(Parr Instruments). No stirring was used. The reaction was
heated to 70 ◦C prior to introduction of hydrogen. To monitor
conversion with time, a series of reactions were performed,
stopped at pre-determined time-intervals. The elevated tempera-
ture, pressure of hydrogen, and presence of succinic acid product
did not affect the stability of the DW, with the material remaining
as a free-flowing powder with similar droplet size throughout
(Fig. 1e).
found to be increasingly aggregated and less flowable, possibly
due to changes in the surface chemistry of the nanoparticles.
From an energy perspective, the DW method presented here
requires just 30 s of blending at 19 000 rpm in addition to a cen-
trifugation step. The centrifugation step is however readily cir-
cumvented since the DW droplets can be coalesced in other ways,
for example by the addition of a small amount of methanol.
Although we have not yet made quantitative comparisons, this is
likely to compare very favourably with the energy input required
to stir the reaction at 1200 rpm for the duration of the reaction.
It is also clear, however, that for this reaction it is necessary to
heat the system for a much longer time to achieve comparable
conversions (Fig. 2). Although this may offset any energy gains
in terms of stirring for this particular reaction, it suggests that the
DW method could be more attractive, for example, in reactions
which occur at lower temperatures. Additionally, for larger scale
reactions which suffer from poor heat exchange, these may be
carried out slower than the intrinsic reaction rate to offset this
disadvantage. In such a case, a DW approach may prove more
energy efficient despite a slower running time.
This catalytic hydrogenation serves as a proof of concept for
the use of a distributed DW gas–liquid interface in multi-phase
fluid reactions. The concept may be applicable to a number of
aqueous-based gas–liquid reactions. For example, more viscous
reaction mixtures require a much greater power input in terms of
stirring and such systems might perform more favorably in a DW
form. Indeed, we have recently prepared DW containing gelling
agents which involve mixtures which are extremely viscous and
hard to stir but which can be blended into stable DW form in as
little as 60 s.14 Additionally, one can envisage a heterogeneous
catalyst that could also act as the solid DW stabilizing agent.
We also suggest that DW may impart greater ease of handling
to some reactants.
The DW mixture was then centrifuged (5000 rpm, 10 min) to
coalesce the droplets and hence to separate the solid and aqueous
components. The conversion of maleic acid to succinic acid was
1
measured using H NMR. Control experiments demonstrated
that this method accurately reflected the composition of the
solution, with no preferential adsorption of either maleic acid
or succinic acid to the silica. The results are shown in Fig. 2.
For comparison, the results for reactions of a bulk, unstirred
aqueous solution of maleic acid with hydrogen in the presence
of Ru/Al2O3 (i.e., not pre-formed into DW) are also shown. We
have also included data for reactions carried out in the absence
of hydrophobic silica, but in an impeller-stirred pressure vessel.
The conditions used here mirror those used elsewhere,12 albeit
with an elevated loading of catalyst (1.63 gL-1). At lower catalyst
loadings (0.065 gL-1), the DW system gave irreproducible
data across a given reaction period, which we ascribe to
uneven distribution of the catalyst across the DW gas–liquid
interface.
In summary, we have used a “dry water” distributed gas–liquid
interface to produce a reaction mixture with a high surface area
to volume ratio. We have demonstrated that this system can be
used to conduct catalytic hydrogenation without the need for
stirring during reaction. This new concept requires considerable
investigation in terms of scale-up potential, but could in time be
applied to specific gas–liquid reactions in order to improve the
overall energy efficiency.
Fig. 2 Kinetic hydrogenation data for unstirred control reaction (ꢀ),
DW (᭹), and for stirred control reaction (ꢀ). All experiments were
carried out at 70 ◦C, 20.7 bar.
Acknowledgements
We thank the EPSRC (EP/G006091/1) for funding, and Wacker
Chemie for providing hydrophobic silica. A. I. C. is a Royal
Society Wolfson Research Merit Award holder.
As can be seen in Fig. 2, a very low conversion of maleic
acid to succinic acid was achieved for an unstirred bulk
mixture of silica, maleic acid solution and catalyst. This is as
expected for a poorly mixed, heterogeneous reaction. Stirring
the system during reaction in the absence of silica results in
fast conversion to succinic acid, as expected.12 Interestingly,
for the DW system, greatly enhanced kinetics of succinic acid
formation were observed over the control, though these kinetics
were somewhat slower than observed for the comparable stirred
reaction.
Notes and references
1 P. D. Vaidya and V. V. Mahajani, Chem. Eng. Sci., 2005, 60, 1881–
1887.
2 P. D. Vaidya and V. V. Mahajani, Ind. Eng. Chem. Res., 2003, 42,
3881–3885.
3 R. P. Fishwick, R. Natividad, R. Kulkarni, P. A. McGuire, J. Wood,
J. M. Winterbottom and E. H. Stitt, Catal. Today, 2007, 128, 108–
114.
4 A. Beenackers and W. P. M. Vanswaaij, Chem. Eng. Sci., 1993, 48,
3109–3139.
5 M. P. Dudukovic, F. Larachi and P. L. Mills, Catal. Rev. Sci. Eng.,
2002, 44, 123–246.
The activity of recovered Ru/Al2O3 was tested by re-blending
recovered catalyst and silica with fresh maleic acid solution,
followed by a 1 h hydrogenation. The resulting samples were
784 | Green Chem., 2010, 12, 783–785
This journal is
The Royal Society of Chemistry 2010
©
Carter, Benjamin O.
