Cross coupling of phenyl groups with alkyl iodides on copper surfaces: A radical mechanism?

Article abstract of DOI:10.1021/jp9608180

The cross coupling of phenyl groups (C₆H₅) with alkyl iodides in adsorbed monolayers on single-crystal Cu(110) and Cu(100) surfaces under ultrahigh-vacuum conditions has been studied by a combination of temperature-programmed reaction techniques and isotope labeling. In these experiments, phenyl groups were generated on the surface by the dissociative adsorption of iodobenzene. Detailed studies in the case of the CH₃I + phenyl reaction to form toluene show that there are two mechanisms for the process: a low-temperature pathway (<160 K; activation energy <10 kcal/mol), in which phenyl groups react directly with CH₃I, and a much higher temperature pathway (～400 K, activation energy ～27 kcal/mol), in which adsorbed methyl groups produced by C-I bond scission couple with coadsorbed phenyl groups. The mechanism of the low-temperature reaction has yet to be fully resolved, but experimental evidence for a radical mechanism is presented. The relevant observations in this regard include the following: (1) when methyl radicals are impinged onto a phenyl precovered Cu(100) surface at 100 K, toluene is produced by an Eley-Rideal mechanism; (2) independent studies have shown that methyl radicals are produced during CH₃I dissociation on Cu(111); (3) the upper limit of 160 K for the phenyl + CH₃I reaction temperature on Cu(110) is quite close to the temperature for C-I bond scission in CH₃I on this surface; and (4) neopentyl iodide, which is much more sterically hindered than CH₃I, also shows a direct reaction with coadsorbed phenyl groups, suggesting that the process is C-I dissociation followed by coupling as opposed to phenyl attack at the C-I bond.

Full text of DOI:10.1021/jp9608180

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
J. Phys. Chem. 1996, 100, 16621-16628
16621
Cross Coupling of Phenyl Groups with Alkyl Iodides on Copper Surfaces: A Radical
Mechanism?
P. W. Kash,† D.-H. Sun, M. Xi, G. W. Flynn,* and B. E. Bent
Department of Chemistry, Columbia UniVersity, New York, New York 10027
ReceiVed: March 18, 1996; In Final Form: July 11, 1996X
The cross coupling of phenyl groups (C6H5) with alkyl iodides in adsorbed monolayers on single-crystal
Cu(110) and Cu(100) surfaces under ultrahigh-vacuum conditions has been studied by a combination of
temperature-programmed reaction techniques and isotope labeling. In these experiments, phenyl groups were
generated on the surface by the dissociative adsorption of iodobenzene. Detailed studies in the case of the
CH3I + phenyl reaction to form toluene show that there are two mechanisms for the process: a low-temperature
pathway (<160 K; activation energy <10 kcal/mol), in which phenyl groups react directly with CH3I, and a
much higher temperature pathway (∼400 K, activation energy ∼27 kcal/mol), in which adsorbed methyl
groups produced by C-I bond scission couple with coadsorbed phenyl groups. The mechanism of the low-
temperature reaction has yet to be fully resolved, but experimental evidence for a radical mechanism is
presented. The relevant observations in this regard include the following: (1) when methyl radicals are
impinged onto a phenyl precovered Cu(100) surface at 100 K, toluene is produced by an Eley-Rideal
mechanism; (2) independent studies have shown that methyl radicals are produced during CH3I dissociation
on Cu(111); (3) the upper limit of 160 K for the phenyl + CH3I reaction temperature on Cu(110) is quite
close to the temperature for C-I bond scission in CH3I on this surface; and (4) neopentyl iodide, which is
much more sterically hindered than CH3I, also shows a direct reaction with coadsorbed phenyl groups,
suggesting that the process is C-I dissociation followed by coupling as opposed to phenyl attack at the C-I
bond.
I. Introduction
dissociation of methyl iodide on Cu(110) and Cu(100) surfaces,
with coadsorbed phenyl groups to form adsorbed toluene.
Bimolecular reactions at the gas-solid interface have tradi-
tionally been classified as occurring by one of two broad classes
of mechanisms. The Langmuir-Hinshelwood mechanism,
which involves a reaction between two chemisorbed species that
have thermally accommodated to the underlying surface,1 has
been the most prevalent mechanism in previous studies. Several
experimental and theoretical studies have, however, also pro-
vided evidence for reactions occurring by an Eley-Rideal
mechanism,2-21 which involves reaction between gas phase and
surface-bound species prior to thermal accommodation of the
gas phase species with the surface.22,23 Although these two
limiting cases provide a framework for the classification of most
reactions occurring on surfaces, there have been a number of
recent reports of a variant of the Eley-Rideal mechanism in
which the “hot” reactant is produced on the surface instead of
in the gas phase. For example, it has been shown that radicals
generated photochemically at surfaces by irradiation of photo-
active monolayers can react with coadsorbed species prior to
thermal accommodation with the surface.24-26 Experimental
evidence for analogous Eley-Rideal reactions in which “hot”
radicals are produced in the thermal dissociation of adsorbed
molecules is less definitive, although recent results19,27-38 have
suggested that these processes may be more prevalent than
previously supposed. Indeed, it is now clear39,40 that the
exothermic dissociation of adsorbed species on surfaces can lead
to “hot” radical fragments, which might react with coadsorbed
species before bonding with and accommodating to the surface.
The current work provides evidence for such a direct Eley-
Rideal reaction of methyl radicals, produced in the thermal
Eley-Rideal reactions of radicals produced in the thermal
dissociation of an adsorbate are uncommon and relatively
unstudied largely because the thermodynamics of adsorbate bond
dissociation reactions generally prevent the formation of free
radical species. As Figure 1 indicates, most chemical bonds
within stable molecules are relatively strong (approximately 80-
100 kcal/mol) compared with the chemical bonds between
surface metal atoms and adsorbate fragments (which are
generally 20-70 kcal/mol). Since entropy changes for adsorp-
tion are usually negative or at most slightly positive, the
thermodynamic driving force for adsorption is generally the
exothermicity of the reaction. Given that the strengths of bonds
to metal surfaces are generally significantly weaker than the
bonds within molecules (as indicated above), an exothermic
dissociative adsorption reaction generally requires making two
bonds to the surface for each bond broken within the dissocia-
tively adsorbed molecule. This general rule of two bonds made
per one bond broken is the reason that both fragments produced
by dissociative adsorption at metal surfaces remain bound to
the surface. We note, also, that this “normal” dissociative
adsorption process corresponds to the microscopic reverse of a
Langmuir-Hinshelwood process, whereas free radical formation
during dissociative adsorption corresponds to the microscopic
reverse of an Eley-Rideal process. The lack of evidence for
Eley-Rideal reactions of radicals produced in adsorbate thermal
dissociation on metal surfaces results largely from the fact that
free radicals are not generally produced in thermal dissociation.
Although this general rule of the production of two surface-
bound species in the dissociation of an adsorbate on a metal
surface accounts for most previously studied surface reactions,
alkyl iodides provide an exception to this rule. Using methyl
iodide adsorbed on copper as a model system, Figure 2 indicates
† NSF Postdoctoral Fellow. Permanent address: Department of Chem-
istry, University of British Columbia, Vancouver, British Columbia, Canada.
* Author to whom correspondence should be addressed.
X Abstract published in AdVance ACS Abstracts, September 1, 1996.
S0022-3654(96)00818-0 CCC: $12.00 © 1996 American Chemical Society

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
16622 J. Phys. Chem., Vol. 100, No. 41, 1996
Kash et al.
In each of these radical-generating dissociative adsorption
processes, there is a branching between radical ejection to the
gas phase and radical adsorption on the surface. There is also
evidence in several cases for surface reactions of these radicals
prior to either thermal accommodation on the surface or
desorption into the gas phase. Examples include ethylene
formation during CH2I2 dissociative adsorption on aluminum
surfaces35,36 and octane formation during 1-C4H9I dissociation
on Cu(111) (ref 34) as well as toluene (C6H5CH3) formation
during dissociative adsorption of CH3I on a Cu(111) surface
precovered with a partial monolayer of adsorbed phenyl (C6H5)
groups.19,33 Analogous reactions of hot atoms or radicals have
also been inferred for systems where atom or free radical
evolution from the surface is not detected concurrently but where
the surface reaction products and kinetics27-32,37,38 are suggestive
of such pathways.
The present work is concerned with the reactions of methyl
iodide and other alkyl iodides with coadsorbed phenyl groups
(C6H5) on Cu(110) and Cu(100) surfaces. Previous studies of
these reactions on a Cu(111) surface have indicated two
mechanisms for this reaction: a high-temperature pathway at
∼350 K involving coupling of surface-bound methyl and phenyl
groups and a low-temperature (<160 K) pathway involving
some type of direct reaction between phenyl groups and the
alkyl iodides.19,20,33 The detection of gas phase methyl radicals
during CH3I dissociation at ∼140 K on Cu(111) (refs 41, 42)
led to the suggestion that the low-temperature toluene formation
pathway involves a free radical mechanism. In the studies
presented here, it is shown that reactions directly analogous to
those described above for Cu(111) also occur on Cu(110) and
Cu(100) surfaces. It is also demonstrated that both the direct
and indirect mechanisms occur not only when CH3I is the
reactant but also when methyl radicals generated with a gas
phase pyrolysis source are impinged onto copper surfaces
precovered by a partial monolayer of phenyl groups. The
kinetics of the low-temperature reaction for both methyl radicals
and methyl iodide have been investigated using a chemical
displacement technique. The results provide further support for
the hypothesis of Eley-Rideal-type processes involving free
radical or hot radical intermediates in the low-temperature cross
coupling reactions of alkyl iodides and adsorbed phenyl groups
on copper surfaces.
Figure 1. Schematic of the two extreme mechanisms for dissociative
adsorption of molecules on metal surfaces. The “normal” mechanism,
depicted by the products in the upper right, results in the formation of
two chemisorbed species by a thermodynamically favorable reverse
Langmuir-Hinshelwood process. The radical pathway, depicted by the
products in the lower right, results in the formation of one chemisorbed
species and one gas phase radical species by a reverse Eley-Rideal
process, which is generally thermodynamically unfavorable.
II. Experimental Section
The experiments were performed in a UHV system at
Columbia University. Details of the Columbia UHV chamber
and of experimental protocol for TPD studies are described
elsewhere.20 Briefly, the Cu(110) and Cu(100) crystals (Mono-
crystals, 99.999%) were cleaned by cycles of Ar+ sputtering
and annealing. Reactants were adsorbed onto the crystal by
back-filling the chamber. Iodobenzene (Aldrich, 99%), iodo-
benzene-d5 (Icon, 97 atom % D), methyl iodide (Aldrich,
99.5%), and toluene (Aldrich, 99.8%) were purified by several
freeze-pump-thaw cycles with liquid nitrogen prior to dosing,
and sample purities were confirmed by in situ mass spectrom-
etry. All exposures are reported in Langmuirs (1 Langmuir (L)
) 1 × 10-6 Torr‚s) and are uncorrected for differing ion gauge
sensitivities. The quadrupole mass spectrometer (QMS) is
installed behind a differentially pumped shield containing a 2
mm diameter aperture. In TPD studies, the sample was held
1-2 mm from the aperture so that only molecules evolved from
the central portion of the 1 cm diameter crystal contribute to
the detected signal. The heating rate in all TPD experiments
was 3 K/s.
Figure 2. Schematic of the two possible dissociative adsorption
pathways (alkyl radical production and metal alkyl formation) for
methyl iodide on Cu(111). Because of the weak CH3-I bond and the
strong Cu-I bond, alkyl radical production during dissociative adsorp-
tion is thermodynamically possible, as discussed in the text.
that the C-I bond energy of 55 kcal/mol in methyl iodide is
unusually weak for a chemical bond.41 In addition, the 60-70
kcal/mol bond energy of I chemisorbed on Cu is unusually
strong for a chemisorption bond.41 Consequently, it is energeti-
cally feasible for C-I bond cleavage in methyl iodide adsorbed
on Cu to result in chemisorbed I and methyl free radicals. In
fact, previous experiments in our laboratory have detected
methyl radicals desorbing into the gas phase at the same surface
temperature as C-I bond cleavage in methyl iodide adsorbed
on Cu(111).41,42 Atom,43 radical,44 and carbene36,45-48 evolution
during the dissociative adsorption of halogen-containing mol-
ecules on other surfaces has also been reported.
Methyl radicals were generated in these studies by pyrolyzing
azomethane using a source similar to the one developed by Peng

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Cross Coupling of Phenyl Groups with Alkyl Iodides
J. Phys. Chem., Vol. 100, No. 41, 1996 16623
et al.49 Briefly, azomethane (synthesized by D.-H. Sun and A.
Gurevich,50 using the procedure of Renaud and Leitch51) was
passed into the vacuum system through a 3 mm i.d. quartz tube
with a 4 cm zone at the tip of the tube heated resistively to
1000 K. Pyrolysis of azomethane in this heated zone produced
a flux of methyl radicals along with some methane, ethane,
nitrogen, hydrogen, and unreacted azomethane. Previous studies
have shown that the species other than methyl radicals do not
stick to copper surfaces held at room temperature.52 In the
studies here, methyl radicals were adsorbed line-of-sight by
positioning the surface approximately 2 cm in front of the dosing
tube. Exposures are reported in Langmuirs where the dose
pressure corresponds to the pressure rise in the chamber during
pyrolysis. For reference, a 1.0 L exposure in these units
corresponds to 80% saturation of the Cu(100) surface with
methyl groups (as determined by calibration TPD experiments
that monitor the ethene coupling product desorption yield).
III. Results and Interpretation
As mentioned in the Introduction, the results presented below
concern the reaction of phenyl groups with coadsorbed methyl
iodide and with methyl radicals incident on the surface from
the gas phase. In both cases the coupling product is toluene.
The studies of the methyl radical reactions were carried out on
a Cu(100) surface, while the studies of the phenyl + CH3I
reaction were carried out on Cu(110). The reason for studies
on two different surfaces is that the Cu(100) sample was
inadvertantly melted during the course of the research, and a
(110) sample was available. However, aside from small kinetic
differences between the two surface geometries, we find no
structure sensitivity of the reaction mechanisms discussed. In
particular, the phenyl + CH3I reactivity we report here for
Cu(110) is directly analogous to that reported previously for
Cu(111), and we therefore expect Cu(100) to behave similarly
with respect to this chemistry.
Figure 3. Temperature-programmed desorption spectra monitoring
toluene evolution at m/e+ ) 91 on Cu(100). The top trace shows results
following addition of 1.0 L of methyl radicals to a surface, which had
been previously exposed to 2.0 L of iodobenzene and briefly annealed
to 200 K to generate adsorbed phenyl groups. The bottom trace shows
results following sequential adsorption of 1.0 L of methyl radicals and
2.0 L of iodobenzene at 100 K. As discussed in the text, the low
temperature toluene desorption peak at 240 K in the top spectrum results
from a direct Eley-Rideal reaction of gas phase methyl radicals, while
the higher temperature peaks at 360 and 380 K in the two spectra result
from a Langmuir-Hinshelwood reaction of surface-bound methyl and
phenyl groups.
IIIA. Adsorbed Phenyl Groups + Gas Phase Methyl
Radicals on Cu(100). Throughout the studies in this paper,
adsorbed phenyl groups (C6H5) have been generated on the
single-crystal copper surfaces by the dissociative adsorption of
iodobenzene (C6H5I). Extensive studies on Cu(111) surfaces
have documented, by high-resolution electron energy loss
spectroscopy,20 near edge X-ray absorption fine structure
measurements,53 work function change measurements,20 and
titration with hydrogen atoms,20 that annealing a monolayer of
iodobenzene to temperatures above 180 K results in C-I bond
scission to form surface phenyl groups that are stable to
temperatures above 300 K. Similar surface chemistry also
appears to be valid for Cu(100) and Cu(110),54 and in the studies
here, iodobenzene monolayers have been annealed to temper-
atures of 200-300 K to generate adsorbed phenyl groups.
The results in Figure 3 demonstrate that gas phase methyl
radicals (CH3) react with adsorbed phenyl groups (C6H5) on
Cu(100) to form toluene (C6H5-CH3) by both the Langmuir-
Hinshelwood and Eley-Rideal mechanisms. Figure 3A (lower
trace) displays a temperature-programmed reaction (TPR)
spectrum for the Langmuir-Hinshelwood pathway, which is
the sole mechanism observed when a Cu(100) surface is first
exposed to 1.0 L of methyl radicals (80% of monolayer
saturation) followed by 2.0 L of iodobenzene (40% of monolayer
saturation). The resulting TPR spectrum shows toluene evolu-
tion at 360 K as a result of phenyl + methyl coupling by a
Langmuir-Hinshelwood mechanism. Since toluene desorbs
from Cu(100) at temperatures below 350 K (see Figure 4 results
described below), we conclude that toluene evolution at 360 K
and above is due to rate-determining coupling of phenyl +
methyl followed by rapid desorption of the product toluene. (For
reference, methyl + methyl coupling to produce ethane occurs
at 400-450 K on Cu(110).55,56) Virtually identical results are
obtained if the order of addition is reversed and methyl radicals
are adsorbed onto a surface precovered at 100 K with a parital
monolayer of iodobenzene, which remains molecularly intact
at this temperature.
To observe the Eley-Rideal mechanism for phenyl + methyl
coupling on Cu(100), the order of reactant addition is reversed
and methyl radicals are impinged onto a Cu(100) surface
precovered with a partial monolayer of phenyl groups. Selected
results of these studies are shown in Figure 3B. In the results
shown, adsorbed phenyl groups were generated on Cu(100) by
exposing the surface to 2.0 L of iodobenzene at 100 K and then
briefly annealing to 200 K to induce C-I bond dissociation.
Subsequent exposure to “1.0 L” of methyl radicals from the
pyrolysis source saturates the surface with methyl groups. Note
that the toluene produced in a subsequent TPR experiment is
evolved from the Cu(100) surface in two distinct temperature
regimes (240 and 380 K). The higher temperature (380 K) peak
reflects that coupling of adsorbed phenyl and methyl groups
analogous to what is shown in Figure 3A when the order of
reactant addition was reversed. The lower temperature (240
K) peak is observed only when methyl radicals are dosed after
phenyl groups have been generated on the surface. (For
example, this peak is also absent when iodobenzene is pread-
sorbed but the surface has not been annealed to dissociate the
C-I bond and generate surface phenyl groups prior to methyl
radical adsorption.) The origin of the 240 K toluene evolution

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
16624 J. Phys. Chem., Vol. 100, No. 41, 1996
Kash et al.
Figure 5. Schematic chemical displacement experiment in which a
relatively large exposure of iodobenzene-d5 is added to a surface initially
covered with molecular toluene (top) or chemisorbed methyl and phenyl
groups (bottom). As indicated, iodobenzene-d5 displaces the weakly
adsorbed toluene molecules from the monolayer to the second or
multilayer, resulting in toluene desorption at 160 K in a subsequent
TPD experiment. As also indicated, iodobenzene-d5 fails to displace
more strongly chemisorbed methyl and phenyl groups.
basis of the TPD results in Figure 4A, 2.0 L of toluene would
be expected to desorb at 200-350 K in the absence of any effect
of coadsorbed iodobenzene. However, as shown by the toluene
TPD spectrum in Figure 4B, the addition of iodobenzene
displaces 100% of the toluene from the monolayer to the
multilayer, where it is evolved at 160 K, consistent with the
multilayer peak temperature in Figure 4A. This displacement
phenomenon is discussed more fully in other publications;57 the
important point for the studies here is that toluene produced in
the phenyl + methyl reaction can be displaced from the
monolayer by iodobenzene.
Figure 4. Temperature-programmed desorption spectra monitoring
toluene molecular desorption at m/e+ ) 91 on Cu(110) following (A)
adsorption of several different exposures of toluene alone and (B)
sequential adsorption of 2.0 L of toluene and 10.0 L of iodobenzene at
100 K. The spectrum in B indicates that iodobenzene quantitatively
displaces toluene from the monolayer on Cu(110).
peak is not obvious from the single experiment in Figure 3B,
but the results below suggest an Eley-Rideal mechanism.
We note first that the 240 K peak temperature for toluene
evolution in Figure 3B is indicative of toluene molecular
desorption from Cu(100). While we have not obtained TPD
spectra for toluene on Cu(100), we have studied toluene
adsorption on Cu(110), where one would expect a similar, if
not stronger, molecule/surface interaction because of the
increased surface corrugation. As shown by the temperature-
programmed desorption (TPD) results in Figure 4A, toluene
desorbs from monolayers on Cu(110) over a wide temperature
range from 200 to 350 K (multilayers are evolved at temper-
atures below 200 K; for example, note the peak temperature of
160 K in Figure 4). The temperature for the onset of desorption
shifts down with increasing surface coverage, and it is reason-
able that toluene evolution in the phenyl + methyl experiments
of Figure 3B should commence near 200 K since the surface is
saturated with phenyl, methyl, and iodine in these studies. That
toluene evolution maximizes at 240 K and ceases for temper-
atures above 300 K in Figure 3B is also reasonable since the
surface remains covered with adsorbed species (CH3, unreacted
phenyl, and iodine) after all toluene has desorbed.
Confirmation that the 240 K desorption temperature in the
experiment of Figure 3B for Cu(100) is determined by the rate
of toluene desorption and that this toluene is actually formed
at temperatures below 160 K comes from chemical displacement
studies. The basis for these experiments is illustrated schemati-
cally in Figure 5 and demonstrated experimentally as shown
by the spectra presented in Figure 4B. As shown in Figure 5,
the idea is to use iodobenzene-d5 to displace weakly adsorbed
molecules such as toluene while leaving strongly chemisorbed
reaction intermediates such as phenyl and methyl on the surface.
The ability of iodobenzene to displace toluene at very low
temperatures is demonstrated by the TPR results in Figure 4B.
In this experiment 2.0 L of toluene has been adsorbed on Cu-
(110) at 100 K followed by 10.0 L of iodobenzene. On the
Figure 6 presents the results for the application of chemical
displacement to the phenyl + methyl coupling reaction. The
TPR spectrum in Figure 6A is directly analogous to that in
Figure 3B except for the fact that 1.0 L rather 2.0 L of
iodobenzene has been adsorbed and annealed to generate phenyl
groups prior to methyl radical adsorption. (It should also be
noted that the studies in Figures 3 and 6 were performed on a
Cu(100) crystal that had been remounted several times in
between, so peak temperatures are probably only comparable
to within 10-20 K.) The experiment in Figure 6B is the same
as that in Figure 6A except that after phenyl generation and
methyl radical exposure, 10 L of iodobenzene-d5 has been added
prior to desorption to displace any product toluene to the
multilayer. (Fully deuterated iodobenzene was used as the
displacing agent to verify that none of the phenyl groups from
the displacing agent are incorporated in the product toluene.)
As shown by the TPR spectrum in Figure 6B, this displacement
procedure shifts the 221 K toluene peak in Figure 6A to the
160 K multilayer desorption temperature (compare Figure 4B).
[The precise yields and temperatures in Figures 6A,B cannot
be compared because the experiments were performed weeks
apart; it should also be noted that the kinetics of the high-
temperature phenyl + methyl coupling reaction are perturbed
(e.g. note the split peak at ∼320 K in Figure 6B) by the presence
of the coadsorbed iodobenzene displacing agent on the surface.]
The shift of the low-temperature toluene peak from 221 to 160
K upon displacement implies that the 221 K toluene peak
temperature in Figure 3A is indicative of the kinetics of toluene
desorption (from the monolayer) and that the toluene is actually
formed below 160 K. Since this temperature is so much lower
than the 350-400 K temperature range for Langmuir-Hin-
shelwood coupling of phenyl and methyl on Cu(100), we