Apparent equilibrium shifts and hot-spot formation for catalytic reactions induced by microwave dielectric heating

Article abstract of DOI:10.1039/a901245a

Microwave dielectric heating of the gas phase decomposition of H₂S catalysed by metal sulfides on a γ-Al₂O₃ support results in significant apparent shifts in the equilibrium constant, which have been attributed to the development of hot-spots in the catalytic beds; X-ray diffraction and electron microscopy measurements have indicated the formation of hot-spots with dimensions of 90-1000 μm and which involve not only the active phase, but also the support.

Full text of DOI:10.1039/a901245a

Apparent equilibrium shifts and hot-spot formation for catalytic reactions
induced by microwave dielectric heating
Xunli Zhang, David O. Hayward and D. Michael P. Mingos*
Department of Chemistry, Imperial College of Science, Technology and Medicine, South Kensington, London, UK
SW7 2AY. E-mail: d.mingos@ic.ac.uk
Received (in Cambridge, UK) 16th February 1999, Accepted 21st April 1999
Microwave dielectric heating of the gas phase decomposition
of H2S catalysed by metal sulfides on a g-Al2O3 support
results in significant apparent shifts in the equilibrium
constant, which have been attributed to the development of
hot-spots in the catalytic beds; X-ray diffraction and electron
microscopy measurements have indicated the formation of
hot-spots with dimensions of 90–1000 mm and which involve
not only the active phase, but also the support.
The acceleration of heterogeneous catalytic processes under
microwave dielectric heating conditions has attracted both
academic and industrial interest.1–5 The majority of systems
studied involve a catalyst which preferentially absorbs the
microwaves relative to the support material and the reasons for
the observed rate enhancements have been the subject of some
Fig. 1 Flow system for the H2S decomposition. 1, Mass flow control unit;
controversy. The measurement of temperature under conditions
2, mass flow controller; 3, temperature display; 4, optical sensor; 5, thermal
couple; 6, pressure gauges; 7, quartz reactors; 8, microwave generator; 9,
forward power meter; 10, reflected power meter; 11, microwave cavity; 12,
where there are strong electric fields is problematical, but may
be overcome, for example, by using optical fibre technology.
Such measurements are still only capable of providing average
temperatures. Both differential coupling abilities of materials
and distribution of electromagnetic fields may result in
localized temperature distribution in catalytic beds, but the
contribution of these effects is difficult to quantify. Stuerga and
Gaillard 6,7 have given a thorough theoretical analysis which has
conclusively demonstrated that specific athermal effects are
implausible given the small energies associated with the
microwave quanta and the classical nature of the heating
phenomenon and have suggested that localized enhancements
of the reaction rates may be responsible for non-isothermal and
heterogeneous kinetic phenomena. The majority of the experi-
mental studies on catalytic reactions under microwave condi-
tions have dwelled on the kinetic aspects and noted that the rates
of reaction and product distributions are more consistent with a
temperature 300-400 K higher than that measured for the bulk
of the catalyst bed. However, for supported metal catalysts,
calculations8,9 have shown that the rate of heat transfer at 1 atm
gas pressure is so high that the formation of hot-spots localised
at the catalytically active sites are implausible. Here we report
significant circumstantial evidence for the formation of hot-
spots in the microwave experiments and have demonstrated for
the first time that these hot-spots are not localised exclusively
on the active catalyst, but also involve the support material. We
have also estimated the dimensions of these hot-spots.
conventional furnace; 13, heating tapes; 14, cold trap; 15, ice–water Dewar;
16, filter; 17, quadrupole mass spectrometer(QMS).
impregnated catalyst based on molybdenum oxide on a g-
alumina support, which was sulfided by a pre-treatment using
an H2S–H2 flow at atmospheric pressure for 2 h, or a
mechanically mixed sample of MoS2 and g-alumina. The
reaction products were analyzed using a quadrupole mass
spectrometer (QMS-200D, European Spectrometry System).
Fig. 2 compares the conversion efficiency for the decomposi-
tion of H2S under microwave and conventional thermal
conditions in the temperature range 500–1100 K for both the
impregnated catalyst and the mechanical mixture of MoS2 and
g-alumina (both 30% by weight MoS2 on g-alumina). Fig. 2 also
illustrates the calculated equilibrium conversion efficiencies
based on thermodynamic data of Kaliodas and Papayannakas.12
It is noteworthy that for such an endothermic reaction the
conversion efficiencies are relatively low and that for the
conventional thermal reaction the experimental observed per-
The catalytic conversion of H2S into hydrogen and sulfur,
which is commercially important for the coal and petrochemical
industry,10,11 has recently been investigated in our laboratories
using parallel microwave and conventional heating conditions.
The experimental set-up is illustrated schematically in Fig. 1
and the reactions were performed under continuous flow
conditions using quartz reactors. The temperature in the
microwave cavity was monitored using an Accufibre optical
fibre thermometer (Model 10 Luxtron) which had its tip at the
centre of the catalyst sample, as shown. The microwaves were
generated at 2.45 GHz using an Electro-Medical Supplies Ltd
generator (0–200 W) and relayed by means of a co-axial cable
to a tunable cylindrical microwave cavity placed around the
reactor. The catalysts used for the study were either an
Fig. 2 H2S conversion vs. temperature with mechanically mixed catalyst A
and impregnated catalyst B
Chem. Commun., 1999, 975–976
975

Table 1 Dielectric properties of the chemicals (25 °C, 2.45 GHz)
mechanically mixed catalysts are respectively ca. 100 and 200
°C higher than the temperature measured by the probe.
X-Ray diffraction measurements on the catalyst before and
after microwave heating have demonstrated some important
differences compared with the parallel experiments done under
thermal conditions. Firstly, some of the alumina undergoes a
phase change from g- to a-alumina—a transition which only
occurs at temperatures above 1273 K.14 It is noteworthy that the
maximum average temperature recorded in the microwave
experiments was 1073 K. The average crystallite size calculated
from the peak widths of five intense peaks is 90 mm. Secondly,
the molybdenum disulfide, which was initially evenly distrib-
uted as 150–170 mm amorphous particles ( electron microscopy
measurements), forms some hexagonal crystals during the
microwave process. Since the melting point of MoS2 is 1458 K,
this is further evidence for the formation of hot-spots involving
both the MoS2 and the alumina support. The electron micros-
copy studies have also indicated that considerable migration of
MoS2 has occurred during the reaction since spheres as large as
1000 mm containing both Al2O3 and MoS2 are formed. Surface
area and size pore measurements have also confirmed that a
considerable reorganisation has occurred and has resulted in a
decrease in the surface area.
Dielectric
Dielectric
Dielectric loss
Chemical(s)
constant, eA loss, eAA
tangent, tan d
g-Al2O3
2.678
2.971
0.01768
0.2694
0.006604
0.09068
MoS2/g-Al2O3
(impregnated)
MoS2+g-Al2O3
(mechanically mixed)
MoS2
3.115
3.331
0.3162
0.6410
0.1015
0.1924
centage conversions are in good agreement with the equilibrium
data. The basic reaction is given by eqn (1)
H2S(g) = H2(g) +1/2S2 (g)
(1)
with negligible production of other gaseous sulfur species.12
Surprisingly, the conversion efficiency for the reaction which
takes place in the microwave cavity is higher than that predicted
on thermodynamic grounds. For the impregnated catalyst the
percentage conversion at a temperature of 800 °C has risen from
ca. 6.5% with conventional heating to 12% when microwaves
are used. The effect is even greater for the mechanically
produced catalyst, where the conversion efficiency has risen to
21% under microwave heating at 800 °C. We have concluded
that these observations are not due to any athermal microwave
effect, but rather that the selective features of microwave
dielectric heating have induced an anisotropic temperature
profile on the catalytic bed which is giving rise to ‘apparent’
shifts in the equilibrium constant for the reaction.
In summary, the preferential heating effects associated with
microwave dielectric heating can give rise to hot-spots which
result not only in rate enhancements, but also in apparent shifts
in the equilibrium constant. These hot-spots probably are larger
than 90 mm and may be as large as 1000 mm and have
temperatures 100-200 K above that of the remainder of the bulk.
The hot-spots also induce a considerable re-organisation of the
catalyst under microwave conditions.
The dielectric measurements on MoS2, the impregnated
catalyst and the support materials, which are summarised in
Table 1, clearly indicate that the catalysts are more ‘lossy’13
than the support materials and will initially preferentially absorb
the microwave energy. The extent to which the heat is then
transferred to the support has not previously been addressed.
While the dielectric loss of g-alumina is temperature dependent,
particularly in the high temperature range, this temperature rise
may be amplified to develop hot-spots. Such hot-spots could not
only give rise to increased reaction rates, but also, if the
residence times of the gases flowing over the catalyst are
relatively long compared to the half-lives for the reaction, then
they could create local shifts in equilibrium constant. To our
knowledge attention has not been drawn previously to large
equilibrium shift effects arising from this phenomenon.
We have argued that these ‘apparent’ equilibrium shift effects
may be enhanced if the catalyst had a higher concentration of
hot-spots, or had a higher temperature of hot-spots. Although,
the maximum rate enhancements are achieved using a well
dispersed catalyst with a high surface area, the hot-spot
phenomenon is accentuated by having a poorly dispersed
catalyst, such as that produced by mechanically mixing MoS2
and g-alumina.
The selective nature of the dielectric heating effect cannot be
reproduced easily using conventional heating methods. This
phenomenon is not only interesting from an academic per-
spective because the results are initially counter-intuitive, but
since it has the potential to increase the conversion efficiencies
of reactions with low yields by an order of magnitude, whilst
maintaining the same operating temperature as a conventionally
heated reactor, it is not without commercial potential.
BP is thanked for endowing DMPM’s Chair and EPSRC are
thanked for financial support.
Notes and references
1 G. Bond and R. B. Moyes, in Microwave Enhanced Chemistry, ed.
H. M. Kingston and S. J. Haswell, ACS Publications, Washington,
1997, p. 551 gives a review of the applications of microwaves in
catalytic chemistry.
2 T. M. Thiebaut, G. Roussy, M. Medjram, L. Seyfield, F. Gasin and J.
Maine, J. Chem. Phys., 1992, 89, 1427.
3 G. Bond, R. B. Moyes and D. A. Whan, Catal. Today, 1993, 17, 427.
4 G. Roussy, E. Marchal, J. M. Thiebaut, A. Kiennemann and G. Maire,
Fuel Process Technol., 1997, 50, 261.
5 G. Roussy, S. Hilaire, J. M. Thiebaut, G. Maire, F.Garin and S. Ringler,
Appl. Catal. A: Gen., 1997, 156, 167.
6 D. Stuerga and P. Gaillard, Tetrahedron, 1996, 52, 5505.
7 D. A. C. Stuerga and P. J. Gaillard, J. Microwave Power Electromagn.
Energy, 1996, 31, 87; 101.
8 W. L. Perry, D. W. Cooke, J. D. Katz, and A. K. Datye, Catal. Lett.,
1997, 47, 1.
9 J. R. Thomas, Catal. Lett., 1997, 49, 137.
10 V. E. Kaloidas and N. G. Papayannakas, Ind. Eng. Chem. Res., 1991, 30,
345.
Provided that the rate determining step for both forward and
back reactions occurs between adsorbed species that are fully
accommodated to the active catalyst temperature the equilib-
rium that is set-up will be characteristic of the hot-spot
temperature and not the average temperature of the catalyst bed
as a whole. By using this principle the hot-spot temperatures
given in the second and third columns of Table 2 were
calculated. The hot-spot temperatures for the impregnated and
11 E. A. Fletcher, J. Noring and J. Murray, Int. J. Hydrogen Energy, 1984,
9, 587.
12 V. E. Kaloidas and N. G. Papayannakas, Int. J. Hydrogen Energy, 1987,
12, 403.
Table 2 Hot-spot temperatures calculated for impregnated catalyst and
mechanical mixture of MoS2 and g-Al2O3
13 A. C. Metaxas and R. J. Meredith, Industrial Microwave Heating, Peter
Peregrines Ltd, London, 1983.
Hot-spot temp./°C
impregnated catalyst
Hot-spot temp./°C
mechanically mixed catalyst
Probe temp./°C
14 Powder Diffraction Files, ‘Inorganic Volume No. PD1S-10iRB, Joint
Committee on Powder Diffraction Standards, Philadelphia, 1967’ Card
10-425, American Society for Testing and Materials, PA.
500
600
700
800
600
700
810
920
650
770
890
1010
Communication 9/01245A
976
Chem. Commun., 1999, 975–976