www.Life cycle assessment of microbial electro synthesis for 1

commercial product generation 2

Tobechi Okoroafor, Ph.D.1; Sue Haile, Ph.D.2; and Sharon Velasquez-Orta, Ph.D.3 3

1 Ph.D. Scholar, School of Engineering, Faculty of Science, Agriculture and Engineering, Newcastle 4

University, Newcastle upon Tyne, NE1 7RU, United Kingdom. Email: t.n.okoroafor1@ncl.ac.uk 5

2 Senior Lecturer, School of Engineering, Faculty of Science, Agriculture and Engineering, Newcastle 6

University, Newcastle upon Tyne, NE1 7RU, United Kingdom. Email: sue.haile@ncl.ac.uk 7

3 Lecturer, School of Engineering, Faculty of Science, Agriculture and Engineering, Newcastle 8

University, Newcastle upon Tyne, NE1 7RU, United Kingdom (corresponding author). Email: 9

sharon.velasquez-orta@ncl.ac.uk 10

Abstract 11

Microbial electrosynthesis (MES) uses microbes and electricity to convert CO2 to high grade chemicals 12

alleviating greenhouse gas emissions. Little is known on the environmental loads associated with the 13

scale up of the technology. Initially, the MES environmental impacts of synthesizing acetic, formic or 14

propionic acids, methanol or ethanol were assessed using the LCA software GaBi and electricity 15

produced 60% from fossil fuels. Results showed that formic acid production had the lowest 16

environmental impact in all eco indicators due to comparatively low energy requirements of its reactor 17

and rectification unit. Second, three different formic acid production methods were compared with 18

MES. Hydrolysis of methyl formate, the main conventional method was shown to be less 19

environmentally harmful than the other three CO2 utilizing technologies analysed when electricity used 20

was generated from fossil fuels except for the impact on climate change. Producing electricity from 21

renewable sources (Hydro, biogas, wind and photovoltaic) made MES able to mitigate climate change 22

having the lowest negative impact on the environment compared to others. Synthesis of products 23

through MES using wind generated electricity could provide considerable benefits and should be 24

considered when MES is industrially applied. 25

Author Keywords: Microbial electrosynthesis, Life cycle assessment, Global warming 26

Introduction 27

Bioelectrochemistry involves the transfer of electrons between a solid electrode and immobilised 28

bacteria. Immobilisation helps reduce the distance between the bacteria and electrode in order to 29

preserve activity (Gooding and Gonçales, 2017). Interest in this science has increased exponentially 30

over the years as researchers become aware of its enormous potential. Bioelectrochemical systems 31

(BES) is a technology that was originally developed for the conversion of wastewater to bioelectricity. 32

The technique uses an anode and cathode electrode, usually separated by a proton exchange 33

membrane (Logan et al., 2006). Oxidation and reduction reactions occur at the anode and cathode, 34

respectively. These redox reactions are driven by catalysts interacting with electrodes connected via 35

an electrical circuit (Das et al., 2019). 36

BES have numerous application and depending on their operation can be classified as microbial fuel 37

cells (MFCs), microbial electrolysis cells (MECs), enzymatic fuel cells (EFCs), microbial solar cells 38

(MSCs) or microbial desalination cells (MDCs) (Santoro et al., 2017). The electron transferred from 39

the electrode to the bacteria could either be direct or indirect through the use of mediators (Rabaey et 40

al., 2010). Microbial electrosynthesis (MES) works similar to MEC which requires external energy to 41

be supplied for the desired bioelectrochemical reaction to occur and has attracted recently a lot of 42

research attention (Jiang and Jianxiong Zeng, 2018; Jiang et al., 2019). MES can be used to produce 43

methane, acetate, formic acid and other higher biofuels from CO2 (Cheng et al., 2009; Nevin et al., 44

2011; Jourdin et al., 2016; Vassilev et al., 2018). It has also been shown recently to have the ability to 45

upgrade biogas by increasing its methane content (Das et al., 2018; Das and Ghangrekar, 2018). 46

Synthesis is usually done using bacteria or enzymes as biocatalyst in the cathode chamber of a BES. 47

However, cell configurations having more than one cathode chamber have been employed with 48

conditions in subsequent chambers set to promote synthesis of different targeted products (Vassilev et 49

al., 2019). As MES uses CO2, technology development can contribute to the European Union 2050 50

emission reduction targets. 51

MES convert CO2 to value added chemicals leading to a reduction in carbon emissions (Finn et al., 52

2012; Bajracharya et al., 2017). However, the MES technology has to be correctly applied to ensure it 53

does not lead to environmental burdens that increase negative environmental impacts. Having this last 54

scenario may mitigate the positive effects of using the technology. For this reason, a comprehensive 55

assessment of other environmental burdens apart from GHGs is required. Researchers have been able 56

to improve productivity and resilience of biocatalysts used for MES. However, after almost a decade, 57

commercial application of the technology has not be exploited (Prévoteau et al., 2020). This however 58

could be in the horizon, as reported chemical yields increase. This current paper evaluates the 59

potential environmental impacts of using MES on an industry scale for the production of chemicals. 60

Life cycle assessment (LCA) was used to analyse the potential environmental impacts of producing 61

acetic, propionic and formic acid, methanol and ethanol using MES. A modelled full-scale production 62

MES process was assessed from cradle to gate. Environmental burdens were characterised using GaBi 63

LCA software. Additionally, MES for the synthesis of formic acid was compared with two CO2 64

utilizing technologies and the main conventional formic acid production route. 65

Methodology 66

Process description, assumptions and products assessed in this paper follow the same general 67

guidelines used in Christodoulou et al. (2017), where the energy requirement, global warming 68

potential and economic feasibility of MES applied on an industrial scale were evaluated for acetic, 69

propionic and formic acids, plus methanol and ethanol production using only natural gas as the energy 70

source. 71

Goal and scope definition 72

This LCA study aimed to analyse and understand the potential environmental impacts of MES 73

through a process based LCA model. The environmental impacts of the process for the synthesis of 74

acetic, propionic and formic acid, methanol and ethanol from cradle to gate was assessed and 75

environmental burdens characterised using the International Life Cycle Data System (ILCD) method 76

in GaBi LCA software. The software was chosen because it has established competency in LCA 77

research (Speck et al., 2015; Speck et al., 2016). The functional unit used was 1,000 tonnes per year 78

(t/yr) of products synthesized from CO2 for a ten year plant life. Environmental burdens associated 79

with decommissioning of the plants were not considered as this was a cradle to gate study. The gas 80

(CO2) was assumed to be captured from a coal fired plant fitted with post combustion capture system. 81

Environmental credit from the CO2 captured is accounted for in the overall calculations. Electricity 82

supplied to the plant is assumed to be from the 2014 estimate of the UK national grid (coal and oil 83

30.5%, natural gas 30.1%, renewables 19.2%, nuclear 19.0%, other 1.2%), of course now the UK grid 84

has decarbonised considerably (Thinkstep, 2016; DBEIS, 2017; DBEIS, 2019). This was selected to 85

illustrate the effect of a fossil fuel dominated grid power production system on the environmental 86

burdens of MES. Potential environmental impacts in this analysis is the net climate change impact 87

calculated by deducting the environmental credit gained from CO2 used by the process. Additionally, 88

this study aims to compare the environmental effects of producing formic acid using MES with that of 89

both abiotic electrochemical reduction and conventional routes. Formic acid was selected because it 90

has been shown to have relatively low global warming potential and high economic value when 91

compared to other chemicals that can be synthesized by the process (Christodoulou et al., 2017). 92

Fig. 1 shows the system boundary of the LCA model. The process is divided into phases to ease 93

analysis including CO2 capture, purification and transportation, product formation, separation and 94

packaging. All MES plant unit operations (mixer, MES reactor, gas separator and rectification unit) 95

and parameters that make up the MES plant remained the same as those described in (Christodoulou 96

et al., 2017). 97

Life Cycle Inventory 98

A conceptual design is implemented in the LCA software GaBi according to a 1,000t/yr commercial 99

plant size (See Appendix Fig.A1-A7 for GaBi flowsheets). Gabi was selected because it is fully 100

compliant with the ISO 14010 and 14044 standards and can efficiently estimate the output of 101

particular processes (in terms of GHG and other environmental burdens produced) (Thinkstep, 2017). 102

Commented [SVO1]: This should be in the past tense isn’t it? Please check all the paper.

Table 1 summarizes the water and energy consumed each year for the products evaluated. Process 103

description and associated assumptions remain the same and have been previously described 104

elsewhere therefore would not be stated here (Christodoulou et al., 2017). The life cycle inventory 105

(LCI) data for energy consumption of each process in MES plant modelling was calculated based on 106

energy data and technical information available from contractors, open literature and the GaBi 107

software (Thinkstep, 2016; Christodoulou et al., 2017). Energy data for the MES reactor took into 108

account both cathode and anode energy load unlike Christodoulou and co-worker where only cathode 109

energy load was evaluated (Christodoulou et al., 2017). Water oxidation (see equation 1) is assumed 110

to occur at the anode for the required protons and electrons while the MES reaction to products at the 111

cathode (Rabaey and Rozendal, 2010; Christodoulou et al., 2017). 112

4𝐻2𝑂 → 2𝑂2 + 8𝐻+ + 8𝑒− Eo = 0.817V vs SHE at pH=7 (1) 113

Table 2 shows the reaction balances and selected MES reactor potentials. The energy values for acetic 114

and formic acid were taken from experimental data while the theoretical electrochemical data was 115

used for propionic acid, methanol and ethanol (Nevin et al., 2011; Marshall et al., 2013). The work of 116

Marshall and co-workers was selected for acetic acid because it showed the long term viability of 117

producing the chemical using MES (Marshall et al., 2013). This indicated that MES can be deployed 118

commercially on a large scale. However, formic acid unlike acetic acid is not widely reported as being 119

synthesized by whole cell biocatalysts. This is because formic acid is a main intermediate in the 120

Wood-Ljungdahl pathway for the synthesis of other chemicals (Oswald et al., 2018). Therefore, the 121

chemical could be used as substrate by other formate consuming bacteria species attached to the 122

biocathode after generation. However, using enzymatic electro-synthesis which uses CO2 like MES, 123

high formic acid productivity has been achieved (Chiranjeevi et al., 2019). Nevin and co-worker 124

showed direct synthesis of formic acid from CO2 in MES at potentials close to its theoretical value (-125

0.430 vs SHE) (Nevin et al., 2011). This work was therefore chosen because detectable formic acid 126

yield making use of whole cell biocatalysts instead of extracted enzymes was achieved. 127

Alongside energy inputs, materials used in manufacturing the MES reactor and other unit operations 128

are accounted for in the assessment. Electrodes for the MES reactors are biotic, similar to the design 129

analysed by Christodoulou and co-workers (Christodoulou et al., 2017). Most MES studies make use 130

of biotic cathodes and noble metals such as Pt as counter electrodes (Jiang and Jianxiong Zeng, 2018). 131

However, biotic anode and cathode MES reactor design have proved effective for electro catalytic 132

CO2 reduction by (Marshall et al., 2013) and (Giddings et al., 2015). Carbon fibre made from 133

polyarcylonitrile was considered as the MES reactor electrode material. Inventory data for the carbon 134

fibre was collated from the work of Shemfe and co-researchers (Shemfe et al., 2018). Electrolytes are 135

needed for electro catalytic reduction of CO2 because it aids the transfer of electrons and protons to 136

and from the electrodes. The MES reactor aqueous medium consisted of the following (per litre of 137

water); 2.5g NaHCO3; 0.25g NH4CL; 0.21g MgCl2.6H2O; 0.03g CaCl2; 0.1g KCl; 0.6g NaH2PO4.H2O 138

and trace quantities of Wolfe vitamin solution and modified Wolfe’s mineral solution (Christodoulou 139

et al., 2017).The database of GaBi and Ecoinvent 3.0 accounted for the chemicals used in the 140

electrolytes. Sodium carbonate, sodium phosphate and potassium carbonate were used as proxies for 141

sodium bicarbonate, sodium dihydrogen phosphate and potassium bicarbonate due to lack of life cycle 142

data. Conversion rates and faradaic efficiencies for the MES reactors were assumed to be 58.8% and 143

69% respectively (Marshall et al., 2013; Christodoulou et al., 2017). 144

Comparison of formic acid production routes 145

Formic acid is a colourless corrosive acid which is totally miscible in water and other numerous polar 146

solvents. It is mainly used in the textile, pharmaceuticals and food industry with the leather and 147

tanning industry being it largest user in 2003 (Reutemann and Kieczka, 2000; Bulushev and Ross, 148

2018).This has however been overtaken recently by its use as an additive and preservative in animal 149

feed which accounts for as at 2019 around 20% of global formic acid usage (Mordor Intelligence, 150

2019). The total amount of formic acid manufactured worldwide is estimated to be 0.95 mega tonnes 151

per year with the price set at between $0.60 and $0.70 per kg in the first half of 2014. Demand for the 152

chemical is set to increase by around 6% in 2019 because of its ever expanding use cases (Bulushev 153

and Ross, 2018). Formic acid can be used to store hydrogen due to its good properties and simple 154

dehydrogenation (Bulushev and Ross, 2018). 155

Conventionally formic acid can be manufactured through oxidation of hydrocarbons, hydrolysis of 156

formamide, production from formates and hydrolysis of methyl formate (Reutemann and Kieczka, 157

2000). In the oxidation of hydrocarbons to produce formic acid, methane and methanol can be 158

employed. Methanol oxidation yield formaldehyde which in turn is oxidized to formic acid in a two 159

step process (see equation 2 and 3) (Andrushkevich et al., 2014). Methane on the one hand is oxidized 160

to formic acid using heterogeneous catalysts. Using methane is advantageous because oxidation 161

occurs at low temperatures (60oC) even though yields are low (Hutchings, 2016). Formic acid 162

production through hydrolysis of formamide was prominent in Europe until the end of the 19 th 163

century. The consumption of ammonia and sulfuric acid coupled with ammonium sulphate generation 164

made the process less economically competitive (Reutemann and Kieczka, 2000). Production from 165

formates usually uses sodium formate (see equation 4) and calcium formate (see equation 5) with 166

acidolysis done with sulfuric or phosphoric acid (Reutemann and Kieczka, 2000). Comparing the 167

different ways formic acid can be produced conventionally, Hydrolysis of methyl formate is currently 168

the main way the chemical is manufactured. The route accounts for around 90% of all formic acid 169

installation production facilities. It occurs in a two stage process where 95% carbon monoxide and 170

30% methanol are initially reacted to produce methyl formate which is then hydrolysed to synthesize 171

formic acid (see equation 6 and 7) (Saavalainen et al., 2017). 172

Oxidation of hydrocarbons: 173

𝐶𝐻3𝑂𝐻 + 0.5𝑂2 → 𝐻𝐶𝐻𝑂 + 𝐻2𝑂 (2) 174

𝐻𝐶𝐻𝑂 + 0.5𝑂2 → 𝐻𝐶𝑂𝑂𝐻 (3) 175

Production from formates: 176

2𝐻𝐶𝑂𝑂𝑁𝑎 + 𝐻2𝑆𝑂4 → 𝑁𝑎2𝑆𝑂4 + 2𝐻𝐶𝑂𝑂𝐻 (4) 177

(𝐻𝐶𝑂𝑂)2𝐶𝑎 + 𝐻2𝑆𝑂4 → 𝐶𝑎2𝑆𝑂4 + 2𝐻𝐶𝑂𝑂𝐻 (5) 178

Hydrolysis of methyl formate; 179

𝐶𝐻3𝑂𝐻 + 𝐶𝑂 → 𝐻𝐶𝑂𝑂𝐶𝐻3 (6) 180

𝐻𝐶𝑂𝑂𝐶𝐻3 + 𝐻2𝑂 → 𝐶𝐻3𝑂𝐻 + 𝐻𝐶𝑂𝑂𝐻 (7) 181

182

183

Electrocatalytically formic acid has been shown to be produced through biotic (Reda et al., 2008; 184

Srikanth et al., 2014) and abiotic catalysts (see Fig. 2) (Gupta et al., 2016). In the case of biotic 185

catalyst the chemical is produced through MES by supplying suitable bacteria or biomolecules 186

electrons and CO2 for synthesis (Srikanth et al., 2017). Abiotically, the chemical can be produced 187

either through homogenous and heterogeneous abiotic catalysts (Gupta et al., 2016). Four ways of 188

manufacturing the chemical are compared in this study using the ILCD (International Life Cycle Data 189

System) method in GaBi. The options consisted of three processes that utilises CO2 (Microbial 190

electrosynthesis (MES), Abiotic electrochemical reduction (AER), Hydrogenation of carbon dioxide 191

(HCD)) and a conventional benchmark (Hydrolysis of methyl formate (HMF)). Comparative analysis 192

between the three CO2 utilizing formic acid production routes listed above and a conventional route 193

was carried out. The functional unit used for this comparison remained 1,000 t/yr of formic acid 194

generated at a commercial grade concentration of between 90% and 99%. The database of Ecoinvent 195

3.0 and Gabi LCA software was used to conduct this comparison with electricity supplied for each 196

plant assumed to be from the 2014 estimate of the UK national grid (coal and oil 30.5%, natural gas 197

30.1%, renewables 19.2%, nuclear 19.0%, other 1.2%) (DBEIS, 2017). Assumptions for the MES 198

plant are the same as those described above while those associated with the AER, HCD and HMF 199

plants are described in the subsections below; 200

Abiotic electrochemical reduction plant 201

Abiotic electrochemical reduction of CO2 follows the same principles as MES without the use of 202

micro-organisms as catalyst. As with MES, CO2 and protons are synthesized to products at the 203

cathode with water oxidized to oxygen and protons at the anode (Tao et al., 2017). It was assumed 204

that the AER plant used similar unit operations (mixer, AER reactor, gas separator and rectification 205

unit) as the MES plant. The anode of the AER reactor was assumed to be platinum while the cathode 206

copper. Platinum was selected because it is the most widely used anode for water oxidation in AER 207

(Endrődi et al., 2017; Evangelisti et al., 2017). As stated previously, copper was chosen as the 208

cathode and therefore the abiotic electro catalyst. Selection was based on its unique ability to produce 209

a wide variety of products specifically hydrocarbons similar to those obtained through MES (Kuhl et 210

al., 2012). Catalytic stability problems leading to gradual degradation due to carbon deposits and 211

other toxic elements was not taken account in this study as it was assumed that catalyst performance 212

does not depreciate over time (Qiao et al., 2014). 213

Electrolytes are needed for electro catalytic reduction of CO2 because it aids the transfer of electrons 214

and protons to and from the electrodes. Electrolytes used for CO2 reduction in AER reactors could be 215

aqueous, ionic or organic (Zhang et al., 2017). This study assumes an aqueous electrolyte (0.1M 216

KHCO3) as it is comparable to those used in MES reactors. Aqueous electrolyte are also cheap due to 217

water usage but has the disadvantage of low CO2 solubility (0.03M in 25oC ) and competing hydrogen 218

evolution reaction reducing faradaic efficiency (Tao et al., 2017). Conversion rates were assumed to 219

be 58.8% for the plant reactor with the unreacted gas recycled back. Cathodic potential (-0.860 vs 220

SHE) and Faradaic efficiency (30%) for product formation in the AER reactor was assumed using 221

experimental results obtained by (Kuhl et al., 2012) and modelled using GaBi (See Appendix Fig.A5. 222

for flowsheet). 223

Hydrogenation of carbon dioxide plant 224

Formic acid production from a HCD plant were evaluated and compared with other ways of 225

generating the chemical. The plant was assumed to be located at Newcastle Upon Tyne, UK. CO2 226

needed for plant operation was assumed to be supplied at 0.1758 GJ per tonne from a coal fired plant 227

30km away (Bhown and Freeman, 2011). The transportation of the gas to the plant site was assumed 228

to be by a diesel powered Euro type truck using GaBi data. Energy required to compress CO2 from 229

atmospheric pressure to that required by the plant is allocated to the HMF plant overall energy duty. It 230

is assumed that the required hydrogen needed to operate the main reactor is supplied by an 231

electrolyser situated onsite. Inventory data needed for modelling of the HCD plant was obtained from 232

the work of (Pérez-Fortes et al., 2016) and the database of Gabi (See Appendix Fig.A6. for 233

flowsheet). 234

Hydrolysis of methyl formate plant 235

Methyl formate hydrolysis was used as the bench mark conventional process because of its wide 236

industrial application (Saavalainen et al., 2017). The process involves two stages with the first being 237

methanol carbonylation for the formation of methyl formate. Methyl formate is then hydrolysed to 238

produce formic acid in the second stage. Methanol and carbon monoxide which are the main raw 239

materials needed by the plant is assumed to be supplied from a production facilities 30 Km from the 240

plant site. Inventory data for the HMF plant was collated from the databases of Ecoinvent 3.0 and 241

Gabi (See Appendix Fig.A7. for flowsheet). 242

Table 3 shows a summary of the LCI of the four formic acid production routes assuming 1,000 t/yr of 243

formic acid is synthesized. The values summarized in the table are grouped into raw materials, 244

product, output and energy consumed by the plants. For the three CO2 utilizing plants unreacted CO2 245

and water are assumed to be perfectly recycled back into the plant with the analysis done for a ten 246

year time frame. 247

It can be seen from Table 3 that the main contributors to climate change and other environmental 248

burdens is expected to be plant electricity and steam usage. Electricity consumed by the three CO2 249

utilizing plants ranged between 14,652 GJ/yr and 7,183 GJ/yr while that of the HMF plant was 1,044 250

GJ/yr. The relative difference is mainly due to electricity being used to undergo formic acid synthesis 251

in the reactors of the MES, AER and HCD plants. The HMF plant (19,500 GJ/yr) showed the highest 252

consumption of steam for the production of 1,000 t/yr of formic acid while the HCD plant (10,030 253

GJ/yr) was the highest for the CO2 utilizing routes. Analysing raw material used it can be observed 254

that both the MES and AER plants (1,094 t/yr) used the most CO2 obtaining the maximum allocated 255

environmental credit associated with CO2 use. The HMF plant does not use CO2 and therefore no 256

environmental credit was given to the process. All plants analysed used comparable process water 257

with the HCD plant using water in an electrolyser to produce H2 which would be feed into its main 258

reactor. 259

Results and discussion 260

MES Plant 261

Climate change 262

The assessment of the MES plant gives an increase in GHG emissions for all the products (acetic acid, 263

formic acid, propionic acid, ethanol and methanol) analysed (see Table 4). This is similar to the values 264

Commented [SVO2]: Check again between past and present tense.

observed by Christodoulou and co-workers for acetic (61,638 t CO2 eqv) and propionic acid (48,797 t 265

CO2 eqv) when natural gas was assumed as the energy source (Christodoulou et al., 2017). However, 266

ethanol (42,100 t CO2 eqv), methanol (33,200 t CO2 eqv) and formic acid (2,200 t CO2 eqv) which 267

previously had a negative value has now increased. This is because total energy load (anode and 268

cathode) was used for this assessment and generation of electricity from the 2014 estimate of the UK 269

national grid produces more than three times as much GHG emissions per GJ of electricity produced 270

than natural gas (0.05 t CO2 eqv per GJ). Table 4 also shows the amount of contribution to climate 271

change by the GHGs. The most prevalent GHG in all the different products is CO2 accounting for up 272

to 99% of emissions with the next potent greenhouse contribution being made by methane due to its 273

high CO2 equivalent value. 274

Fig. 3 shows the climate change contributions of the different unit operations of the MES plant for 275

acetic, propionic acid and formic acid, methanol and ethanol. Acetic and propionic acids relatively 276

high amount of contribution to climate change is mainly due to their energy intensive rectification 277

unit. Rectification of acetic (50,700 t CO2 eqv) and propionic acid (44,234 t CO2 eqv) contributed 278

75% and 65% respectively to climate change. Formic acid rectification contributed 55% while that of 279

methanol and ethanol contributed less than 6% to climate change. Evaluating the five MES reactors it 280

was observed that only formic acid (-170.77 t CO2 eqv) had a negative climate change impact. The 281

MES reactors of ethanol and methanol contributed significantly to climate change accounting for 92 282

and 90% respectively. This can be attributed to the large amount of energy needed for MES (380,371 283

GJ for ethanol and 286,479 GJ for methanol) compared to other unit operations. Acetic acid (16,000 t 284

CO2 eqv) and propionic acid (22,100 t CO2 eqv) MES reactors contributed less than 30% to climate 285

change. Results indicates that without supporting unit operations standalone MES reactors (16,000 t 286

CO2 eqv for acetic acid, 22,100 t CO2 eqv for propionic acid, 39,095 t CO2 eqv for ethanol and 30,200 287

t CO2 eqv for methanol) still has a net positive global warming potential. Formic acid (-170.77 t CO2 288

eqv) is a notable exception as it was observed to be the best performing reactor having a negative 289

global warming potential. 290

Other Environmental Burdens 291

Eight midpoint indicators were selected for a more detailed analysis from thirteen indicators of the 292

ILCD methods. The indicators chosen included ozone depletion (OD, in kg R11-Eqv), human toxicity 293

cancer effects (HT, in CTUh), particulate matter (PM, in PM2.5 Eqv), ionising radiation (IR, in U235 294

Eqv), photochemical ozone formation (POF, in kg NMVOC Eqv), acidification (AC, in Mole of H+ 295

Eqv), freshwater europhication (FE, in kg P Eqv) and ecotoxicity (EC, in CTUe). The remaining 296

midpoint indicator results and raw data for the MES plants can be seen in Appendix Fig. A8 and 297

Table A1 respectively. The results are displayed relative to the maximum value in each of the 298

midpoint impact category. 299

Data obtained from the analysis shows that all MES plants has overall negative impact on the 300

environment in all the impact category. This indicates that the environmental benefits of CO2 301

utilization cannot compensate for the generation of other environmental burdens. Fig. 4 and Fig. 5 302

shows that formic acid contributes the lowest in the selected impact categories as its relative values to 303

the maximum is always less than 30%. Ethanol was observed to have the maximum value in all of the 304

selected impact categories differing from results obtained for climate change (See Fig. 3). Propionic 305

and acetic acid gave very similar results in all the selected impact categories as they were found to 306

always have a relative value more than 50% to the maximum (ethanol). 307

The cradle to gate environmental impacts are further assessed in terms of plant unit operations. Grid 308

electricity used for synthesis of products in the MES reactors had relatively large environmental 309

impact contributions in most impact categories. Assessing ethanol which had the most energy 310

intensive MES reactor it was observed that synthesis of the chemical takes up between 94.5% and 311

98.2% in the selected impact categories. Similar results were seen in methanol and formic acid which 312

alongside ethanol had relatively low rectification energy duty. Assessing formic acid specifically it 313

was observed that its MES reactor takes up 88.9% of AC, 81.8% of EC, 76.7% of FE, 80.2% of HT, 314

92.3% of IR, 76.1% of OD, 86.3% of PM and 87.7% of POF. Process steam needed to heat the 315

reboilers of acetic and propionic producing MES plants contributed on average 15% in the selected 316

impact categories. Analysing acetic acid which uses the most energy for rectification it was observed 317

that the unit operation takes up 20.4% of AC, 9.7% of EC, 4.9% of FE, 14.9% of HT, 1.0 % of IR, 318

16.1% of OD, 14.1% of PM and 34.6% of POF. This was found to be always less than its MES 319

reactor in all impact categories except climate change. Rectification of formic acid, ethanol and 320

methanol contributed less than 4% in all impact categories apart from climate change. All other unit 321

operations (CO2 capture, mixer and gas separator) contribute less than 5% in all selected impact 322

categories for each product analysed. Therefore, it can be suggested that the hot spot of the modelled 323

MES plants in terms of unit operation are the MES reactors. 324

Flue gas from the gas separator and incineration of filtered biomass could also contribute to the 325

environmental burdens of the MES plants. As the selectivity of CO2 to products (88%) and faradaic 326

efficiency (69%) of the MES reactors are not 100% there may be undesired by-products formed. This 327

would have to be vented off to the atmosphere alongside oxygen as CO2 is being recycled back to the 328

MES reactor. The impurities in the flue gas would most likely be composed of hydrogen (over 95%) 329

and methane (below 5%). The use of 2-bromoethanesulfonate in biofilm development should help 330

suppress methanogenic bacteria growth hence its expected minimal quantity (Christodoulou et al., 331

2017). Vented Hydrogen gas can affect the upper atmosphere due to a build-up of water vapour. This 332

can cause ozone depletion and its undesired consequences (Tromp et al., 2003). Methane on the other 333

hand contributes to climate change and POF environmental burdens (Sadhukhan et al., 2014). 334

Incineration of biomass would take place twice each year and it is away from the plants; hence 335

environmental burdens associated with its transportation needs to be taken into account 336

(Christodoulou et al., 2017). Municipal waste incineration is known to contribute about 0.36 t CO2 337

eqv/ton to global warming (Havukainen et al., 2017). This value should be what is expected when the 338

filtered bacteria cells are incinerated. As the MES plants would be incinerating just a few grams of 339

bacteria cells per year, this waste treatment method should not significantly affect climate change and 340

other environmental burdens results of the MES plants. 341

Sensitivity Analysis 342

MES plant 343

A sensitivity analysis has been carried out by exploring two MES plant parameters changes. CO2 344

conversion and faradaic efficiency were examined to see the environmental impact variations 345

resulting from the changes in these parameters. Alternative scenarios different to the base case where 346

these parameters are set at 40% and 100% respectively were assessed. The scenario analysis results of 347

the different CO2 conversion and efficiency are shown in Fig. 6 for climate change and Appendix 348

Fig.A9 for other environmental burdens. 349

From Fig. 6, it was observed that when the CO2 conversion and faradaic efficiency are reduced a 350

higher environmental impact in terms of climate change is seen. This can be explained by the change 351

in the energy requirements for the MES reactor and gas separator. An energy requirement change of 352

between 25-30% is seen in the MES reactor for each of the products analysed when the CO2 353

conversion and faradaic efficiency is changed to the selected percentages. In terms of the gas 354

separator a CO2 conversion of 100% eliminates the need for the unit operation as oxygen can be 355

vented off to the atmosphere directly from the MES reactor. A reduction to 40% increases the energy 356

requirement by more than two times in the case of all products analysed. It can be seen from Fig. 6 357

that only the global warming potential of formic acid (-734 t CO2 eqv) becomes negative when the 358

CO2 conversion and faradaic efficiency is set at 100%. Comparatively low rectification and MES 359

reactor energy requirements are the main reason for this change yielding a positive result. This 360

indicates that low synthesis and rectification energy requirements are necessary for good conversion 361

and efficiency values seen in experimental research to yield positive global warming reduction for a 362

commercial grade MES plant. Sensitivity analysis results for other environmental impact burdens 363

(Appendix Fig. A9.) indicated that for all selected impact categories a variation of between 5% and 364

40% occurred. Ethanol at 40% CO2 conversion and efficiency as with the base scenario was observed 365

to have the maximum value in all (HT,IR,OD,PM,AC,EC FE, and POF) of the selected impact 366

categories. The base scenario was surpassed marginally by acetic acid and methanol at 40% CO2 367

conversion and faradaic efficiency. Formic acid was observed to still have the lowest effect on the 368

environment for all impact categories even though the parameters were adjusted to the worst-case 369

scenario. 370

Comparison of formic acid production routes 371

Climate change 372

An assessment of four formic acid producing plants was done using GaBi. The assessment gives a 373

positive impact in terms of GHG emissions for the all the formic acid producing plants analysed (see 374

Table 5). However, the MES plant (2,120 t CO2 eqv) was found to emit the least amount of GHGs 375

about eight times lower than the HMF plant (18,400 t CO2 eqv). Interestingly a CO2 utilizing plant 376

had a higher global warming potential than the conventional plant analysed. This shows that the 377

environmental credit associated with CO2 use may not be sufficient to rival conventional processes. 378

Table 5 also shows the amounts of GHGs emitted by each of the formic acid producing routes. The 379

most prevalent GHG in all the four plants analysed was CO2 with the next potent GHG being 380

methane. The HCD plant emitted the most CO2 (28,806 t CO2 eqv) alongside the other GHGs 381

assessed. Emission amounts were between 44% and 83% that of the HMF plant. It was observed that 382

the HMF plant emitted the least amount of sulphur hexafluoride (1.02E-11 t CO2 eqv) and nitrous 383

oxide (0.15 t CO2 eqv) while the MES plant carbon dioxide (12,221 t CO2 eqv) and methane (28 t 384

CO2 eqv). Comparing the three carbon utilizing routes, MES released the least amount of all GHGs, 385

significantly lower than the AER and HCD plants. As MES had the lowest global warming potential 386

amongst the plants analysed and released comparably low amounts of each GHGs, producing formic 387

acid through this route in terms of climate change is beneficial over the routes analysed. 388

Other Environmental Burdens 389

Eight midpoint indicators were selected for a more detailed analysis from thirteen indicators of the 390

ILCD methods. The indicators chosen included ozone depletion (OD, in kg R11-Eqv), human toxicity 391

cancer effects (HT, in CTUh), particulate matter (PM, in PM2.5 Eqv), ionising radiation (IR, in U235 392

Eqv), photochemical ozone formation (POF, in kg NMVOC Eqv), acidification (AC, in Mole of H+ 393

Eqv), freshwater europhication (FE, in kg P Eqv) and ecotoxicity (EC, in CTUe). The remaining 394

midpoint indicator results and raw data for the formic acid producing plants can be seen in Appendix 395

Fig. A10 and Table A3 respectively. The results are displayed relative to the maximum value in each 396

of the midpoint impact category. 397

Data obtained from the analysis shown in Fig. 7 and Fig. 8 indicates that the CO2 utilizing HCD plant 398

contributes the most in all the selected impact categories. It was on average more than 20% higher 399

than the next worst plant for all the impact categories analysed. The conventional formic acid 400

production plant (HMF) performed better than all assessed plants in six (AC,EC, POF,HT,IR and PM) 401

of the selected ILCD impact categories. This may be due to the use of mostly steam (19,500 GJ/yr) 402

instead of electricity (1,044 GJ/yr) for energy. Steam usage in the HMF contributes 42.5% to AC, 403

17.1% to EC, 45.2% to POF, 27.9% to HT, 4.1% to IR and 34.7% to PM. The HMF plant was also 404

seen to be marginally better than the HCD plant in the EF environmental impacts category. Using 405

MES for the generation of formic acid had the lowest environmental impact in the EF and OD impact 406

categories. The plant was observed to always have a relative value in each impact category less than 407

60% of the maximum which was the HCD plant. Results presented here showed than the use of CO2 408

as raw material does not guarantee environmental benefit as other factors such as amount of energy 409

needed for production should be considered. The use of an electrolyser in the HCD plant places large 410

energy burden on the plant and if eliminated could reduce energy consumption by around 92% (Pérez-411

Fortes et al., 2016). This makes the use of MES particularly attractive over other CO2 utilizing routes 412

as it used comparatively low amounts of energy (7,183 GJ/yr) hence the relatively low environmental 413

impact. However in most impact categories this was not sufficient to show value over the 414

conventional process. 415

Environmental impacts are further assessed in terms of raw materials. For the HMF plant, carbon 416

monoxide production using synthetic gas contributes 42% to EC, 84.2% to EF, 39.7% to HT and 417

69.4% to OD. It is the highest contributor in these impact categories and in all the impact category 418

analysed was consistently above 15%. Comparing this with the environmental burdens of capturing 419

CO2 from a coal fired plant for MES. It was observed that the environmental burdens did not exceed 420

2.6% in all impact categories other than climate change. Overall in MES, electricity for synthesis 421

contributed the highest in all the impact categories. This shows that CO2 capture energy is not the 422

main environmental hot spot for implementation of MES technology in the manufacturing of formic 423

acid on a commercial scale. 424

Looking at the market, formic acid is sold at different concentrations. These concentration vary 425

between 85 - 99 weight% with 85% formic acid concentration being the most traded (Pérez-Fortes et 426

al., 2016). Assessing the MES plant, a change in formic acid concentration delivered to the end users 427

would specifically affect the rectification unit of the plant. The concentration of formic acid after 428

synthesis in the MES reactor is 45% hence the need for rectification. Energy required for the 429

rectification of the chemical contributes 55% to climate change but only less than 4% in the other 430

selected impact categories. This is because natural gas is used to supply the needed energy instead of 431

the grid. Any change in the formic acid concentration delivered to the end user by the plant would 432

therefore affect mainly climate change. Comparative results shown here already indicates that the 433

MES plant is more beneficial than all the other plants analysed in this environmental category. 434

Therefore, a concentration change should not affect results presented here. In the case of other 435

environmental burdens due to the relatively insignificant effect of the rectification unit a change in 436

formic acid concentration to 85% would not be able to make the MES plant rival the conventional 437

plant in the AC,EC, POF,HT,IR and PM impact categories. 438

Production using different energy sources 439

Different electricity sources were evaluated for the production of formic acid in MES, AER, HCD and 440

HMF plants. A mix of fossil fuel (coal, oil and gas) and renewable sources (Hydro, biogas, wind and 441

photovoltaic) were chosen. Environmental impacts in terms of climate change and other burdens are 442

analysed in the subsections below. 443

Climate Change 444

Table 6 shows the global warming potential associated with using fossil fuel and renewable sources to 445

generate electricity for the MES, AER, HCD and HMF plants. The MES plant global warming 446

potential remained positive when powered by coal (9,940 t CO2 eqv) and oil (9,367 t CO2 eqv) but 447

turned negative when natural gas (-615 t CO2 eqv) was used. This is consistent with results obtained 448

by Christodoulou and co-researchers as natural gas is often seen as a cleaner form of fossil fuel based 449

electricity source and has been shown to emit less GHGs than coal and oil (Jaramillo et al., 2007; 450

Burnham et al., 2012; Christodoulou et al., 2017).. Negative global warming potential are also 451

recorded when renewable energy sources were used. This shows that the choice of energy source for 452

the synthesis of formic acid using biocatalysts is important. However, renewable energy should be 453

favoured as it consistently had a lower global warming potential than natural gas. As mentioned 454

previously the UK grid is decarbonising quickly (DBEIS, 2019). Using renewable energy (hydro, 455

biogas, wind and photovoltaic) decreased the global warming potential in the base scenario (2014 UK 456

national grid) by on average more than 9,000 t CO2 eqv while that of natural gas by 2,735 t CO2 eqv. 457

Comparing coal (9,940 t CO2 eqv) and oil (9,367 t CO2 eqv) with the base scenario (2,120 t CO2 eqv) 458

it was observed that there is no climate change benefit to their use. The global warming potential of 459

the conventional route analysed (HMF plant) remained positive even though renewable energy 460

sources are used. This is because 71.1% of its global warming potential value is accounted to the 461

steam (10,030 GJ/yr) used by the plant. 462

Looking at the HCD plant it was observed to have negative global warming potential when nuclear (-463

1,272 t CO2 eqv), hydro (-1,178 t CO2 eqv) and wind (-1,240 t CO2 eqv) are used. However, it had a 464

positive global warming potential when biogas (6,390 t CO2 eqv) and photovoltaic (5,590 t CO2 eqv) 465

are used as electricity sources. This was however between 60% and 98% lower than values seen for 466

fossil fuels usage. Results outlined here show that there is climate change benefit of using renewable 467

energy source to power formic acid synthesis plants that utilizes CO2. 468

Other Environmental Burdens 469

For the assessment of other environmental burdens associated with producing formic acid from MES 470

using other sources of electricity, the same impact categories as used above was employed. The 471

complete midpoint results for the MES are shown in Appendix Fig. A11. 472

Fig. 9A and Fig. 9B shows the results for other environmental impact burdens when electricity source 473

differed from that of the base scenario (2014 UK national grid). It was observed that for all the impact 474

categories using coal, oil, nuclear, biogas and photovoltaic means of electricity generation was higher 475

than the base scenario for the MES plant. Coal and oil had more negative environmental impact in 476

five (AC, EC, FE, HT, PM and POF) of the eight selected categories. Nuclear, biogas and 477

photovoltaic also had more negative impact in three (EC, HT and PM), one (FE) and three (EC, HT 478

and OD) midpoint indicators respectively. Natural gas and the remaining renewable energy sources 479

(Hydro and wind) consistently had lower negative environmental impact than the base scenario. 480

Results obtained here are comparable to the other CO2 utilizing plant (HCD plant) analysed (see 481

Appendix Fig. A12.). Electricity generation by wind turbines shows good promise when both climate 482

change and other environmental burdens are assessed. The technology was consistently lower than the 483

base scenario for the CO2 utilizing plants in all ILCD impact categories. 484

The same impact category was used to assess the other environment burdens of using different 485

electricity sources in the conventional plant (see Appendix Fig. A13.). It was observed that coal and 486

oil was higher than the base case in five of the eight selected impact categories. Environmental 487

burdens associated with the use of nuclear, biogas and photovoltaic were higher in up to three impact 488

categories. This showed similar trend to what was observed for the CO2 utilizing plants. However in 489

the AC, EC, FC and POF impact categories environmental burdens did not go lower than 60% the 490

value of the base scenario differing from what was observed in the MES plant. Generation of 491

electricity was also found to be lower than the base scenario in all ILCD impact categories. Wind 492

energy usage for formic acid production has been shown to be environmentally better than the 2014 493

estimate of the UK national grid for both CO2 utilizing and conventional plants. UK is an island 494

nation therefore energy from both onshore and offshore wind turbines can be relatively easy to 495

harness when compared to landlocked countries although cost have to also be considered. However 496

offshore wind farms should be favoured as it has been proven marginally beneficial in terms of global 497

warming (Kaldellis and Apostolou, 2017). The UK government has made great strides in its low 498

carbon generation initiative as energy production from wind turbines and other renewables now 499

account for the majority of the national grid power (DBEIS, 2019). 500

Fig. 10 compares the four plants analysed if electricity used was generated using only wind turbines. 501

Values displayed are relative to the maximum value in each impact category. It was observed that 502

wind energy usage relegated the HMF plant to having comparatively the worst impact on the 503

environment than all CO2 utilizing routes. This could be because the impact of a change to a more 504

environmentally friendly source of electricity is relatively small for the HMF plant as steam is mostly 505

used. The HCD plant which previously had the highest impact in most categories is on average 55% 506

better when wind energy is used instead of the base scenario. These analysis and results suggest that 507

MES should be deployed on an industrial level favouring renewable electricity generation. 508

509

Future Outlook 510

MES is being considered as a conversion route of CO2 to chemicals that can alleviate greenhouse gas 511

emissions alongside other environmental burdens (Das et al., 2019). However, the current industrial 512

application of MES is still constrained by low chemical synthesis rates, high energy demand and high 513

capital expenditures (Prévoteau et al., 2020). 514

During the decade long research into MES, the maximum synthesis rate of acetic acid, the most 515

frequently produced chemical, is still in the region of 685 g/m2/d (Jourdin et al., 2015). Industrial 516

production will require this value to increase at least 100 times to reach commercial viability. Another 517

problem associated with industrialization is product selectivity, as currently targeted chemicals are 518

usually manufactured along with other side products. This is especially the case when a range of 519

chemicals are synthesized using homoacetogens (Prévoteau et al., 2020) or mixed cultures. Good 520

selectivity of products can be obtained by carefully managing the pH of the cathodic medium and the 521

use of inhibitors such as 2-bromoethanesulfonate to prevent loss of electrons to methanogenic bacteria 522

(Christodoulou et al., 2017). However, this may be impractical in a full scale industrial plant as it adds 523

environmental and financial burdens to the system.. 524

The source and amount of energy used in MES is of utmost importance for the sustainability and 525

economic viability of the technology. Environmental burdens associated with electricity usage are 526

highly dependent on the faradaic efficiency and applied voltage of the process (Jung et al., 2020). Cell 527

design would need to be improved to enhance these parameters and reduce energy loss. The use of 528

renewable energy instead of fossil fuels may increase the sustainability of the system. Different 529

sources of intermediate energy supply could be a way of dealing with interruptions from renewable 530

energy sources (del Pilar Anzola Rojas et al., 2018). For example by combining solar, wind and/or 531

biomass energy sources. Having a combined renewable system, could significantly increase the 532

reliability and sustainability of the technology when scaling up. Also industrial MES reactors may 533

require improvements in the ionic membranes used and a more stable anode electrode than the biotic 534

electrodes used in this study. However, the use of noble metals as anodes will add to environmental 535

burdens and cost. 536

Environmental assessment of MES done in this study made use of operating parameters and reaction 537

efficiencies from experimental results using lab scale reactors. These values may be too optimistic as 538

they are based on the parameters not deviating from those of expected future commercial scale up of 539

the technology. As there are numerous unsolved challenges which have been outline above, the long 540

term performance in a real industrial environment needs to be evaluated. This would enable data from 541

practical large scale operation to be used in future sustainability assessments. As MES have had a 542

relative brief time to technologically mature, limited pilot scale systems are available. 543

Notwithstanding, over the last decade the process has made great stride towards commercialization 544

and with continuous research industrial scale plants that can mitigate climate change and other 545

environmental burdens can be achieved. 546

Conclusions 547

To facilitate the commercial application of microbial electrosynthesis (MES) a cradle to gate life 548

cycle assessment was performed in this study. Environmental impact variations were performed for 549

MES plants capable of producing acetic, propionic and formic acid, ethanol and methanol. The results 550

showed that formic acid production was always more than 70% lower than the maximum product in 551

the climate change, ozone depletion (OD), human toxicity cancer effects (HT), particulate matter 552

(PM), ionising radiation (IR), photochemical ozone formation (POF), acidification (AC), freshwater 553

europhication (FE) and ecotoxicity (EC) environmental impact categories. The low environmental 554

impacts were mainly due to the lower energy requirement of its reactor and rectification unit. The 555

reduction of greenhouse gas emissions due to MES of formic acid can only be achieved at high 556

conversion and faradaic efficiencies using electricity from fossil fuel dominated grid power. 557

Depending on the product generated, conversion and faradaic efficiencies there can be climate change 558

benefit of using MES for synthesis of chemicals. 559

Life cycle assessment (LCA) of four types of formic acid production routes, a MES, an abiotic 560

electrochemical reduction (AER), a hydrogenation of carbon dioxide (HCD) and hydrolysis of methyl 561

formate (HMF) plant were also developed in this study. The use of HMF to produce formic acid was 562

shown to be environmentally beneficial than the three CO2 utilizing technologies in six (AC,EC, 563

POF,HT,IR and PM) of the nine impact categories analysed when electricity used was generated from 564

the fossil fuel dominant 2014 estimate of the UK national grid. MES had the lowest environmental 565

impact in the climate change, FE and OD impact categories always having a relative value in each 566

impact category less than 60% of the maximum. Assessment of other sources of energy showed that 567

renewable energy helps reduce climate change in both CO2 utilizing and conventional plants. 568

However, in the case of biogas and photovoltaic energy generation environmental burdens shifted to 569

the EC and OD impact categories. Generation of electricity through wind turbines is of particular 570

interest as it had the ability to reduce environmental burdens in all impact categories analysed when 571

compared with the grid. This enabled the MES, AER and HCD to be more environmental friendly 572

than the conventional plant (HMF) in all impact categories. Synthesis of formic acid through MES 573

using wind generated electricity provides benefits and should be employed when MES is industrially 574

applied. 575

As there is still need for more research on the industrial application of MES, this LCA study provided 576

an initial assessment serving as a basis for future LCA studies on large scale application of MES and 577

other types of bioelectrochemical systems. Based on the conclusions, the production of formic acid is 578

of interest and is the best suited product for MES to provide environmental benefits if applied 579

industrially. 580

Data Availability 581

All data, models, or code generated or used during the study appear in the submitted article. 582

Acknowledgements 583

The authors are thankful to the Petroleum Technology Development Fund (Nigeria) for financial 584

funding. 585

References 586

Andrushkevich, T.V., Popova, G.Y., Danilevich, E.V., Zolotarskii, I.A., Nakrokhin, V.B., Nikoro, 587

T.A., Stompel, S.I. and Parmon, V.N. (2014) 'A new gas-phase method for formic acid production: 588

Tests on a pilot plant', Catalysis in Industry, 6(1), pp. 17-24. 589

Bajracharya, S., Srikanth, S., Mohanakrishna, G., Zacharia, R., Strik, D.P. and Pant, D. (2017) 590

'Biotransformation of carbon dioxide in bioelectrochemical systems: State of the art and future 591

prospects', Journal of Power Sources. 592

Bhown, A.S. and Freeman, B.C. (2011) 'Analysis and Status of Post-Combustion Carbon Dioxide 593

Capture Technologies', Environmental Science & Technology, 45(20), pp. 8624-8632. 594

Blanchet, E., Duquenne, F., Rafrafi, Y., Etcheverry, L., Erable, B. and Bergel, A. (2015) 'Importance 595

of the hydrogen route in up-scaling electrosynthesis for microbial CO2 reduction', Energy & 596

Environmental Science, 8(12), pp. 3731-3744. 597

Bulushev, D.A. and Ross, J.R. (2018) 'Towards Sustainable Production of Formic Acid', 598

ChemSusChem, 11(5), pp. 821-836. 599

Burnham, A., Han, J., Clark, C.E., Wang, M., Dunn, J.B. and Palou-Rivera, I. (2012) 'Life-Cycle 600

Greenhouse Gas Emissions of Shale Gas, Natural Gas, Coal, and Petroleum', Environmental Science 601

& Technology, 46(2), pp. 619-627. 602

CEAE (2014) Redox half reaction reductions potentials and free energies (Accessed: 20/07/2016). 603

Cheng, S., Xing, D., Call, D.F. and Logan, B.E. (2009) 'Direct biological conversion of electrical 604

current into methane by electromethanogenesis', Environmental Science and Technology, 43(10), pp. 605

3953-3958. 606

Chiranjeevi, P., Bulut, M., Breugelmans, T., Patil, S.A. and Pant, D. (2019) 'Current trends in 607

enzymatic electrosynthesis for CO2 reduction', Current Opinion in Green and Sustainable Chemistry, 608

16, pp. 65-70. 609

Christodoulou, X., Okoroafor, T., Parry, S. and Velasquez-Orta, S.B. (2017) 'The use of carbon 610

dioxide in microbial electrosynthesis: Advancements, sustainability and economic feasibility', Journal 611

of CO2 Utilization, 18, pp. 390-399. 612

Das, S., Chatterjee, P. and Ghangrekar, M.M. (2018) 'Increasing methane content in biogas and 613

simultaneous value added product recovery using microbial electrosynthesis', Water Science and 614

Technology, 77(5), pp. 1293-1302. 615

Das, S., Das, S., Das, I. and Ghangrekar, M.M. (2019) 'Application of bioelectrochemical systems for 616

carbon dioxide sequestration and concomitant valuable recovery: A review', Materials Science for 617

Energy Technologies, 2(3), pp. 687-696. 618

Das, S. and Ghangrekar, M. (2018) 'Value added product recovery and carbon dioxide sequestration 619

from biogas using microbial electrosynthesis', Indian Journal of Experimental Biology, 56(1), pp. 620

470-478. 621

DBEIS, D.f.B., Energy and Industrial Strategy) (2017) 'Digest of United Kingdom energy statistics, 622

2017'. 623

DBEIS, D.f.B., Energy and Industrial Strategy) (2019) 'Digest of United Kingdom energy statistics, 624

2019'. 625

del Pilar Anzola Rojas, M., Zaiat, M., Gonzalez, E.R., De Wever, H. and Pant, D. (2018) 'Effect of the 626

electric supply interruption on a microbial electrosynthesis system converting inorganic carbon into 627

acetate', Bioresource Technology, 266, pp. 203-210. 628

Endrődi, B., Bencsik, G., Darvas, F., Jones, R., Rajeshwar, K. and Janáky, C. (2017) 'Continuous-629

flow electroreduction of carbon dioxide', Progress in Energy and Combustion Science, 62, pp. 133-630

154. 631

Evangelisti, S., Tagliaferri, C., Brett, D.J.L. and Lettieri, P. (2017) 'Life cycle assessment of a 632

polymer electrolyte membrane fuel cell system for passenger vehicles', Journal of Cleaner Production, 633

142, pp. 4339-4355. 634

Finn, C., Schnittger, S., Yellowlees, L.J. and Love, J.B. (2012) 'Molecular approaches to the 635

electrochemical reduction of carbon dioxide', Chemical Communications, 48(10), pp. 1392-1399. 636

Giddings, C.G.S., Nevin, K.P., Woodward, T., Lovley, D.R. and Butler, C.S. (2015) 'Simplifying 637

microbial electrosynthesis reactor design', Frontiers in microbiology, 6, pp. 468-468. 638

Gooding, J.J. and Gonçales, V.R. (2017) 'Recent advances in the molecular level modification of 639

electrodes for bioelectrochemistry', Current Opinion in Electrochemistry, 5(1), pp. 203-210. 640

Gupta, K., Bersani, M. and Darr, J.A. (2016) 'Highly efficient electro-reduction of CO2 to formic acid 641

by nano-copper', Journal of Materials Chemistry A, 4(36), pp. 13786-13794. 642

Havukainen, J., Zhan, M., Dong, J., Liikanen, M., Deviatkin, I., Li, X. and Horttanainen, M. (2017) 643

'Environmental impact assessment of municipal solid waste management incorporating mechanical 644

treatment of waste and incineration in Hangzhou, China', Journal of Cleaner Production, 141, pp. 453-645

461. 646

Hutchings, G.J. (2016) 'Methane Activation by Selective Oxidation', Topics in Catalysis, 59(8), pp. 647

658-662. 648

Jaramillo, P., Griffin, W.M. and Matthews, H.S. (2007) 'Comparative Life-Cycle Air Emissions of 649

Coal, Domestic Natural Gas, LNG, and SNG for Electricity Generation', Environmental Science & 650

Technology, 41(17), pp. 6290-6296. 651

Jiang, Y. and Jianxiong Zeng, R. (2018) 'Expanding the product spectrum of value added chemicals in 652

microbial electrosynthesis through integrated process design—A review', Bioresource Technology, 653

269, pp. 503-512. 654

Jiang, Y., May, H.D., Lu, L., Liang, P., Huang, X. and Ren, Z.J. (2019) 'Carbon dioxide and organic 655

waste valorization by microbial electrosynthesis and electro-fermentation', Water Research, 149, pp. 656

42-55. 657

Jourdin, L., Freguia, S., Flexer, V. and Keller, J. (2016) 'Bringing High-Rate, CO2-Based Microbial 658

Electrosynthesis Closer to Practical Implementation through Improved Electrode Design and 659

Operating Conditions', Environmental Science & Technology, 50(4), pp. 1982-1989. 660

Jourdin, L., Grieger, T., Monetti, J., Flexer, V., Freguia, S., Lu, Y., Chen, J., Romano, M., Wallace, 661

G.G. and Keller, J. (2015) 'High Acetic Acid Production Rate Obtained by Microbial Electrosynthesis 662

from Carbon Dioxide', Environmental Science & Technology, 49(22), pp. 13566-13574. 663

Jung, S., Lee, J., Park, Y.-K. and Kwon, E.E. (2020) 'Bioelectrochemical systems for a circular 664

bioeconomy', Bioresource Technology, p. 122748. 665

Kaldellis, J.K. and Apostolou, D. (2017) 'Life cycle energy and carbon footprint of offshore wind 666

energy. Comparison with onshore counterpart', Renewable Energy, 108, pp. 72-84. 667

Kuhl, K.P., Cave, E.R., Abram, D.N. and Jaramillo, T.F. (2012) 'New insights into the 668

electrochemical reduction of carbon dioxide on metallic copper surfaces', Energy & Environmental 669

Science, 5(5), pp. 7050-7059. 670

Logan, B.E., Hamelers, B., Rozendal, R., Schröder, U., Keller, J., Freguia, S., Aelterman, P., 671

Verstraete, W. and Rabaey, K. (2006) 'Microbial Fuel Cells:  Methodology and Technology', 672

Environmental Science & Technology, 40(17), pp. 5181-5192. 673

Marshall, C.W., Ross, D.E., Fichot, E.B., Norman, R.S. and May, H.D. (2013) 'Long-term operation 674

of microbial electrosynthesis systems improves acetate production by autotrophic microbiomes', 675

Environmental science & technology, 47(11), pp. 6023-6029. 676

Mordor Intelligence (2019) Formic acid market - growth, trends, and forecast (2020 - 2025). India. 677

[Online]. Available at: https://www.mordorintelligence.com/industry-reports/formic-acid-market. 678

Nevin, K.P., Hensley, S.A., Franks, A.E., Summers, Z.M., Ou, J., Woodard, T.L., Snoeyenbos-West, 679

O.L. and Lovley, D.R. (2011) 'Electrosynthesis of Organic Compounds from Carbon Dioxide Is 680

Catalyzed by a Diversity of Acetogenic Microorganisms', Applied and Environmental Microbiology, 681

77(9), pp. 2882-2886. 682

Oswald, F., Stoll, I.K., Zwick, M., Herbig, S., Sauer, J., Boukis, N. and Neumann, A. (2018) 'Formic 683

Acid Formation by Clostridium ljungdahlii at Elevated Pressures of Carbon Dioxide and Hydrogen', 684

Frontiers in Bioengineering and Biotechnology, 6(6). 685

Pérez-Fortes, M., Schöneberger, J.C., Boulamanti, A., Harrison, G. and Tzimas, E. (2016) 'Formic 686

acid synthesis using CO2 as raw material: Techno-economic and environmental evaluation and 687

market potential', International Journal of Hydrogen Energy, 41(37), pp. 16444-16462. 688

Prévoteau, A., Carvajal-Arroyo, J.M., Ganigué, R. and Rabaey, K. (2020) 'Microbial electrosynthesis 689

from CO2: forever a promise?', Current Opinion in Biotechnology, 62, pp. 48-57. 690

Qiao, J., Liu, Y., Hong, F. and Zhang, J. (2014) 'A review of catalysts for the electroreduction of 691

carbon dioxide to produce low-carbon fuels', Chemical Society Reviews, 43(2), pp. 631-675. 692

Rabaey, K., Largus Angenent, Uwe Schroder and Keller, J. (eds.) (2010) Bioelectrochemical Systems: 693

from extracellular electron transfer to biotechnological application. London, UK: IWA Publishing. 694

Rabaey, K. and Rozendal, R.A. (2010) 'Microbial electrosynthesis — revisiting the electrical route for 695

microbial production', Nat Rev Micro, 8(10), pp. 706-716. 696

Reda, T., Plugge, C.M., Abram, N.J. and Hirst, J. (2008) 'Reversible interconversion of carbon 697

dioxide and formate by an electroactive enzyme', Proceedings of the National Academy of Sciences, 698

105(31), pp. 10654-10658. 699

Reutemann, W. and Kieczka, H. (2000) 'Formic Acid', in Ullmann's Encyclopedia of Industrial 700

Chemistry. Wiley-VCH Verlag GmbH & Co. KGaA. 701

Saavalainen, P., Turpeinen, E., Omodara, L., Kabra, S., Oravisjärvi, K., Yadav, G.D., Keiski, R.L. 702

and Pongrácz, E. (2017) 'Developing and testing a tool for sustainability assessment in an early 703

process design phase – Case study of formic acid production by conventional and carbon dioxide-704

based routes', Journal of Cleaner Production, 168, pp. 1636-1651. 705

Sadhukhan, J., Ng, K.S. and Hernandez, E.M. (2014) Biorefineries and chemical processes: design, 706

integration and sustainability analysis. John Wiley & Sons. 707

Santoro, C., Arbizzani, C., Erable, B. and Ieropoulos, I. (2017) 'Microbial fuel cells: From 708

fundamentals to applications. A review', Journal of Power Sources, 356, pp. 225-244. 709

Shemfe, M., Gadkari, S., Yu, E., Rasul, S., Scott, K., Head, I.M., Gu, S. and Sadhukhan, J. (2018) 710

'Life cycle, techno-economic and dynamic simulation assessment of bioelectrochemical systems: A 711

case of formic acid synthesis', Bioresource Technology, 255, pp. 39-49. 712

Speck, R., Selke, S., Auras, R. and Fitzsimmons, J. (2015) 'Choice of Life Cycle Assessment 713

Software Can Impact Packaging System Decisions', Packaging Technology and Science, 28(7), pp. 714

579-588. 715

Speck, R., Selke, S., Auras, R. and Fitzsimmons, J. (2016) 'Life Cycle Assessment Software: 716

Selection Can Impact Results', Journal of Industrial Ecology, 20(1), pp. 18-28. 717

Srikanth, S., Alvarez‐Gallego, Y., Vanbroekhoven, K. and Pant, D. (2017) 'Enzymatic 718

electrosynthesis of formic acid through carbon dioxide reduction in a bioelectrochemical system: 719

effect of immobilization and carbonic anhydrase addition', ChemPhysChem, 18(22), pp. 3174-3181. 720

Srikanth, S., Maesen, M., Dominguez-Benetton, X., Vanbroekhoven, K. and Pant, D. (2014) 721

'Enzymatic electrosynthesis of formate through CO2 sequestration/reduction in a bioelectrochemical 722

system (BES)', Bioresource Technology, 165, pp. 350-354. 723

Sutter, J. (2007) Life Cycle Inventories of chemicals. Swiss Centre for Life Cycle Inventories. 724

Tao, M., Qun, F., Hengcong, T., Zishan, H., Mingwen, J., Yunnan, G., Wangjing, M. and Zhenyu, S. 725

(2017) 'Heterogeneous electrochemical CO 2 reduction using nonmetallic carbon-based catalysts: 726

current status and future challenges', Nanotechnology, 28(47), p. 472001. 727

Thinkstep (2016) GaBi - Product Sustainability Software [Computer program]. 728

Thinkstep (2017) Overview of GaBi LCA Software. Available at: http://www.gabi-software.com/uk-729

ireland/index/ (Accessed: 27/09/2017). 730

Tromp, T.K., Shia, R.-L., Allen, M., Eiler, J.M. and Yung, Y.L. (2003) 'Potential Environmental 731

Impact of a Hydrogen Economy on the Stratosphere', Science, 300(5626), p. 1740. 732

Vassilev, I., Hernandez, P.A., Batlle Vilanova, P., Freguia, S., Krömer, J.O., Keller, J., Ledezma, P. 733

and Virdis, B. (2018) 'Microbial Electrosynthesis of Isobutyric, Butyric, Caproic acids and 734

corresponding Alcohols from Carbon dioxide', ACS Sustainable Chemistry & Engineering. 735

Vassilev, I., Kracke, F., Freguia, S., Keller, J., Krömer, J.O., Ledezma, P. and Virdis, B. (2019) 736

'Microbial electrosynthesis system with dual biocathode arrangement for simultaneous acetogenesis, 737

solventogenesis and carbon chain elongation', Chemical Communications, 55(30), pp. 4351-4354. 738

Zhang, L., Zhao, Z.-J. and Gong, J. (2017) 'Nanostructured Materials for Heterogeneous 739

Electrocatalytic CO2 Reduction and their Related Reaction Mechanisms', Angewandte Chemie 740

International Edition, 56(38), pp. 11326-11353. 741

742

743

744

745

746

747

748

749

750

751

752

753

754

755

756

757

Table 1. Energy and water consumption of assessed acetic, formic, propionic acids, methanol and ethanol MES plants 758

Parameter Acetic

acid

Formic

acid

Propionic

acid

Methanol Ethanol

Energy consumed capturing and

processing CO2 (GJ/yr)

379.68 247.68 461.65 355.80 494.89

Process water consumed (m3/yr) 740.81 483.28 930.52 1319.66 1448.42

Energy consumed by MES plant

(rectification unit, bioreactor and

gas separator) (GJ/yr)

95457.09 8844.97 92211.75 31779.49 41051.51

Energy consumed for packaging

(GJ/yr)

126.70 109.00 134.67 167.92 168.58

759

760

761

762

763

764

765

766

767

768

769

770

771

772

773

774

775

776

777

Table 2. Reaction balances and cell potential for CO2 reduction into acetic, formic and propionic acids, methanol and 778 ethanol MES plants 779

Product Overall reaction Anode

potential

(V vs.

SHE)

Cathode

potential

(V vs.

SHE)

MES

reactor

Potential

(V vs.

SHE)

References

Acetic acid 2𝐶𝑂2 + 4𝐻2𝑂𝐵𝑖𝑜𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡→ 𝐶𝐻3𝐶𝑂𝑂𝐻 + 2𝐻2𝑂 + 2𝑂2 0.817 -0.393 -1.210 (Marshall et

al., 2013)

Formic acid 𝐶𝑂2 + 𝐻2𝑂𝐵𝑖𝑜𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡→ 𝐻𝐶𝑂𝑂𝐻 + 0.5𝑂2

0.817 -0.400 -1.217 (Nevin et

al., 2011;

CEAE,

2014)

Propionic

acid

3𝐶𝑂2 + 7𝐻2𝑂𝐵𝑖𝑜𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡→ 𝐶𝐻3𝐶𝐻2𝐶𝑂𝑂𝐻 + 4𝐻2𝑂 + 3.5𝑂2

0.817 -0.290 -1.107 (CEAE,

2014)

Methanol 𝐶𝑂2 + 3𝐻2𝑂𝐵𝑖𝑜𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡→ 𝐶𝐻3𝑂𝐻 + 𝐻2𝑂 + 1.5𝑂2 0.817 -0.390 -1.207 (CEAE,

2014)

Ethanol 2𝐶𝑂2 + 6𝐻2𝑂𝐵𝑖𝑜𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡→ 𝐶𝐻3𝐶𝐻2𝑂𝐻 + 3𝐻2𝑂 + 3𝑂2

0.817 -0.335 -1.152 (Blanchet et

al., 2015)

780

781

782

783

784

785

786

787

Table 3. Life cycle inventory of the formic acid production routes analysed 788

Unit MES Plant AER Plant HCD Plant HMF

Plant

Reference (Pérez-

Fortes et

al., 2016)

(Sutter,

2007)

Raw

material

CO2 t/yr 1,094 1,094 830 -

H2O t/yr 483 483 560 600

CO t/yr - - - 614

CH3OH t/yr - - - 40

Product

HCOOH t/yr 1,000 1,000 1,000 1,000

Output

O2 t/yr 348 348 480 -

Energy

CO2 capture

energy

GJ/yr 192 192 146 -

Plant

electricity

GJ/yr 7,183 11,989 14,652 1,044

Steam GJ/yr 1,771 1,771 10,030 19,500

Total energy GJ/yr 8,954 13,761 24,682 20,544

789

790

791

792

793

794

795

796

Table 4. Greenhouse gases emissions from acetic acid, propionic acid, formic acid, ethanol and methanol MES plant (1,000 797 t/yr) modelled using GaBi (Electricity from 2014 UK national grid) compared with emissions using natural gas as electricity 798

Products

Greenhouse gases GWP (Electricity

from UK national

Grid) (t CO2 eqv)

GWP (Electricity

from burning

natural gas) (t CO2

eqv)

(Christodoulou et

al., 2017)

Carbon

dioxide (t)

Methane

(t)

Nitrous

Oxide

(t)

Sulfur

Hexafluoride

(t)

1 t CO2

eqv

25 t CO2

eqv

298 t

CO2

eqv

22800 t CO2

eqv (x 10-7)

Acetic

Acid

81,203 118 0.90 1.03 67,800 61,638

Propionic

Acid

84,171 135 1.11 1.08 67,700 48,797

Ethanol 59,703 144 1.42 1.52 42,100 -7,526

Methanol 45,636 109 1.32 1.15 33,200 -5,577

Formic

Acid

12,305 28 0.27 0.32 2,200 -4,428

799

800

801

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

Table 5. Greenhouse gases emissions from MES, AER, HCD and HMF plants (1,000t/yr) modelled using GaBi (Electricity 817 from 2014 UK national grid) 818

Plant

Greenhouse gases GWP (Electricity

from UK

national Grid) (t

CO2 eqv)

Carbon

dioxide (t)

Methane

(t)

Nitrous

Oxide

(t)

Sulfur

Hexafluoride

(t)

1 t CO2

eqv

25 t CO2

eqv

298 t

CO2

eqv

22800 t CO2

eqv (x 10-7)

MES 12,221 28 0.27 0.32 2,120

AER 19,397 46 0.45 0.51 9,820

HCD 28,806 60 0.56 0.61 22,200

HMF 17,378 33 0.15 0.10 18,400

819

820

821

822

823

824

825

826

827

828

829

830

831

832

833

834

835

836

837

838

Table 6. Global warming potential from the eight different electricity sources analysed for production of 1,000t/yr formic 839 acid using MES, HCD and HMF plants 840

Plant

GWP (t CO2 eqv)

Coal Gas Oil Nuclear Hydro Biogas Wind Photo

voltaic

MES 9,940 -615 9,367 -9,600 -9,550 -5,780 -9,581 -8,687

AER 22,700 5,309 21,797 -9,530 -9,453 -3,220 -9,500 -8,024

HCD 37,932 16,755 36,786 -1,271 -1,178 6,390 -1,240 559

HMF 19,500 18,032 19,448 16,759 16,766 17,300 16,800 16,900

841

842

843

844

845

846

847

848

849

850

851

852

853

854

855

856

857

858

859

860

Fig. 1. System boundary for 1,000t/yr MES plant for formic, acetic and propionic acids, methanol and ethanol production 861

862

863

864

865

866

867

868

869

870

871

872

873

874

875

876

877

878

879

880

Fig. 2. Biotic and abiotic electro catalytic formic acid production 881

882

883

884

885

886

887

888

889

890

891

892

893

894

895

Fig. 3. Climate change contributions of MES plant for the production of acetic, propionic and formic acids, methanol and 896 ethanol 897

898

899

900

901

902

903

904

905

906

907

908

909

910

Fig. 4. Life cycle environmental burdens (Human toxicity, ionizing radiation, ozone depletion and particulate matter) of the 911 MES plant for the production of acetic, propionic and formic acids, ethanol and methanol using ILCD method. Results are 912 displayed relative to the maximum value in each impact category. 913

914

915

916

917

918

919

920

921

922

923

924

925

926

Fig. 5. Life cycle environmental burdens (Acidification, ecotoxixity, eutrophication and photochemical ozone formation) of 927 the MES plant for the production of acetic, propionic and formic acids, ethanol and methanol using ILCD method. Results 928 are displayed relative to the maximum value in each impact category. 929

930

931

932

933

934

935

936

937

938

939

940

941

942

943

944

Fig. 6. Scenario analysis of CO2 conversion and faradaic efficiency 945

946

947

948

949

950

951

952

953

954

955

956

957

958

Fig. 7. Life cycle environmental burdens (Acidification, ecotoxixity, eutrophication and photochemical ozone formation) of 959 MES, AER, HCD and HMF plants using ILCD method. Results are displayed relative to the maximum value in each impact 960 category. 961

962

963

964

965

966

967

968

969

970

971

972

973

974

975

Fig. 8. Life cycle environmental burdens (Human toxicity, ionizing radiation, ozone depletion and particulate matter) of 976 MES, HCD and HMF Plants using ILCD method. Results are displayed relative to the maximum value in each impact 977 category. 978

979

980

Fig. 9. Life cycle environmental burdens of the MES plant for eight different electricity sources using ILCD method. A) 981 Acidification, ecotoxixity, eutrophication and photochemical ozone formation B) Human toxicity, ionizing radiation, ozone 982 depletion and particulate matter. Results are displayed relative to the value of the UK national grid in each impact category. 983

984

Fig. 10. Life cycle environmental burdens of the MES, AER, HCD and HMF plant using wind as the electricity source. A) 985 Acidification, ecotoxixity, eutrophication and photochemical ozone formation B) Human toxicity, ionizing radiation, ozone 986 depletion and particulate matter. Results are displayed relative to the maximum value in each impact category. 987

988