Alkali metals on nanoporous carbon: new solid-base catalysts
Mark G. Stevens and Henry C. Foley*
Center for Catalytic Science and Technology, Department of Chemical Engineering, University of Delaware, Newark, Delaware,
USA
Caesium entrapped in nanoporous carbon is not pyrophoric,
is thermally stable towards desorption up to 773 K, but
retains its ability to produce hydrogen from water, and is a
strong basic catalyst, providing a greater than 9:1 ratio of
the less stable cis-but-2-ene over the trans isomer in the
isomerization of but-1-ene at 273 K.
formed predominantly in base catalysis, although trans-but-
2-ene is more stable thermodynamically.
Samples containing 3–42 mass% Cs in CMS‡ were synthe-
sized. The thermal stability of these materials was demonstrated
by heating them in vacuum to temperatures as high as 773 K; the
Cs did not desorb from the CMS. In contrast Cs is lost rapidly
from Cs/graphite compounds, which exfoliate readily.5
Materials produced from alkali metals and carbon have been of
interest to researchers for many years. Graphite intercalation
compounds of alkali metals have been extensively studied1 and
shown to be active basic catalysts in reactions such as side-chain
alkylation,2 but-1-ene isomerization3 and amine synthesis.4
Unfortunately, these materials were found to be pyrophoric, had
relatively low surface area, and exfoliated at reaction tem-
perature.5 A catalytically active caesium fulleride, C20Cs6, has
been reported, but it was found to be highly air-sensitive.6
Carbogenic molecular sieves (CMS) are chemically inert
solids that contain nanopores,7,8 formed from aromatic nano-
domains, arranged chaotically in space. Here, we show that
alkali metals introduced into a CMS structure adsorb in these
nanopore spaces, and that curvature effects within the CMS
pores provide for a much higher thermal stability, and that these
materials are exceptionally active base catalysts in double-bond
migration.
Results of magnetic susceptibility and EPR measurements
show a significant population of unpaired electrons exists in
these materials. When a 17% Cs/CMS sample was exposed to
100% humidity at 288 K it absorbed 850 mg g21 of water. When
submerged in oxygen-free water, a 9.2% Cs sample produced
sufficient hydrogen to indicate that 26% of the Cs was still
active for water reduction.§ Hence, the Cs is active and
accessible. Yet, unlike metallic Cs and graphite Cs compounds,
Cs/CMS is not pyrophoric, even at 40% Cs loading.
A wholly nanoporous pyrolized polyvinylchloride–poly-
vinyldichloride carbon impregnated with 14% Cs showed no
catalytic activity for but-1-ene isomerization¶ at 298 K, but did
show activity at higher temperatures (18% conversion at 523
K). When transport porosity is added to the CMS, making the Cs
in the nanopores more accessible, the reaction took place at low
temperatures (Table 1). The catalysts were found to be stable,
with a reaction turnover of > 30 (30 mol of but-2-ene produced
per mol of Cs) before deactivation. To determine the global,
first-order rate constant for the reaction we diluted the butene
feed to 33 mol% but-1-ene in argon. From differential
conversion experiments¶ on a 9% Cs sample we determined an
apparent first-order rate constant (ka) of 0.07 s21 g21 and a
cis:trans ratio of 9.1 at 273 K. This combination of high
bacisity with reduction activity and the lack of pyrophoricity
makes Cs/CMS extraordinarily interesting and useful.
CMS samples were prepared pyrolitically from poly-
vinylchloride–polyvinyldichloride-, poly(furfuryl alcohol)- and
poly(furfuryl alcohol)–poly(ethylene glycol) mixtures.7–10
A
poly(furfuryl alcohol)–poly(ethylene glycol) mixture (PFA–
PEG) produces material that forms a nanopore structure similar
to the PFA carbons, with a pore-size distribution centred around
0.5 nm. However, the addition of PEG results in the formation
of transport porosity11 with a second distribution centred around
10 nm, that allows access to catalytic sites contained in the
nanopores.12 Cs was chosen as the catalytic metal because it is
the most electropositive element and its covalent atomic
diameter (0.48 nm) is close to the nanopore size of the CMS (ca.
0.5 nm). Since Cs is strongly electropositive, a significant
portion of the Cs may be ionized (crystallographic diameter
0.34 nm), which should add to the compound’s stability.
The synthesis of the Cs/CMS material was achieved by
vapour-phase deposition.† The base-catalysed isomerization of
but-1-ene to but-2-ene, which begins with the abstraction of the
allylic proton to form the cis or trans form of the allyl anion,13
was used to probe for catalytic activity. Since the cis form of this
intermediate is more stable than the trans form, cis-but-2-ene is
Footnotes
† After a sample of the CMS was heat-treated in vacuum at 450 °C for 24
h, Cs vapour was introduced into the carbon at moderate temperature
(350 °C) at its vapour pressure (ca. 10 mmHg). The temperature was held
at 350 °C under static vacuum for 24 h.
‡ Galbraith Laboratories, Inc. provided metal analysis via plasma emission
spectroscopy for the determination of alkai-metal loading.
§ A 2 g sample of catalyst was submerged in 20 ml of deionized, oxygen-
free, water, without exposure to air. Gas evolved. The gas was quantified
and determined to be hydrogen by gas chromatography. The water and
catalyst were titrated with 0.0125 m sulfuric acid to calculate the amount of
Cs present as CsOH.
¶ We constructed a computer-controlled, tubular, flow reactor. For the
initial experiments: the reactor was loaded in an inert atmosphere with 4.75
g of catalyst. Technical grade but-1-ene (Matheson gas company) flowed at
15 cm3 min21 at atmospheric pressure. For the kinetic experiments: the
reactor contained 3.0 g of catalyst. Technical grade but-1-ene flowed at 20
cm3 min21, mixed with 40 cm3 min21 argon at atmospheric pressure.
Analysis was performed via gas chromatography.
Table 1 Average conversion on 4.5 g catalyst and cis/trans ratio of the
product gas for the isomerization of 15 cm5 min21 butene to cis- and trans-
but-2-ene at 298 K and 1 bar. Six datapoints (gas chromatography) were
collected from the flow reactor over a 2 h period and averaged
Catalyst
Conversion (%)
cis/trans ratio
3% Cs-PFA–PEG
10% Cs-PFA–PEG
11% Cs-PFA–PEG
18% Cs-PFA–PEG
21% Cs-PFA–PEG
13%
75%
60%
90%
87%
2.3
3.3
4.9
2.8
3.7
References
1
N. Bartlet and B. W. McQuillan, in Intercalation Chemistry, ed. M. S.
Whittingham and A. J. Jacobson, Academic Press, New York, 1982,
pp. 19–50 and references therein.
2 W. E. Foster, US Pat. 3,160,670, 1964.
Chem. Commun., 1997
519
Stevens, Mark G.
