Towards selective electrochemical conversion of glycerol to 1,3-propanediol

Article abstract of DOI:10.1039/c8ra00711j

1,3-propanediol (1,3-PD) is a bulk chemical with myriad applications in polymers, lubricants, cosmetics, foods industries and in the synthesis of heterocyclic compounds. Current commercial production of 1,3-PD involves a thermocatalytic process using acrolein (DuPont) and ethylene oxide (Shell) as starting feedstock. These feedstocks are petroleum-based and there are many efforts at using glycerol as low cost biomass-derived feedstock for 1,3-PD production. A number of catalyst designs and bacterial & fungal strains are being explored for respective catalytic and fermentation routes to glycerol-to-1,3-PD. However, the electrochemical method received little attention for the purpose. In this work, in order to explore the possibility of using partly refined glycerol byproduct of biodiesel production as feedstock, we investigated conversion and 1,3-PD selectivity of glycerol electrolysis in chloride media. We demonstrated selective glycerol-to-1,3-PD conversion using Pt or RuO₂-based dsa as anode and Zn or Pb as cathode in NaCl and KCl at pH 1. This electrochemical glycerol-to-1,3-PD conversion is not only green, it is a potential process network loop between biodiesel production and chlor-alkali industry.

Full text of DOI:10.1039/c8ra00711j

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Towards selective electrochemical conversion of
glycerol to 1,3-propanediol†
Cite this: RSC Adv., 2018, 8, 10818
Olusola O. James, *ab Waldemar Sauter and Uwe Schroder*
a
a
¨
1,3-propanediol (1,3-PD) is a bulk chemical with myriad applications in polymers, lubricants, cosmetics,
foods industries and in the synthesis of heterocyclic compounds. Current commercial production of 1,3-
PD involves a thermocatalytic process using acrolein (DuPont) and ethylene oxide (Shell) as starting
feedstock. These feedstocks are petroleum-based and there are many efforts at using glycerol as low
cost biomass-derived feedstock for 1,3-PD production. A number of catalyst designs and bacterial &
fungal strains are being explored for respective catalytic and fermentation routes to glycerol-to-1,3-PD.
However, the electrochemical method received little attention for the purpose. In this work, in order to
explore the possibility of using partly refined glycerol byproduct of biodiesel production as feedstock, we
investigated conversion and 1,3-PD selectivity of glycerol electrolysis in chloride media. We
Received 23rd January 2018
Accepted 1st March 2018
2
demonstrated selective glycerol-to-1,3-PD conversion using Pt or RuO -based dsa as anode and Zn or
DOI: 10.1039/c8ra00711j
Pb as cathode in NaCl and KCl at pH 1. This electrochemical glycerol-to-1,3-PD conversion is not only
green, it is a potential process network loop between biodiesel production and chlor-alkali industry.
rsc.li/rsc-advances
available as a byproduct of the oleochemical industry and enjoys
wide application as moisturizer in cosmetic products. Increasing
1
Introduction
Despite the current low price of petroleum, non-oil producing biodiesel productions in recent times has brought about an
1
2,13
nations have concerns about the long term security of the oversupply of glycerol for its traditional applications.
Hence,
petroleum supply. Moreover, climate and global warming there are efforts at channeling the oversupply of glycerol towards
remain key drivers to replace the current petroleum based serving as feedstock to valuable C chemicals. Examples of
transportation and organic chemical industries with renewable derivable high value C chemicals from glycerol include:
1–3
3
3
4,5
14–16
resources. Presently, the entire transportation sector depends propylene glycols, hydroxyacids, and diacid.
Among the
almost exclusively on petroleum. Also, petroleum provides the possible glycerol derivable C chemicals, 1,3-propanediol (1,3-
3
feedstocks and the energy inputs for the petrochemical industry PD) is of special interest because it is a bulk chemical that has
for production of basic and ne organic chemicals. The wide applications in polymers, lubricants, cosmetics, foods and
17
economy of the petrochemical industry is built around the high- in the synthesis of heterocyclic compounds.
energy density, abundance and low cost of petroleum. These set
The traditional routes of commercial synthesis of 1,3-PD use
of attributes present formidable economic challenges for acrolein (DuPont) and ethylene oxide (Shell) as starting feed-
adoption of biofuels and renewable substitutes to petroleum in stock. The DuPont route involves hydration of acrolein to 3-
the transportation sector. However, conversion of biomass- hydroxypropionaldehyde (3-HPA), followed by hydrogenation of
derived feedstocks into high value commodity chemicals is 3-HPA to produce 1,3-PD. While shells route involves hydro-
gaining interest as a more plausible path to reduce dependence formylation of ethylene oxide to 3-hydroxypropanal and subse-
6–9
18
on petroleum. And where possible, use of renewable energy quent hydrogenation to 1,3-PD. These traditional commercial
input for biomass conversion will be an added step towards routes to 1,3-PD depend exclusively on petroleum-derived
waning reliance on fossil fuels.
Glycerol has been identied as an important biomass-derived cialized a glucose-based process route to 1,3-PD. The process
feedstocks. DuPont and Tate & Lyle developed and commer-
10,11
feedstock to a variety of commodity chemicals.
It is readily involves fermentation of corn syrup using genetically modied
e. coli with addition of vitamin B . This bioconversion route to
1
2
1
,3-PD is reported to consume 40% less energy than the tradi-
tional petroleum-based route. However, the need for vitamin
12 addition has an impact on the economics of the glucose-
a
Institute of Environmental and Sustainable Chemistry, Technical University,
Hangenring 30, 38106, Braunschweig, Germany. E-mail: uwe.schroeder@
B
tu-braunschweig.de
b
19
Chemistry Unit, Kwara State University, Malete, P.M.B. 1530, Ilorin, Nigeria. E-mail: based process route to 1,3-PD. Moreover, this bio-process
olusola.james@kwasu.edu.ng
route to 1,3-PD involved glycerol as an intermediate, which
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/c8ra00711j
1
10818 | RSC Adv., 2018, 8, 10818–10827
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
further motivates interest in glycerol to 1,3-PD considering chambers were separated via a cation exchange membrane
oversupply of glycerol from biodiesel production.
(fumasep FKE, Fumatech, Germany).
Although the glycerol-to-1,3-PD fermentation process is
being explored, the productivity so far obtained is lower
compared to the glucose-based fermentation. In recent times
there are a number reports on heterogeneous catalytic hydro-
2
.4 Analysis
Quantitative analyses were performed by high performance
liquid chromatography (HPLC), by means of a refractive index
20–22
genolysis of glycerol to 1,3-propanediol (1,3-PD).
This route
requires high temperature and pressure conditions and
hydrogen supply which at present is most economically
produced via steam reforming of natural gas or a petroleum
feedstock. In this study, we examined electrochemical conver-
sion of glycerol to 1,3-PD. Since 1,3-PD is a bulk chemical with
many applications, electrochemical glycerol-to-1,3-PD conver-
sion is not only a green and sustainable route to 1,3-PD but also
it can serve as a means of storing excess renewable electricity.
(
RI) detector (Spectrasystem P4000, Finnigan Surveyor RI Plus
Detector, Fischer Scientic) equipped with HyperREZ XP
carbohydrate H + 8 mm (S/N: 026/H/012-227) column. Sulphuric
acid (0.005 N, ow rate 0.5 ml min ) was used as eluent. The
column was operated at different temperature depending on
sample composition, the refractory index detector was operated
ꢀ1
ꢁ
at 42 C. Glycerol conversions and products selectivities are
calculated as follows:
moles of glycerol consumed
Glycerol conversion ðmol%Þ ¼
2
Experimentals
moles of starting glycerol
ꢂ 100
2.1 Chemicals
All chemicals used in this study were of analytical grade. For
qualitative and quantitative analysis, reference materials and
solvents were used as purchased, without purication. Glycerol,
dihydroxyacetone, glyceric acid, glyceraldehyde, acetol, 1,3-
propanediol, acetone, 2-propanol and 1-propanol.
moles of product formed
Product selectivity ðmol%Þ ¼
moles of glycerol consumed
ꢂ 100
3
Results and discussion
2.2 Electrode materials
3.1 Electrode selectivity requirements
AgCl, sat. KCl electrode was used (SE11, Sensortechnik Meins- It is interesting to note that 3-hydroxypropionaldehyde (3-HPA)
berg, Germany), 0.197 vs. standard hydrogen electrode (SHE) as is a common intermediate in 1,3-PD production by conven-
reference. All used electrodes are listed in Table 1 with their tional Dupont commercial synthesis of 1,3-PD from acrolein,
respective purity and surface area.
glycerol metabolism (fermentation) and heterogeneous catalytic
glycerol hydrogenolysis (Scheme 1). The fermentation and the
catalytic hydrogenolysis conversion routes followed the same
tandem dehydration and hydrogenation steps to 1,3-PD. This
common feature in the routes seems to be an indication of 3-
HPA as a critical intermediate for glycerol to 1,3-PD conversions,
and the electrochemical method may be no exception. Ther-
mocatalytic hydrogenolysis of glycerol oen involves tandem
dehydration and hydrogenation. These thermochemical reac-
tion steps are difficult to replicate electrochemically, especially
the dehydration step. Hence direct electro-hydrogenolysis of
glycerol appears electrochemically inaccessible. Moreover, as
shown in the reaction mechanism of the fermentation method,
dihydroxyacetone (DHA) intermediate could constitute a drain
on 1,3-PD productivity in an electrochemical method. Thus,
none of the routes in Scheme 1 is replicable for achieving
electrochemical conversion of glycerol to 1,3-PD.
2.3 Electrochemical procedure
All electrochemical reactions were conducted under potentiostatic
control using a potentiostat/galvanostat SP50 (Bio-Logic SAS, Claix,
Frankreich). All experiments used a three-electrode conguration
and were stirred continuously with a magnetic stirrer. The undi-
vided cell is a single 40 ml chamber cylindrical glass cell with
a 1 cm neck window tted with a Teon cork stopper bearing and
the anode and cathode terminals are 2 cm apart. The divided cell is
a 40 ml each, two-chamber H-type glass cell. The anode and
cathode terminals are 10 cm apart, the anode and the cathode
Table 1 Purity and properties of the electrodes used
Electrode
surface (cm )
2
The Clemmensen reduction of carbonyl groups is well
documented in organic chemistry for achieving conversion of
Electrode material
Purity
a
23
Lead (Pb)b
99.999
99.99
99.9
11
11
11
11
a carbonyl group to a methylene bridge. In the conventional
Zinc (Zn)
Clemmensen reduction, zinc is typically used as a stoichio-
metric reductant. However, in cathodic Clemmensen-type
reduction electrons are the reductant while zinc plays a cata-
lytic role. The cathodic Clemmensen-type reduction has been
reported for other metallic (lead, cadmium and mercury) elec-
Platinum
Ruthenium based
dimensional stable
c
anode (RUA)
a
b
ChemPUR, Germany. Umicore Galvanotechnik GmbH, Germany.
2
4–27
c
trodes;
and it has been exploited for biomass-to-biofuel
Ti Anode Fabricators Pvt. Ltd, Chennai – 126, India.
28,29
production.
Thus, it is a potent electrochemical step for
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 10818–10827 | 10819

View Article Online
RSC Advances
Paper
Scheme 1 Mechanistic representation of reaction route to 1,3-Propanediol via glycerol fermentation, catalytic glycerol hydrogenolysis, and
conventional Dupont commercial synthesis of 1,3-PD from acrolein.
achieving a conversion of a carbonyl group to a methylene potential of creating a network loop of electrochemical chlor-
group. So, tandem selective oxidation of glycerol to DHA, and alkali cell, glycerol production (from biodiesel production)
Clemmensen-type cathodic reduction of DHA seems a plausible and electrochemical glycerol-to-1,3-PD conversion. Thus, for
reaction route for achieving glycerol-to-1,3-PD electrochemical simplicity, we are inclined to explore carrying out the glycerol
conversion. However, there are other pathways that compete oxidation in chloride media without the complication of
with Clemmensen mechanism in cathodic reductions of alpha- a TEMPO catalyst. Combination of glyceraldehyde selectivity
hydroxycarbonyl compounds. Carbonyl-to-hydroxyl or dehy- and stability in chloride media requirements place stringent
droxylation of alpha-hydroxyl group(s) to a carbonyl group had criteria on selection of the anode for achieving the electro-
been identied as competing pathways to Clemmensen mech- chemical conversion. Potential candidates that can meet the
anism in cathodic reductions of alpha-hydroxycarbonyl selectivity requirement are platinum, gold and ruthenium
3
0,31
compounds.
tion active metal electrodes (Pb and Zn) in acidic media (HCl, is prone to corrosion in chloride media and platinum is
SO and H PO ) showed that electroreduction of DHA fol- expensive and may be cost prohibitive for large scale applica-
lowed the dehydroxylation path. On the other hand, the dehy- tion. RuO -dsa is already an established anode in the chlor-
2
Preliminary studies using Clemmensen reduc- based dimensionally stable anode (RuO -dsa). The Gold anode
H
2
4
3
4
2
33,34
droxylation path favours electroreduction of glyceraldehyde to alkali industry.
,3-PD in neutral and acidic pH. This is in line with deduction dsa is selective to chlorine formation and aqueous chlorine
from the fermentation route that a selective electrochemical species are an potential oxidant for glycerol oxidation. Hence,
glycerol-to-1,3-PD process should avoid DHA as an interme- RuO -dsa was investigated rst to ascertain anodic glyceralde-
diate. Therefore, selective anodic oxidation of glycerol to glyc- hyde selectivity in our quest towards selective electrochemical
In chloride media, ruthenium oxide-based
1
2
eraldehyde is critical to achieve high 1,3-PD productivity.
The nature of electrolyte is an important variable in elec-
conversion of glycerol to 1,3-PD.
trochemical conversions. Anodic oxidation can via direct elec- 3.2 Glycerol electrolysis using RuO -dsa anode and Pb & Zn
2
tron transfer from the substrate to the anode or indirectly cathode
through an anodically generated oxidant species. Regardless of
In order to ascertain anodic glyceraldehyde selectivity we
the mechanism of the anodic oxidation, the main target at the
investigated the oxidation of glycerol in chloride media using
anode is to achieve selective glycerol-to glyceraldehyde oxida-
RuO -dsa anode. Fig. 1 shows glycerol conversions and product
2
tion. Glycerol is made up of two primary and one secondary
hydroxyl groups. Selective glycerol to glyceraldehyde oxidation
selectivities in a divided cell. Although the ideal anodic result is
100% glyceraldehyde specicity, the results obtained are less
will entails preferential oxidation one of the primary hydroxyl
groups of glycerol over its secondary hydroxyl group. Preferen-
tial chemical oxidation of a primary hydroxyl group over
a secondary hydroxyl group had been reported using aqueous
hypochlorite as oxidant and 2,2,6,6-tetramethylpiperidinyl-1-
oxy (TEMPO), catalyst. And using appropriate anode hypo-
chlorite and other aqueous chlorine species can be anodically
generated. Moreover, the use of chloride media has the
than the ideal glyceraldehyde selectivities. The glyceraldehyde
selectivities obtained are sufficiently high enough to feed the
coupled cathodic dehydroxylation to 1,3-PD. The results of the
tandem glycerol oxidation and cathodic dehydroxylation in
chloride media over Pb and Zn electrodes are presented in Fig. 2
and 3 respectively. The neutral media displayed lower glycerol
conversions compared to the HCl medium. Also 1,3-PD selec-
tivity is higher in NaCl than in KCl medium. While higher
32
10820 | RSC Adv., 2018, 8, 10818–10827
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 1 Conversion and selectivity of oxidation of 0.25 M glycerol in a divided cell using Ti–RuO
2
working electrode and Pt counter electrode in
ꢁ
0
.5 M chloride solutions (potential 2.5 V vs. AgCl/Cl sat. KCl, temperature 25 C).
glycerol conversions were obtained in HCl medium, the Zn ZnCl . So the Pb electrode benets from protective surface
2
cathode vs. RuO -dsa anode gave higher glycerol conversion coverage of PbCl which reduced the rate of Pb corrosion in the
2
2
than the Pb cathode vs. RuO
2
-dsa anode. The detailed reason for reaction environment.
this observation is not clear at the moment. However, it is
The observed mass gain of Pb electrode suggests adsorption
observed that the Zn electrode was corroded as indicated by of edduct or counter ion (chloride ion) on the Pb electrode
about 8% mass loss and a high zinc concentration was found in surface. Zn lacks similar surface protection and may become
the reaction cell content aer the electrolysis. Although mass impractical as a cathode for a long time or large scale electrol-
gain of 0.15% was observed for the Pb electrode, ICP-OES ysis in such a highly corrosive reaction environment. The
analysis of the reaction cell content also indicates corrosion of cathodic redeposition restores the Pb electrode much more
the Pb electrode (Tables S1a and b†). As shown in Fig. 4 efficiently than the Zn electrodes. Although the corrosion
increased surface roughness was observed for the Zn electrode reaction is a drain on the current efficiency of the glycerol
while that of Pb became smoother aer the electrolysis. This is electrolysis, it may be benecial to renewal of the electrode
an indication of surface reconstruction during the course of the surface which may account for the higher glycerol conversion
cathode half reaction. The increased roughness of the Zn elec- obtained in the electrolysis using the Zn cathode.
trode aer the electrolysis suggests dissolution and re-
Purication of crude glycerol (from biodiesel synthesis or
deposition of Zn as dendrite-like micro-crystals. Although triglyceride saponication) usually involves an initial acidica-
35
dissolution of Pb and Zn in acidic media is thermodynamic tion to pH 1. Direct usage of the partially rened glycerol up to
downhill, moreover the chlorine oxidant species are corrosive to this stage will constitute a cheap feedstock for glycerol elec-
Pb and Zn. PbCl has a smaller solubility product constant than trolysis to 1,3-PD. Encouraged by higher 1,3-PD productivity of
2
Fig. 2 Conversion and selectivity of oxidation of 0.25 M glycerol in an undivided cell using Pb working electrode and Ti–RuO
in 0.5 M chloride solutions (potential ꢀ1.8 V vs. AgCl/Cl sat. KCl, temperature 25 C).
2
counter electrode
ꢁ
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 10818–10827 | 10821

View Article Online
RSC Advances
Paper
Fig. 3 Conversion and selectivity of oxidation of 0.25 M glycerol in a undivided cell using Zn working electrode and Ti–RuO
in 0.5 M chloride solutions (potential 2.5 V vs. AgCl/Cl sat. KCl, temperature 25 C).
2
counter electrode
ꢁ
Zn compared to Pb in neutral media, we tested the tolerance of equalize over the course of the electrolysis. This may be
Zn to pH 1. There was no signicant difference in the mass of Zn explained by a proposed pathway for the glycerol electrolysis
plate electrode before and aer 10 h of electrolysis. With depicted in Scheme 2.
reference to the neutral media glycerol conversion was unaf-
fected by lowering the pH of the media to pH 1 using the Pb and
the Zn electrode. However, with the Zn electrode, 1,3-PD selec- 3.3 Glycerol electrolysis using Pt anode vs. Pb and Zn
tivity was enhanced at pH 1. Similar enhancement of 1,3-PD cathode
productivity was not observed for Pb electrode (Fig. S3 and S4†).
Despite initial high glyceraldehyde and low DHA selectivities of
glycerol oxidation obtained in Fig. 1, selectivities of reduced
products of DHA (acetol) and glyceraldehydes (1,3-PD) tend to
Electrochemical conversion of glycerol has been a subject of
many studies over the past three decades. Fuel cell application
was the main driver of earlier studies on electrooxidation of
glycerol. Large proportions of the studies were carried in
Fig. 4 Confocal micrograph image of the surface of the Zn and Pb as working electrode before and after electrolysis 0.25 M glycerol in an
ꢁ
undivided cell with Ti–RuO
2
counter electrode in 0.5 M HCl solution (potential ꢀ1.8 V sat'd AgCl/Cl, Temperature 25 C).
10822 | RSC Adv., 2018, 8, 10818–10827
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
alkaline media. Few of the studies were in acidic media. platinum surface. The oxides platinum surface increase glyc-
Generally, glycerol oxidation typically shows higher activity in eraldehyde adsorption on the electrode and promotes further
36
alkaline media. Glycerate, dihydroxyacetone, and tartronate oxidation to glyceric acid. This tendency nonetheless, for
are reaction products over noble metals in alkaline media. The a head-to-head comparison with RuO2 dsa, we carried out
pathway of glycerol oxidation in alkaline media over a gold glycerol oxidation in chloride media using platinum at 2.5 V vs.
electrode had been exploited for electricity-chemical cogenera- Ag/AgCl. Expectedly glyceric acid selectivities obtained were
37
tion. Selective electrochemical oxidation of glycerol at plat- considerable; glycerol oxidation using platinum gave similar
inum has enjoyed focus of several theoretical and experimental glyceraldehyde selectivities in neutral pH but higher glyceral-
reports. In acidic media, the pathway of glycerol oxidation over dehyde selectivity in HCl compared to that obtained using RuO
platinum surface is via the primary hydroxyl group, and modi- dsa. DHA selectivity was lower using platinum than RuO dsa,
2
2

cation of the platinum surface with Bi or Sb can tune the especially in KCl (Fig. S4†). For both platinum and RuO dsa,
2
38,39
pathway towards the secondary hydroxyl group.
material balances of the reaction with time are attributable to
It has been shown that electrooxidation of glycerol in acidic undesired over oxidation of glycerol to CO
media over Pt is selective to glyceraldehyde at potentials below products that were not monitored.
2
and other side
40
oxygen evolution. The glyceraldehyde selectivity was attributed
The high glyceraldehyde selectivity obtained in electro-
to selective dehydrogenation of a primary hydroxyl group of oxidation of glycerol in HCl using platinum did not translate
glycerol oxidation via weakly adsorbed OH species on platinum into expected high 1,3-PD selectivity in the corresponding
and fast desorption of glyceraldehyde from Pt surface. This is glycerol electrolysis when platinum (as anode) was coupled with
a desirable selectivity for achieving the objective of selective lead or zinc (as cathode). Notably, the electrolysis in HCl gave
glycerol-to-1,3-PD electrochemical conversion. However, glyc- very low 1,3-PD selectivity and glycerol conversion was lowest
erol conversions are very low at the potentials below oxygen using zinc (Fig. S5 and S6†). As explained in Section 3.2 with
2
evolution. High glycerol conversion is obtainable at potentials RuO dsa, zinc is subject to severe corrosion in HCl. However,
above oxygen evolution. But high glycerol conversion could be platinum is less selective to chloride oxidation due to compet-
compromised with glyceraldehyde selectivity. This is because at itive adsorption between chloride and glycerol on the platinum
potential above oxygen evolution, stable oxides form on the surface. The amount of chlorine species produce during the
Scheme 2 Possible pathways of tandem glycerol oxidation and cathodic reduction (Pb and Zn).
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 10818–10827 | 10823

View Article Online
RSC Advances
Paper
Fig. 5 Conversion and selectivity of oxidation of 0.25 M glycerol in a undivided cell using Pb working electrode and Pt counter electrode in 0.5 M
ꢁ
chloride solutions acidified with HCl to pH 1 (potential ꢀ1.8 V vs. AgCl/Cl sat. KCl, temperature 25 C).
electrolysis in HCl is lower using platinum than with RuO dsa. equally not simple. Hence, detailed understanding of the
2
Nonetheless, the zinc electrode corrodes rapidly and its rede- results presented in Fig. 5 and 6 is not clear at the moment.
position competes with the desired cathode dehydroxylation of However, effect of pH change at near surface of the cathodes is
the glyceraldehyde intermediate. As with RuO
2
dsa considerable an important factor. Effect of local pH changes at the surface of
1
,3-PD selectivities were obtained with platinum in neutral electrodes during electrolysis that are accompanied by
41
media (Fig. S5 and S6†). Glycerol electrolysis at pH 1 was also hydrogen or oxygen evolution had been advocated. Youngkook
investigated using platinum anode with Pb and Zn cathode. As and Koper used the effect of local pH changes at the electrode
shown in Fig. 5 and 6, Pt vs. Pb exhibited greater selectivity to surface to explain the mechanism of cathodic reduction of
42
1,3-PD in NaCl, while Pt vs. Zn is more suited to 1,3-PD selec- glucose on metal electrodes in Na SO solution. Kwon and
2 4
tivity in KCl. The proposed scheme of anticipated series of Koper advanced that hydrogen evolution via the reaction: 1/
ꢀ
reaction steps of the glycerol electrolysis at pH 1 towards the 2H O + 2e / 1/2H + OH , increases the local surface cathode
2
2
desired selective glycerol-to-1,3-PD conversion is represented in pH. The authors added that surface hydroxide ions promoted
Scheme 3. Although the dynamics of reactions at the anodic isomerisation of glucose from the cyclic (inactive) form to the
component of the electrolysis is well anticipated, the dynamics open chain structures (aldehyde and enol) that are active
of reactions at the cathodic component of the electrolysis is towards cathodic transformations.
Fig. 6 Conversion and selectivity of oxidation of 0.25 M glycerol in a undivided cell using Zn working electrode and Pt counter electrode in 0.5 M
ꢁ
chloride solutions acidified with HCl to pH 1 (potential ꢀ1.8 V vs. AgCl/Cl sat. KCl, temperature 25 C).
10824 | RSC Adv., 2018, 8, 10818–10827
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Scheme 3 Proposed scheme of reaction steps of the glycerol electrolysis at pH 1 in chloride media.
A similar electrode local pH changes effect may be invoked to condition is achieved at pH 1. It is interesting that the
explain cathodic transformations of glyceraldehyde interme- compromise may allows for direct usage of crude glycerol from
diate media aspect of the present glycerol electrolysis in chlo- biodiesel synthesis as cheap feedstock for a commercial elec-
ride media. But with the addition, that glyceraldehyde exists in trochemical glycerol-to-1,3-PD process. Most commercial bio-
solution as equilibrium between hydrated (inactive) form and diesel production uses alkalis (NaOH, KOH, which are
free (active) form. Driving the equilibrium towards the cathodic produced by chlor-alkali process), as catalysts. The crude
active form involves dehydration which is traditionally an acid glycerol byproduct of the biodiesel process contains: the alkali
catalysed transformation. Cathodic hydrogen evolution causes catalyst, methanol and other coproducts such as diglycerides
the increase of the local pH at the cathode surface. The local and soap. Acidication of the crude glycerol to pH 1 will bring
high pH condition promotes another equilibrium loop in the about precipitation of free fatty acid/soap and diglycerides
reaction media between glyceraldehyde, DHA and an enol which can allow for the separation of these impurities by
intermediate (Scheme 4). Free glyceraldehyde and its enol forms decantation. The resulting glycerol still contains methanol,
are dehydroxylated to 1,3-PD or reduced back to glycerol. Low which may be recovered by distillation. The glycerol will
1,3-PD selectivities obtained in the glycerol electrolyses in HCl require further purication steps to obtain cosmetic or phar-
are indication to importance of local high pH condition to maceutical grade glycerol. However, this partly rened glycerol
achieving high 1,3-PD selectivity. Also it suggests that the enol is suitable for use in the proposed electrochemical glycerol-to-
form is cathodically more active than the free glyceraldehyde 1,3-propandiol process in this study. Also, most chlor-alkali
form. At the same time the tendency towards full isomerisation installations use dimensionally stable anodes. One of the
to DHA creates a drain on 1,3-PD selectivity.
electrodes investigated in this study. If a biodiesel production
While low pH conditions favour glycerol-to-glyceraldehyde, plant is sited near a chlor-alkali installations, it is possible that
the anodic half of the glycerol electrolysis, subsequent the proposed glycerol-to-1,3-PD may be able to share facilities
glyceraldehyde-to-1,3-PD transformation requires a subtle with chlor-alkali installations. Thus like in the petrochemical
local cathodic acid-base equilibrium. It appears a compromise industry, a network loop between chlor-alkali plant and
Scheme 4 Proposed mechanism of cathodic transformation of glyceraldehyde.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 10818–10827 | 10825

View Article Online
RSC Advances
Paper
Fig. 7 Schematic representation of potential process loop connecting chlor-alkali cell, biodiesel synthesis and glycerol-to-1,3-PD electrolysis.
biodiesel plant can be envisaged (Fig. 7). Thus, the glycerol-to-
Conflicts of interest
1
,3-PD process can be integrated into an existing chlor-alkali
plant, thereby representing a sustainable and green route to There are no conicts to declare.
,3-PD.
1
Acknowledgements
4
Conclusion
Olusola O. James acknowledged Alexander von Humbolt
Foundation, Germany, for Goerg Forster Postgraduate Fellow-
ship Award. O. O. James is also grateful to Mr V. P. Jekakumar,
Managing Director, Ti Anode Fabricators Pvt. Ltd, Chennai –
We have examined glycerol electrolysis in chloride media.
Selective glycerol-to-1,3-PD transformation requires a selective
glycerol-to-glyceraldehyde anodic half reaction coupled to
a subsequent selective to glyceraldehyde-to-1,3-PD cathodic half
126, India, for providing the DSA anode used in this study at
zero cost.
2
reaction. RuO -based dsa and platinum satised the glyceral-
dehyde selectivity and the stability in chloride as selective
electrode for the anodic half of the electrolysis. Lead and zinc
also displayed cathodic dehydroxylation activity of the hydrox-
ycarbonyl intermediates. Although high glyceraldehyde selec-
tivity was obtained in HCl, it is unsuitable for the cathodic local
acid–base dynamics. The purely acidic medium gave high
glycerol conversion and low 1,3-PD selectivity, while high 1,3-PD
selectivity was obtained in the neutral chloride media but at low
glycerol conversion. An optimized conversion and selectivity
was obtained in acidied chloride media (pH 1), which may
allow for direct usage of crude glycerin from biodiesel synthesis
for the electrolysis. This study showed that glycerol electrolysis
in chloride media is a potential green and sustainable industrial
route to 1,3-PD production.
References
1 G. M. Richard and R. S. Steven, Philos. Trans. R. Soc., A, 2014,
372, DOI: 10.1098/rsta.2013.0179.
2 I. Chapman, Energy Policy, 2014, 64, 93–101.
3 B. Roger and B. Yongmei, Energy Policy, 2015, 86, 880–890.
4 N. Stefan, H. Hans and G. Henrike, Biotechnol. J., 2007, 2,
1505–1513.
5 K. H. Thushara, Y. Zahira and N. B. Narayanan, Renewable
Sustainable Energy Rev., 2015, 42, 1234–1244.
6 L. Tao, F. R. Luis, N. S. Yogendra, C. H. Alan and K. C. Ting,
Bioresour. Technol., 2014, 156, 256–266.
7 J. J. S. Adrie, Chem. Rev., 2014, 114(3), 1871–1908.
10826 | RSC Adv., 2018, 8, 10818–10827
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
8
Z. Yanan, H. Guiping and C. B. Robert, Bioresour. Technol., 25 S. Sherlock Jr, L. F. Deditiusc and W. A. Pyhrr, J. Electrochem.
014, 157, 28–36. Soc., 1935, 68(1), 321–327.
U. K. Esra, P. T. Antoine and L. Yu, J. Chem. Technol. 26 S. Sherlock Jr, R. W. Benoliel, L. R. Lyons and W. H. Pahl, J.
Biotechnol., 2015, 90, 1364–1379. Electrochem. Soc., 1941, 79(1), 83–91.
0 M. Anitha, S. K. Kamarudin and N. T. Koi, Chem. Eng. J., 27 S. Sherlock Jr, E. I. Onstott and F. H. Baastad, J. Electrochem.
016, 295, 119–130. Soc., 1955, 102(3), 113–116.
1 L. Xiaolan, G. Xumeng, C. Shaoqing and L. Yebo, Bioresour. 28 P. Nilges, T. R. dos Santos, F. Harnisch and U. Schroder,
Technol., 2016, 215, 144–154. Energy Environ. Sci., 2012, 5, 5231–5235.
2 A. G. Q. Cesar, J. R. C. Christian and A. C. Joao Jr, Renewable 29 X. Le, Z. Zhiyong, Q. Ji, J. C. David, Q. Yang, M. W. Kayla and
Sustainable Energy Rev., 2013, 27, 475–493. L. Wenzhen, ChemSuSChem, 2013, 6(4), 674–686.
3 C. Rosaria, D. P. Cristina, R. Michele and P. Mario, Eur. J. 30 M. Fedoro nˇ ko, J. Konigstein and K. Linek, Collect. Czech.
Lipid Sci. Technol., 2014, 116, 1432–1439. Chem. Commun., 1965, 30, 4297–4306.
4 S. K. Pei, K. A. Mohamed and M. A. W. D. Wan, Renewable 31 M. Fedoro nˇ ko, Chem. Pap., 1987, 41(6), 767–776.
2
9
1
1
1
1
1
1
1
1
2
Sustainable Energy Rev., 2016, 63, 533–555.
5 M. Andreas, A. Udo, G. Inaki and L. A. Pedro, Eur. J. Lipid Sci.
Technol., 2013, 115, 9–27.
32 R. Siedlecka, J. Skarzewski and J. Młochowski, Tetrahedron
Lett., 1990, 31(15), 2177–2180.
33 C. Ruiyong, T. Vinh, S. Bernd, N. Harald, K. Jurgen,
B. Andreas, W. Rainer and H. Rolf, Z. Phys. Chem., 2013,
227, 651–666.
6 E. S. Vasiliadou and A. A. Lemonidou, WIREs Energy Environ.,
2015, 4, 486–520.
7 L. Hongjuan, X. Yunzhen, Z. Zongming and L. Dehua, 34 R. K. Karlsson and A. Cornell, Chem. Rev., 2016, 116(5), 2982–
Biotechnol. J., 2010, 5(11), 1137–1148.
3028.
1
1
8 A. K. George, Clean: Soil, Air, Water, 2008, 36(8), 648–651.
9 S. Yaqin, et al., 1,3-Propanediol, in Bioprocessing of
35 X. Yang, X. Guomin and V. Arvind, Ind. Eng. Chem. Res., 2013,
52, 14291–14296.
Renewable Resources to Commodity Bioproducts, ed. 36 K. Youngkook, J. P. S. Klasas and M. T. M. Koper,
Virendra, S. B. and Akihiko, K., John Wiley & Sons, 1st edn,
014.
0 S. Daolai, Y. Yasuhiro, S. Satoshi and U. Wataru, Appl. Catal.,
B, 2016, 193, 75–92.
ChemCatChem, 2011, 3(7), 1176–1185.
37 X. Le, Z. Zhiyong, W. Zhichao and L. Wenzhen,
ChemCatChem, 2012, 4(8), 1105–1114.
38 K. Youngkook, B. Yuvraj, S. Ioannis, R. Paramaconi and
M. T. M. Koper, ACS Catal., 2012, 2(5), 759–764.
2
2
2
1 S. P. Samudrala, B. Ponnala, P. K. Vanama, K. D. Deepa,
R. P. Selvakannan, K. B. Suresh, L. K. Mannepalli and 39 L. Seonhwa, J. K. Hyung, J. L. Eun, K. Youngmin, N. Yuseong,
V. R. C. Komandur, ACS Sustainable Chem. Eng., 2016, 4,
212–1222.
W. H. George and B. K. Won, Green Chem., 2016, 18, 2877–
2887.
1
2
2
2
2 Z. Wei, Z. Yujun, W. Yue, W. Shengping and M. Xinbin, 40 R. S. Goncalves, W. E. Triaca and T. Rabockai, Anal. Lett.,
ChemCatChem, 2016, 8, 3663–3671. 1985, 18(8), 957–973.
3 W. B. Motherwell and C. J. Nutley, Contemp. Org. Synth., 41 A. T. Kuhn and C. Y. Chan, J. Appl. Electrochem., 1983, 13,
1994, 1, 219–241.
189–207.
4 S. Sherlock Jr, J. Electrochem. Soc., 1932, 62(1), 177–182.
42 K. Youngkook and M. T. M. Koper, ChemSusChem, 2013, 6,
455–462.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 10818–10827 | 10827