Heterogeneous catalysis in an oscillatory baffled flow reactor

Article abstract of DOI:10.1039/c3cy00282a

The first demonstration of heterogeneous catalysis within an oscillatory baffled flow reactor (OBR) is reported, exemplified by the solid acid catalysed esterification of organic acids, an important prototypical reaction for fine chemicals and biofuel synthesis. Suspension of a PrSO₃H-SBA-15 catalyst powder is readily achieved within the OBR under an oscillatory flow, facilitating the continuous esterification of hexanoic acid. Excellent semi-quantitative agreement is obtained between OBR and conventional stirred batch reaction kinetics, demonstrating efficient mixing, and highlighting the potential of OBRs for continuous, heterogeneously catalysed liquid phase transformations. Kinetic analysis highlights acid chain length (i.e. steric factors) as a key predictor of activity. Continuous esterification offers improved ester yields compared with batch operation, due to the removal of water by-product from the catalyst, evidencing the versatility of the OBR for heterogeneous flow chemistry and potential role as a new clean catalytic technology. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3cy00282a

Catalysis
Science & Technology
View Article Online
View Journal | View Issue
PAPER
Heterogeneous catalysis in an oscillatory baffled
flow reactor†
Valentine C. Eze,a Anh N. Phan,a Cyril Pirez,b Adam P. Harvey,*a Adam F. Leeb and
Karen Wilson*b
Cite this: Catal. Sci. Technol., 2013,
3, 2373
The first demonstration of heterogeneous catalysis within an oscillatory baffled flow reactor (OBR) is
reported, exemplified by the solid acid catalysed esterification of organic acids, an important prototypical
reaction for fine chemicals and biofuel synthesis. Suspension of a PrSO3H-SBA-15 catalyst powder is
readily achieved within the OBR under an oscillatory flow, facilitating the continuous esterification of
hexanoic acid. Excellent semi-quantitative agreement is obtained between OBR and conventional stirred
batch reaction kinetics, demonstrating efficient mixing, and highlighting the potential of OBRs for
continuous, heterogeneously catalysed liquid phase transformations. Kinetic analysis highlights acid chain
length (i.e. steric factors) as a key predictor of activity. Continuous esterification offers improved ester
yields compared with batch operation, due to the removal of water by-product from the catalyst,
evidencing the versatility of the OBR for heterogeneous flow chemistry and potential role as a new clean
catalytic technology.
Received 23rd April 2013,
Accepted 24th June 2013
DOI: 10.1039/c3cy00282a
www.rsc.org/catalysis
of fatty acids to fatty acid alkyl esters integral to current biodiesel
technologies.5 Furthermore, esterification pre-treatments of bio-
A Introduction
The application of heterogeneous catalysis1 and process intensifi-
cation2 in chemical synthesis are cornerstones of Green Chemistry.
Together, these offer improved use of raw materials and energy
efficiency, as well as safer operation through smaller system
footprints3 and reductions in work-up steps and associated waste
production. A particular challenge for industrial chemists is the
replacement of mineral and inorganic acids widely used in fine,
speciality and bulk chemical manufacturing processes, ranging
from alkylation, acylation, and dehydration to condensation and
esterification reactions. The esterification of organic acids is a
simple reaction, but one of great significance due to its widespread
application in the production of fine and speciality chemicals,
including the manufacture of intermediates for fragrances,
lubricants, fabrics and polymers.4 Esterification is also a key
step in biofuels production, with the acid catalysed conversion
oils derived from fast pyrolysis of biomass are a potential route
to improved oil stability via elimination of short chain acids
and deoxygenation.6 Tightening legislation on the emission of
hazardous pollutants produced through existing homogeneously
catalysed processes is stimulating the quest for alternative
heterogeneous routes, which offer several intrinsic advantages,
most notably ease of product separation and catalyst re-use, and
continuous flow versus batch operation.7
Various process engineering solutions have been considered for
continuous esterification, including the use of fixed bed processes
for fatty acids,8 microchannel-flow reactors,9 pervaporation
methods,10,11 or reactive distillation12–14 for acetic acid esterifica-
tion with butanol or ethanol. Continuous reactors must be carefully
designed to utilise the full potential of the associated hetero-
geneous catalyst; plug flow is a desirable reactor characteristic,
as it allows tighter control of product composition and thereby
reduces downstream separation processes and capital and
running costs. However, conventional plug flow reactors are
unsuitable for slow reactions as they necessitate very high
length:diameter ratios to achieve the required mixing. Such
designs are problematic due to their large footprints and
pumping duties, and control difficulties. The Oscillatory
Baffled Reactor (OBR) circumvents these problems, offering
good mixing and plug flow by oscillating the reaction fluid
through orifice plate baffles.15 In an OBR, mixing is decoupled
a School of Chemical Engineering and Advanced Materials, Faculty of Science,
Agriculture and Engineering, Newcastle University, Newcastle upon Tyne,
NE1 7RU, UK. E-mail: adam.harvey@ncl.ac.uk; Fax: +44 (0)191 222 5292;
Tel: +44 (0)191 222 7268
b Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Cardiff,
CF10 3AT, UK. E-mail: wilsonk5@Cardiff.ac.uk; Fax: +44 (0)29 208 74030;
Tel: +44 (0)29 208 70827
† Electronic supplementary information (ESI) available: Experimental setup,
details of catalyst preparation and characterisation and additional reaction
profiles. See DOI: 10.1039/c3cy00282a
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2373--2379 2373

View Article Online
Paper
Catalysis Science & Technology
from the net flow: this allows reactor designs that are very
different from conventional plug flow reactors. OBR mixing is
also scalable, meaning that the longer reaction times can be
achieved at industrial scales. Previous work16,17 has shown that
vortical mixing in the OBR is an effective, controllable method
of uniformly suspending solid particles. Such reactors have been
shown to be effective for homogeneously catalysed transesterifica-
tions employing NaOH, but surprisingly have never been exploited
for heterogeneous catalysis.18–22 Here we report the first such Fig. 1 Catalyst suspension in an OBR (a) without oscillation; and (b) with
oscillation of 8 mm amplitude and 4.5 Hz frequency (Reo = 2400).
demonstration, in which a solid acid catalyst is incorporated within
an OBR to facilitate the esterification of organic acids to short chain
and fatty methyl esters, pertinent to fine chemicals synthesis and
The oscillatory frequency (4.5 Hz) and amplitude (8 mm) corre-
biofuels production.
sponding to oscillatory Reynolds numbers (Reo) of 2490, 2400,
2330 and 2340 for the esterification of propanoic, hexanoic, lauric
and palmitic acids at 30 : 1 methanol to acid molar ratio, and 1934
for the hexanoic acid esterification at 6 : 1 methanol to acid molar
ratio were chosen in this work. The slight variations in the Reo at
similar oscillatory conditions were due to the differences in the
B Experimental
i. Catalyst synthesis and characterisation
Propylsulfonic acid derivatised SBA-15 (PrSO3H-SBA-15) silica was
prepared via grafting of mercaptopropyl trimethoxysilane onto
viscosities of the methanol–acid mixtures; the viscosities at 30 : 1
preformed SBA-15, and subsequent oxidation with 30% H2O2 as
methanol to acid molar ratio were 0.346, 0.358, 0.371 and
previously reported.23,24 Full details of the catalyst characterisation
0.372 mPa s for propanoic, hexanoic, lauric and palmitic acid,
are published elsewhere,24 and a summary of the physiochemical
properties are presented in Table S1 (ESI†).
respectively and 0.445 mPa s at 6 : 1 methanol to hexanoic
acid molar ratio. The acid conversions were independent of the
mixing intensity at the various oscillatory Reynolds numbers used
(Reo Z 676). Oscillatory mixing provided effective mixing of the
reaction solution and uniform catalyst suspension, as shown in
ii. Oscillatory baffled reactor
A schematic of the OBR is shown in Fig. S1 (ESI†). It comprises a
5 mm i.d. by 340 mm long integral baffled tube reactor, with
Fig. 1. Without an oscillatory flow catalyst particles settled near
2.5 mm diameter baffles equally spaced every 7.5 mm along the
tube length. The total reactor volume was B10 ml. The OBR was
mounted vertically inside a water bath (Ecoline, LUADA E100),
set at the desired temperature of 40, 50 or 60 1C. An output valve
the orifice constrictions along the integral baffled tube (Fig. 1a).
However, under oscillatory flow the catalyst remained uniformly
suspended (Fig. 1b) under conditions similar to those for solid
suspension in larger reactors.22,25
fitted with a 0.45 mm filter was connected at the top of the baffled
tube reactor through which samples were collected, whilst the
base of the reactor was connected to Confluent PVM syringe
pumps (Eurodyne Ltd) to provide the desired oscillation (fre-
oscillatory flow was performed using 5 wt% of a PrSO3H-SBA-15
quency and amplitude) for efficient fluid mixing and particle
suspension and net flow. The residence time distribution for the
which was suspended in 10 mL of anhydrous methanol (Sigma
reactor is shown in Fig. S2 (ESI†), which shows that high degrees
of plug flow behaviour can be achieved for this design. The use
Reaction commenced upon continuous dispensing of hexanoic
of oscillatory flow mixing to suspend particles has been reported
iii. Catalytic esterification
Hexanoic acid (Sigma Aldrich 99%) esterification under continuous
mesoporous solid acid catalyst (based on the organic acid mass),
Aldrich 99.8%) in the reactor at 60 1C.
acid at 0.113 mL minꢀ1 and methanol at 0.22 mL minꢀ1 into the
baffled reactor from reservoirs held at 60 1C, corresponding to net
flow Reynolds number (Ren) of 2.52 at the oscillatory Reynolds
number (Reo) of 1934. The esterifications employed a 6 : 1 molar
ratio of methanol to hexanoic acid, with 156.4 mg catalyst (5 wt%)
suspended in the baffled tube reactor. Three confluent PVM syringe
pumps (Eurodyne Ltd) were used: one providing oscillation and
the other two providing the net flows of methanol and hexanoic
acid into the reactor. The residence time (t) for the continuous
by various authors.22,25 The oscillation frequency was obtained
by adjusting the speed of the piston movement to 4.5 Hz whilst
the amplitude (8 mm) was determined by the amount of volume
dispensed. The net and oscillatory flow Reynolds numbers
(Ren and Reo) were calculated as shown in eqn (1) and (2):
Ren = ruD/m
(1)
Reo = 2pfoxorD/m
(2)
r: density of the reactant mixture (kg mꢀ3); u: reactants net flow esterification of the hexanoic acid at 6 : 1 molar ratio was 30 min.
(axial) velocity (m sꢀ1); m: dynamic viscosity of the reactant Steady-state continuous screening at ramped 30 min and 60 min
mixture (Pa s); D: internal diameter of the baffled tube (m); fo: residence (reaction) times was performed adopting the same
oscillatory frequency (Hz); xo: centre-to-peak amplitude of procedure but at a 30: 1 molar ratio of methanol to hexanoic
oscillation (m).
acid, with 82 mg of catalyst (10 wt%) suspended. The flows of the
When the reactor was operated in batch mode, oscillatory reactants into the reactor were 0.333 mL minꢀ1 (Ren of 3.24) at
mixing was induced without net flow through the reactor. 30 min residence time and 0.167 mL minꢀ1 (Ren of 1.62) at the
c
2374 Catal. Sci. Technol., 2013, 3, 2373--2379
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
60 min residence time. The estimated mixing intensity at the
30 : 1 molar ratio corresponds to the oscillatory Reynolds num-
bers (Reo) of 2490 due to the lower viscosity of the reaction
mixture. Approximately 0.2 mL of samples was collected at
various time intervals during the continuous esterification
experiments and stored in a freezer at ꢀ20 1C prior to analyses
with GC. Screening of process conditions and a range of other
C3–C16 organic acids was performed under batch conditions
using a slightly modified reactor set-up, in which the filter
syringe was replaced with an output valve for sampling. 5 wt%
PrSO3H-SBA-15 was suspended in a proportionate amount of
methanol (9.0, 8.5, 8.0 and 7.6 mL for propanoic (99.5%),
hexanoic (99%), lauric (98%) and palmitic acid (98%) respectively,
all from Sigma-Aldrich) to achieve a 30 : 1 molar ratio of methanol to
organic acid. The reactor was pre-heated to 60 1C before the desired
amount of acid (0.56 g, 0.82 g, 1.31 g and 1.62 g of propanoic,
hexanoic, lauric and palmitic acid respectively) was transferred into
the reactor from reservoirs heated at 60 1C (or 63 1C for palmitic
acid). Reactions were followed by periodically withdrawing B0.2 mL
of the reaction mixture through 1.0 mL syringe fitted with a 0.45 mm
syringe filter. The samples were similarly stored in a freezer
at ꢀ20 1C prior to the GC analyses. The same procedure was
subsequently used during hexanoic acid esterification to investigate
the effects of varying methanol : acid molar ratios, catalyst loading,
water addition and reaction temperature.
Fig. 2 Continuous esterification in OBR at 30 : 1 molar ratio of methanol : hexanoic
acid, 60 1C, Reo of 2400 (8 mm and 4.5 Hz), 10 wt% PrSO3H-SBA-15 and ramped
residence times (t) of 30 min (Ren = 3.24) and 60 min (Ren = 1.62).
Fig. 3 compares hexanoic acid esterification in the OBR with
that obtained during the initial stages of reaction within a
conventional stirred batch reactor. Excellent semi-quantitative
agreement is obtained for the first 120 min reaction, demonstrating
efficient mixing within the OBR and highlighting its potential
for heterogeneously catalysed liquid phase transformations.
The calculated TOF of 38 hꢀ1 is lower than that observed for
the same residence time (t = 30 min) in Fig. 2, reflecting the
lower methanol : acid ratio of 6 : 1 employed in this instance.
Fig. 3 also shows a decrease in methyl hexanoate production for
C Results and discussion
i. Continuous flow solid acid catalysed esterification
The feasibility of solid acid catalysed hexanoic acid esterification
with methanol under continuous flow was first investigated at
two residence times (t) of 30 min and 60 min in the OBR.
Hexanoic acid esterification at 30: 1 methanol to hexanoic acid
molar ratio, 60 1C and 10 wt% catalyst was run continuously at a
residence time of 30 min for 105 min (3.5t), before the residence
time was immediately changed to 60 min, for a period of
120 min (2t), by adjusting the net flows of the reactants.
Fig. 2 shows the cumulative turnover number (TON) during
reaction, initially with t = 30 min, from which it is apparent that
stable hexanoic acid conversion is achieved after B40 min, with
the associated Turnover Frequency (TOF) of 57 hꢀ1 comparing
favourably with the 31 hꢀ1 reported previously under stirred batch
conditions with the same catalyst.24 The TOFs were calculated as
the TON of the catalyst per unit time. TON is the number of moles
of the reactant (hexanoic acid) converted per mole of the catalytic
active sites, i.e. TOF = TON/time.
Doubling the residence time (i.e. t = 60 min) increased
hexanoic acid conversion from 15 to 20%, with the cumulative
TON continuing to increase linearly yielding a TOF of 38 hꢀ1
.
However, the conversion and TOF fell slightly against time
(running time) at this longer (60 min) residence time, attributed
to reversible water poisoning (see later). It should be noted that
the above reaction conditions were selected so as to ensure the
intrinsic reaction kinetics dominate any possible mass-transport
limitations; steady state conversions can be boosted simply by
raising the reaction temperature.
Fig. 3 Cumulative TON from the PrSO3H-SBA-15 catalysed continuous esterifi-
cation in OBR at 30 : 1 molar ratio of methanol : hexanoic acid, 60 1C, Reo of 2400
(8 mm and 4.5 Hz), 10 wt% and residence time (t) of 30 min (Ren = 3.24), as
compared to a stirred batch reactor at reaction conditions of 6 : 1 methanol to
acid molar ratio, 60 1C and 5 wt% catalyst.
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2373--2379 2375

View Article Online
Paper
Catalysis Science & Technology
the OBR after 120 min reaction (with the TOF falling to 25 hꢀ1),
indicating partial catalyst deactivation, although XPS of fresh
and hot-filtered spent catalysts (Table S1, ESI†) showed no loss
in surface S, hence leaching can be discounted. The most likely
origin of this deactivation is water accumulation at the acid
site, displacing the equilibrium towards the hexanoic acid
reactant. The effect of deliberately spiking reactions with a
controlled amount of water was thus subsequently investigated.
Water poisoning is a common problem in the H2SO4 and
SAC-13 catalysed esterification of acetic acid with methanol at
60 1C.26,27 The effect of water on acrylic acid esterification by
butanol over Cs2.5H0.5PW12O40, SO4–ZrO2, Amberlyst-15, Nafion-H
and H3PW12O40 has also been reported;28 the equilibrium
conversions decreased by 24%, 99%, 69%, 48%, and 42%
respectively upon addition of one equivalent of water.
Sulphated zirconia was the most severely impaired, probably
due to leaching of sulphate ions into the reaction mixture. Such
water poisoning is problematic in esterification, since water is
the major by-product. In the case of hydrophilic solid acid
catalysts, irreversible water adsorption at the active sites is
widely believed to cause deactivation, driving the search for
hydrophobic catalysts able to expel water from catalyst surface.29
A key advantage of the OBR is that it allows a wide range of
experimental parameters to be explored in a single experiment,
in contrast to batch reactions wherein e.g. separate experiments
must be conducted to compare initial rates under different
conditions. The impact of water on esterification was therefore
investigated for a 30: 1 methanol : acid ratio using 10 wt%
catalyst in order to maximise conversion, by injecting 6 mmol
H2O into the OBR after 150 min reaction. This volume equates to
the amount of water generated after a similar operating time
during a comparable PrSO3H-SBA-15 catalysed batch esterifica-
tion. Fig. 4a reveals that water addition impacted upon hexanoic
acid conversion, with the TOF falling from 52 hꢀ1 to 37 hꢀ1. In
contrast, a parallel spiking experiment using the same catalyst
under batch conditions within the OBR evidenced a 51% activity
loss (Fig. 4b). These results suggest on-stream deactivation
associated with the accumulation of reactively formed water is
less significant under continuous flow than under batch conditions.
We attribute the improved tolerance of PrSO3H-SBA-15 towards
high water content feed-streams within the OBR, to the continuous
addition of fresh hexanoic acid and methanol, assisting water
removal and helping to drive esterification.
Fig. 4 Impact of water on hexanoic acid esterification with methanol. (a) OBR
under continuous flow at a residence time (t) of 60 min, 30 : 1 methanol to acid
molar ratio, 10 wt% catalyst, Ren of 1.62 and Reo of 2400 (8 mm, 4.5 Hz)
and (b) stirred batch reactor at 30 : 1 methanol to acid molar ratio and 5 wt%
PrSO3H-SBA-15 catalyst.
reaction, however after 120 min, hexanoic conversion increased
with methanol concentration, reaching 30% conversion for the
30: 1 mix, but only 15% for the 1 : 1 mix, after 150 min reaction
time. This reflects the reversible nature of esterification, which
requires a large excess of alcohol (methanol) to drive the reaction
towards product formation (methyl hexanoate).
ii. OBR operation in batch mode
The option to operate an OBR under batch as well as continuous
flow conditions permits detailed kinetic studies. The effect of
catalyst mass on reaction rate was first investigated (Fig. S3, ESI†),
and revealed first-order kinetics, thus verifying the excellent OBR
mixing characteristics and absence of bulk diffusion limitations
for the short chain hexanoic acid.
RCOOH + CH3OH 2 RCOOH3 + H2O
(3)
The impact of MeOH : acid ratio on PrSO3H-SBA-15 catalysed
Fig. 5 also shows that hexanoic acid esterification begins to
hexanoic acid esterification was subsequently explored (Fig. 5), plateau after 180 min reaction at low methanol concentrations, with
and demonstrates the feasibility of operation at low MeOH a limiting conversion around 18%. The calculated TON (shown inset
concentrations. Acid conversion was constant for MeOH : acid after 3 h reaction) likewise increase with methanol : acid ratio
molar ratios between 1 : 1 and 30 : 1 during the first 60 min of as the equilibrium is displaced forward.
c
2376 Catal. Sci. Technol., 2013, 3, 2373--2379
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Good agreement is also observed with values reported for oleic
acid esterification with ethanol at a 120 : 1 ethanol/acid molar
ratio and 0.1 mol% SnCl2 catalyst,31 and esterification of
propanoic acid with n-amyl alcohol using Amberlyst-15.32 Liu
et al. reported activation energies of 53 and 52 kJ molꢀ1 for the
H2SO4 and SAC-13 catalysed esterification of acetic acid with
methanol, respectively.26
iii. Effect of carboxylic acid chain length – kinetic analysis:
Fig. 7 shows reaction profiles for the esterification of a series of
C3–C16 organic acids with methanol in the OBR. The first point
to note is how the initial rate decreases with increasing chain
length, in agreement with our previous reports.24 Pseudo first
order kinetic analysis of these reaction profiles was possible
due to the large excess of alcohol (30 : 1 methanol : acid).
The kinetic rate eqn (4) is as shown below:
ꢀrA = k1CA = k2CB0CA = k3CcatCB0CA (k1 = k2CB0) (k2 = k3Ccat
)
Fig. 5 Hexanoic acid esterification as a function of MeOH : acid molar ratio in an
OBR under batch operation at oscillations of 8 mm amplitude and 4.5 Hz
frequency, 60 1C, 5 wt% catalyst: (E) 30 : 1 (Reo = 2400), (m) 6 : 1 (Reo = 1934),
(’) 2 : 1 (Reo = 1441) and (K) 1 : 1 (Reo = 1063). Inset shows corresponding TON
after 180 min reaction.
(4)
dCA CA0dXA
¼
ꢀ
¼ k1CA0ð1 ꢀ XAÞ
(5)
dt
dt
Fig. 6 shows the impact of reaction temperature on the
esterification of hexanoic acid at 1 : 1 methanol/acid molar
ratio, revealing a strong dependence of conversion upon
reaction temperature. Corresponding initial rates (shown in
Table S2, ESI†) were used to calculate the apparent activation
energy for esterification as 46 ꢁ 3 kJ molꢀ1. This is in good
agreement with the work of Berrios and co-workers,30 who
reported activation energies of 51 kJ molꢀ1 for the esterification
of sunflower oil FFA (oleic acid) with methanol at a molar
ratio of methanol to acid of 60 : 1 and 5 wt% H2SO4 catalyst.
ꢀln(1 ꢀ XA) = k1t
(5a)
CA = concentration of acid/L molꢀ1; CB0 = original concen-
tration of the methanol/L molꢀ1; Ccat = catalyst loading in the
reactant mixture/g Lꢀ1; k3 = actual rate constant for esterifica-
tion/L molꢀ1 sꢀ1 g(cat) L.
ꢀ1
The pseudo-first-order rate constants (k1) for the esterifica-
tion of each acid at 30 : 1 methanol/acid molar ratios (Table 1)
were obtained from plots of ꢀln(1 ꢀ XA) versus time, and
were subsequently used to calculate the other rate constants
(k2 and k3).
Fig. 6 Effect of temperature on hexanoic acid esterification in the OBR under
batch operation at oscillations of 8 mm amplitude and 4.5 Hz frequency, 1 : 1
methanol to acid molar ratio, 5 wt% catalyst, and Reo of 760 at 40 1C, 886 at 0 1C
and 1063 at 60 1C.
Fig. 7 Effect of chain length on carboxylic acid conversion during batch esterifica-
tion in the OBR, at a molar ratio of methanol to acid of 30 : 1, reaction
temperature of 60 1C and 5 wt% PrSO3H-SBA-15 and oscillatory Reynolds
numbers (Reo) of 2490–2340.
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2373--2379 2377

View Article Online
Paper
Catalysis Science & Technology
Table
1
Rate constants for the RSO3H-SBA-15 catalysed esterification of
carboxylic acids with methanol in an OBR
Acid
reactant
Catalyst loading/
g(cat) Lꢀ1
k2/L
k3/L molꢀꢀ11
k1/sꢀ1
molꢀ1 sꢀ1
sꢀ1 g(cat)
L
Propanoic 2.83
Hexanoic 4.38
7.50 ꢂ 10ꢀ05 3.42 ꢂ 10ꢀ06 1.21 ꢂ 10ꢀ06
4.00 ꢂ 10ꢀ05 1.90 ꢂ 10ꢀ06 4.35 ꢂ 10ꢀ07
8.33 ꢂ 10ꢀ06 4.22 ꢂ 10ꢀ07 6.41 ꢂ 10ꢀ08
3.33 ꢂ 10ꢀ06 1.78 ꢂ 10ꢀ07 2.08 ꢂ 10ꢀ08
Lauric
6.58
Palmitic 8.55
30 : 1 MeOH : acid ratio, 5 wt% PrSO3H-SBA-15.
Table 1 also shows that the actual rate constants for the
reactions (k3) decrease with the molecular weight of the acid or the
length acid chain in accordance with batch reactor studies.24,33
Charton34,35 proposes that the rates of acid-catalyzed esterifica-
tion are solely a function of steric effects (6), in which n is the
van der Waals radii associated with a particular substituent
and c and h are constants.
Fig. 8 Plot of log10(TOF) versus organic acid chain length for esterification with
methanol in an OBR under batch operation, and in stirred batch reactors using
SAC-1333 and PrSO3H-SBA-1524 catalysts.
log10(rate) = cn + h
(6)
Liu et al. have shown that this modified form of the Taft
equation can be used to plot a linear relationship between
log10(TOF) and n for C1–C8 acids.33 However data on n for acids
>C8 is not available to extend this correlation. For a straight
chain acid it is reasonable to assume that chain length
or molecular weight will be representative of the steric bulk,
and thus may correlate with TOF. Fig. 8 plots log10(TOF) for
C3–C16 organic acids measured in the OBR versus chain length,
revealing the extension of this linear relationship across the
entire alkyl acid series.
It is particularly interesting to note the excellent agreement
between stirred batch results from Liu using the silica–
nafion composite SAC-13,33 and our own work employing
PrSO3H-SBA-15.24 This correlation demonstrates that (in the
absence of diffusion limitations), the rate of organic acid
esterification is a simple function of alkyl chain length (l),
independent of reactor configuration (7).
Taft parameter and rates for C2–C8 acids,37 however >C8 the
Taft parameter remains constant.35 Steric and bulk mixing
effects are expected to be more significant for longer chains,
wherein conformational changes are likely in order to minimise
the interfacial energy between the hydrophobic alkyl chain and
polar methanol solvent. For homogeneous catalysts, the impact
of such a change in fatty acid ‘bulkiness’ may be negligible,
however for a nanoporous catalyst it is likely that such conforma-
tional changes would hamper in-pore diffusion, leading to poorer
performance for a conventional mesoporous PrSO3H-SBA-15.
Our future work will thus address the use of pore-expanded
materials, and solid acids possessing hierarchical pore networks,
within the OBR; such factors significantly impact upon TOFs in
stirred batch reactors.
D Conclusions
log10(TOF) = ꢀ0.09l + 2.5
(7)
Here we report the first use of an OBR to conduct a hetero-
geneously catalysed reaction, namely solid acid catalysed
It is interesting to note that literature activation energies for organic acid esterification, widely employed in fine chemicals
H2SO4 catalysed esterification appear independent of acid and biofuel synthesis. Solid catalyst particles are uniformly
chain length,31,32,36 likewise illustrating the importance of suspended when an oscillatory flow is imposed upon the net
steric factors in controlling reaction rates. A general decrease flow. Operation under flow conditions enables variable residence
in esterification rate with alkyl chain length is of course times in a wide range of ratios of Reo to Ren to obtain plug flow,
expected, and previously reported in C2–C8 acid esterification without the need to alter reactor dimensions or compromising
catalysed by H2SO4 and SAC-13.33 Such decreases are attributed mixing. The present study on sulfonic acid catalysed esterifica-
to a combination of polar and steric influences of the alpha tion also illustrates the capability of OBRs for rapid and efficient
substituent on the carboxylic group. Inductive effects may continuous screening of solid catalysts, permitting detailed
reduce reactivity; although the carboxylate electron density kinetic studies and rapid optimisation of reaction conditions
increases with alkyl chain length, favouring initial protonation, in a single experiment. Continuous flow operation significantly
the electrophilicity of the carbonyl simultaneously decreases, thus lowers the detrimental effect of water upon equilibrium-limited
disfavouring subsequent alkoxy insertion.33 For chain lengths >C4, esterification, by preventing its accumulation in the catalyst
electronic properties are not believed to significantly change,38 zone, as occurs for conventional stirred batch reactors. Steady
hence subsequent variations in rate are attributed to steric or state reaction conditions can be readily optimised using meso-
mixing effects; the former predictable by the Taft equation. scale OBRs, thereby minimising feedstock usage and process
Previous research has shown a linear correlation between times. In summary, the enhanced characteristics of OBR versus
c
2378 Catal. Sci. Technol., 2013, 3, 2373--2379
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
17 M. E. Fabiyi and R. L. Skelton, Process Saf. Environ. Prot.,
2000, 78, 399–404.
batch reactors offers exciting possibilities for alternative clean
chemical processes employing heterogeneous catalysts.
18 A. P. Harvey, M. R. Mackley and T. Seliger, J. Chem. Technol.
Biotechnol., 2003, 78, 338–341.
19 A. N. Phan, A. P. Harvey and V. Eze, Chem. Eng. Technol.,
2012, 35, 1214–1220.
20 A. N. Phan, A. P. Harvey and M. Rawcliffe, Fuel Process.
Technol., 2011, 92, 1560–1567.
21 M. Zheng, R. L. Skelton and M. R. Mackley, Process Saf.
Environ. Prot., 2007, 85, 365–371.
Acknowledgements
We thank the EPSRC for funding (EP/F063423/2, EP/F067372/1
and EP/G007594/2). K.W. acknowledges The Royal Society for
the award of an Industry Fellowship and A.F.L. thanks the
EPSRC for the award of a Leadership Fellowship.
22 N. Reis, A. P. Harvey, M. R. Mackley, A. A. Vicente and
J. A. Teixeira, Chem. Eng. Res. Des., 2005, 83, 357–371.
Notes and references
1 P. T. Anastas, M. M. Kirchhoff and T. C. Williamson, Appl. 23 J. P. Dacquin, A. F. Lee, C. Pirez and K. Wilson, Chem.
Catal., A, 2001, 221, 3–13. Commun., 2012, 48, 212–214.
2 C. Jimenez-Gonzalez and D. J. C. Constable, Green Chemistry 24 C. Pirez, J.-M. Caderon, J.-P. Dacquin, A. F. Lee and
´
´
and Engineering: A Practical Design Approach, John Wiley &
Sons, 2011.
3 C. Ramshaw, Green Chem., 1999, 1, G15–G17.
K. Wilson, ACS Catal., 2012, 2, 1607–1614.
25 M. R. Mackley, K. B. Smith and N. P. Wise, Chem. Eng. Res.
Des., 1993, 71, 649–656.
4 H. E. Hoydonckx, D. E. De Vos, S. A. Chavan and P. A. Jacobs, 26 Y. Liu, E. Lotero and J. G. Goodwin Jr, J. Catal., 2006, 242,
Top. Catal., 2004, 27, 83–96. 278–286.
5 K. Wilson and A. F. Lee, Catal. Sci. Technol., 2012, 2, 884–897. 27 Y. Liu, E. Lotero and J. G. Goodwin Jr, J. Mol. Catal. A: Chem.,
6 J. J. Wang, J. Chang and J. A. Fan, Energy Fuels, 2010, 24,
3251–3255.
2006, 245, 132–140.
28 X. Chen, Z. Xu and T. Okuhara, Appl. Catal., A, 1999, 180,
7 C. Wiles and P. Watts, Green Chem., 2012, 14, 38–54.
261–269.
8 Y. Cheng, Y. Feng, Y. Ren, X. Liu, A. Gao, B. He, F. Yan and 29 J.-P. Dacquin, H. E. Cross, D. R. Brown, T. Duren,
J. Li, Bioresour. Technol., 2012, 113, 65–72.
9 A. A. Kulkarni, K.-P. Zeyer, T. Jacobs and A. Kienle, Ind. Eng.
Chem. Res., 2007, 46, 5271–5277.
J. J. Williams, A. F. Lee and K. Wilson, Green Chem., 2010,
12, 1383–1391.
´
´
30 M. Berrios, J. Siles, M. A. Martın and A. Martın, Fuel, 2007,
10 O. de la Iglesia, R. Mallada, M. Menendez and J. Coronas,
Chem. Eng. J., 2007, 131, 35–39.
86, 2383–2388.
31 A. Cardoso, S. Neves and M. Da Silva, Energies, 2008, 1,
11 S. Assabumrungrat, W. Kiatkittipong, P. Praserthdam and
S. Goto, Catal. Today, 2003, 79, 249–257.
12 C. Buchaly, P. Kreis and A. Gorak, Ind. Eng. Chem. Res., 2012,
51, 896–904.
13 I. K. Lai, Y.-C. Liu, C.-C. Yu, M.-J. Lee and H.-P. Huang,
Chem. Eng. Process., 2008, 47, 1831–1843.
79–92.
32 B. Erdem and M. Cebe, Korean J. Chem. Eng., 2006, 23,
896–901.
33 Y. Liu, E. Lotero and J. G. Goodwin Jr, J. Catal., 2006, 243,
221–228.
34 M. Charton, J. Am. Chem. Soc., 1975, 97, 1552–1556.
14 A. Singh, R. Hiwale, S. M. Mahajani, R. D. Gudi, J. Gangadwala 35 M. Charton, J. Org. Chem., 1976, 41, 2217–2220.
and A. Kienle, Ind. Eng. Chem. Res., 2005, 44, 3042–3052.
15 X. Ni, M. R. Mackley, A. P. Harvey, P. Stonestreet,
36 A. Cardoso, R. Augusti and M. Da Silva, J. Am. Oil Chem. Soc.,
2008, 85, 555–560.
¨
M. H. I. Baird and N. V. R. Rao, Chem. Eng. Res. Des., 37 J. Lilja, D. Y. Murzin, T. Salmi, J. Aumo, P. Maki-Arvela and
2003, 81, 373–383. M. Sundell, J. Mol. Catal. A: Chem., 2002, 182–183, 555–563.
16 A. P. Harvey, M. R. Mackley and P. Stonestreet, Ind. Eng. 38 H. Fujimoto, Y. Mizutani, J. Endo and Y. Jinbu, J. Org.
Chem. Res., 2001, 40, 5371–5377.
Chem., 1989, 54, 2568–2573.
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol., 2013, 3, 2373--2379 2379