalysis:
Fundamentals and Practical Applications; Pa a´ l, Z.; Menon, P. G., Eds.;
Marcel Dekker: New York, 1988; pp 425-447.
(63) Kemball, C.; Kempling, J. C. Proc. R. Soc. London A 1972, 329,
391.
(64) Anderson, J. R.; Avery, N. R.; J. Catal. 1966, 5, 446.
(65) Garin, F.; Gault, F. G. J. Am. Chem. Soc. 1975, 97, 4466.
(66) McKervey, M. A.; Rooney, J. J.; Samman, N. G. J. Catal. 1973,
30, 330.
(
(
(
50) Grubbs, R. H.; Miyashita, A. J. Am. Chem. Soc. 1978, 100, 7418.
51) Cooper, N. J.; Green, M. L. H. J. Chem. Soc, Dalton Trans. 1979,
1
121.
(52) Schrock, R. R. Acc. Chem. Res. 1979, 12, 98.
53) Cross, R. J. in The Chemistry of the Metal-Carbon Bond; Hartley,
(
F. R., Patai, S., Eds.; Wiley: New York, 1985; Vol. 2; pp 559-624.
54) Churchill, M. R.; Wasserman, H. J.; Turner, H. W.; Schrock, R. R.
J. Am. Chem. Soc. 1982, 104, 1710.
55) McDermott, G. A.; Mayr, A. J. Am. Chem. Soc. 1987, 109, 580.
(
(
(67) Karpinski, Z.; Guczi, L. J. Chem. Soc., Chem. Commun. 1977, 563.

12746 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Zaera and Tjandra
it has been shown that R- and γ-hydride elimination steps
display similar rates,68 in contrast with the case of Ni(100),
where selective R-hydride elimination is seen instead.
this surface until reaching temperatures around 400 K, at which
point isobutene is formed. Isotope labeling experiments indicate
that the chemisorbed neopentyl intermediates dehydrogenate
below or about 150 K to neopentylidene, and that this neopen-
tylidene is the one that converts to the olefin. The experimental
data also point to the selective scission of the R-â C-C bond,
and to primary kinetic isotope effects in the neopentylidene
conversion to isobutene with deuterium substitution at either
the R or γ positions. All this suggests a concerted step involving
the CR-Câ bond and two hydrogen atoms, and rules out most
of the decomposition pathways reported in the literature for alkyl
and carbene ligands in organometallic compounds.
5
. Conclusions
The chemistry of neopentyl iodide on Ni(100) surfaces has
been characterized by temperature-programmed desorption and
X-ray photoelectron spectroscopies, and is summarized in
Scheme 1. The same as in the case of most other alkyl iodides,
the first reaction that takes place on this system is the scission
of the C-I bond, presumably to produce neopentyl surface
groups. At high coverages about half of these neopentyl
moieties undergo an early reductive elimination step with surface
hydrogen (or deuterium) to neopentane, which is detected in
the gas phase below or around 180 K. The other half remains
on the surface, and no other hydrocarbon products desorb from
Acknowledgment. Financial support for this research was
provided by a grant from the National Science Foundation
(CHE-9530191).
(68) Janssens, T. V. W.; Jin, G.; Zaera, F. J. Am. Chem. Soc., submitted
for publication.
JA9628056