phile on a co-adsorbed electrophile. The resultant surface
intermediate then undergoes b-H elimination to produce the
coupled product, for example, an amide. The electrophile
can be either externally introduced or, in principle, formed
secondary oxidation product of dimethylamine (methyl iso-
cyanate), was also observed in small amounts, which is con-
sistent with our previous studies.[6b] Further oxidation of di-
methylformamide was not observed, likely due to the ab-
sence of excess surface oxygen. In contrast, when dimethyla-
mine and methanol are sequentially dosed at 120 K, a signif-
icant amount of the coupling occurs near 235 K (Figure 1B).
C2H5N is still the dominant product, along with a small
amount of methyl isocyanate. Little of the possible oxidation
products of methanol, such as formaldehyde, formic acid
and methyl formate, or the combustion product, CO2, is de-
tected following dosing at either temperature, indicating
that excess adsorbed atomic oxygen preferentially attacks
the adsorbed dimethylamide under these conditions.
in situ on O/AuACHTUNGTRENNUNG(111), for example, through selective remov-
al of the acidic hydrogen of an alcohol followed by b-H
elimination (Scheme 1, dashed box).[6a] Because alcohols are
in general more available and less costly than aldehydes, it
is desirable to design processes with alcohols as the raw ma-
terials.
Herein, we report a mechanistic study on the selectivity
control on the coupling of dimethylamine and methanol
(ethanol) on O/AuACHTUNGTRENNUNG(111) to form dimethylformamide (dime-
thylacetamide), which inherently competes with oxidative
dehydrogenation and secondary oxidation of the amine. The
mechanistic insight sheds light on the guiding principles for
achieving optimum selectivity for such coupling reactions.
Significantly more methanol is co-adsorbed at 120 K than
150 K, as evidenced by the observation that the amount of
the unreacted methanol that desorbs is about 25 times
higher following adsorption at 120 K vs. 150 K (Figure 1).
This difference is expected, as the temperature of desorption
Results
of methanol from the multilayer on AuACTHUNGTERNNU(G 111) has been re-
ported to be 134 K.[11] In contrast, the amount of unreacted
dimethylamine increases only by about a factor of 1.5.
Therefore, the different dosing temperatures result in signifi-
cantly different relative concentrations of the preadsorbed
reactants, which apparently affects the product distribution.
Indeed, for a fixed initial oxygen adatom coverage of
0.1 ML, the amount of dimethylformamide formed corre-
lates well with the absolute amount of adsorbed methanol
rather than the relative ratio of adsorbed methanol and di-
methylamine. Similar ratios of methanol-to-dimethylamine
Selective coupling to form dimethylformamide requires an
excess of adsorbed methanol. This effect is reflected in a
strong dependence of the product distribution for coupling
on the temperatures at which reactants are introduced to O/
Au
ACHTUNGTRENNUNG
actants are dosed onto O/AuAHCTNUGTRENNNUG
0.1 ML) at 150 K. The dominant product results from the
oxygen-assisted dehydrogenation of dimethylamine, forming
C2H5N and water at approximately 250 K (Figure 1A).[6a]
We were unable to differentiate the two possible isomers of
C2H5N, CH3N=CH2 and ethylenimine, due to the unavaila-
bility of authentic samples and reference fragmentation pat-
terns. However, from the work of Angelici on gold particles
in solution, we infer that the product is the imine.[10] The
condensed on O/AuACTHNUTRGNEUNG(111) were created at surface tempera-
tures of both 120 and 150 K with varying absolute amounts.
For example, both 6 L of dimethylamine and 0.06 L of meth-
anol dosed at 120 K and 0.24 L of dimethylamine and 6 L of
methanol dosed at 150 K result in a methanol-to-dimethyla-
mine ratio of approximately 1 desorbing from the surface;
however, the absolute amount of methanol or dimethyla-
mine initially adsorbed when dosed at 120 K is higher by a
factor of approximately 15. The yield of the coupling prod-
uct, dimethylformamide, increases almost linearly as the
methanol-to-dimethylamine ratio rises when reactants are
introduced at 120 K (Figure 2A, black solid trace). In con-
trast, little coupling occurs when reactants are introduced at
150 K, independent of the methanol to dimethylamine ratio
(Figure 2A, gray dotted trace). The amount of dimethylfor-
mamide formed correlates well with the amount of adsorbed
methanol when both reactants were introduced at 120 K
(Figure 2B). Similar results were observed for dimethyla-
mine and ethanol; a small amount of coupling product, di-
methylacetamide, was detected when dimethylamine and
ethanol were dosed onto O/AuACHTNUGTRENUNG(111) (qO =0.1 ML) at 150 K,
whereas a significant amount of dimethylacetamide was
formed with reactants co-adsorbed at 120 K (Figure 3).
X-ray photoelectron spectroscopy suggests that less
atomic oxygen remains after reaction of 6 L of methanol
and dimethylamine with 0.1 ML adsorbed oxygen at 120 K
compared with 150 K (Figure 4). A single O (1 s) peak at
Figure 1. A) No dimethylformamide (m/z=73) is observed when dime-
thylamine and methanol are introduced to O/Au
ACHTUNGTERN(NGNU 111) (qO =0.1 ML) at
150 K; whereas, B) coupling of dimethylamine and methanol occurs
when the reactants are introduced to O/AuACTHNUTRGNEN(UG 111) (qO =0.1 ML) at 120 K.
The exposure for all reactants is 6 L. The heating rate is 5 KsÀ1. Surface
oxygen was prepared by ozone exposure at 200 K.
2314
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2313 – 2318