Versatile and scalable synthesis of cyclic organic carbonates under organocatalytic continuous flow conditions

Article abstract of DOI:10.1039/c9cy01659g

The benchmark route for the preparation of cyclic organic carbonates starts from toxic, volatile and unstable epoxides. In this work, cyclic organic carbonates are prepared according to alternative sustainable and intensified continuous flow conditions from the corresponding 1,2-diols. The process utilizes dimethyl carbonate (DMC) as a low toxicity carbonation reagent and relies on the organocatalytic activity of widely available and cheap organic ammonium and phosphonium salts. Glycerol is selected as a model substrate for preliminary optimization with a library of homogeneous ammonium and phosphonium salts. The nature of the anion dramatically influences the catalytic activity, while the nature of the cation does not impact the reaction. Upon optimization, glycerol carbonate is obtained in 95% conversion and 79% selectivity within 3 min residence time at 180 °C (11 bar) with 3.5 mol% of tetrabutylammonium bromide as the organocatalyst. A straightforward liquid-liquid extraction procedure enables both the purification of glycerol carbonate and the recycling of the homogeneous catalyst. The conditions are amenable to refined and crude bio-based glycerol, although conversions are lower in the latter case. Control experiments suggest that water present in the crude samples induces significant hydrolysis of glycerol carbonate. The reaction conditions are then successfully applied on a wide variety of substrates, affording the corresponding cyclic carbonates in overall good to excellent yields (20 examples, 45-95%). The substrate scope notably encompasses bio-based starting materials such as glycerol ethers and erythritol-derived diols. In-line NMR is featured as a qualitative analytical tool for real-time reaction monitoring. The scalability of this carbonation procedure on glycerol is assessed in a commercial pilot-scale silicon carbide continuous flow reactor of 60 mL internal volume. Glycerol carbonate is obtained in 76% yield, corresponding to a productivity of 13.6 kg per day.

Full text of DOI:10.1039/c9cy01659g

Volume 9 Number 24 21 December 2019 Pages 6819–7052
Catalysis
Science &
Technology
rsc.li/catalysis
ISSN 2044-4761
PAPER
Jean-Christophe M. Monbaliu et al.
Versatile and scalable synthesis of cyclic organic carbonates under
organocatalytic continuous flow conditions

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Versatile and scalable synthesis of cyclic organic
carbonates under organocatalytic continuous flow
conditions†
Cite this: Catal. Sci. Technol., 2019,
9, 6841
Romaric Gérardy, a Julien Estager,b Patricia Luis,c
a
Damien P. Debecker d and Jean-Christophe M. Monbaliu
*
The benchmark route for the preparation of cyclic organic carbonates starts from toxic, volatile and unsta-
ble epoxides. In this work, cyclic organic carbonates are prepared according to alternative sustainable and
intensified continuous flow conditions from the corresponding 1,2-diols. The process utilizes dimethyl car-
bonate (DMC) as a low toxicity carbonation reagent and relies on the organocatalytic activity of widely
available and cheap organic ammonium and phosphonium salts. Glycerol is selected as a model substrate
for preliminary optimization with a library of homogeneous ammonium and phosphonium salts. The nature
of the anion dramatically influences the catalytic activity, while the nature of the cation does not impact
the reaction. Upon optimization, glycerol carbonate is obtained in 95% conversion and 79% selectivity
within 3 min residence time at 180 °C (11 bar) with 3.5 mol% of tetrabutylammonium bromide as the
organocatalyst. A straightforward liquid–liquid extraction procedure enables both the purification of glyc-
erol carbonate and the recycling of the homogeneous catalyst. The conditions are amenable to refined
and crude bio-based glycerol, although conversions are lower in the latter case. Control experiments sug-
gest that water present in the crude samples induces significant hydrolysis of glycerol carbonate. The reac-
tion conditions are then successfully applied on a wide variety of substrates, affording the corresponding
cyclic carbonates in overall good to excellent yields (20 examples, 45–95%). The substrate scope notably
encompasses bio-based starting materials such as glycerol ethers and erythritol-derived diols. In-line NMR
is featured as a qualitative analytical tool for real-time reaction monitoring. The scalability of this carbon-
ation procedure on glycerol is assessed in a commercial pilot-scale silicon carbide continuous flow reactor
of 60 mL internal volume. Glycerol carbonate is obtained in 76% yield, corresponding to a productivity of
13.6 kg per day.
Received 29th May 2019,
Accepted 10th October 2019
DOI: 10.1039/c9cy01659g
rsc.li/catalysis
their preparation: (a) the carbonation of 1,2-diols (2) and (b)
the carboxylation of epoxides (3) (Fig. 1). The latter is the
Introduction
Cyclic organic carbonates, or 1,3-dioxolan-2-ones (1), are bio-
degradable and non-toxic compounds with a low vapor pres-
sure. A broad range of industrial applications for cyclic or-
ganic carbonates have been reported, including their use as
solvents, as electrolyte carriers in batteries, and as building
blocks for the preparation of various chemicals and
polymers.1–6 Two main strategies are typically reported for
benchmark route for the preparation of functionalized cyclic
organic carbonates and appears to be the most sustainable
a Center for Integrated Technology and Organic Synthesis, Research Unit MolSys,
University of Liège, B-4000 Liège (Sart Tilman), Belgium.
E-mail: jc.monbaliu@uliege.be; Web: http://www.citos.uliege.be
b Certech, Rue Jules Bordet, Zone Industrielle C, B-7180 Seneffe, Belgium
c Materials & Process Engineering (iMMC-IMAP), UCLouvain, B-1348 Louvain-la-
Neuve, Belgium
d Institute of Condensed Matter and Nanosciences, UCLouvain, B-1348 Louvain-la-
Neuve, Belgium
† Electronic supplementary information (ESI) available: Characterization of com-
pounds, details of the fluidic reactors and components, additional experimental
data and procedures. See DOI: 10.1039/c9cy01659g
Fig.
1 Main strategies for the preparation of cyclic organic
carbonates. R1, R2 = alkyl, aryl.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6841

View Article Online
Paper
Catalysis Science & Technology
since it directly utilizes CO2 as a C1 platform.7,8 However, ep-
oxides are in general toxic, volatile and unstable.9 By contrast,
1,2-diols have a low toxicity profile, a low vapor pressure and
are generally stable. Direct carboxylation of 1,2-diols with CO2
is currently not synthetically viable since it typically requires
harsh conditions, elaborate catalysts and toxic additives to at-
tain poor yields.10,11
Reactive carbonation reagents are required to convert
1,2-diols into cyclic carbonates (Table 1); typical examples in-
clude enthalpy-activated reagents such as phosgene,12 alkyl
chloroformates13 and carbonyldiimidazole (CDI).14 The utili-
zation of phosgene on the industrial scale raises safety con-
cerns. Alkyl chloroformates and CDI are far less toxic, but are
typically derived from phosgene, expensive and water-sensi-
tive.15,16 Other reagents such as carbon monoxide in combi-
nation with an oxidant,17,18 urea,19 diphenyl carbonate,20 and
dialkyl carbonates21–24 are reported as well; yet, these strate-
gies suffer from some drawbacks (Table 1).
In contrast to higher carbonates, dimethyl and diethyl car-
bonates (DMC and DEC, respectively) have attracted consider-
able attention as a class of “green” solvents and reagents for
organic transformations. DMC and DEC are widely available,
non-toxic liquids, and they are not sourced from
phosgene.25–27 The carbonation of 1,2-diols with DMC or
DEC releases 2 molar equivalents of methanol or ethanol, re-
spectively, i.e. non-corrosive by-products that can be easily
recycled and utilized as solvents, reagents or fuels.
One of the most privileged substrates for carbonation
with DMC/DEC is glycerol (4a)28–33 and the methodology
was scarcely extended to other functionalized 1,2-diol
substrates.21–24 Numerous homogeneous and heterogeneous
catalysts are described for the carbonation of glycerol with
DMC/DEC,28–33 including organocatalysts. Organocatalysts
have a bright forecast for the sustainable upgrading of bio-
based chemicals.34 Typical organocatalysts described for the
carbonation of 4a with DMC/DEC include amines, amidines,
guanidines, N-heterocyclic carbenes, phosphazenes, enzymes,
organic salts and ionic liquids.21,28–33 However, one of the
main drawbacks precluding the utilization of organocatalysts
for the upgrading of low-cost, high-volume chemicals, such
as 4a, is their elevated cost.
Despite the many assets of continuous flow processes for
the manufacturing of commodities and fine chemicals,35,36
only a few reports described the continuous flow carbon-
ation of 4a with DMC/DEC (Table 2).21,37–42 De Souza and
coworkers described
a process involving 1.5 mol% of
diazabicycloIJ5.4.0)undec-7-ene (DBU) as a catalyst, 3 equiv. of
DMC and solvent-free conditions. Glycerol carbonate (1a)
was obtained with 80% conversion and 82% selectivity
within 8 min at 120 °C.38 Selva et al. reported a catalyst-free
procedure for 1a under high pressure and temperature con-
ditions (230–250 °C, 50 bar) using methanol as a solvent. 10
equiv. of DMC and 15 min of residence within the reactor
were necessary to reach conversion in the 78–94% range and
83–92% selectivity. The methodology was also successfully
extended to ethylene and propylene glycol (2b, 2c).40
Baxendale and coworkers described a process using 3 wt%
of tetraethylammonium pipecolinate and 3.5 equiv. of DMC
for the synthesis of carbonate 1a. Ethanol was used as the
solvent and 30 min of residence at 140 °C was required to
obtain 1a with 90% conversion and 85% selectivity.41
Monbaliu and coworkers reported a solvent-free procedure
Table 1 Comparison of common reagents utilized for the carbonation of 1,2-diols
Classification
Reagent
Pros
Cons
Extremely toxic
Gaseous
Ref.
12
Toxic
13
14
Activated, phosgene-sourced
No catalyst required
Water-sensitive
Water-sensitive
Expensive
Atom economy
Cheap and widely available
CO2-Sourced
Extremely toxic
Gaseous
Pd catalyst + oxidant required
Corrosive and gaseous by-product (NH3)
Catalyst required
17, 18
19
Non activated
Cheap and widely available
Widely available
Solvent required for flow processing
Toxic by-product (PhOH)
20
No catalyst required
Widely available
Solvent required for flow processing
Activated, non phosgene-sourced
Catalyst required
21–24
Low toxicity by-products (ROH)
CDI = carbonyldiimidazole. R = Me or Et.
6842 | Catal. Sci. Technol., 2019, 9, 6841–6851
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Table 2 Homogeneous continuous flow processes from the prior Art for the carbonation of glycerol (4a) into glycerol carbonate (1a) with DMC
Paper
Solvent
Neat
DMC
Catalyst
Conditions
Conv. (%)
80
Selec. (%)
82
Scale-up
None
1,2-Diol substrate scope
None
120 °C,
8 min
De Souza (2015)38
Selva (2016)40
3 equiv.
10 equiv.
(1.5 mol%)
None
230–250 °C,
15 min
MeOH
78–94
83–92
None
3 substrates
140 °C,
30 min
2.1 kg per
day
Baxendale (2018)41
EtOH
3.5 equiv.
90
85
None
(3 wt%)
135 °C,
2 min
8.1 kg per
day
Monbaliu (2019)21
This work
Neat
Neat
3 equiv.
94
95
87
79
9 substrates
20 substrates
(1 mol%)
NBu4Br
(3.5 mol%)
180 °C,
3 min
13.6 kg per
day
2.25 equiv.
utilizing 1 mol% of 2-tert-butyl-1,1,3,3-tetramethylguanidine Experimental section
and 3 equiv. of DMC. Glycerol carbonate (1a) was obtained
within 2 min at 135 °C with 87% selectivity and 94% conver-
Chemicals
Ethanediol, 1,2-propanediol, 1,2,3-propanetriol, 3-methoxy-1,2-
propanediol, 3-tert-butoxy-1,2-propanediol, 3-allyloxy-1,2-
propanediol, 3-(o-tolyloxy)-1,2-propanediol, 3-(dimethylamino)-
1,2-propanediol, 3-morpholino-1,2-propanediol, 1,2-
butanediol, 2,3-butanediol, 3-butene-1,2-diol, 1,4-anhydro-
erythritol, cis-1,2-cyclohexanediol, trans-1,2-cyclohexanediol,
sion. The methodology was also successfully extended to
8 other simple substrates upon minor adjustment of the
conditions.21 However, these studies did not demonstrate
the applicability of the method for the preparation of func-
tionalized carbonates and relied on expensive organo-
catalysts or extreme conditions.
1,2-octanediol, tetramethylammonium
bromide, tetra-
Our involvement in scalable, intensified and organo-
catalyzed carbonation processes led us to consider cheap,
widely available and low-toxicity43 homogeneous organic salts
from the ammonium and phosphonium types. The related
prior Art contains one patent filed in 1992, where the carbon-
ation of glycerol was effected in batch at 120 °C in the pres-
ence of DMC and tetrabutylammonium bromide. Glycerol
carbonate (1a) was obtained with 92% conversion and 92%
selectivity after 6 h under these conditions. The methodology
was also extended to ethylene (2b) and propylene (2c) gly-
col.44 In 2014, Selva reported carbonate phosphonium salts
as efficient catalysts for the carbonation of 2b and 2c, which
was conducted in batch at 90 °C.45
ethylammonium bromide, tetrabutylammonium bromide,
tetra-n-octylammonium bromide, tetramethylphosphonium
bromide,
tetrabutylphosphonium
bromide,
tetra-
phenylphosphonium bromide, tetrabutylammonium chlo-
ride, tetrabutylammonium iodide, tetrabutylammonium ace-
tate,
tetrabutylammonium
2-hydroxybenzoate,
tetra-
tetrabutylammonium
trifluoromethanesulfonate,
butylammonium tetrafluoroborate, tetrabutylammonium
hexafluorophosphate, dimethyl carbonate, 2,3-epoxy-1-
propanol and dimethyl sulfoxide were obtained from com-
mercial sources and used as received. Refined and crude bio-
based glycerol samples were provided by Cargill™ and Valtris
Champlor SAS. Methyl ((2-oxo-1,3-dioxolan-4-yl)methyl) car-
In this work, we report an efficient organocatalytic process
for the preparation of cyclic organic carbonates. The proce-
dure is unique since (a) it involves intensified continuous-
flow conditions (i.e. solvent-free or highly concentrated me-
dia, short reaction times, low catalytic loadings and low ex-
cess of DMC), (b) it utilizes cheap, air-stable and benign or-
ganic salts of the ammonium and phosphonium-type as
catalysts, and (c) it is extended to a wide range of functional-
ized diols (20 examples). A straightforward liquid–liquid ex-
traction procedure for the isolation of glycerol carbonate and
for the recycling of the homogeneous catalyst is proposed.
Refined and crude bio-based glycerol samples are evaluated
as well, giving promising results. The process also features
low-field in-line NMR analysis for qualitative real-time reac-
tion monitoring, and scalability is assessed in a pilot-scale
continuous flow reactor.
bonate,
propanediol, 3,3′-(butane-1,4-diylbisIJoxy))bisIJpropane-1,2-diol)
and 3,3′-((furan-2,5-diylbisIJmethylene))bisIJoxy))bisIJpropane-
3-propargyloxy-1,2-propanediol,
3-furfuryloxy-1,2-
1,2-diol) were prepared following literature procedures (ESI†).
Experimental setup (microscale)
Microfluidic continuous flow reactors consisted of stainless
steel (SS) coils (1.58 mm o.d. × 500 μm i.d.) equipped with SS
nuts, ferrules and unions (VALCO). In experiments involving
crude glycerol, the reactor consisted of a SS coil (3 mm o.d. ×
2.1 mm i.d.) equipped with reducing unions (1/8″ to 1/16″,
Swagelok). The reactors were thermoregulated with
a
Heidolph™ MR Hei-Tec® oil bath equipped with a Pt-1000
temperature sensor. Sections of the reactors that were not
subjected to high temperatures were constructed from PEEK
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6843

View Article Online
Paper
Catalysis Science & Technology
or PFA tubing (1.58 mm o.d. × 750 μm i.d.) equipped with
PEEK/ETFE connectors and ferrules (IDEX/Upchurch Scien-
tific). Feed solutions were handled with high force Chemyx
Fusion 6000 syringe pumps equipped with SS syringes and
Dupont Kalrez O-rings or with Knauer Azura P 4.1S HPLC
pumps. Downstream pressure was regulated using a dome-
type back pressure regulator (Zaiput Flow Technologies BPR-
10) connected to a nitrogen cylinder to set the working pres-
sure (11 bar). In-line NMR reaction monitoring was carried
out using a 43 MHz Spinsolve™ carbon NMR spectrometer
from Magritek®.
4-(morpholinomethyl)-1,3-dioxolan-2-one, 4,5-dimethyl-1,3-di-
oxolan-2-one, cis-hexahydro-1,3-benzodioxol-2-one, 2-hydroxy-
cyclohexyl methyl carbonate, tetrahydrofuroij3,4-d]ij1,3]dioxol-
2-one, 4,4′-((butane-1,4-diylbisIJoxy))bisIJmethylene))bisIJ1,3-di-
oxolan-2-one) and 4,4′-(((furan-2,5-diylbisIJmethylene))bisIJoxy))-
bisIJmethylene))bisIJ1,3-dioxolan-2-one) yields were determined
by 1H NMR spectroscopy using mesitylene as the internal
standard.
Typical run 1: microscale synthesis of glycerol carbonate (1a)
from commercial glycerol
The syringe pump used to deliver the solution of glycerol (4a)
containing 3.5 mol% of NBu4Br was set to 67.2 μL min−1 (1
equiv. of 4a) and the syringe pump used to deliver neat DMC
was set to 154 μL min−1 (2.25 equiv.). Both streams were mixed
through a PEEK T-mixer, and the biphasic mixture entered a
1/16″ o.d. SS coil (0.67 mL, 3 min of residence) heated at 180
°C. The outlet of the reactor was connected to a dome-type back
pressure regulator set at 11 bar (Fig. 5). After equilibration,
the reactor effluent was collected, diluted with EtOH, and an-
alyzed by GC/FID (4a conv. = 95%, 1a yield = 75%, see Fig. 5).
Experimental setup (pilot-scale)
Mesofluidic experiments were carried out in a Corning®
Advanced-Flow™ G1 SiC reactor featuring 6 fluidic modules
(FMs) connected in series (10 mL per FM). Feed and collec-
tion lines consisted of PFA tubing (1/4″ o.d., 0.047 inch wall
thickness) equipped with PFA or SS Swagelok connectors and
ferrules. Feed solutions were handled with Corning dosing
lines (HMP gear pumps). The reactor was thermoregulated
with a LAUDA Integral XT 280 thermostat using LAUDA
Therm 180 silicone oil. A temperature probe was positioned
at a critical position along the thermofluid flow path for tem-
perature monitoring. Downstream pressure was regulated
using a dome-type back pressure regulator (Zaiput Flow Tech-
nologies BPR-1000) connected to a nitrogen cylinder to set
the working pressure (11 bar).
Typical run 2: microscale synthesis of glycerol carbonate (1a)
from crude bio-based glycerol
Crude bio-based glycerol was filtered prior to use. The syringe
pump used to deliver the solution of crude bio-based glycerol
(4a) containing 3.5 mol% of NBu4Br was set to 212 μL min−1
(1 equiv. of 4a) and the syringe pump used to deliver neat
DMC was set to 411 μL min−1 (2.25 equiv.). Both streams
were mixed through a PEEK T-mixer, and the biphasic mix-
ture entered a 1/8″ o.d. SS coil (1.87 mL, 3 min of residence)
heated at 180 °C. The outlet of the reactor was connected to
a dome-type back pressure regulator set at 11 bar. After equil-
ibration, the reactor effluent was collected, diluted with
EtOH, and analyzed by GC/FID (4a conv. = 30%, 1a yield =
29%, see Table 5, entry 3).
Analytical methods
Conversion and yield were determined by gas chromatogra-
phy coupled with flame ionization detection (GC/FID) using a
Shimadzu GC-2014 gas chromatograph. Selectivity is defined
as the ratio between the yield and the conversion. Where
specified, yields were determined by high field 1H NMR
spectroscopy using mesitylene as the internal standard. Struc-
tural identity was confirmed by 1H and 13C NMR spectro-
scopy conducted on a high field (400 MHz) Bruker Avance
spectrometer in CDCl3 or d4-MeOD (ESI†). The chemical
shifts are reported in ppm relative to TMS as the internal
standard or to the solvent residual peak. GC conversions were
determined using external calibration curves established with
commercial standards of ethanediol, 1,2-propanediol, 1,2,3-
propanetriol, 3-methoxy-1,2-propanediol, 1,2-butanediol, and
3-butene-1,2-diol. GC yields were determined using external
calibration curves established with commercial standards of
1,3-dioxolan-2-one, 4-methyl-1,3-dioxolan-2-one, 4-hydroxy-
Typical run 3: microscale synthesis of 4-((o-tolyloxy)methyl)-
1,3-dioxolan-2-one (1l)
The HPLC pump used to deliver the feed solution of
mephesin (2l, 3 M) and NBu4Br (0.105 M) in DMSO was set
to 222 μL min−1 (1 equiv. of 2l). The HPLC pump used to de-
liver neat DMC was set to 168 μL min−1 (3 equiv.). Both
streams were mixed through a PEEK T-mixer, and the mix-
ture entered a 1/16″ o.d. SS coil (1.17 mL, 3 min of residence)
heated at 180 °C. The outlet of the reactor was connected to
a dome-type back pressure regulator set at 11 bar (Fig. 5). Af-
ter equilibration, the reactor effluent was collected and ana-
methyl-1,3-dioxolan-2-one,
4-methoxymethyl-1,3-dioxolan-2-
one, 4-ethyl-1,3-dioxolan-2-one, 4-vinyl-1,3-dioxolan-2-one, 2,3-
epoxy-1-propanol and a synthesized sample of methyl ((2-oxo-
1,3-dioxolan-4-yl)methyl) carbonate. 4-Hexyl-1,3-dioxolan-2-
one, 4-(tert-butoxymethyl)-1,3-dioxolan-2-one, 4-((allyloxy)-
1
lyzed by H NMR with mesitylene as the internal standard (1l
yield = 86%, see Fig. 5).
methyl)-1,3-dioxolan-2-one,
dioxolan-2-one, 4-((furan-2-ylmethoxy)methyl)-1,3-dioxolan-2-
one, 4-((o-tolyloxy)methyl)-1,3-dioxolan-2-one,
4-((prop-2-yn-1-yloxy)methyl)-1,3-
Typical run 4: pilot-scale synthesis of glycerol carbonate (1a)
The pump used to deliver the preheated (60 °C) feedstock so-
lution of glycerol (4a) containing 3.5 mol% of NBu4Br was
4-((dimethylamino)methyl)-1,3-dioxolan-2-one,
6844 | Catal. Sci. Technol., 2019, 9, 6841–6851
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
set to 10.9 g min−1 (1 equiv. of 4a), and the pump used to de-
liver the feedstock solution of DMC was set to 21.4 g min−1
(2.25 equiv.). Both streams came into contact in the first FM
of the Corning® Advanced-Flow™ G1 SiC reactor operated at
180 °C, and further reacted in the 5 other FMs, also operated
at 180 °C (2 min residence time, 60 mL total internal volume,
see Fig. 7). The outlet of the reactor consisted of a 1/4″ o.d.
PFA cooling loop immersed in a water bath and connected to
a dome-type back pressure regulator set at 11 bar. After equil-
ibration, the reactor effluent was collected, diluted with EtOH
and analyzed by GC/FID (4a conv. = 95%, 1a yield = 76%).
Fig.
2
Schematic continuous flow setup for the carbonation of
glycerol (4a) with dimethyl carbonate.
studied the upgrading of glycerol into glycidol using glycerol
carbonate as the reaction intermediate.47,52–54
Initial conditions involved a residence time of 3 min, a re-
actor temperature of 160 °C, 2.25 equiv. of DMC, and 1
mol% of organic salt catalyst. A control experiment without a
catalyst yielded a mere 5% glycerol conversion (Table 3, entry
1). Tetrabutylammonium bromide served as a reference cata-
lyst,44 and afforded glycerol carbonate (1a) with 63% conver-
sion and 95% selectivity (entry 2). Within the ammonium
bromides, short or long alkyl chains gave overall similar re-
sults in terms of conversion (56–63%) and selectivity (90–
95%, entries 2–4). Tetraoctylammonium bromide was insolu-
ble in both feeds and could not be assessed. Similarly, phos-
phonium bromides of various bulkiness gave glycerol conver-
sion in the 58–59% range and a selectivity varying between
87 and 91% (entries 5–7).
The nature of the anion impacted significantly the cata-
lytic activity. Changing the anion from bromide to chloride
had only a minor effect on both conversion (59%) and selec-
tivity (89%, entry 8). Tetrabutylammonium iodide and tetra-
butylammonium 2-hydroxybenzoate were insoluble in both
feeds and could not be assessed.
Results and discussion
Screening of organic salts as catalysts
The investigation began by screening different organic salts
for the model carbonation of neat glycerol (4a) with DMC un-
der continuous flow conditions. Several complex and expen-
sive ionic liquids/organic salts are already described as effi-
cient catalysts for this reaction in batch46–51 and flow.41 Only
cheap and commercially available catalytic materials were
considered in the present work (Table 3). Organic salts fea-
turing a hydroxide anion are in general air-sensitive, and
were thus not considered. For solubility reasons, the catalysts
were typically loaded within the glycerol feedstock solution.
Neat DMC was handled as a separate feed since it forms a
heterogeneous mixture with 4a. Both liquid feeds were mixed
through a T-mixer and reacted in a heated stainless-steel (SS)
capillary coil (Fig. 2). A back pressure regulator (BPR) was
inserted downstream the SS coil to maintain the reactor un-
der a counterpressure of 11 bar. Quantitative analyses were
carried out by off-line GC/FID analysis of the crude reactor ef-
fluent (ESI,† section S2.2). Under these conditions, glycerol
carbonate (1a) was the major product, although the reaction
mixture also contained minor side-products including
glycidol (3a) and methyl ((2-oxo-1,3-dioxolan-4-yl)methyl) car-
bonate (5a). Although 5a has no commercial application, 3a
is a high-value added building block, and several reports
Tetrabutylammonium acetate [NBu4IJOAc)] gave results
similar to those of NBu4Br, with 58% conversion and 88% se-
lectivity (entry 9). On the other hand, non-coordinating an-
−
−
ions such as triflate (CF3SO3 , OTf), tetrafluoroborate (BF4 )
−
and hexafluorophosphate (PF6 ) were barely active as catalysts
(entries 10–12). Similar observations were reported by Sairi
et al. in relation to the catalytic efficiency of various ionic liq-
uids for the conversion of 4a into 1a with DEC. Increasing
the hydrogen-bonding ability of the anion improved the activ-
ity of the catalyst. For instance, the authors reported that
1-butyl-3-methylimidazolium tetrafluoroborate gave a mere
<5% glycerol conversion, while 1-ethyl-3-methylimidazolium
acetate gave 94% conversion under the same conditions (120
°C, 2 h, 0.5 mol% of catalyst, and 2 equiv. of DEC).48 Gu and
coworkers reported that the catalytic activity of
hydroxypropyl-tripentylammonium hydroxide and halides for
the conversion of 4a into 1a increased with the basicity of the
anion. The yield of 1a as a function of the anion nature
followed the order: OH− (83%) > Cl− (21%) = Br− (21%) > I−
(11%), under identical reaction conditions (80 °C, 90 min,
0.9 mol% of catalyst, and 2 equiv. of DMC).49 Reports by
Kelkar and Selva on the carbonation of alcohols with DMC
catalyzed by ionic liquids suggested that both the nature of
the anion and of the cation dictate the catalytic activity. A
Table 3 Evaluation of organic salts as catalysts for the conversion of
glycerol (4a) into glycerol carbonate (1a)
Entrya
Catalyst
4a conv.b (%)
1a selec.b (%)
1
2
3
4
5
6
7
8
—
5
87
95
92
90
91
87
88
89
88
—
—
99
NBu4Br
NMe4Br
NEt4Br
63
56
61
59
59
58
59
58
0
PMe4Br
PBu4Br
PPh4Br
NBu4Cl
NBu4IJOAc)
NBu4IJOTf)
NBu4BF4
NBu4PF6
9
10c
11c
12c
0
5
a
b
Process conditions are described in Fig. 2. Conversion and yield
c
were determined by GC/FID. The catalyst is insoluble in glycerol
and was dissolved in the DMC feedstock solution.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6845

View Article Online
Paper
Catalysis Science & Technology
nucleophilic and electrophilic activation was proposed, where
the anion and the cation activate the alcohol and carbonyl
moieties, respectively.53,55 By contrast, Gu and coworkers
showed by FTIR the absence of interaction between a
propyl-tripentylammonium cation and the carbonyl moiety
of DMC.49 Among the organic salts evaluated in the present
work, the nature of the anion dramatically influenced the
catalytic activity, while the nature of the cation did not sig-
nificantly influence the outcome of the reaction. The
highest yield of 1a (60%) was obtained with tetra-
butylammonium bromide.
compounds with hydrogen bonding ability or basic sites. It is
therefore expected that an increase in the catalytic loading
leads to an increase in selectivity for side-product 3a. The se-
lectivity for methyl ((2-oxo-1,3-dioxolan-4-yl)methyl) carbonate
(5a) remained low (in the 0.1–1.2% range), regardless of the
conditions utilized. While slightly higher yields were
obtained with a 5 mol% loading of NBu4Br, 3.5 mol% gave a
good balance between conversion, selectivity and cost-effi-
ciency, and it was defined as the optimized value for further
optimization.
Next, the influence of the residence time on the reaction
was investigated (ESI,† section S2.3.1). An increase from 3 to 5
min merely improved the conversion of glycerol (4a),
irrespective of the DMC/4a molar ratio. Similarly, the selectiv-
ity for 1a was not affected by the residence time or the excess
of DMC. The selectivity for glycidol (3a) decreased with the in-
crease in residence time, which might be rationalized by deg-
radation reactions of 3a, such as oligomerization.56 The selec-
tivity for compound 5a followed an opposite trend, slightly
increasing with the residence time.21 A residence time of 3
min was selected as the optimized value, giving the highest
productivity for 1a (0.78 mol per day with 2.25 equiv. of DMC).
The influence of the temperature (150–180 °C) was also
evaluated (ESI,† section S2.3.2). Increasing the temperature
improved the conversion of glycerol (4a), but had no signifi-
cant effect on the selectivity for glycerol carbonate (1a) and
for glycidol (3a).52,54 By contrast, the selectivity for side-
product 5a slightly increased with the temperature.21
Process optimization
The parameters influencing the carbonation of glycerol (4a)
with DMC were next evaluated. The optimization concerned
the residence time, the temperature and the loading of NBu4-
Br, using different molar ratios between DMC and 4a.
Preliminary trials were carried out at 160 °C with a resi-
dence time of 3 min and an increasing DMC/4a ratio (1.4, 1.7
or 2.25). The effect of the catalytic loading (1, 2, 3.5 or 5
mol%) on the reaction was investigated. This parameter af-
fected both conversion and selectivity (Fig. 3). Increasing the
molar amount of NBu4Br in the glycerol feed from 1 to 5
mol% raised the conversion from 45 to 68% (with 1.4 equiv.
of DMC), from 50 to 74% (with 1.7 equiv. of DMC), and from
63 to 85% (with 2.25 equiv. of DMC). Although it had a posi-
tive impact on the conversion, the increase in NBu4Br loading
decreased the selectivity for glycerol carbonate (1a). It de-
clined from 93 to 84% (with 1.4 equiv. of DMC), from 92 to
82% (with 1.7 equiv. of DMC), and from 95 to 80% (with 2.25
equiv. of DMC). The decrease in selectivity for 1a indicates
the increasing formation of other side-products. The forma-
tion of glycidol (3a) increased with larger catalytic loadings.
The selectivity for 3a evolved with the catalytic loading from
1.8 to 7.4% (with 1.4 equiv. of DMC), from 2.9 to 8.5% (with
1.7 equiv. of DMC), and from 1.1 to 7.8% (with 2.25 equiv. of
DMC). A similar impact of the catalytic loading on the selec-
tivity for 1a and 3a was observed by others using various
ionic liquids as catalysts for the reaction between glycerol
and DMC/DEC.47,48,52–54 The carbonation of glycerol with
DMC/DEC toward 1a and the decarboxylation of 1a into
glycidol are consecutive reactions that are both catalyzed by
To sum up this round of optimizations, the highest pro-
ductivity for glycerol carbonate (95% conversion, 79% selec-
tivity, 0.85 mol per day) was obtained at 180 °C within 3 min
of residence in the presence of 3.5 mol% of NBu4Br and 2.25
equiv. of DMC.
Purification
Although glycerol carbonate can be purified by vacuum distil-
lation, the low volatility of 1a (b.p. = 145–148 °C at 0.2 mmHg
(ref. 41)) makes the procedure energy-intensive. In our hands,
vacuum distillation of a crude reactor effluent yielded only
glycidol (3a) through NBu4Br-catalyzed decarboxylation of
1a.47,52–54 An alternative purification procedure was sought
and a liquid–liquid extraction procedure was devised57 to
Fig. 3 Evolution of (a) glycerol conversion and (b) glycerol carbonate selectivity as a function of the NBu4Br loading and the excess of dimethyl
carbonate. Conditions: T = 160 °C, residence time = 3 min, P = 11 bar.
6846 | Catal. Sci. Technol., 2019, 9, 6841–6851
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Table 4 Optimization of the purification of glycerol carbonate (1a) by liquid–liquid extraction
Amount of compound extracted in the organic solvent (%)
Organic
solvent
Aqueous
phaseb
Volume of
organic solvent
Entrya
1a
3a
4a
5a
NBu4Br
1
2
3
4
5
6
7
8
9
MIBK
NaCl 1 wt%
H2O
H2O
NaCl 1 wt%
H2O
H2O
H2O
H2O
H2O
H2O
3 × 5 mL
3 × 5 mL
3 × 5 mL
3 × 5 mL
3 × 5 mL
3 × 5 mL
1 × 3 mL
3 × 5 mL
3 × 5 mL
3 × 5 mL
90
79
47
60
53
34
20
15
29
8
85
76
11
0
0
0
0
0
0
0
0
>99.9
95
98
59
16
28
98.7
94
83
8
AcOPr
AcOBu
AcOBu
CH2Cl2
CHCl3
CHCl3
Et2O
c
—
—
>99.9
>99.9
>99.9
>99.9
81
40
68
50
c
71
42
43
37
51
11
MTBE
Toluene
5
1.6
10
0
a
AcOBu = n-butyl acetate, AcOPr = n-propyl acetate, Et2O = diethyl ether, MIBK = methyl isobutyl ketone, MTBE = methyl tert-butyl ether. See
ESI, section S2.4 for the detailed procedure. b Volume = 1 mL. c Not determined.
separate unconverted 4a, side-products 3a and 5a, and NBu4-
Br from the target compound 1a. model solution
Eventually, 1a was obtained in 60% yield (based on glycerol)
after removal of MIBK under reduced pressure. The target
compound was contaminated with traces of 3a (3.1 mol%
with respect to 1a), 4a (2.7 mol% with respect to 1a), 5a
(0.6 mol% with respect to 1a) and NBu4Br (1.9 mol% with
respect to 1a).
Overall, our methodology not only enabled the purifica-
tion of glycerol carbonate without vacuum distillation or col-
umn chromatography, but was also amenable to the partial
recycling and purification of the homogeneous catalyst.
A
containing the above-mentioned compounds was prepared,
and several extraction solvents were screened (Table 4 and
ESI,† section S2.4). Glycerol (4a) was quantitatively retained
in the aqueous phase, irrespective of the organic solvent uti-
lized. Only MIBK was able to extract a small fraction (11%) of
4a from the aqueous phase (entry 1). By contrast, side-
product 5a was efficiently extracted (68–99.9%) by most sol-
vents (entries 1–7, 9) except for diethyl ether and toluene
(40% and 50% extracted, entries 8 and 10, respectively). In
general, oxygenated solvents, such as n-propyl acetate, n-butyl
acetate, diethyl ether and MTBE, were slightly more efficient
for the extraction of glycerol carbonate (1a) and glycidol (3a)
than that of NBu4Br (entries 2–4, 8–9). Chlorinated solvents
gave the most interesting results, as they were able to selec-
tively extract NBu4Br, while retaining most of 1a in the aque-
ous phase (entries 5–7). Upon optimization of the extraction
volumes, chloroform enabled extraction of 83% of tetra-
butylammonium bromide, while retaining 80% of 1a in the
aqueous phase (entry 7).
Based on these results, we applied the methodology on a
crude reactor effluent (ESI,† section S2.5). At first, excess
DMC and by-product MeOH were removed under reduced
pressure. Then, the crude was diluted with water and
washed with CHCl3 to remove side-product 5a as well as
NBu4Br. The chloroform phase could be further processed
to recycle the catalyst. The aqueous phase was then loaded
with 1 wt% of NaCl, and glycerol carbonate was extracted
with MIBK, enabling its separation from unconverted 4a.
Extension to refined and crude bio-based glycerol
Glycerol is a by-product from the soap, fatty acid and biodie-
sel industries, and there is a significant difference in price
between crude and refined glycerol. In 2013–2014, prices of
crude and refined glycerol were about 240 USD per ton and
900–965 USD per ton, respectively.31 Refined glycerol remains
the privileged substrate in the primary literature, but there is
considerable economic interest in the direct upgrading of
crude 4a. In this context, we assessed our conditions on sam-
ples of refined and crude bio-based glycerol (Table 5). As
expected, the sample of refined bio-based glycerol afforded
conversion (96%) and selectivity (78%) similar to those of the
commercial sample utilized for the process optimization (en-
tries 1–2). Samples of crude 4a containing various amounts
of water, methanol, salts and organic matter (ESI,† Tables S4
and S5) yielded the target compound 1a with a significantly
lower conversion (30 and 44%, entries 3 and 4). On the other
hand, the selectivity for glycerol carbonate (1a) was excellent
Table 5 Assessment of bio-based glycerol samples for the preparation of glycerol carbonate (1a)
Entrya
Glycerol supplier
Glycerol grade (purity)
4a conv.b (%)
1a selec.b (%)
1
2
3
4
Honeywell
Cargill
Cargill
Commercial
95
96
30
44
79
78
98
71
Refined bio-based (99.5%)
Crude bio-based (85%)
Crude bio-based (84%)
Valtris Champlor SAS
a
b
Process conditions: temperature = 180 °C, residence time = 3 min, 3.5 mol% of NBu4Br, 2.25 equiv. of DMC, P = 11 bar. Conversion and
yield were determined by GC/FID.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6847

View Article Online
Paper
Catalysis Science & Technology
significantly contributes to decreasing the conversion with
crude glycerol. Overall, the process would require some re-
optimization to attain high conversion of 4a, but is viable for
the upgrading of crude bio-based glycerol.
Substrate scope
With the best conditions and catalyst in hand, the scope of
the reaction was extended to a large variety of 1,2-diols (ESI,†
Table S6). Simple 1,3-dioxolan-2-ones, such as propylene (1c)
and 1,2-butylene (1d) carbonates, were obtained with unsatis-
factory conversion of the starting material (77 and 79%,
respectively). The process conditions were thus slightly
adapted. Increasing the DMC/diol molar ratio from 2.25 to 3
gave excellent results on diol 2c (conv. = 95%, selec. = 98%)
and 2d (conv. = 98%, selec. = 92%). These adapted conditions
were next applied to a wide range of 1,2-diols 2b–t (Fig. 5).
Liquid diols 2b–d, 2f–k, 2m–o and 2r, s were processed under
neat conditions, while feed solutions of solid diols 2e, 2l, 2p,
2q and 2t were prepared in DMSO (1–4 M) to ensure continu-
ous operation.
Fig. 4 Continuous flow setup for the assessment of the stability of
glycerol carbonate (1a) in the presence of water.
with the sample from Cargill (98%, entry 3) and good with
the sample from Valtris Champlor SAS (71%, entry 4). In the
latter case, the lower selectivity could not be ascribed to the
formation of volatile side-products, since only 4a and 1a were
detected by GC/FID.
Water is present in significant amounts in the crude sam-
ples of glycerol and most likely influences the reaction in a
negative way. It probably (a) competes with the alcohol moie-
ties of 4a in binding with the anion of the catalyst, (b) ham-
pers the mass transfer between glycerol and the dimethyl car-
bonate phase (especially in the presence of salts), and (c)
hydrolyses glycerol carbonate (1a) back into glycerol (4a). A
consequence of the hydrolysis of 1a is the formation of CO2
as a gaseous by-product, decreasing the residence time of the
liquid phase within the reactor. An important evolution of
gas was indeed released after the back pressure regulator.
Methanol is present in low amounts in the samples of crude
glycerol assessed and likely does not influence the equilib-
rium of the reaction significantly. The detrimental effect of
water on the conversion of glycerol into glycerol carbonate
was also reported by other authors.41,52,58,59 To evaluate the
significance of 1a hydrolysis on the outcome of the reaction,
several control experiments were performed under the opti-
mized conditions with glycerol carbonate as the substrate in
the presence of increasing amounts of water (Fig. 4 and
Table 6). Reacting 1a with DMC in the absence of water
yielded methyl ((2-oxo-1,3-dioxolan-4-yl)methyl) carbonate (5a)
as the main product (21% yield, entry 1). Increasing the
amount of water in the feedstock solution of 1a from 2.5 to 10
wt% decreased the yield of 5a from 12% to 0% (entries 2–4).
By contrast, increasing the water content from 2.5 wt% to
20 wt% triggered the hydrolysis of 1a into glycerol up to 43%
(entries 2–5). These results suggest that the hydrolysis of 1a
Very high to excellent yields (82–93%) were obtained for
the preparation of 1,3-dioxolan-2-ones of types 1b–e with in-
creasing carbon chain length. 3-Butene-1,2-diol (2f), which
can be sourced from bio-based erythritol,60 was somehow less
reactive and gave vinyl ethylene carbonate (1f) in 68% yield.
A two-fold increase in the residence time merely improved
the yield (71%). The scope also encompassed a wide range of
glycerol ethers 2g–l, which have a promising forecast as ver-
satile renewable building blocks.56 Methyl (1g, 91%),
tert-butyl (1h, 86%), allyl (1i, 95%), propargyl (1j, 78%),
furfuryl61 (1k, 85%) and o-tolyl (1l, 86%) ethers of glycerol
carbonate were obtained in high to excellent yields. The prep-
aration of furfuryl ether 1k in high yield required a longer
residence (9 min). Introduction of a dimethylamino or
morpholino group vicinal of the 1,2-diol pattern did not alter
its reactivity, and carbonates 1m and 1n were obtained in
82% and 92% yields, respectively. A previous report utilizing
urea as a carbonation reagent reported only 15% yield for
carbonate 1m,19 while another study described the synthesis
of compound 1n in 99% yield upon utilization of DMC as the
reagent.24 2,3-Butanediol (2o) was found to be less reactive
than the parent 1,2-butanediol (2d) under these conditions,
and 4,5-dimethyl-1,3-dioxolan-2-one (1o) was obtained in 45%
yield.21 A further increase of the residence time to 6 min
slightly increased the yield to 51%. The presence of two sec-
ondary alcohols that are not cycle-strained in a relative cis-
configuration (see below) most likely accounted for the lower
reactivity of diol 2o toward DMC.55 cis-Cyclohexanediol (2p)
was converted into the target cyclic carbonate 1p in good
yield (73%). On the other hand, a control experiment on
trans-cyclohexanediol (2q) yielded methyl carbonate 1q as the
main product in 46%, emphasizing that a relative cis-
configuration is required for the reaction to proceed on cycle-
strained diols. An analogous observation was made by Beller
et al. upon reacting trans-cyclopentanediol with urea, which
Table 6 Assessment of the stability of glycerol carbonate (1a) in the
presence of dimethyl carbonate and an increasing amount of water
Water contentb
(wt%)
1a conv.c
(%)
5a yieldc
(%)
4a yieldc
(%)
Entrya
1
2
3
4
5
0
2.5
5
10
20
28
14
4.6
9
21
12
1.2
0
0
0
2.5
9
43
43
0
a
b
Process conditions are described in Fig. 4. Relative to glycerol
carbonate. c Conversion and yield were determined by GC/FID.
6848 | Catal. Sci. Technol., 2019, 9, 6841–6851
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Fig. 5 Substrate scope for the carbonation of 1,2-diols 4a, 2b–t toward cyclic carbonates 1a–t. Unless otherwise specified, the diol + catalyst feed-
stock solution was pumped neat. Yields were determined by GC/FID or by 1H NMR with mesitylene as the internal standard (isolated yields are indi-
b
c
d
cated in parentheses). a 2.25 equiv. of DMC. The diol was pumped as a 4 M solution in DMSO. Residence time = 9 min. 2l was pumped as a 3 M
solution in DMSO. e 3.5 mol% of NBu4Br and 3 equiv. of DMC per 1,2-diol functional group. f 2t was pumped as a 1 M solution in DMSO.
yielded the corresponding carbamate as the main product.19
Bio-based anhydroerythritol (2r), bearing also a vicinal diol in
a relative cis configuration, was carbonated into the corre-
sponding bicyclic 1,3-dioxolan-2-one 1r in excellent 95% yield.
Lastly, we considered the preparation of dicarbonates 1s and
1t that are potential bio-based monomers for polyurethanes.3–5,8
Preliminary experiments (ESI,† Table S6) on diol 2s evidenced
that a residence time of 9 min was necessary to afford
dicarbonate 1s in good yield (73%). Applying the same condi-
tions to 2t yielded the hydroxymethylfurfural-derived
dicarbonate 1t in 69%.
process monitoring. The spectra acquired during the carbon-
ation of 2c into propylene carbonate (1c) are depicted in
Fig. 6. The doublet at 0.8 ppm was used as a probe of the
concentration of 2c, while the doublet at 1.2 ppm and the
multiplet at 3.7–4.8 ppm enabled the monitoring of carbonate
1c. GC/FID was utilized as a complementary tool to obtain
In-line NMR monitoring
Real-time analysis for pollution prevention is one of the
twelve principles of green chemistry.62 In-line NMR has re-
cently emerged as a versatile tool for real-time analytical mon-
itoring of continuous processes.21,63 In this context, a low-
field in-line benchtop NMR spectrometer was inserted down-
stream the reactor, and the temperature of the reactor was
progressively decreased from 180 to 100 °C. These control ex-
periments were implemented to mimic a possible process fail-
ure, and thus evaluate the potential for real-time qualitative
Fig. 6 Representative low-field in-line NMR spectra obtained during the
carbonation of 2c into 1c in the 100–180 °C range. The doublet at 0.8
ppm was used to monitor the variation of 2c concentration (light grey).
The doublet at 1.2 ppm and the multiplet at 3.7–4.8 ppm were used to
monitor the variation of 1c concentration (dark grey). Conditions: resi-
dence time = 3 min, P = 11 bar, 3.5 mol% of NBu4Br, 3 equiv. of DMC.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6849

View Article Online
Paper
Catalysis Science & Technology
straightforward liquid–liquid extraction protocol. Refined
and crude bio-based glycerol were also evaluated as sub-
strates, and promising results were obtained. Control experi-
ments on glycerol carbonate suggested that hydrolysis of glyc-
erol carbonate becomes significant when water is present in
the medium at a high level. By contrast to the prior Art, the
methodology was extended to a wide palette of substrates (20
examples, broad molecular diversity) that were converted in
overall good to excellent yields (45–95%). The efficiency of
low-field in-line NMR for real-time analytical monitoring of
the process was exemplified. The scalability of the procedure
for the carbonation of glycerol was next documented in a
pilot-scale continuous flow reactor of 60 mL internal volume,
and similar results to those on the microscale were obtained.
The pilot-scale reactor yielded a daily productivity of 115 mol
(13.6 kg) of glycerol carbonate.
Fig. 7 Photograph of the Corning® Advanced Flow™ G1 SiC reactor.
Key: H = inlet and outlet for the heat exchanging fluid; I1 = inlet for
DMC; I2 = inlet for preheated glycerol and NBu4Br; FM = fluidic module.
quantitative results (ESI,† Table S6) and similar trends were
observed. However, sample preparation and a prolonged anal-
ysis time were required for GC/FID, while in-line NMR analy- Conflicts of interest
sis provided real-time qualitative monitoring of the carbon-
ation procedure. This approach was also extended to the
There are no conflicts to declare.
carbonation of diols 2b and 2d (ESI,† Fig. S8 and S9).
Acknowledgements
Scale-up
This work was supported by the European Regional Develop-
The scalability of this procedure was assessed in a commer-
cial pilot-scale continuous flow reactor (Corning® Advanced
Flow™ G1 SiC reactor, Fig. 7) of 60 mL internal volume. The
reactor featured 6 silicon carbide (SiC) fluidic modules (FMs)
of 10 mL internal volume each, connected in series and inte-
grated with heat exchangers. Glycerol (4a) was selected as the
model substrate, since 1a is the most industrially relevant tar-
ment Fund (ERDF) and Wallonia within the framework of the
program “Wallonie-2020.EU” (INTENSE4CHEM, project no.
699993-152208). The authors thank Thomas Toupy for the
experimental support. Cargill and Valtris Champlor SAS are
gratefully acknowledged for providing bio-based glycerol
samples.
get. The same process conditions as those on the microscale Notes and references
(180 °C, 3 min, 3.5 mol% of NBu4Br, and 2.25 equiv. of DMC)
1 J. H. Clements, Ind. Eng. Chem. Res., 2003, 42, 663–674.
2 B. Schäffner, F. Schäffner, S. P. Verevkin and A. Börner,
Chem. Rev., 2010, 110, 4554–4581.
3 H. Blattmann, M. Fleischer, M. Bähr and R. Mülhaupt,
Macromol. Rapid Commun., 2014, 35, 1238–1254.
4 L. Maisonneuve, O. Lamarzelle, E. Rix, E. Grau and H.
Cramail, Chem. Rev., 2015, 115, 12407–12439.
5 J. Datta and M. Włoch, Polym. Bull., 2016, 73, 1459–1496.
6 W. Guo, J. E. Gómez, À. Cristòfol, J. Xie and A. W. Kleij,
Angew. Chem., Int. Ed., 2018, 57, 13735–13747.
were initially applied, and similar results (4a conversion =
98%, 1a selectivity = 74%) were obtained, corresponding to a
9.0 kg per day (76.4 mol per day) productivity for glycerol car-
bonate (1a). In order to further increase the productivity, the
residence time was lowered to 2 min, and excellent results
were obtained as well (4a conversion = 95%, 1a selectivity =
80%). This data point corresponds to a daily productivity of
115 mol (13.6 kg) in carbonate 1a, demonstrating the poten-
tial application to the industrial scale.
7 H. Büttner, L. Longwitz, J. Steinbauer, C. Wulf and T.
Werner, Top. Curr. Chem., 2017, 375, 50.
Conclusion
This work reports a robust and low-footprint continuous flow
procedure for the carbonation of various 1,2-diols toward
5-membered cyclic organic carbonates (1,3-dioxolan-2-ones).
Dimethyl carbonate is utilized as a low-toxicity reagent, and
tetrabutylammonium bromide as a cheap and widely avail-
able homogeneous catalyst. The conditions feature a low cat-
alytic loading (3.5 mol%), short reaction times (3–9 min) and
solvent-free or highly concentrated media. The reaction con-
ditions were optimized on the model synthesis of glycerol
carbonate, affording the target compound in 79% selectivity
with 95% conversion. Glycerol carbonate was purified and
8 F. De La Cruz-Martínez, J. Martínez, M. A. Gaona, J.
Fernández-Baeza, L. F. Sánchez-Barba, A. M. Rodríguez, J. A.
Castro-Osma, A. Otero and A. Lara-Sánchez, ACS Sustainable
Chem. Eng., 2018, 6, 5322–5332.
9 G. Sienel, R. Rieth and K. T. Rowbottom, Epoxides in
Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH
Verlag GmbH & Co. KGaA, Weinheim, Germany, 2000, DOI:
10.1002/14356007.a09_531.
10 N. Kindermann, T. Jose and A. W. Kleij, Top. Curr. Chem.,
2017, 375, 15.
11 J.-Y. Li, Q.-N. Zhao, P. Liu, D.-S. Zhang, Q.-W. Song and K.
Zhang, Greenhouse Gases: Sci. Technol., 2018, 8, 803–838.
the homogeneous catalyst was partially recovered by
a
6850 | Catal. Sci. Technol., 2019, 9, 6841–6851
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
12 R. S. Hanslick, W. F. Bruce and A. Mascitti, Org. Synth.,
1953, 33, 74–75.
13 W. Guerin, A. K. Diallo, E. Kirilov, M. Helou, M. Slawinski,
J.-M. Brusson, J.-F. Carpentier and S. M. Guillaume,
Macromolecules, 2014, 47, 4230–4235.
14 X. M. Yu, H. Han and B. S. J. Blagg, J. Org. Chem., 2005, 70,
5599–5605.
15 S. Böhm and M. Beth-Hübner, Chloroformic Esters in
Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH
Verlag GmbH & Co. KGaA, Weinheim, Germany, 2006, DOI:
10.1002/14356007.a06_559.pub2.
37 M. G. Álvarez, M. Plíšková, A. M. Segarra, F. Medina and F.
Figueras, Appl. Catal., B, 2012, 113–114, 212–220.
38 D. O. Nogueira, S. P. de Souza, R. A. C. Leão, L. S. M. Miranda
and R. O. M. A. de Souza, RSC Adv., 2015, 5, 20945–20950.
39 R. A. C. Leão, S. P. de Souza, D. O. Nogueira, G. M. A. Silva,
M. V. M. Silva, M. L. E. Gutarra, L. S. M. Miranda, A. M.
Castro, I. I. Junior and R. O. M. A. de Souza, Catal. Sci.
Technol., 2016, 6, 4743–4748.
40 S. Guidi, R. Calmanti, M. Noè, A. Perosa and M. Selva, ACS
Sustainable Chem. Eng., 2016, 4, 6144–6151.
41 S. Van Mileghem, W. M. De Borggraeve and I. R. Baxendale,
Chem. Eng. Technol., 2018, 41, 2014–2023.
16 K. M. Engstrom, A. Sheikh, R. Ho and R. W. Miller, Org.
Process Res. Dev., 2014, 18, 488–494.
17 D. M. Pearson, N. R. Conley and R. M. Waymouth, Adv.
Synth. Catal., 2011, 353, 3007–3013.
42 M. A. do Nascimento, L. E. Gotardo, R. A. C. Leão, A. M. de
Castro, R. O. M. A. de Souza and I. Itabaiana Jr., ACS Omega,
2019, 4, 860–869.
18 Y. Chen, C. A. Hone, B. Gutmann and C. O. Kappe, Org.
Process Res. Dev., 2017, 21, 1080–1087.
19 M. Peña-López, H. Neumann and M. Beller, Eur. J. Org.
Chem., 2016, 3721–3727.
43 For a review on the toxicity of ionic liquids and organic
salts, see: K. S. Egorova and V. P. Ananikov, ChemSusChem,
2014, 7, 336–360.
44 R. A. Grey, US Pat. 5091543, 1992.
20 E. R. Baral, J. H. Lee and J. G. Kim, J. Org. Chem., 2018, 83,
11768–11776.
45 M. Selva, A. Caretto, M. Noè and A. Perosa, Org. Biomol.
Chem., 2014, 12, 4143–4155.
21 Z. Wang, R. Gérardy, G. Gauron, C. Damblon and J.-C. M.
Monbaliu, React. Chem. Eng., 2019, 4, 17–26.
22 K. M. Tomczyk, P. A. Guńka, P. G. Parzuchowski, J. Zachara
and G. Rokicki, Green Chem., 2012, 14, 1749–1758.
23 G. D. Yadav and P. S. Surve, Catal. Sci. Technol., 2013, 3,
2668–2676.
24 D. Nakatake, Y. Yokote, Y. Matsushima, R. Yazaki and T.
Ohshima, Green Chem., 2016, 18, 1524–1530.
25 P. Tundo, M. Musolino and F. Aricò, Green Chem., 2018, 20,
28–85.
46 C. Chiappe and S. Rajamani, Pure Appl. Chem., 2012, 84, 755–762.
47 M. K. Munshi, P. S. Biradar, S. M. Gade, V. H. Rane and
A. A. Kelkar, RSC Adv., 2014, 4, 17124–17128.
48 Z. I. Ishak, N. A. Sairi, Y. Alias, M. K. T. Aroua and R. Yusoff,
Chem. Eng. J., 2016, 297, 128–138.
49 E. Elhaj, H. Wang and Y. Gu, Mol. Catal., 2019, 468, 19–28.
50 Y. Yi, Y. Shen, J. Sun, B. Wang, F. Xu and R. Sun, Chin. J.
Catal., 2014, 35, 757–762.
51 X. Wang, P. Zhang, P. Cui, W. Cheng and S. Zhang, Chin. J.
Chem. Eng., 2017, 25, 1182–1186.
26 G. Fiorani, A. Perosa and M. Selva, Green Chem., 2018, 20,
288–322.
27 M. Selva, A. Perosa, D. Rodríguez-Padrón and R. Luque, ACS
Sustainable Chem. Eng., 2019, 7, 6471–6479.
52 S. M. Gade, M. K. Munshi, B. M. Chherawalla, V. H. Rane
and A. A. Kelkar, Catal. Commun., 2012, 27, 184–188.
53 M. K. Munshi, S. M. Gade, V. H. Rane and A. A. Kelkar, RSC
Adv., 2014, 4, 32127–32133.
28 J. R. Ochoa-Gómez, O. Gómez-Jiménez-Aberasturi, C.
Ramírez-López and M. Belsué, Org. Process Res. Dev.,
2012, 16, 389–399.
29 M. O. Sonnati, S. Amigoni, E. P. Taffin de Givenchy, T.
Darmanin, O. Choulet and F. Guittard, Green Chem.,
2013, 15, 283–306.
54 Y. Zhou, F. Ouyang, Z.-B. Song, Z. Yang and D.-J. Tao, Catal.
Commun., 2015, 66, 25–29.
55 M. Selva, M. Noè, A. Perosa and M. Gottardo, Org. Biomol.
Chem., 2012, 10, 6569–6578.
56 M. Sutter, E. Da Silva, N. Duguet, Y. Raoul, E. Métay and M.
Lemaire, Chem. Rev., 2015, 115, 8609–8651.
30 W. K. Teng, G. C. Ngoh, R. Yusoff and M. K. Aroua, Energy
Convers. Manage., 2014, 88, 484–497.
31 P. U. Okoye and B. H. Hameed, Renewable Sustainable
Energy Rev., 2016, 53, 558–574.
32 A. Galadima and O. Muraza, Waste Biomass Valorization,
2017, 8, 141–152.
33 S. Christy, A. Noschese, M. Lomelí-Rodriguez, N. Greeves
and J. A. Lopez-Sanchez, Curr. Opin. Green Sustain. Chem.,
2018, 14, 99–107.
34 D. Liu and E. Y.-X. Chen, Green Chem., 2014, 16, 964–981.
35 R. Gérardy, N. Emmanuel, T. Toupy, V.-E. Kassin, N. N.
Tshibalonza, M. Schmitz and J.-C. M. Monbaliu, Eur. J. Org.
Chem., 2018, 2301–2351.
36 R. Gérardy, R. Morodo, J. Estager, P. Luis, D. P. Debecker
and J.-C. M. Monbaliu, Top. Curr. Chem., 2019, 377, 1.
57 J. R. Ochoa-Gómez, O. Gómez-Jiménez-Aberasturi, C. Ramírez-
López and B. Maestro-Madurga, Green Chem., 2012, 14, 3368–3376.
58 V. P. Indran, A. S. H. Saud, G. P. Maniam, Y. H. Taufiq-Yap
and M. H. A. Rahim, Waste Biomass Valorization, 2017, 8,
1049–1059.
59 M. Tudorache, A. Negoi, B. Tudora and V. I. Parvulescu,
Appl. Catal., B, 2014, 146, 274–278.
60 N. N. Tshibalonza, R. Gérardy, Z. Alsafra, G. Eppe and
J.-C. M. Monbaliu, Green Chem., 2018, 20, 5147–5157.
61 E. Dolci, G. Michaud, F. Simon, B. Boutevin, S. Fouquay and
S. Caillol, Polym. Chem., 2015, 6, 7851–7861.
62 P. T. Anastas and J. C. Warner, Green chemistry: theory and
practice, Oxford University Press, New York, 1998.
63 P. Giraudeau and F.-X. Felpin, React. Chem. Eng., 2018, 3,
399–413.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6841–6851 | 6851