Microbial electrosynthesis of butyrate from carbon dioxide

Article abstract of DOI:10.1039/c4cc10121a

This work proves for the first time the bioelectrochemical production of butyrate from CO₂ as a sole carbon source. The highest concentration of butyrate achieved was 20.2 mMC, with a maximum butyrate production rate of 1.82 mMC d^-1. The electrochemical characterisation demonstrated that the CO₂ reduction to butyrate was hydrogen driven. Production of ethanol and butanol was also observed opening up the potential for biofuel production.

Full text of DOI:10.1039/c4cc10121a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Microbial electrosynthesis of butyrate from
carbon dioxide†
Cite this: Chem. Commun., 2015,
1, 3235
5
R. Ganigue,‡ S. Puig,*‡ P. Batlle-Vilanova, M. D. Balaguer and J. Colprim
´
Received 18th December 2014,
Accepted 9th January 2015
DOI: 10.1039/c4cc10121a
www.rsc.org/chemcomm
This work proves for the first time the bioelectrochemical production of authors have attempted to upgrade it to higher value products.
butyrate from CO as a sole carbon source. The highest concentration of Sharma and co-workers investigated the biocatalysed reduction
2
butyrate achieved was 20.2 mMC, with a maximum butyrate production of acetic and butyric acid to bioalcohols and mid-chain fatty acids
À1
rate of 1.82 mMC d . The electrochemical characterisation demon- via direct electron transfer at À0.65 V vs. standard hydrogen
strated that the CO reduction to butyrate was hydrogen driven. electrode (SHE). They observed the transformation of those into a
2
Production of ethanol and butanol was also observed opening up the mixture of products, including 0.8 mM of methanol, 0.2 mM of
potential for biofuel production.
ethanol, 0.4 mM of propanol, 0.6 mM of butanol and 0.2 mM of
acetone, as well as lower amounts of propionic and caproic
7
The depletion of fossil resources, their ever-increasing high price acids. In a similar way, Steinbusch et al. (2010) studied the
and the negative environmental impacts derived from their use are bioelectrochemical ethanol production through mediated acetate
forcing the transition to more sustainable energy and chemical reduction with methyl viologen. They reached 13.5 mM of
8
production models, based on renewable and carbon-neutral alcohols as well as C4 compounds. Finally, Eerten-Jansen et al.
commodity chemicals and fuels. Carbon dioxide (CO
transformed into added-value products mainly by chemical reduce acetate, obtaining 6.8 mM of caproate and 3.0 mM of
transformations, photochemical, chemical and electrochemical butyrate as main products. The present work proves for the first
2
) can be (2013) poised the cathode potential at À0.9 V vs. SHE to biologically
9
reductions, biological conversions, reforming, and inorganic time the bioelectrochemical concomitant production of acetate and
1
transformations. Current CO mitigation and conversion techniques butyrate from CO . Butyrate is an industrial feedstock with many
2
2
require extremely large surface and volumes, energy intense applications in the pharmaceutical and chemical industries, and can
2
10
processing steps and/or chemicals and expensive catalysts.
be converted into fuels through esterification.
Two experiments were performed to prove the bioelectro-
chemical production of butyrate from CO . These were conducted in
In such systems, carboxydotrophic microorganisms are harnessed to 240 mL two-chambered H-type BES. A cathode made of commercial
Microbial electrosynthesis has been recently postulated as a
3
promising approach to transform CO
2
into value-added compounds.
2
s
2
fixate CO into products via the Wood–Ljungdahl pathway, using carbon cloth (NuVant’s ELAT LT2400W, FuelCellsEtc, USA), with
electrical current as a driving force. This concept was first proven by an area of 9 cm and an area to volume ratio of 0.075 cm mL , was
2
2
À1
Nevin et al. (2010), who transformed CO into acetate using pure used as a working electrode. An Ag/AgCl (+0.197 V vs. SHE, model
2
4
cultures, up to a concentration of 2 mM. Two subsequent studies RE-5B, BASI, United Kingdom) was also placed in the cathodic
from Marshall and co-workers increased, using mixed cultures, the chamber as a reference electrode, whereas a titanium rod
5,6
À2
acetate titer to 28.5 mM and 175 mM, respectively.
To date, acetate has been the sole product of CO
(Ti plus 50 g m Pt, Magneto, The Netherlands) served as a
2
reduction in counter electrode in the anodic compartment. Both chambers
BioElectrochemical Systems (BESs). However, the autotrophic were filled with 120 mL of mineral medium similar to ATCC
2
1
production of acetate is not very attractive from the economic 1754 (Tanner et al., 1993), containing 2-bromoethanesulfonate
standpoint due to its low market price. In this light, several to inhibit methanogenesis (see ESI,† Table S1). These compart-
ments were separated by a cationic exchange membrane (CMI-
7
000, Membranes International Inc., USA) and stirred to avoid
LEQUIA, Institute of the Environment, University of Girona, Campus de Montilivi,
Girona, Catalonia, E-17071, Spain. E-mail: sebastia@lequia.udg.cat
mass transfer limitations. The cathode compartment had two
butyl-rubber sampling ports. Finally, the cells were wrapped with
a coil of plastic tubing connected to a thermostatic bath to
control the operational temperature. Temperatures for first and
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cc10121a
These authors contributed equally.
‡
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 3235--3238 | 3235

View Article Online
Communication
ChemComm
second experiments were 35.5 Æ 2.5 1C and 33.9 Æ 1.1 1C,
respectively.
The biocathode was poised at a potential of À0.8 V vs. SHE
using a potentiostat (BioLogic, Model SP50, France), based on a
three-electrode configuration. Software from the same producer
(EC-Lab v10.37) was used to run simultaneous multitechnique
electrochemistry routines, which included cyclic voltammetry
(
CV) and chronoamperometry (CA). The parameters for CV are
À1
as follows: scan rate: 1 mV s ; E
i
f
= À0.8 V vs. SHE; E = 0.0 V vs.
SHE. For CA, the parameters were Ecathode (E) = À0.8 V vs. SHE.
Initially, a control experiment was performed in 120 mL airtight
serum bottles to prove that no bacterial growth or metabolite
production occurred when reducing power was not provided.
Bottles were filled with 100 mL of mineral medium and 5 mL of
enriched carboxydotrophic mixed culture from a syngas (32% CO,
3
2% H
2 2 2
, 8% CO and 28% N ) fermenting lab-scale reactor capable
of producing two-carbon (C2) and four-carbon (C4) organic acids
11
and alcohols. The inoculum was dominated by species of the
genus Clostridium (see ESI,† Table S2). Bottles were regularly
sparged with CO up to a pressure of 2 atm, and incubated at
2
37 1C for a period of 60 days. During that period, no increase in
optical density or concentration of products was observed.
Subsequently, a BES H-type system was inoculated using 5 mL of
the same enriched carboxydotrophic mixed culture from a syngas
Fig. 1 (A) Current density, total Coulombs supplied and total accumulated
concentration of products in the BES. (B) Accumulated concentration of the
different products and pH in the biocathode. Black inverted triangles on the
fermenting lab-scale reactor, and operated as described. Pure CO
2
(Praxair, Spain) was bubbled regularly every 2–3 days for a period of
2
top part of the figure indicate sampling and subsequent flushing with CO .
5
minutes to ensure substrate availability (at 33.9 1C, concentration
at saturation of 26.9 mM of carbon (mMC) as CO ). Liquid samples
2
were taken periodically from both anodes and cathodes for the reducing power, butyrate production may be favoured over acetate
13
monitoring of liquid products concentration and pH. The volume by acetogens because of a slightly higher pK
withdrawn was substituted with fresh mineral medium. The concen- To validate that butyrate can be produced from CO
tration of organic acids and alcohols was analysed using an Agilent a second experiment was conducted with the same set-up and
890A gas chromatograph (Agilent, USA) equipped with a DB-FFAP conditions. Fig. 2 shows that current demand exponentially
a
(4.82 vs. 4.76).
2
in a BES,
7
column and a flame ionisation detector. Gas samples were increased throughout the experiment, with current densities
À2
periodically drawn for the analysis of hydrogen, carbon dioxide, increasing from 3 to 20 A m . The concentration of total
methane, oxygen and nitrogen (H , CO , CH , O , N ) in the added-value products increased in parallel, reaching 104 mMC
2
2
4
2
2
biocathode headspace. Samples were analysed using a second on day 34. The rate of carbon fixation into products was linear
À1
channel of the GC equipped with an HP-Molesieve column and a from day 7 to day 27, around 4.1 mMC d . Initially, acetate was
thermal conductivity detector (TCD).
the only compound produced, and its maximum concentration –
Next, Fig. 1 depicts the total accumulated concentration of around 55 mM – was reached on day 20. The evolution of pH in
products, coulombs supplied, together with the product speciation the biocathode was linked to the production of organic acids,
and pH at the cathode during a 19 days experiment. The demand and it decreased from pH 7 on day 7 to pH 6.35 on day 20. At
of electrons increased linearly from day 0 until the end of the that point, butyrate was detected, together with ethanol. The
experiment. A total of 7000 coulombs were supplied to the system, simultaneous presence of both compounds in the fermentation
À2
with a current demand of around 6.3 A m on day 17. At the broth opens up a new potential route for butyrate production.
end of the experiment the total concentration of products was Clostridium kluyveri, which had been detected in the parent
1
1
5
5 mMC. Microbial electrosynthesis commenced with the syngas fermentation reactor, can produce butyrate by the
production of acetate, and its concentration increased almost combination of acetate and ethanol, by the so-called chain
1
4,15
linearly throughout the experiment, reaching a maximum concen- elongation reactions.
In this light, it is unclear which is
tration of 40 mMC. Butyrate was first observed on day 3, and during the true mechanism governing butyrate production, or whether
À1
the following 9 days was produced at a rate of 1.49 mMC d , up to both metabolic routes occurred in parallel.
a maximum concentration of 15 mMC. Butyrate is one of the end
The production of butyrate subsequently led to the decrease
products of Clostridium carboxidivorans/Clostridium ragsdalei, of the concentration of acetate and the pH, which reached
bacteria present in the inoculum, and could have been directly values of pH 4.63 on day 34. In parallel, the concentration of
produced from CO
2
via the Wood–Ljungdahl pathway coupled to ethanol and butanol increased to 30.8 mMC and 7.3 mMC,
1
2
acetyl-CoA reduction. Under acidic conditions and excess of respectively. It is hypothesized that the decrease of pH, coupled
3
236 | Chem. Commun., 2015, 51, 3235--3238
This journal is ©The Royal Society of Chemistry 2015

View Article Online
ChemComm
Communication
likely that production rates were limited by the low solubility of
CO and H , being the mass transfer of these compounds the key
2
bottleneck of syngas fermentation technology. In this light, MES
of butyrate in BES has the potential to circumvent this barrier by
the in situ supply of reducing power, either in the form of
2
electrons or H .
Turnover CVs were periodically carried out to characterise
electrochemically both BESs. Fig. 3 presents two representative
CVs performed under turnover conditions, one corresponding
to the control (abiotic) test and the second one when the
biofilm had fully developed, producing organic acids (acetate
and butyrate) and bioalcohols (ethanol and butanol) from CO2.
A sudden increase of the current demand was observed on the
abiotic test at a cathode potential of À0.65 V vs. SHE. This CV
2
shape is typically linked to the catalytic production of H using
16
non-precious metals (i.e. carbon cloth) as electrodes. During the
production of organic acids and alcohols, hydrogen production
started before the control CV, at À0.55 V vs. SHE. The production
of H at higher cathode potential and the higher current indicated
2
that its production was partially biocatalysed, decreasing the
16
important energy losses associated to purely electrolytic reduction.
In this respect, it is important to bear in mind that proton availability
plays a critical role in H production, and hence the reducing power
availability for bacterial utilization. The lower the pH, the higher the
production, and electron availability. This phenomenon explains
2
Fig. 2 (A) Current density, total Coulombs supplied and total accumulated
concentration of products in the BES. (B) Accumulated concentration of the
different products and pH in the biocathode. White inverted triangles on the
H
2
top part of the figure indicate CO
2
flushing and black triangles sampling.
the exponential increase of the coulombs supplied and current
density during the second experiment, linked to the decrease in
media pH. Besides, the higher reducing power availability likely
to the high concentrations of organic acids, and the increase of favoured the production of more reduced end-products than acetate,
available reducing power (as will be later discussed) led to the such as butyrate, ethanol or butanol.
re-assimilation of acetate and butyrate and their conversion
The coulombic efficiencies of the experiments were calculated
into ethanol and butanol. Homoacetogens obtain their energy based on the current demand and final concentration of products,
by proton-gradient driven phosphorylation. If high concentrations of and were 28% and 32%, respectively. Two main electron sinks of
unionised organic acids penetrate the bacterial membrane, they can the system were hypothesized: (i) protons were reduced to hydrogen
cause the proton gradient between inside and outside to collapse. at the electrode surface, consuming electrons (Fig. 3). In this light,
The conversion of acids to solvents is one of the mechanisms that part of the electrons supplied may have been lost as un-utilized
Clostridium sp. utilise to prevent further pH decrease during fermen- hydrogen, diffusing out of the cathode through the connectors and
2
tations. In this respect, ethanol production started when the concen- the membrane, or stripped during the periodic CO flushing. These
17
tration of acetate was 40 mMC, corresponding to an unionised acetic phenomena have been extensively described in the literature; and
acid concentration of 0.91 mM. As a consequence, pH increased
from pH 6.32 to 6.45 from days 17 to 20. Although higher acetic acid
concentrations and similar pH values were reached by Marshall et al.
6
(
2013), no butyrate or ethanol production was reported. This could
be explained by two reasons: (i) the lower amount of reducing power
supplied in that study, which could have limited solventogenesis;
and (ii) differences in the bacterial community.
When comparing the performance of the BES in the present
study, to the syngas fermentation experiments performed by S ´a nchez
and co-workers using the same biocatalyst, overall production of C4
compounds was fairly similar (around 30% of the total products),
although more alcohols were produced in the syngas study. Besides,
rates of CO transformation to products reached in the present study
2
À1
À1
(
2.9 mMC d and 4.1 mMC d , for the first and second experi-
ment, respectively) were comparable to the maximum observed in
the syngas fermentation study (2.4–4.9 mMC d ). Given the high
, it is The scan rate was 1 mV sÀ1
.
reducing power availability in the syngas – 32% CO and 32% H
2
Fig. 3 Turnover cyclic voltammograms of an abiotic control (synthetic
medium) (black) and organic acids and alcohols production biofilm (grey).
À1 11
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 3235--3238 | 3237

View Article Online
Communication
ChemComm
(ii) the oxidation of water was likely the main reaction in the anode, Marie Curie Career Integration Grants (PCIG13-GA-2013-618593).
which led to the production of O . Oxygen can permeate to the P. B.-V is supported by a project grant from the Catalan Govern-
2
cathode through the cationic exchange membrane, being subse- ment (2014 FI-B1 00119).
quently reduced to water again or used for oxidizing organic
18
compounds, with the associated consumption of electrons. Gas
Notes and references
sample analysis showed low H
0–1% v/v). On the contrary, high dissolved oxygen (DO) concentra-
2
concentrations in the headspace
(
1
M. Mikkelsen, M. Jørgensen and F. C. Krebs, Energy Environ. Sci.,
2010, 3, 43–81.
À1
2
tions (8 mgO L ) were measured in the anode compartment at
2
3
4
R. S. Haszeldine, Science, 2009, 325, 1647–1652.
the end of the experiment, revealing this second potential sink as
the main contributor to electron loses from the system.
K. Rabaey and R. A. Rozendal, Nat. Rev. Microbiol., 2010, 8, 706–716.
K. P. Nevin, T. L. Woodard, A. E. Franks, Z. M. Summers and
D. R. Lovley, mBio, 2010, 1, e00103.
Finally, it is important to consider that separation and
recovery of fermentation products, even from highly specific
pure cultures, can account for over 60% of the total production
5
6
7
8
9
C. W. Marshall, D. E. Ross, E. B. Fichot, R. S. Norman and H. D. May,
Appl. Environ. Microbiol., 2012, 78, 8412–8420.
C. W. Marshall, D. E. Ross, E. B. Fichot, R. S. Norman and H. D. May,
Environ. Sci. Technol., 2013, 47, 6023–6029.
M. Sharma, N. Aryal, P. M. Sarma, K. Vanbroekhoven, B. Lal,
X. D. Benetton and D. Pant, Chem. Commun., 2013, 49, 6495–6497.
K. J. J. Steinbusch, H. V. M. Hamelers, J. D. Schaap, C. Kampman
and C. J. N. Buisman, Environ. Sci. Technol., 2010, 44, 513–517.
M. C. A. A. Van Eerten-Jansen, A. Ter Heijne, T. I. M. Grootscholten,
K. J. J. Steinbusch, T. H. J. A. Sleutels, H. V. M. Hamelers and
C. J. N. Buisman, ACS Sustainable Chem. Eng., 2013, 1, 513–518.
1
9
costs. To date, a number of technologies have been developed
for the separation of organic acids from fermentation broths
(
both online and offline), including liquid:liquid extraction and
10
electrodialysis. Recently, Andersen and co-workers developed a
processing pipeline to transform carboxylates into fine chemicals
by combining Membrane Electrolysis (ME) and Biphasic Esterifica-
20
tion (BE). This is of special interest because MES of butyrate from 10 M. Dwidar, J.-Y. Park, R. J. Mitchell and B.-I. Sang, Sci. World J.,
2
012, 471417.
1 P. S ´a nchez, R. Ganigue, L. Ba n˜ eras and J. Colprim, Chem. Eng. J.,
014, under review.
12 J. Daniell, M. K o¨ pke and S. D. Simpson, Energies, 2012, 5, 5372–5417.
CO could be coupled to this concept, with organic acids produced
2
1
in the cathode being extracted and concentrated in the anode, prior
to esterification.
2
1
1
1
1
3 A. S. G o¨ ssner, F. Picardal, R. S. Tanner and H. L. Drake, FEMS
Microbiol. Lett., 2008, 287, 236–242.
4 R. K. Thauer, K. Jungermann, K. Decker and P. P. H. Pi, Bacteriol.
Rev., 1977, 41, 809.
5 M. T. Agler, B. A. Wrenn, S. H. Zinder and L. T. Angenent, Trends
Biotechnol., 2011, 29, 70–78.
6 P. Batlle-Vilanova, S. Puig, R. Gonzalez-Olmos, A. Vilajeliu-Pons,
This study demonstrates for the first time the bioelectro-
chemical CO
characterisation demonstrated that the CO
2
transformation to butyrate. The electrochemical
reduction to buty-
2
rate was hydrogen driven. Production of ethanol and butanol
was also observed at low pH values and high concentrations of
undissociated organic acids, opening up the potential for the
L. Ba n˜ eras, M. D. Balaguer and J. Colprim, Int. J. Hydrogen Energy,
2
014, 39, 1297–1305.
bioelectrochemical production of biofuels from CO
2
as a sole
1
7 A. W. Jeremiasse, H. V. M. Hamelers and C. J. N. Buisman, Bioelec-
carbon source. Future work should aim to increase product
trochemistry, 2010, 78, 39–43.
titers and coulombic efficiency of the system by minimising 18 M. C. A. A. Van Eerten-jansen, A. Ter Heijne, C. J. N. Buisman and
H. V. M. Hamelers, Int. J. Energy Res., 2012, 809–819.
9 I. Bechthold, K. Bretz, S. Kabasci, R. Kopitzky and A. Springer, Chem.
Eng. Technol., 2008, 31, 647–654.
anodic oxygen production and its diffusion to the cathode.
1
The authors would like to thank the Spanish Ministry (Best-
Energy, CTQ2011-23632) for its financial support in this study. 20 S. J. Andersen, T. Hennebel, S. Gildemyn, M. Coma, J. Desloover,
J. Berton, J. Tsukamoto, C. V. Stevens and K. Rabaey, Environ. Sci.
Technol., 2014, 48, 7135–7142.
1 R. S. Tanner, L. M. Miller and D. Yang, Int. J. Syst. Bacteriol., 1993,
LEQUIA has been recognised as a consolidated research group by
the Catalan Government (2014-SGR-1168). R.G. gratefully acknowl-
2
edges support from Beatriu de Pin ´o s fellowship (BP-2011-B) and FP7
43, 232–236.
3
238 | Chem. Commun., 2015, 51, 3235--3238
This journal is ©The Royal Society of Chemistry 2015