A pathway for NH addition to styrene promoted by gold

Article abstract of DOI:10.1002/anie.200602876

(Figure Presented) Going for gold! The synthesis of aziridines using heterogeneous gold catalysts has an unanticipated potential. Chemisorbed atomic oxygen is used to activate ammonia, producing NH bound to the gold surface (see scheme). Addition of NH ac

Full text of DOI:10.1002/anie.200602876

Angewandte
Chemie
Surface Chemistry
DOI: 10.1002/anie.200602876
A Pathway for NH Addition to Styrene Promoted
by Gold**
Xingyi Deng, Thomas A. Baker, and Cynthia M. Friend*
Gold-based heterogeneous catalysts have surprising potential
[
1–4]
for low-temperature oxidation processes,
including alcohol
[
5–8]
[9,10]
oxidation,
oxidation,
direct synthesis of hydrogen peroxide,
CO
[
11,12]
[11,13]
and olefin epoxidation.
These systems
have potential for a substantial positive impact on the
environment and economy because of their high selectivity
[
14]
and also the low temperature at which they operate. Hence,
a substantial amount of effort has been directed to further
improve the performance of heterogeneous gold catalysts and
[
15–18]
to understand the origin of their catalytic activity.
Herein, we investigate the addition of NH to styrene using
heterogeneous Au because of the potential importance of
these three-membered N-containing rings in organic synthesis
as a building block for biologically active molecules and for
[
19–21]
use in antitumor and antibiotic applications.
Previously,
we demonstrated that the Au(111) surface promotes oxida-
tion reactions that also occur on catalysts with high surface
areas at higher pressure, once oxygen has been adsorbed onto
[
22]
the surface, thereby establishing that Au(111) is a good
model for understanding the molecular-level detail of het-
erogeneous Au-based oxidation catalysis. Hence, reactions
that occur on Au(111) provide a guide for the type of
reactions that may be induced by heterogeneous gold
catalysts.
Nitrene addition to styrene is promoted by a gold surface,
which indicates that such reactions should be possible with
heterogeneous Au catalysts. To our knowledge, this is the first
report of gold-catalyzed functionalization of an olefin with
NH in a heterogeneous system, which may have advantages
and mechanistic differences when compared to aziridination
[
23,24]
in solution,
for amination.
or in other methods of activating ammonia
[
25]
Aziridines are structurally analogous to epoxides and may
be formed by addition of a nitrene group to the olefin.
Recently, we showed that oxidized Au(111) promotes styrene
[
22]
epoxidation. This observation motivated us to explore the
[
*] X. Deng, T. A. Baker, Prof. C. M. Friend
Department of Chemistry and Chemical Biology
Harvard University
Cambridge, MA 02138 (USA)
Fax: (+1)617-496-8410
E-mail: cfriend@deas.harvard.edu
Prof. C. M. Friend
Division of Engineering and Applied Sciences
Harvard University
Cambridge, MA 02138 (USA)
[
**] This work was supported financially by the US Department of
Energy, Basic Energy Sciences (grant no. FG02-84-ER13289) and the
National Science Foundation Graduate Fellowship.
Angew. Chem. Int. Ed. 2006, 45, 7075 –7078
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7075

Communications
analogous aziridination of styrene on Au(111). Indeed,
ing species detected upon heating; NO, N O, and NO are not
2 2
aziridination of styrene does occur on Au(111) covered with
detected (data not shown). Furthermore, there is no detect-
NH . By precovering the surface with NH, we are able to
able formation of O , which would occur at around 550 K
x
2
study the elementary steps important in catalytic processes.
This is a well-established approach for studying other
reactions important in catalysis using the tools of surface
from oxygen-atom recombination if there were residual
adsorbed oxygen present. The reactivity of the oxygen layer
towards NH and other species depends on the method used
3
[
26–28]
science.
for oxidation.
Reactive NHx (x = 1,2) species are formed from the
To accumulate NH and remove OH and H O, as well as
x
2
reaction of NH3 with chemisorbed oxygen on Au(111)
molecular NH3 from the surface, we exposed the oxygen-
(
Figure 1). By selecting the appropriate conditions, we are
covered Au(111) surface (q = 0.2 monolayer (ML)) to a
O
larger dose of NH at 190 K (a temperature that is high
3
enough to prevent adsorption of molecular NH and low
3
enough to keep NH on the surface; Figure 1b). After this
x
treatment, NH reformation from NH , was observed as a
3
x
peak at 220 K and N formation at 425 K. Water formed from
2
the reaction of NH with oxygen desorbs from the surface
3
during the reaction and therefore there is no well-defined
H O peak—only an increased water background was
2
observed. This desorption is concomitant with ammonia
desorption, possibly because of the displacement of water
by ammonia from other surfaces. Since this method exclu-
sively produces NHx on the surface, it was used in all
experiments described later.
Addition of NH to styrene was observed following the
reaction of styrene on Au(111) covered with NH at 150 K
x
(
1
3
Figure 2). Specifically, a product with a parent ion at m/z =
19 and the most intense peak at m/z = 118was evolved at
90 K and tentatively identified as 2-phenylaziridine. A small
Figure 1. Temperature-programmed reaction spectra following the
adsorption of NH on O-covered Au(111) (q =0.2 ML). When NH is
exposed to the O-covered surface at 110 K (a), all oxygen reacts with
3
O
3
amount of H O was also produced in the reaction, possibly
because of the presence of some residual oxygen on the
surface following exposure to ammonia. Notably, no oxygen-
NÀH bonds to form water on heating above 175 K. Exposure of NH
2
3
to the same surface at 190 K (b) leads to removal of all oxygen as
water, so as to generate a surface which contains mainly adsorbed
NH . The NH disproportionates to evolved gaseous NH and ad-
x
x
3
sorbed N upon heating to higher temperatures. The adsorbed nitrogen
atoms combine to form N at 425 K. The oxygen-covered surface was
2
prepared by O decomposition at 200 K. I=mass spectroscopy inten-
3
sity.
able to remove all the oxygen through the evolution of water.
In temperature-programmed reaction experiments, water is
evolved at 210 K if NH is exposed to O-covered Au(111) at
3
1
10 K; however, when dosed with an excess of ammonia and
the surface maintained at 190 K, the oxygen is removed
through water evolution (Figure 1a,b). Notably, ammonia
does not react on the clean Au(111) surface, but desorbs
molecularly at 155 K (data not shown). The NH formed on
x
the surface undergoes disproportionation upon heating
(
1
Figure 1). Molecular ammonia desorbs from the surface at
60 K. A second NH3 peak at 210 K is ascribed to the
reformation of NH from disproportionation of NH species.
3
x
Adsorbed nitrogen atoms are also formed, which ultimately
lead to N2 formation at 425 K. No H is evolved; rather
2
hydrogen is removed either as water during formation of NHx
or by reformation of NH . All these observations show that
3
NÀH bonds are activated by the reaction of ammonia with
Figure 2. Temperature-programmed reaction spectra following the
adsorption of styrene on NH -covered Au(111) at 150 K. The reaction
x
oxygen-covered Au(111), analogous to reactions reported
yields 2-phenylaziridine (390 K), exclusively. Neither oxidation nor
nitrilation products were observed. Some residual oxygen reacts with
NH and forms H O at 250 K. No N desorption was observed, which
[
29–31]
previously on Ag and Cu surfaces,
and to recent reports
[
32]
for Au(111) oxidized using an oxygen plasma.
x
2
2
All adsorbed oxygen is converted into water through the
indicates a high overall activity since all the NH has been consumed
x
reaction with NÀH bonds. Water is the only oxygen-contain-
in the reaction.
7076 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7075 –7078

Angewandte
Chemie
addition products (CO, CO , epoxide, or organic acids) were
detected. Remaining styrene desorbed from the surface at
Table 1: Mass spectrometer fragmentation patterns of 2-phenylaziridine
2
from the reaction of styrene on a Au(111) surface covered with NH and
x
also reactions of isomers with C H N stoichiometry.
8
9
2
00 K (multilayers) and 285 K (monolayer). A residue con-
[
a]
taining carbon and nitrogen remained on the surface after
reaction of styrene and NH . NO and CO were produced
m/ 2-Phenyl-
Indoline N-Phenyl-
4-Amino-
styrene
[
b]
[b]
z
aziridine
aziridine
x
2
after exposing the residue on the surface to ozone and
subsequent heating. The fact that NO was produced from
post-oxidation of the residue, but no N was observed after
2
9
9
1
1
119
120
1
2
40
3
45 (30)
5 (3)
45 (29)
100 (100) (9)
50 (63)
4 (6)
(100)
(10)
(0)
(24)
(5)
(15)
(58)
(100)
(8)
reaction of styrene with NH , indicates that the nitrogen is
x
bound to carbon in the residue.
17 36
18 100
We unequivocally established that the product formed at
1
5
3
90 K contains nitrogen by using NH as our starting reagent
6
0
(40)
(4)
3
(
Figure 3). The parent ion shifts from m/z = 119 to 120 when
[a]These data rule out the possibility of the product being indoline, N-
phenylaziridine, or 4-aminostyrene. [b] The numbers in the parentheses
were obtained from the NIST database.
source of nitrogen is compelling evidence that the product has
a benzyl group (C H CH -), which is consistent with the 2-
6
5
2
phenylaziridine structure. Furthermore, this result rules out
N-phenyl aziridines and the imine, C H C(CH )=NH as the
6
5
3
product because the strong m/z = 91 signal contains nitrogen
in this isomer. Although the mass spectrum for 2-phenyl-
aziridine is not available from the NIST database for direct
comparison, our observations all strongly point to it being the
product. The other imine C H CH C(H)=NH cannot be
6
5
2
completely ruled out.
We suggest that the formation of 2-phenylaziridine is the
result of the transfer of NH to styrene on Au(111) by analogy
with the corresponding epoxidation reaction on this sur-
Figure 3. Mass spectrometer fragmentations of 2-phenylaziridine from
1
4
the reaction of styrene on a Au(111) surface covered with a) NH or
x
[
22]
1
5
face. Spectroscopic studies are underway to identify the
b) NH . The shift of +1 for the parent ion (m/z 119 to 120), as well
as the most intense peak (m/z 118 to 119) when using NH , is a
x
1
5
NH species present on the surface during reaction. We will
3
x
strong indication that the product contains nitrogen.
show in a subsequent report that styrene nitrilation (forming
benzo- and benzyl nitrile) occurred only when nitrogen atoms
[
33]
were pre-adsorbed on Au(111). This observation indicates
that the presence of NH (NH and/or NH ), not nitrogen
1
5
the reaction of styrene on the surface covered with NH was
x
x
2
performed. Notably, a shift of the m/z signal by + 1 is also
observed for all the observed mass fragments which contain
nitrogen; for example, the most intense fragmentation [MÀ1]
shifts from m/z = 118to 119.
atoms, is necessary for aziridination.
Our results indicate that 2-phenylaziridine is formed from
the Au-promoted reaction of styrene and NH . In this
x
reaction, oxygen is used to activate the NH and to form
3
The stoichiometry (C H N) of the product was deter-
NH species on the surface. By adjusting the preparation
8
9
x
mined from the parent ion signal (m/z = 119) in conjunction
conditions, oxygen is consumed by the NH and forms H O,
3
2
with the shifts in the signals found in the experiments using
which desorbs, and NH remains exclusively on the Au(111)
surface.
x
1
5
NH . The intensities of the major fragmentation ions of the
3
product from the experiment using our mass spectrometer
were compared with the literature values of selected ion
intensities of isomers with C H N stoichiometry (Table 1).
The aziridination of organic molecules is generally a
complex process usually promoted by a homogeneous cata-
lyst. Our study clearly shows the promise of the aziridina-
[
24]
8
9
The use of literature values to eliminate possible products
based on the yields of the fragment ions is validated by the
good agreement with the measured intensities of the fragment
ions of indoline.
tion of a C=C bond by using a heterogeneous gold system,
which may provide a convenient and more efficient route. The
key features are the NÀH bonds so as to form the desired
nitrene and a C=C bond. Notably, the temperature needed for
We assign the product formed in the reaction as 2-
phenylaziridine based on analysis of its fragmentation pattern
this process is low, which is significant for the selective
synthesis of molecules important in the pharmaceutical
industry. Nevertheless, there are still unanswered questions
including the identification of the product by other means to
completely rule out other products like imines, the generality
of the process for more complex nitrenes and olefins, the
degree of selectivity, whether these reactions can be carried
(
Table 1). The fragments of the product we observed do not
match any of those C H N isomers cited by the National
8
9
Institute of Standards and Technology (NIST). In addition,
the presence of the m/z = 91 fragment and the absence of a
1
5
mass shift for this fragment when NH was used as the
3
Angew. Chem. Int. Ed. 2006, 45, 7075 –7078
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7077

Communications
out stereoselectively, and whether this process can be carried
[8] D. I. Enache, J. K. Edwards, P. Landon, B. Soslona-Espriu, A. F.
Carley, A. A. Herzing, M. Watanabe, C. J. Kiely, D. W. Knight,
G. J. Hutchings, Science 2006, 311, 362.
out catalytically. Although this study was carried out using a
single crystal and under ultrahigh-vacuum conditions, it is
reasonable that this pathway would also apply for high-
pressure conditions using supported nanoscopic Au clusters,
especially since the related oxidation reactions on gold occur
under both low- and high-pressure conditions.
[
9] P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely, G. J. Hutch-
ings, Chem. Commun. 2002, 2058.
[
10] P. Landon, P. J. Collier, A. F. Carley, D. Chadwick, A. J. Pap-
worth, A. Burrows, C. J. Kiely, G. J. Hutchings, Phys. Chem.
Chem. Phys. 2003, 5, 1917.
[
[
[
[
11] M. Haruta, Catal. Today 1997, 36, 153.
12] G. C. Bond, D. T. Thompson, Gold Bull. 2000, 33, 41.
13] T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 1998, 178, 566.
14] C. T. Campbell, Science 2004, 306, 234.
Experimental Section
All experiments were performed in a stainless-steel ultrahigh-vacuum
À10
[15] M. Valden, X. Lai, D. W. Goodman, Science 1998, 281, 1647.
[16] M. S. Chen, D. W. Goodman, Science 2004, 306, 252.
[
chamber with a base pressure of < 1 ” 10 Torr, as described
[
34]
previously. The chamber was equipped with a quadrupole mass
spectrometer (UTI model 100C) which was used for temperature-
programmed reaction spectroscopy (TPRS), an Auger electron
spectrometer (AES), and low-energy electron diffraction optics
17] B. Yoon, H. Hakkinen, U. Landman, A. S. Worz, J. M. Anto-
nietti, S. Abbet, K. Judai, U. Heiz, Science 2005, 307, 403.
18] M. D. Hughes, Y. J. Xu, P. Jenkins, P. McMorn, P. Landon, D. I.
Enache, A. F. Carley, G. A. Attard, G. J. Hutchings, F. King,
E. H. Stitt, P. Johnston, K. Griffin, C. J. Kiely, Nature 2005, 437,
[
(
LEED).
The Au(111) surface was prepared by cycles of sputtering of Ar
ions (1000 eV, 1.3 mA) at 300 K, followed by annealing at 900 K for
min, and then at 700 K for 30 min. This procedure was repeated
+
1132.
[
19] K. Dvorakova, C. M. Payne, M. E. Tome, M. M. Briehl, T.
McClure, R. T. Dorr, Biochem. Pharmacol. 2000, 60, 749.
20] F. Brown, Vaccine 2001, 20, 322.
21] T. B. Sim, S. H. Kang, K. S. Lee, W. K. Lee, H. Yun, Y. K. Dong,
H. J. Ha, J. Org. Chem. 2003, 68, 104.
5
until no impurities were detected using AES.
[
[
Ozone was produced by electrical discharge and was trapped in
silica gel (3–8mesh, Fisher Scientific Co.) at À788C using a mixture of
dry ice and ethanol. The Au(111) surface at 200 K was exposed to O ,
3
À8
[22] X. Deng, C. M. Friend, J. Am. Chem. Soc. 2005, 127, 17178.
[
with a pressure rise of 1 ” 10 Torr over 20 s, which resulted in an
approximate 0.2 ML. The resulting oxygen atom coverage was
determined by comparing the integrated area of the O -desorption
23] D. A. Evans, M. M. Faul, M. T. Bilodeau, J. Am. Chem. Soc.
994, 116, 2742.
24] P. Müller, C. Fruit, Chem. Rev. 2003, 103, 2905.
1
2
[
peak relative to a saturation coverage (ca. 1 ML) which had been
[
35,36]
[25] J. Zhao, A. S. Goldman, J. F. Hartwig, Science 2005, 307, 1080.
[26] J. M. Gottfried, K. Christmann, Surf. Sci. 2004, 566, 1112.
[
[28] J. T. Roberts, R. J. Madix, J. Am. Chem. Soc. 1988, 110, 8540.
[
determined previously.
Following the deposition of oxygen, the
1
5
surface was exposed to NH (Matheson, anhydrous grade) or NH3
3
27] J. T. Roberts, A. J. Capote, R. J. Madix, Surf. Sci. 1991, 253, 13.
(Cambridge Isotope Laboratories, 98%). The NH was deposited at
3
À10
1
90 K for 2 min with a pressure rise of 1 ” 10 Torr. Finally, styrene
29] B. Afsin, P. R. Davies, A. Pashuski, M. W. Roberts, Surf. Sci.
1991, 259, L724.
30] X. C. Guo, R. J. Madix, Surf. Sci. 2002, 501, 37.
31] B. Afsin, P. R. Davies, A. Pashusky, M. W. Roberts, D. Vincent,
Surf. Sci. 1993, 284, 109.
(
Alfa Aesar, 99.5%) was deposited at 150 K for 30 s with a pressure
À10
rise of 1 ” 10 Torr and was used after cycles of freeze-pump-thaw
purification, with the purity checked by mass spectrometry.
All temperature-programmed reaction spectra (TPRS) were
taken by a computer-controlled UTI 100c mass spectrometer, as
[
[
[
37]
[32] J. Gong, R. A. Ojifinni, T. S. Kim, J. M. White, C. B. Mullins, J.
Am. Chem. Soc. 2006, 128, 9012.
[
[
[
[
described in detail previously. The crystal was biased at À100 V
during the collection of temperature-programmed reaction data, so as
to avoid an electron-induced reaction from the mass-spectrometer
filament. The temperature was measured with a K-type (Chromega/
Alomega) thermocouple and radiative heating was used to achieve
the temperature ramp to 650 K. The heating rate for TPRS was
33] X. Deng, C. M. Friend, unpublished results.
34] M. K. Weldon, C. M. Friend, Surf. Sci. 1994, 310, 95.
35] N. Saliba, D. H. Parker, B. E. Koel, Surf. Sci. 1998, 410, 270.
36] B. K. Min, A. R. Alemozafar, D. Pinnaduwage, X. Deng, C. M.
Friend, J. Phys. Chem. B 2006, ASAP DOI:10.1021/jp0616213.
relatively linear, in the range 150–600 K with an average of about
À1
[37] L. J. Deiner, J. G. Serafin, C. M. Friend, S. G. Weller, J. A.
6
Ks
.
Levinson, R. E. Palmer, J. Am. Chem. Soc. 2003, 125, 13252.
Received: July 19, 2006
Published online: October 2, 2006
Keywords: ammonia · aziridination · gold ·
.
heterogeneous catalysis · surface chemistry
[
[
1] G. C. Bond, D. T. Thompson, Catal. Rev. Sci. Eng. 1999, 41, 319.
2] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Science 2003,
301, 935.
[
3] J. E. Bailie, G. J. Hutchings, Chem. Commun. 1999, 2151.
[
4] C. Mohr, H. Hofmeister, M. Lucas, P. Claus, Chem. Ing. Tech.
1999, 71, 869.
[5] S. Biella, L. Prati, M. Rossi, Inorg. Chim. Acta 2003, 349, 253.
[6] S. Biella, L. Prati, M. Rossi, J. Catal. 2002, 206, 242.
[7] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C. J. Kiely, G. J.
Hutchings, Phys. Chem. Chem. Phys. 2003, 5, 1329.
7078 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7075 –7078