Pressure-accelerated azide-alkyne cycloaddition: Micro capillary versus autoclave reactor performance

Article abstract of DOI:10.1002/cssc.201403034

Pressure effects on regioselectivity and yield of cycloaddition reactions have been shown to exist. Nevertheless, high pressure synthetic applications with subsequent benefits in the production of natural products are limited by the general availability o

Full text of DOI:10.1002/cssc.201403034

DOI: 10.1002/cssc.201403034
Full Papers
Pressure-Accelerated Azide–Alkyne Cycloaddition: Micro
Capillary versus Autoclave Reactor Performance
Svetlana Borukhova,[a] Andreas D. Seeger,[b] Timothy Noꢀl,[a] Qi Wang,[a] Markus Busch,[b] and
Volker Hessel*[a]
Pressure effects on regioselectivity and yield of cycloaddition
reactions have been shown to exist. Nevertheless, high pres-
sure synthetic applications with subsequent benefits in the
production of natural products are limited by the general avail-
ability of the equipment. In addition, the virtues and limita-
tions of microflow equipment under standard conditions are
well established. Herein, we apply novel-process-window
(NPWs) principles, such as intensification of intrinsic kinetics of
a reaction using high temperature, pressure, and concentra-
tion, on azide–alkyne cycloaddition towards synthesis of Rufi-
namide precursor. We applied three main activation methods
(i.e., uncatalyzed batch, uncatalyzed flow, and catalyzed flow)
on uncatalyzed and catalyzed azide–alkyne cycloaddition. We
compare the performance of two reactors, a specialized auto-
clave batch reactor for high-pressure operation up to 1800 bar
and a capillary flow reactor (up to 400 bar). A differentiated
and comprehensive picture is given for the two reactors and
the three methods of activation. Reaction speedup and conse-
quent increases in space–time yields is achieved, while the pro-
cess window for favorable operation to selectively produce Ru-
finamide precursor in good yields is widened. The best condi-
tions thus determined are applied to several azide–alkyne cy-
cloadditions to widen the scope of the presented methodolo-
gy.
Introduction
High temperature, pressure, concentration, and/or addition of
a catalyst can maximize the reaction kinetics of most chemical
reactions.[1] Provided that the activation energy of a particular
reaction is positive (Ea >0), its reaction rate constant will in-
crease at higher temperatures. Similarly, as long as the activa-
perature processing in microflow reactors resulted in numer-
ous successful cases,[6] high-pressure operations, having a po-
tential of accelerating reactions and directing selectivity, still
constitute to a mystery nowadays. Up to now, pressure effects
in microflow have been observed (i) under sub- or supercritical
conditions and (ii) in enhancement of interfacial mass transfer
for gas–liquid reactions.[7] In the case of pressure impact on
a reaction with a negative activation volume as envisaged in
this study, notable impact of the capillaries or microchips use
was mostly achieved under non-continuous, stop-flow condi-
tions.[7b]
¼
tion volume (DV <0) of a reaction is negative, increasing
pressure will facilitate the reaction.[2] Thus, application of rela-
tively high temperature and pressure on a reactive system may
lead to a chemical intensification.[3] Novel process windows
(NPWs) is a concept that embraces the opportunities present-
ed by chemical and process-design intensification.[3] Reducing
the size of equipment to a microscale leads to process intensi-
fication because of enhanced heat and mass transfer.[4] Contin-
uous microflow operation is a very desirable method in phar-
maceutical industry because of a possible high degree of con-
trol over reaction parameters, a higher safety, a reduced
manual handling, a flexibility of production volume, an easier
reproducibility, a possibility of reaction telescoping, and a po-
tential for an integrated purification.[5] Even though high-tem-
Most of the powerful high-pressure installations can sustain
a pressure of up to 3 GPa (30000 bar) and are based on a dia-
mond anvil cell or piston–cylinder-type reactors.[8] Such reac-
tors supply new information on physical properties of matter
as well as on reaction dynamics and mechanism. High fabrica-
tion cost, limited volume, high energy cost, and constrained
safety are, however, the main reasons for not being widely ap-
plicable on a larger scale. To the best of our knowledge indus-
trially used high pressure reactors have a range of 5–2000 bar
and require special precautions, advanced safety control instal-
lation, regular check of tightened areas for leaks, and constant
monitoring during the operation. In microflow, high pressure
applications fall into a much narrower space of only 5–
600 bar.[7e,9] A few hundred bars can be applied and sustained
within a microflow reactor of any desired volume by using
HPLC pumps equipped with an appropriate back-pressure reg-
ulator. It is expected that the handling of flow reactors at high
pressures is generally easier.[7b] Flow reactors allow high pres-
sures to be easily reached because of the smaller internal di-
[a] S. Borukhova, Dr. T. Noꢀl, Q. Wang, Prof. Dr. V. Hessel
Department of Chemical Engineering and Chemistry
Micro Flow Chemistry and Process Technology
Eindhoven University of Technology
Den Dolech 2, 5612AZ, Eindhoven (The Netherlands)
E-mail: v.hessel@tue.nl
[b] Dr. A. D. Seeger, Prof. Dr. M. Busch
Technische Chemie III
Technische Universitꢁt Darmstadt
Alarich-Weiss-Straße 8, 64287 Darmstadt (Germany)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201403034.
ChemSusChem 0000, 00, 0 – 0
1
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
mensions and smaller overall size, low numbers of tightening
areas, and variable volume for production.
copper, serves as one of the best examples of a ‘perfect’ reac-
tion termed as ‘click’ reaction.[15] In the last decade, click reac-
tion became a synthetic tool with a special emphasis on the
use of combinatorial chemistry to yield natural products on
the way of drug development.[16] Copper-catalyzed cycloaddi-
tion of alkyl azides and terminal alkynes results in 1,4-substitut-
ed 1,2,3-triazole, which is the building block of many natural
products.[17] One of the bestselling 200 drugs of recent years is
an antiepileptic drug, 1,2,3-triazole-
As mentioned above, reactions with negative activation
volume constitute a class of reactions facilitated by high pres-
sure.[2] Thus, cycloaddition and condensation reactions, reac-
tions proceeding via cyclic transition state (such as Cope and
Claisen rearrangements), reactions involving the formation of
dipolar transition states (such as electrophilic aromatic substi-
tutions), and reactions with a steric hindrance can be influ-
enced by high pressure. Moreover, based on the difference in
volume occupied by a product, the distribution of reaction
products can be altered. One of the most extensively studied
class of reactions under high pressure are [4+2] Diels–Alder cy-
cloadditions because of their wide application in general and
their change in activation volume being the second most neg-
ative (ꢀ25 to ꢀ50 mLmolꢀ1).[10] High pressure was shown to
direct regioselectivity of cycloaddition due to the difference in
volume of regioisomers, changes in electronic demand, and
steric hindrance.[11] Moreover, the combination of catalysis and
high pressure was demonstrated to have a synergistic effect
when a Lewis acids was used to catalyze the cycloaddition of
a pyrrole derivative with an electron-rich diene.[12] Finally, high
pressure affects the reaction medium by affecting its physical
properties, such as boiling and melting points, density, viscosi-
ty, dielectric constant, compressibility, conductivity, and surface
tension;[2,8a] however, this is out of the scope of the present
study.
Rufinamide (Scheme 1).[18] The pro-
duction process was initially devel-
oped by Novartis and is now realiz-
ed by Eisai Ltd. under the commer-
cial names Inovelon and Banzel.
Scheme 1. Rufinamide struc-
The anticonvulsant is used in the
treatment of seizures associated
with the Lennox–Gastaut syn-
ture.
drome of patients older than 4 years.[19] We have previously
published a study on the combination of high concentration
and high temperature, where relatively unreactive enol ether
was used to synthesize the crystalline Rufinamide precursor
under solvent-free conditions in a microcapillary reactor.[6a]
Synthesis of Rufinamide completely based on continuous flow
starting from 2,6-difluorobenzyl bromide and methyl propio-
late was recently reported by Jamison et al.[20]
Herein, we focus on the optimization of the 1,3-dipolar cy-
cloaddition of 2,6-dilfuorobenzyl azide and methyl propiolate,
which leads to the 4-substituted 1,2,3-triazole, Rufinamide pre-
cursor (Scheme 2). Separation of 1,5-cycloadduct is a require-
Several reactions have been performed in microreactors
under high pressure and stop–flow regime. The nucleophilic
aromatic substitution reaction of p-halonitrobenzenes with
cyclic amines has been investigated in a microcapillary under
batch conditions at pressures up to 600 bar.[7b] Rate enhance-
ments by a factor of 2.7, 1.7, and 1.5 were observed for pyrroli-
dine, piperidine, and morpholine, respectively. The Diels–Alder
reaction of 2- and 3-furylmethanol with maleimides, performed
under elevated pressure, demonstrated that high pressure in-
creases the rate of the 2-furylmethanol reaction with malei-
mides, which is less reactive than 3-furylmethanol under at-
mospheric conditions. A larger negative change in the reaction
volume of the formation of the exo product in comparison to
the endo product resulted in a slight increase in the amount of
exo product formed. An increase of the reaction rate of the
Diels–Alder reaction of cyclopentadiene with phenylmaleimide
by a factor of 14 was observed upon increasing the pressure
to 150 bar in a high-pressure glass microreactor. Razzaq et al[6e]
and Tilstam et al.[13] reported multiple high-pressure, high-tem-
perature acceleration of reaction rates of Newman–Kwart and
Claisen rearrangements, a Fischer indole synthesis, and nucleo-
philic substitution.
Scheme 2. 1,3-dipolar cycloaddition to Rufinamide precursor.
ment in an industrially applied process when performed with-
out a copper catalyst. We investigate the effect of pressure on
the regioselectivity to maximize the yield of the desired 1,4-cy-
cloadduct. Moreover, we look into the effect of the synergy be-
tween high pressure and catalyst on the reaction outcome.
Additionally, the performance of the high-pressure autoclave
reactor and the flow reactor, two specialized apparatuses built
for the current study, are compared in the light of the herein
mentioned advantages and disadvantages. Finally, the best
thus determined flow conditions are applied to a wider scope
of azide–alkyne cycloadditions.
Because of their lower activation volumes, [3+2] Huisgen cy-
cloadditions are less studied under high pressure.[14] [3+2]
Huisgen cycloaddition takes place when 1,3-dipole reacts with
a dipolarophile to form five-membered cyclic compounds.
Azides are a class of 1,3-dipoles, and their reaction with termi-
nal alkynes results in a mixture of 1,4- and 1,5-cycloadducts,
unless selectivity is directed by a catalyst in favor of a single
cycloadduct.[14a] Azide–alkyne cycloaddition, when catalyzed by
Results and Discussion
The reaction of interest was studied using three systems differ-
ent in operating pressure and temperature limits as well as in
operation method. Details on the reaction methods can be
found in the Experimental Section. We performed our investi-
gations in three stages:
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
2
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
1. Batch experiment in a stirred glass round bottom flask
under uncatalyzed conditions.
plication factors of subsequent reaction rates lead to a calcula-
tion of an activation volume of ꢀ21.13 mLmolꢀ1, which is con-
sistent with published results for 1,3- dipolar cycloadditions.[2b]
Calculations of activation volume are described in the Support-
ing Information. Switching our focus to the temperature effect
we performed the reaction at 1200 bar in batch reactor again
for 24 h, under the same conditions). Lower yields due to the
decomposition of 2,6-difluorobenzyl azide were obtained at
higher temperatures, that is, 66% and 2% at 175 and 2508C,
respectively.
2. High-pressure and -temperature (HPHT) experiments in
a non-stirred autoclave reactor under uncatalyzed condi-
tions.
3. HPHT experiments in a HPHT flow reactor under uncata-
lyzed and catalyzed conditions.
Pressure and temperature windows in standard and auto-
clave batch reactors
We performed the cycloaddition in a stirred round bottom
glass flask at 908C and 0.25m of 2,6-difluorobenzyl azide in N-
methyl-2-pyrrolidone (NMP). After 24 h 55%yield of 1,4-cyclo-
adduct was obtained, giving rise to 60% yield in 6 days. The
product distribution remained the same throughout the reac-
tion with the ratio of the 1,4-/1,5-cycloadducts of 2.8:1. To de-
termine the activation volume and pressure effect on 1,3-dipo-
lar cycloaddition of 2,6-difluorobenzyl azide to methyl propio-
late, we performed high-pressure experiments in the non-
stirred autoclave batch reactor. Performing the reaction at
500 bar and otherwise same conditions resulted in a yield of
72% of the 1,4-cycloadduct with increased preference to the
desired product, giving a ratio of 3.6:1 for the 1,4- and 1,5-cy-
cloadducts, as shown in Figure 1. Increasing pressure further
resulted in a close to linear increase in yield. At 1800 bar, the
upper limit of our investigations, a final yield of 84% of the
1,4-cycloadduct, was achieved. Thus, an overall increase of
30% in yield of the 1,4-cycloadduct was observed when com-
pared to stirred batch conditions under atmospheric pressure.
Regioselectivity increased in favor of the 1,4-cycloadduct with
a 1,4-/1,5-cycloadduct regioisomeric ratio from 2.8 at atmos-
pheric conditions to 6.3 at 1800 bar. Taking the reaction rate
constant at 500 bar as reference and calculating relative multi-
Pressure, temperature, and concentration windows in micro-
capillary reactor
Based on the availability of equipment, namely pumps and
back-pressure regulators (BPRs), the maximum pressure reacha-
ble in our home-built microflow setup is 400 bar. Experience
with batch experiments showed that reaction kinetics at rela-
tively low temperature is relatively slow. Although possible,
long reaction times in flow are not desirable. Thus, the uncata-
lyzed azide–alkyne cycloaddition under interest was studied in
the HPHT microcapillary flow setup at a shorter reaction time
than in batch. A residence time of 30 min and pressures up to
400 bar were allowed for the reaction at 908C. Under atmos-
pheric pressure and same concentration of 0.25m as in the
batch experiment, 48% yield of 1,4-cycloadduct was obtained
in 30 min. An increase of pressure to 400 bar resulted in 58%
yield of the desired regioisomer. No significant change in prod-
uct distribution was observed with increasing pressure: the
product distribution increased from 3.5 at 1 bar to 3.6 at
400 bar. Increasing the temperature showed that the highest
yield of 80% of the 1,4-cycloadduct under the given the condi-
tions could be obtained at 1408C and 400 bar.
Next, we investigated the effect of residence time at various
reaction temperatures. Figure 2 shows that the time effect is
diminished at temperatures higher than 1408C, resulting in
a full conversion of 2,6-difluorobenzyl azide at all investigated
residence times. Decrease in the yield is observed at tempera-
tures higher than 1408C and residence times of more than
10 min. This observation can be explained by the decomposi-
tion of 2,6-difluorobenzyl azide at high temperatures and
longer contact times, similar to our previous study.[3b]
Reactions performed at higher concentrations have been
proven to proceed in a safer manner in flow than in batch,
with an additional benefit of faster kinetics. In this case, how-
ever, the temperature was kept constant at 908C for the sake
of comparison with previously performed batch experiments
and reducing the competing pressure and temperature effects.
Increasing the concentration of 2,6-difluorobenzyl azide to 0.5
and 1.0m was possible and speeded up the kinetics as shown
in Figure 3. Throughout the whole investigated pressure range,
an increase in yield of approximately 10% was found for each
concentration. The highest yield of 81% of the desired 1,4-cy-
cloadduct was obtained at 1.0 molLꢀ1, 908C, and 400 bar at
30 min residence time.
Figure 1. Pressure effect on the yield of uncatalysed azide–alkyne cycloaddi-
tion of 2,6-difluorobenzyl azide and methyl propiolate (2 equiv) when ex-
periments were performed in high pressure autoclave reactor at 908C,
0.25m for 24 h. Blue line represents yield ratio of 1,4- to 1,5 cycloadduct.
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
3
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
lary-based flow system. We performed the copper-catalyzed
cycloaddition by using 1.0 mol% of the homogeneous copper
catalyst (1,10-phenanthroline)bis(triphenylphosphine)–copper(I)
nitrate dichloromethane adduct ([Cu(phen)(PPh3)2]NO3). The
choice of the catalyst was based on its recently exhibited supe-
rior performance[21] resulting in 96% yield of the 1,4-cycload-
duct in 3 min when phenyl acetylene and phenyl azide reacted
under solvent-free conditions at room temperature in batch.
We performed cycloaddition of methyl propiolate with 2,6-di-
fluorobenzyl azide in NMP (0.25m) at 908C. Only 5% of 1,4-cy-
cloadduct was obtained after 90 min, whereas the formation of
1,5-cycloadduct was not observed. To perform the reaction in
flow, an extra pump was used to introduce the copper catalyst
into the stream of 2,6-difluorobenzyl azide before mixing with
methyl propiolate. Premixing was reported to result in the for-
mation of copper–alkyne aggregated complexes.[22] Methyl
propiolate was introduced into the mixed stream through
a High Pressure IMM mixer (with Reynolds number of 55) con-
structed for high pressure and based on flow lamination/hy-
drodynamic focusing (Figure 4). Increasing the concentration
to 0.5m under otherwise same conditions with no pressure ap-
plied and 1 min residence time, 7% yield of 1,4-cycloadduct
was obtained. We studied the reaction with increased azide
concentration of 0.5m in the same range of process conditions
as in the uncatalyzed case. Figure 5 shows rapid decrease in re-
gioselectivity with increasing temperature, which speeds up
the reaction as demonstrated by the increase of the 1,4-cyclo-
adduct yield. The highest yield obtained was 77% of the de-
sired cycloadduct, with a 1,4-/1,5-cycloadduct ratio of 4.2 at
1608C in 5 min and in the presence of the copper catalyst.
A slight increase in yield due to pressure is observed at moder-
ate temperatures, which is less pronounced than for the unca-
talyzed reaction in flow.
Figure 2. Temperature effect on the yield of 1,4-cycloadduct in [3+2] cyclo-
addition, when performed at 400 bar, 0.25m and 5, 10, 20, and 30 min resi-
dence time.
Figure 6 gives an overview of all the findings reported
above with the boundaries of the process windows resulting
in higher than 70% yield of the desired 1,4-cycloadduct. The
pressure–temperature window for the uncatalyzed autoclave
reactor is large, showing the good potential of this reactor and
in general the wide flexibility commonly acknowledged for
batch operation. It allows operation in pressure regions in
which a flow reactor cannot be used at this point of time due
to technological limitations. Moreover, a change in regioselec-
tivity in favor of the desired regioisomer can be achieved. Yet,
long processing times are needed, which results in a drop in
productivity, thus requiring also more energy per given unit of
manufactured product. Here, the much shorter processing
times of the uncatalyzed flow reactor provide a good alterna-
tive, even more if safety under high pressure and energy mini-
mization is an issue: activation by pressure is somewhat help-
ful, yet the big activation boost comes from the temperature
and concentration flexibility. However, the small pressure effect
is not intrinsic to flow, but the pressure range utilizable simply
is lower for flow reactors at this time of technological develop-
ment.
Figure 3. Pressure and concentration effect on yield of 1,4-substituted 1,2,3-
triazole at different concentrations (0.25, 0.5, and 1.0 molLꢀ1) for experi-
ments performed at 1–400 bar in a high pressure flow setup at 908C and
30 min reaction time.
Concentration, pressure, and temperature windows in mi-
crocapillary reactor with catalyst
As mentioned in the Introduction, combination of catalysis and
high pressure may have a synergistic effect.[12] When catalyzed
by copper the azide–alkyne cycloaddition constitutes a class of
Click reactions. According to mechanistic studies, copper di-
rects the regioselectivity of cycloaddition. The directing power
of copper varies for each different combination of catalyst and
reactants based on their individual reactivity.[15] Thus, 100% re-
gioselective cycloaddition is not always guaranteed. To test the
merits of reaction activation and regioselectivity control by
a catalyst, we tested a copper catalyst in a HPHT microcapil-
Figure 6 also shows the process window coordinates for the
best yields obtained for the three processing types. The activa-
tion in the flow reactor is mostly temperature based and pres-
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
4
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 4. HPHT microcapillary-based flow system with added stream of homogeneous [Cu(phen)(PPh3)2]NO3 as catalyst. The setup consists of a stainless steel
(SS) capillary, three HPLC pumps, T- and high-pressure IMM mixers, heating and cooling oil baths, sample loop connected to a six-port valve, and one BPR.
sure-based for the autoclave reactor. The uncatalyzed reaction
presents a case where both modes are used and this presents
the core information of this paper. An additional difference is
the time needed to achieve the yield, a 48-fold reduction in
residence time is given for the uncatalyzed flow reactor com-
pared to the uncatalyzed autoclave reactor.
Use of a catalyst is justified at lower temperatures and
longer residence times as only 1,4-cycloadduct is formed. Re-
gioselectivity is rapidly reduced as the temperature increases,
being highest at 258C when no formation of 1,5-cycloadduct is
detected; the ratio of the 1,4-/1,5-cycloadduct is reduced to 64
at 608C and rapidly falls to 4.6 at 1208C. An overview of the
best conditions for flow in terms of yield of the 1,4-cycload-
duct at 1608C and 400 bar for 5 min shows that the merit of
using a catalyst is an increase of only 1% in the yield of the
desired product. The use of catalyst adds an additional down-
stream operation within the production process. Separation of
toxic metal is required prior to the final stages of pharmaceuti-
cals’ production. It is evident that the flow operation demands
the use of higher temperatures to achieve best yields.
Figure 5. Temperature effect on the yield of 1,4-cycloadduct and regioselec-
tivity expressed as a ratio of 1,4- to 1,5-cycloadduct in the presence of
[Cu(phen)(PPh3)2]NO3 as catalyst at various temperatures (60–2008C) and
pressures (1–400 bar).
Although our primary synthetic target was the Rufinamide
precursor, we applied the best conditions (1408C and 400 bar)
to other substrates. The results shown in Table 1 imply that the
more electron-deficient the dipolarophile is, the higher is its
activity towards 1,3-dipolar cycloaddition. The opposite is true
for the dipole: azidobenzene is the most active among the in-
vestigated azides because of the higher conjugation and
higher electron density over the azide dipole.
Conclusions and Outlook
Novel process windows (NPW) principles were applied to the
1,3-dipolar cycloaddition to yield the Rufinamide precursor.
The conditions for the activation of the reaction and the regio-
selectivity towards the 1,4-cycloadduct, the desired precursor,
served as two foci of the study. Concerning both aspects,
merits of high pressure, high temperature, high concentration,
Figure 6. Process windows for the yields of desired regioisomer above 70%
in high pressure autoclave reactor, uncatalyzed, and catalyzed processes in
flow reactor.
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
5
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
methods resulted in a yield of
81% of the desired 1,4-cycload-
duct in 30 min.
Table 1. 1,2,3-triazoles synthesized under catalyst-free conditions in microcapillary flow reactor in 30 min resi-
dence time at 1408C and either 10 or 400 bar.[a]
Entry
Azide
Alkyne
Product
Total yield[b] [%]
1,4-/1,5-cycloadduct
Copper-based catalysis was in-
vestigated as an additional acti-
vation method for the cycloaddi-
tion in the flow reactor. 76–77%
yield at various individual tem-
perature–pressure combinations
for 5 min was obtained. Thus, al-
though not optimized, the inves-
tigated process window showed
that a fivefold acceleration is
possible when compared with
uncatalyzed flow operation.
However, formation of the unde-
sired 1,5-cycloadduct occurred
even in the presence of the cata-
lyst.
(1,4-regioisomer)
10 bar
400 bar
10 bar
400 bar
1
99
99
2.5
2.5
2
3
50
46
62
59
1.2
1.1
1.4
0.9
4
5
6
99
58
72
99
65
78
3.6
1.1
1.5
3.5
1.3
1.4
7
41
49
2.1
2.1
To provide
a differentiated,
comprehensive picture, process
windows maps of favorable op-
eration conditions (pure 1,4 iso-
meric yield >70%) are given.
The best conditions were ap-
plied to a wider selection of
azides and alkynes.
8
99
33
23
99
42
24
3.2
1.4
1.5
2.1
1.8
1.5
9
10
The results obtained in this
paper have also provided some
insight into the often claimed
easiness of pressure operation
[a] Reaction conditions: azide (1 equiv, 0.25m), alkyne (ene) (2 equiv, 0.5m) in NMP at 1408C for 30 min resi-
dence time. [b] Yields were calculated by using H NMR spectroscopy with the use of 1,3,5-trimethoxy benzene
as internal standard.
1
using
microflow
reactors.
Indeed, data collection was
much faster with the microflow
and catalyst were compared (four out of six NPW conditions as
given in Ref. [3]). In addition, a comparison was made between
a home-built high pressure autoclave and a high pressure and
high temperature microcapillary flow reactors. The reaction
was on the order of several days when carried out in a round
bottom flask under atmospheric conditions, resulting in 60%
yield after 6 days. In contrast, when the high-pressure auto-
clave reactor was used, speedup of the reaction kinetics and
improvement in regioselectivity were observed. Increasing the
concentration could speed up the reaction further; however,
due to the possible decomposition of the azide and subse-
quent pressure buildup we did not pursue this option, which
also would never be industrially viable. Use of such a reactor
on production scale even within given limits is still questiona-
ble due to the fabrication costs and safety issues.
setup because of faster heating, shorter reaction times, and
easier sampling due to the installed sample loop. In addition,
the much larger volume used for the autoclave reactor restrict-
ed exploration of temperatures above a certain limit. Thus, the
range of information, with regard to the expansion of the pro-
cess windows, was better for the microflow reactor. The combi-
nation of catalysis and harsh conditions was slightly advanta-
geous when compared to the uncatalyzed process, where the
presence of a catalyst required extra downstream operation,
thus the overall benefit is higher when no catalyst is used with
harsh conditions.
The consequences of the process design of our reaction in-
tensification will be reported in a separate paper, including
cost analysis and life-cycle assessment: Process-design intensi-
fication in flow improves cost, sustainability, and energy.[3,23]
The design of continuous microflow-based processes enables
either the use of new types of process integration or process
simplification. Maximum impact is typically gained for entirely
new chemical transformations only realizable in flow. Those,
when combined in sequence as in multi-step syntheses, aim at
compactness and, therefore, bring special attention to slow re-
actions that need to be intensified. In our case, the cycloaddi-
The flow reactor, on the other hand, can be scaled up by in-
creasing the number of reactors and smart scale-out (small
widening of characteristic dimensions) without a major loss of
performance. Operation in a capillary flow reactor allowed to
increase temperatures up to the superheated range, making
use of pressures up to 400 bar and increase the reactants’ con-
centrations. The combined action of these three activation
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
6
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
a higher pressure. One of the pumps was auxiliary and was used
to fill the syringe pump and the reactor with a reacting mixture
whereas the syringe pump was used to apply pressure. The pres-
sure was measured by a transducer located between the syringe
pump and the reactor. A rupture disc with the upper limit of
2500 bar was inserted at another outlet of the reactor as a barrier
with an emergency outlet. In case of a pressure build up higher
than 2500 bar, the reactor contents would be sucked in from the
reactor interior, thus minimizing the risk of the experiments. The
components of the setup were connected as shown in Fig-
ure 7right. Compounds bearing an azide group should be handled
with caution. In a previous study we studied the decomposition of
2,6-difluorobenzyl azide using differential scanning calorimetry.[6a]
To avoid any incidents caused by a possible nitrogen evolution
upon decomposition of organic azide in the autoclave reactor ,we
decided to keep the upper limit of concentration at 0.25m.
tion to synthesize the 1,2,3-triazole precursor of Rufinamide
constitutes such a reaction.
A key to process-design intensification is a high space–time
yield, allowing the preparation of very compact reactors, which
facilitates or even enables system integration in electronics in-
dustry. Based on this motivation we calculated space–time
yields for the three types of operation (and two types of reac-
tors) investigated (Table 2). There is a never-completed discus-
Table 2. Space–time yield for the three types of operation (and two
types of reactors).
Operation type
Space–time yield [molLꢀ1 hꢀ1
]
uncatalyzed autoclave reactor
uncatalyzed flow reactor
catalyzed flow reactor
0.0084
4.56
4.62
The third platform was a HPHT microcapillary-based flow system
(Figure 8), with operating limits of 400 bar and 3008C. The setup
consisted of a SS capillary (500 mm inner diameter, 10 m long), two
HPLC pumps (Knauer 1000 series), heating and cooling oil baths
(Lauda), sample loop connected to a six-port valve (Vici Valco), and
one Bronkhorst BPR. HPLC pumps could be operated at pressures
up to 400 bar whereas the BPRs kept the system at the set pres-
sure. The SS microcapillary reactor was heated in the oil bath with
upper temperature limit of 3008C. A second bath was used to cool
the reacting mixture with the goal of quenching the reaction and
preparing the stream for safe sample collection. The valve was
used as a sample collection loop so that the dead volume of the
BPRs would not contribute to the residence time of reactants. The
BPR was selected to be combined with a control valve applicable
at relatively low flow rates, 2–20 mLminꢀ1. The residence time was
manipulated by changing the flow rate and length of the capillary
tubing. Due to the lower internal volume (2 mL) and faster proc-
essing, 1.00m concentration of 2,6-difluorobenzyl azide was kept
as an upper limit. Thus, the existing p–T process window of the
standard batch reactor could be considerably widened by the use
of more advanced reactors, such as autoclave and microflow reac-
tors.
sion among the microreactor engineering community whether
the inner or outer volume of the reactor is to be taken as refer-
ence. We used the inner volume (i.e., the fluid reaction
volume) as reference to calculate the space–time yield. It is evi-
dent in both projections that the flow reactor is more produc-
tive than the autoclave reactor. The autoclave reactor, although
offering a higher activation by pressure, has a limited and low
volume. Naturally, our data are lab based and production reac-
tors will behave somewhat different; however, this will not
change the overall message. The compactness of the flow re-
actor resembles Ramshaw’s first definition of process intensifi-
cation: ’to shrink down the plant’.[24]
Experimental Section
Operating platforms—Process windows of reactors under
study
The reaction of interest was studied using three systems different
in operating pressure and temperature limits, as well as in opera-
tion method. The first system was the standard round bottom flask
under atmospheric conditions that is limited in terms of tempera-
ture by the boiling point of the
Synthesis
Methyl 1-(2,6-difluorobenzyl)-1H-1,2,3-triazole-4-carboxylate in
high pressure autoclave reactor: 2,6-difluorobenzyl azide (2.1 g,
solvent.
The second system was a high
pressure autoclave reactor setup
with operating limits of 2500 bar
and 3008C. The 14 mL reactor was
constructed according to the sche-
matic representation shown in Fig-
ure 7left. No copper or ruthenium-
containing material was used in
the construction of the reactor.
Two heating jackets surrounded
the upper and lower halves of the
reactor. A thermocouple was in-
serted into one of the four inlets,
Figure 7. Left: Autoclave reactor with 14 mL internal volume and operating limits of 2500 bar and 3008C. Four
outlets: inlet, sample collection outlet, thermocouple, and connection to the safety line with rupture disk with-
standing 2500 bar. Right: Autoclave batch high-pressure setup consisting of (from left to right) feed tank, auxiliary
manual pump (250 bar), three valves, manual syringe pump (3000 bar), pressure transducer, and autoclave reactor
with the tip located in the middle
of the reactor. Two pumps were
needed to deliver the reacting
mixture and for the generation of surrounded with electrical heating jackets.
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
7
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
residence time. The temperature for cooling was kept at 208C re-
gardless of the temperature of the reaction mixture. After flushing
three reactor volumes to ensure steady-state data collection, three
samples were collected. 250 mL was collected each time within the
loop and pushed into a GC-vial using a syringe filled with an aceto-
nitrile/water (50:50) mixture. The contents of the GC-vial were then
analyzed using a GC (for 2,6-difluorobenzyl azide conversion) and
a HPLC (for the determination of the yield); calculations are based
on internal standard. Notes: The thermal stability of the internal
standard was investigated prior to the experiments.
Methyl 1-(2,6-difluorobenzyl)-1H-1,2,3-triazole-4-carboxylate in
flow reactor (catalyzed): 2,6-difluorobenzyl azide (71.0 g,
420 mmol) and 1,3,5-trimethoxybenzene (internal standard; 14.1 g,
84 mmol) were weighed and diluted to 240 mL using NMP;
[Cu(phen)(PPh3)2]NO3 (3.7 g, 4.2 mmol) was diluted to 240 mL using
NMP whereas methyl propiolate (35.3 g, 420 mmol) was diluted to
420 mL. Solutions were pumped using three pumps to be mixed in
a T-mixer: 2,6-difluorobenzyl azide and catalyst mix first, later
a stream of methyl propiolate was introduced to result in a 0.5m
solution of 2,6-difluorobenzyl azide and a 0.5m solution of methyl
propiolate. Reactants reacted in the reaction zone of 2 mL internal
volume. Flowrates were adjusted based on the desired residence
time. The reaction zone was heated inside the oil bath while pres-
sure was applied through a BPR. A sample loop was installed be-
tween the cooling zone and the BPR to allow sample collection
with accurate residence times. The temperature for cooling was
kept at 208C regardless of the temperature of the reaction mixture.
After flushing three reactor volumes to ensure steady-state data
collection, three samples were collected. 250 mL were collected
each time within the loop, pushed out into a GC-vial using syringe
filled with an acetonitrile/water (50:50) mixture. Contents of the
GC-vial were then analyzed using a GC (for 2,6-difluorobenzyl azide
conversion) and a HPLC (for the determination of the yield); calcu-
lations are based on internal standard.
Figure 8. HPHT microcapillary-based flow system with operating limits of
400 bar and 3008C. The setup consists of a SS capillary, two HPLC pumps,
heating and cooling oil baths, sample loop connected to a six-port valve,
and one BPR.
12.5 mmol) and methyl propiolate (2.1 g, 25 mmol) were dissolved
in N-methyl-2-pyrrolidone (NMP; 50 mL). The solution was filled
into the feed tank and pumped through the pipelines into the re-
actor. After the first drop leaving the reactor, the system was
sealed, the last valve before the reactor was closed. The valve be-
tween auxiliary pump and the feed syringe was opened and the
auxiliary pump was used to pump the solution from the feed tank
into the syringe. After the syringe pump was filled, the valve was
closed and the one next to the reactor was opened. By turning the
‘steering wheel’ of the syringe, the piston located within applied
pressure. The turning was stopped at 200 bar below the desired
set-point. The resulting mixture was then heated to 908C, and the
resulting increase in pressure (if any) was measured. Upon heating
the reactor, extra pressure buildup was observed. Finally, the pres-
sure was adjusted to the final set-point and the reactor was left for
24 h. Demineralized water (10 mL) was added to a collected
sample (14 mL), and the mixture was extracted using ethyl acetate
(4ꢂ10 mL). The collected organic phase was washed using demi
water (2ꢂ10 mL) and brine (1ꢂ10 mL) and dried over MgSO4. All
volatiles were evaporated under vacuum at 508C. For a further pu-
rification the precipitate was recrystallized from MeOH (1:10). mp
Acknowledgements
We would like to thank Flowid B.V. for lending us HPIMM mixer
to be used in our investigations, Jan Volkers and Guus Honcoop
(Knauer) for their help with setting up the pumps, and Erik van
Herk for helping with building the HPHT flow setup. Funding by
the Advanced European Research Council Grant “Novel Process
Windows—Boosted Micro Process Technology” (Grant number:
267443) is kindly acknowledged.
1
136–1378C; H NMR (400 MHz, [D6]DMSO): d=8.85 (s, 1H), 7.50 (t,
J=16 Hz, 1H), 7.16 (m, J=12 Hz, 2H), 5.73 (s, 2H), 3.81 ppm (s,
3H); 13C (100 MHz, [D6]DMSO): d=162.49 (d, J=8 Hz), 160.99,
160.00 (d, J=7 Hz), 138.94, 132.30 (t, J=10 Hz), 129.89, 112.38 (m,
J=12 Hz), 111.26 (t, J=19 Hz), 51.24, 41.79 ppm (t, J=3 Hz);
19FNMR (400 MHz, [D6]DMSO): d=114.05 ppm (s, 2F); HRMS calcu-
lated for C11H9F2N3O2Na 276.0561, found 276.0566 (M+Na+).
1H chemical shifts are reported in ppm downfield from tetrame-
thylsilane (TMS), whereas 13C chemical shifts are reported down-
field from TMS with the resonance of the [D6]DMSO as the internal
standard.
Keywords: 1,3-dipolar cycloaddition
·
high pressure
·
microflow chemistry · novel process windows · rufinamide
Methyl 1-(2,6-difluorobenzyl)-1H-1,2,3-triazole-4-carboxylate in
flow reactor: 2,6-difluorobenzyl azide (5.1 g, 30 mmol) and propio-
phenone as internal standard (0.8 g 6 mmol) were weighed and di-
luted to 60 mL using NMP; methyl propiolate (5.0 g, 60 mmol) was
diluted to yield 60 mL. The solutions were pumped using two
pumps to be mixed in a T-mixer, resulting in 0.25m solution of 2,6-
difluorobenzyl azide and 0.5m solution of methyl propiolate. The
reactants reacted in a reaction zone of 2 mL internal volume. The
flow rates were adjusted based on the desired residence time. The
reaction zone was heated inside the oil bath while pressure was
applied using a BPR. A sample loop was installed between the
cooling zone and the BPR to allow sample collection with accurate
[1] I. Chorkendorff, J. W. Niemantsverdriet in Concepts Mod. Catal. Kinet.,
Wiley-VCH, 2005, pp. 23–78.
[2] a) K. Matsumoto, H. Hamana, H. Iida, Helv. Chim. Acta 2005, 88, 2033–
2234; b) R. Bini, M. Ceppatelli, M. Citroni, V. Schettino, Chem. Phys.
2012, 398, 262–268.
[3] a) V. Hessel, Chem. Eng. Technol. 2009, 32, 1655–1681; b) S. Borukhova,
V. Hessel in Process Intensif. Green Chem. (Eds.: K. Boodhoo, A. Harvey),
Wiley, Chichester, UK 2013, pp. 91–156; c) T. Razzaq, T. N. Glasnov, C. O.
Kappe, Chem. Eng. Technol. 2009, 32, 1702–1716; d) V. Hessel, D. Kra-
lisch, N. Kockmann, T. Noꢀl, Q. Wang, ChemSusChem 2013, 6, 746–789.
[4] R. L. Hartman, J. P. McMullen, K. F. Jensen, Angew. Chem. Int. Ed. 2011,
50, 7502–7519; Angew. Chem. 2011, 123, 7642–7661.
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
8
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[5] a) L. Malet-Sanz, F. Susanne, J. Med. Chem. 2012, 55, 4062–4098; b) P. T.
Baraldi, V. Hessel, Green Process Synth. 2012, 1, 149–167; c) T. Noꢀl, S. L.
Buchwald, Chem. Soc. Rev. 2011, 40, 5010–5029.
d) J. Matsuo, S. Sasaki, H. Tanaka, H. Ishibashi, J. Am. Chem. Soc. 2008,
130, 11600–11601.
[13] U. Tilstam, T. Defrance, T. Giard, M. D. Johnson, Org. Process Res. Dev.
2009, 13, 321–323.
[6] a) S. Borukhova, T. Noꢀl, B. Metten, E. de Vos, V. Hessel, ChemSusChem
2013, 6, 2220–2225; b) H. Kobayashi, B. Driessen, D. J. G. P. van Osch, A.
Talla, S. Ookawara, T. Noꢀl, V. Hessel, Tetrahedron 2013, 69, 2885–2890;
c) T. Illg, V. Hessel, P. Lçb, J. C. Schouten, ChemSusChem 2011, 4, 392–
398; d) T. Illg, P. Lçb, V. Hessel, Biol. Med. Chem. 2010, 18, 3707–3719;
e) T. Razzaq, T. N. Glasnov, C. O. Kappe, Eur. J. Org. Chem. 2009, 1321–
1325; f) N. Kockmann, M. Gottsponer, B. Zimmermann, D. M. Roberge,
Chem. Eur. J. 2008, 14, 7470–7477.
[7] a) F. Benito-Lopez, R. M. Tiggelaar, K. Salbut, J. Huskens, R. J. M. Egber-
ink, D. N. Reinhoudt, H. J. G. E. Gardeniers, W. Verboom, Lab Chip 2007,
7, 1345–1351; b) W. Verboom, Chem. Eng. Technol. 2009, 32, 1695–
1701; c) Y. Zhao, G. Chen, C. Ye, Q. Yuan, Chem. Eng. Sci. 2013, 87, 122–
132; d) A. Leclerc, M. Alamꢃ, D. Schweich, P. Pouteau, C. Delattre, C. de
Bellefon, Lab Chip 2008, 8, 814–817; e) F. Trachsel, C. Hutter, P. R. von
Rohr, Chem. Eng. J. 2008, 135, S309–S316; f) J. Kobayashi, Y. Mori, S. Ko-
bayashi, Chem. Commun. 2005, 2567–2568.
[8] a) W. Grochala, R. Hoffmann, J. Feng, N. W. Ashcroft, Angew. Chem. Int.
Ed. 2007, 46, 3620–3642; Angew. Chem. 2007, 119, 3694–3717; b) A.
Sharma, J. H. Scott, G. D. Cody, M. L. Fogel, R. M. Hazen, R. J. Hemley,
W. T. Huntress, Science 2002, 295, 1514–1516; c) A. Y. Rulev, H. Kotsuki,
J. Maddaluno, Green Chem. 2012, 14, 503.
[9] a) J. Keybl, K. F. Jensen, Ind. Eng. Chem. Res. 2011, 50, 11013–11022;
b) R. M. Tiggelaar, F. Benito-Lꢄpez, D. C. Hermes, H. Rathgen, R. J. M. Eg-
berink, F. G. Mugele, D. N. Reinhoudt, A. van den Berg, W. Verboom,
H. J. G. E. Gardeniers, Chem. Eng. J. 2007, 131, 163–170; c) N. Lorber, F.
Sarrazin, P. Guillot, P. Panizza, A. Colin, B. Pavageau, C. Hany, P. Maestro,
S. Marre, T. Delclos, C. Aymonier, P. Subra, L. Prat, C. Gourdone, E.
Mignard, Lab Chip 2011, 11, 779–787.
[14] a) R. Huisgen, Proc. Chem. Soc. 1961, 357–396; b) R. Huisgen, Angew.
Chem. Int. Ed. Engl. 1963, 2, 565–598; Angew. Chem. 1963, 75, 604–
637; c) G. T. Anderson, J. R. Henry, S. M. Weinreb, J. Org. Chem. 1991, 56,
6946–6948; d) V. Melai, A. Brillante, P. Zanirato, J. Chem. Soc. Perkin
Trans. 2 1998, 2447–2450; e) J.-C. Fan, J. Liang, Y. Wang, Z.-C. Shang,
THEOCHEM 2007, 821, 145–152; f) H. Elamari, I. Jlalia, C. Louet, J. Her-
scovici, F. Meganem, C. Girard, Tetrahedron: Asymmetry 2010, 21, 1179–
1183.
[15] a) H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. Int. Ed. 2001, 40,
2004–2021; Angew. Chem. 2001, 113, 2056–2075; b) H. C. Kolb, K. B.
Sharpless, Drug Discovery Today 2003, 8, 1128–1137; c) L. D. Pachꢄn,
J. H. van Maarseveen, G. Rothenberg, Adv. Synth. Catal. 2005, 347, 811–
815; d) J. E. Moses, A. D. Moorhouse, Chem. Soc. Rev. 2007, 36, 1249–
1262.
[16] J. Wang, G. Sui, V. P. Mocharla, R. J. Lin, M. E. Phelps, H. C. Kolb, H.-R.
Tseng, Angew. Chem. Int. Ed. 2006, 45, 5276–5281; Angew. Chem. 2006,
118, 5402–5407.
[17] a) C. D. Smith, I. R. Baxendale, S. Lanners, J. J. Hayward, S. C. Smith, S. V.
Ley, Org. Biomol. Chem. 2007, 5, 1559–1561; b) A. R. Bogdan, K. James,
Chem. Eur. J. 2010, 16, 14506–14512; c) A. C. Varas, T. Noꢀl, Q. Wang, V.
Hessel, ChemSusChem 2012, 5, 1703–1707; e) L. Wang, S. Peng, L. J. T.
Danence, Y. Gao, J. Wang, Chem. Eur. J. 2012, 18, 6088–6093.
[18] M. Baumann, I. R. Baxendale, S. V. Ley, N. Nikbin, Beilstein J. Org. Chem.
2011, 7, 442–495.
[19] M. E. Lemmon, E. H. Kossoff, Curr. Treat. Options Neurol. 2013, 15, 519–
528.
[20] P. Zhang, M. G. Russell, T. F. Jamison, Org. Process Res. Dev., 2014, 18,
1567–1570.
[21] D. Wang, M. Zhao, X. Liu, Y. Chen, N. Li, B. Chen, Org. Biomol. Chem.
2012, 10, 229–231.
[22] J. E. Hein, V. V. Fokin, Chem. Soc. Rev. 2010, 39, 1302–1315.
[23] a) V. Hessel, I. Vural Gꢅrsel, Q. Wang, T. Noꢀl, J. Lang, Chem. Eng. Technol.
2012, 35, 1184–1204; b) I. Vural-Gꢅrsel, Q. Wang, T. Noꢀl, V. Hessel, J. T.
Tinge, Ind. Eng. Chem. Res. 2013, 52, 7827–7835.
[10] a) H. S. P. Rao, R. Murali, A. Taticchi, H. W. Scheeren, Eur. J. Org. Chem.
2001, 2869–2876; b) L. G. Jenner, J. Phys. Org. Chem. 1999, 12, 619–
625; c) Z. Shi, W. Liang, J. Luo, S. Huang, B. M. Polishak, X. Li, T. R. Youn-
kin, B. a. Block, A. K.-Y. Jen, Chem. Mater. 2010, 22, 5601–5608; d) R. E.
Martin, F. Morawitz, C. Kuratli, A. M. Alker, A. I. Alanine, Eur. J. Org. Chem.
2012, 47–52; e) L. Minuti, A. Temperini, E. Ballerini, J. Org. Chem. 2012,
77, 7923–7931; f) H. Chen, B.-B. Ni, F. Gao, Y. Ma, Green Chem. 2012, 14,
2703–2705.
[24] C. Rosenfeld, C. Serra, C. Brochon, V. Hessel, G. Hadziioannou, Chem.
Eng. J. 2008, 135, S242–S246.
[11] A. Vidis, G. Laurenczy, E. Kuesters, G. Sedelmeier, P. J. Dyson, J. Phys. Org.
Chem. 2007, 20, 109–114.
[12] a) A. Chrꢃtien, I. Chataigner, S. R. Piettre, Tetrahedron 2005, 61, 7907–
7915; b) Y. Misumi, K. Matsumoto, Angew. Chem. Int. Ed. 2002, 41,
1031–1033; Angew. Chem. 2002, 114, 1073–1075; c) P. Kwiatkowski, M.
Asztemborska, J. Jurczak, Tetrahedron: Asymmetry 2004, 15, 3189–3194;
Received: September 22, 2014
Published online on && &&, 0000
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
9
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
S. Borukhova, A. D. Seeger, T. Noꢀl,
Q. Wang, M. Busch, V. Hessel*
Press to access! The potential of pres-
sure in chemical intensification of intrin-
sic kinetics of 1,3-dipolar cycloaddition
is investigated along with high temper-
ature and concentration effects. Two re-
actors are compared, a specialized auto-
clave batch reactor for high-pressure
operation up to 1800 bar and a capillary
flow reactor for up to 400 bar. Reaction
speedup and increases in space-time
yields are reached while widening pro-
cess windows of favorable operation to
selectively produce Rufinamide precur-
sor in good yields.
&& – &&
Pressure-Accelerated Azide–Alkyne
Cycloaddition: Micro Capillary versus
Autoclave Reactor Performance
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
10
ꢁ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!