Accepted Manuscript

Ferric ion pair mediated biomass redox flow fuel cell and related chemical re-
action kinetics study

Xihong Zu, Lele Sun, Jian Gong, Xiaochun Liu, Yangxi Liu, Xu Du, Wei Liu,
Lifen Chen, Guobin Yi, Weigang Zhang, Wenjing Lin, Weizhao Li, Yulin Deng

PII: S1385-8947(18)30548-5
DOI: https://doi.org/10.1016/j.cej.2018.03.190
Reference: CEJ 18797

To appear in: Chemical Engineering Journal

Received Date: 9 January 2018

Revised Date: 30 March 2018
Accepted Date: 31 March 2018

Please cite this article as: X. Zu, L. Sun, J. Gong, X. Liu, Y. Liu, X. Du, W. Liu, L. Chen, G. Yi, W. Zhang, W. Lin,
W. Li, Y. Deng, Ferric ion pair mediated biomass redox flow fuel cell and related chemical reaction kinetics study,
Chemical Engineering Journal (2018), doi: https://doi.org/10.1016/j.cej.2018.03.190

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ferric ion pair mediated biomass redox flow fuel cell and
related chemical reaction kinetics study
Xihong Zua, Lele Suna, Jian Gongb, Xiaochun Liua,Yangxi Liua, Xu Dub, Wei Liub, Lifen

Chena, Guobin Yia,*, Weigang Zhanga, Wenjing Lina, Weizhao Lia, Yulin Dengb,*

a
School of Chemical Engineering and Light Industry, Guangdong University of Technology,

Guangzhou, Guangdong, 510006, P.R. China

b
School of Chemical & Biomolecular Engineering, Georgia Institute of Technology，500

10th Street N.W., Atlanta, GA 30332-0620, USA

Corresponding author: yulin.deng@ipst.gatech.edu; ygb702@163.com.

Abstract: Here, a novel and low cost redox flow fuel cell that directly converts raw biomass

to electricity at low temperature without utilizing any noble metal catalyst is reported. In

this cell, Fe3+ ion, which has strong oxidizing power and is chemical stable, was directly

utilized as an oxidant for biomass in anolyte. The reduced Fe2+ acted as charge carriers,

transferring the electrons to the anode of the cell and being converted back to Fe3+

simultaneously. The degradation chemistry of the biomass from the oxidation reaction and

the apparent oxidation kinetics were investigated using glucose as the biomass model

compound. The results showed that glucose was decomposed to small organic molecules

and even CO2, and the reaction rate was significantly affected by temperature and content of

reactant. The apparent activation energy of the redox reaction is 127.85 kJ·mol-1 based on

the linear relation equation at the initial stage of the reaction. And the preference conditions

of the redox reaction are 100～110 ℃, 0.2～0.5 mol·L-1 glucose and the mol ratio of

1
C6H12O6/ FeCl3 from 1:10 to 1:12. Meantime, a high potential (VO2)2SO4 solution was

synthesized as catholyte which was regenerated by O2 in the present of small amount of

HNO3. The power density and discharge current density of the biomass redox flow fuel cell

could reach 125.7 mW•cm-2 and 314.2 mA•cm-2, respectively, which was about 755 times

higher than the reported microbial fuel cell by utilizing glucose as substrate. This work

provides a promising way for converting abundant natural biomasses, such as crop straws

and bagasse, into electricity at low temperature (80 ℃～110 ℃).

Keywords: redox flow fuel cell; biomass; reaction kinetics; ferric chloride; redox couple

1. Introduction

Renewable energy sources have been drawing an extensive attention for less consumption of

fossil fuels and reducing environmental issues by employing different innovative energy

conversion technologies [1-3]. Biomass is a clean and renewable energy resource with

abundant reserves and low cost derived from various agriculture crops and residues, forestry

residues, and other carbonaceous wastes. The efficient exploitation of biomass is a

promising option to solve problems on energy resources and ecological environment [4].

Therefore, the development of high-efficiency energy conversion technologies is very

important for the production of electricity [5], biofuels [6], activated carbons [7] and basic

chemicals [8]. Among various energy conversion technologies, biomass-based fuel cells are

promising technologies for conversion of the stored chemical energy of biomass into

electricity [9-11].

Recently, biomass-based fuel cells have received much attention, including solid oxide

2
fuel cells, microbial fuel cells, photo fuel cells and redox flow fuel cells. Among these, solid

oxide fuel cell technology is a conventional technology for converting the stored chemical

energy of biomass into electricity via an integrated biomass gasification and fuel cell system.

The advantages of this technology include excellent fuel flexibility, greater tolerance toward

contaminants, and the high conversion efficiency of fuel-to-electricity [12-15]. However, the

operation temperature of solid oxide fuel cells is very high (>450℃) , and the biomass has to

be gasified into gas or pyrolyzed into biochar at high temperature (500～1,000℃) before

using as fuels [16-19]. Although microbial fuel cells can convert organic substrates into

electricity at low temperature through bio-oxidation process, the electric power density is

very low because of the limited electrogenic activity and lifetime of the anodic

microorganisms, electron transfer between biofilm and the anode, and the poor cathodic

performance owing to kinetic sluggishness of cathodic oxygen reduction reaction [20-23].

You et al developed a microbial fuel cell by using metal-organic-framework-derived

electrocatalysts for oxygen reduction reaction, and its maximum power density was 0.1665

mW•cm-2 when glucose is used as the substrate [24]. Miran et al reported a microbial fuel

cell by utilizing orange peel as a carbon source, and the maximum current density and

power density were about 0.0847 mA•cm-2 and 0.0359 mW•cm-2, respectively [25]. Li et al

reported a dual chamber microbial fuel cell which can convert Cr(VI) into nontoxic Cr(III)

in wastewater and generate bioelectricity. Its maximum power density can achieve 0.1222

mW•cm-2 [26]. Photo fuel cells can simultaneously convert solar and bioenergy into

electricity by the photocatalytic biomass oxidation on a photoanode combined with Pt as

dark cathode for oxygen reduction reaction. It can be clearly seen that the noble metal

3
catalyst was utilized, and the electric power density of photo fuel cells is very low (<0.082

mW•cm-2). Also, the kinds of biomass as fuels are limited [27,28]. Among various biomass

fuel cells, redox flow fuel cells can directly convert biomass into electricity at low

temperature mediated by redox couples [29-32]. In our previous work, we present a

solar-induced hybrid fuel cell that utilized H3PMo12O40 (Mo6+) as the photocatalyst to

oxidize cellulose while being reduced from Mo 6+ to Mo5+, and then Mo5+ changed back to

Mo6+ again by releasing electron on anode plate. And the released electrons passed through

an external circuit to produce electricity [31]. Afterward, we further reported a noble-metal

free and liquid-catalyst fuel cell for converting biomass into electricity directly and

efficiently. In this fuel cell, biomass oxidation and oxygen reduction were catalyzed by two

kinds of polyoxometalates solutions in anode and cathode, respectively. And the power

density was 34 mW•cm-2 when starch was utilized as fuels [32]. Ding et al reported a

biomass fuel cell system by using phosphomolybdic acid as an electron mediator and proton

carrier to achieve an integration of wheat straw pretreatment for ethanol and conversion of

biomass into electricity [33]. In their study, POM was used as the mediate oxidation agent

and charge transfer agent. Although POM is considered as a low cost catalyst, its high

molecular weight and large molecular size limits its charge transfer rate in anode [31-33].

Besides, the electric power density still needs to be improved for practical application, and

the reaction kinetics of biomass in the redox flow fuel cell system needs to be studied.

In this work, the simple ion redox couples with low cost and high solubility were

developed to serve as both anode and cathode redox couples in the direct biomass redox

flow fuel cell as shown in scheme 1. The anodic half-cell was mediated by the Fe3+/Fe2+

4
redox couple which was paired with /VO2+ redox couple at the cathodic half-cell.

Biomass was oxidized in the anolyte tank by FeCl3, and the electrons were simultaneously

stored in Fe2+. The anolyte with Fe2+ was cyclically pumped through the anode for releasing

electrons to generate electricity and Fe3+. The electrons passed through the external circuit

and were captured by at the cathode to give VO2+. In order to regenerate ions in

the catholyte for stable electricity output by maintaining high electric potential, O2 was

bubbled in the catholyte to oxidize the VO2+ ions into in the presence of nitric acid as

catalyst at 80℃. Besides, the apparent reaction kinetics of biomass oxidized by FeCl3 was

also discussed.

Scheme1 a) Schematic illustration of the direct biomass redox flow fuel cell based on two simple ion redox couples. b)

Working principle of the redox flow fuel cell.

5
2. Materials and methods

2.1 Materials

Glucose (C6H12O6), molybdenum(Ⅵ) oxide (MoO3, 99.5%), vanadium(Ⅴ) oxide (V2O5,

99.6%), hydrogen peroxide (H2O2, 35%), hydrochloric acid (HCl, 36%), phosphoric acid

(H3PO4, 85%) and sulfuric acid (H2SO4, 98%) were all from Alfa Aesar. Ferric chloride with

6-hydrate (FeCl3•6H2O), sodium acetate (CH3COONa), acetic acid (CH3COOH, >99.7%)

and 1,10-phenanthroline monohydrate (C₁ ₂ H₈ N₂ ·H₂ O) were from VWR.

Fe(NH4)2(SO4)2•6H2O was from Ward’s Science. Nafion® 115 membrane (127 micrometers

thick) was from DuPont.

2.2 Kinetic experiments of the redox reactions between biomass and FeCl3 in anolyte

The calculated amount of glucose, FeCl3•6H2O and HCl were dissolved in DI water and

then diluted to a certain volume. Then the mixed solution was poured into a three-neck flask,

and heated to a certain temperature under stirring for reacting different times. The

decomposed products of glucose were measured by the Micro GC and GC-MS. The reaction

rate was studied by measuring the concentration of Fe2+ in anolyte.

Besides, different kinds of biomasses were also used as fuels, including bagasse, orange

peel and corn stalks. 2.0 g biomass and 27.0 g FeCl3•6H2O were added into 50 mL acid

solution. Then the mixed solution was poured into a flask with three necks, and heated to

110 ℃ under stirring for reacting different times.

2.3 Analysis of redox rate between FeCl3 and biomass in anolyte

Phenanthroline spectrophotometric method was used to measure the concentration of Fe2+

during the redox reaction between FeCl3 and biomass. To obtain the calibration curve, 0.1 g

6
1,10-phenanthroline monohydrate was dissolved in 100 mL DI water by stirring and heating

to 80 ℃. 136 g sodium acetate and 120 mL acetic acid were dissolved in DI water and

diluted to 500 mL as buffer solution. 0.0365 g Fe(NH4)2(SO4)2•6H2O was dissolved in 50

mL 1 M HCl solution and then diluted to 500 mL as standard Fe2+ solution (0.01 mg/mL).

Then, a series of standard samples was prepared by accurately pipetting calculated volumes

of standard Fe2+ solution, 10 ml buffer solution and 10 mL 1,10-phenanthroline solution into

the 100 mL volumetric flask and diluting to the mark with DI water. Then the standard

samples were measured using the UV-Visible Spectrophotometer to get the calibration

curve between the absorbency and concentration of Fe2+. According to the above method,

the concentration of Fe2+ in different reaction solution was measured and used for studying

the apparent reaction kinetics.

Besides, the reduction degree (r) of Fe3+ to Fe2+ in the redox experiments was defined as

the number of mole electrons accepted by per mole Fe3+. In other words, it is the mole ratio

of Fe2+ generated in the anolyte to the initial mole number of Fe3+ in the anolyte as

described in the equation (1):

(1)

2.4 Preparation of catholyte

Two kinds of catholytes were synthesized in this study. One is (VO2)2SO4 solution which

was synthesized by gradually dissolving 20.0 g V2O5 and 75 mL H2SO4 (98%) into 525 mL

DI water to generate the transparent solution with yellow color. The other one is

high-vanadium Mo-V-phosphoric heteropoly acid aqueous solution with modified

composition H12P3Mo18V7O85 (noted as POM) was synthesized according to the literature

7
[32]. The typical synthesis method included two steps. Step 1 was that 25.48 g V2O5 was

added into 1430 mL DI water with ice bath. Then 171 mL H2O2 was added into the V2O5

solution gradually and stirred for 1 h. After that 2.66 mL H3PO4 was added into the solution

and kept reacting for another 3 h. Step 2 was that 103.68 g MoO3 was added into 1000 mL

DI water with stirring. Then 5.61 mL H3PO4 was added into the solution and heated to

boiling. After that, the prepared step1 solution was added into the step 2 solution gradually

under boiling state. And the mixed solution was kept boiling until MoO3 dissolved

thoroughly. Then the solution was evaporated to 135 mL to obtain the POM solution with

the concentration of 0.3 mol·L-1.

2.5 Assembly of direct biomass redox flow fuel cell and test methods

The direct biomass redox flow fuel cell mainly included two all carbon graphite

electrodes separated by Nafion 115 membrane as shown in scheme 1. The inner face of the

all carbon graphite electrodes has a serpentine flow channel with 2 mm wide, 2 mm deep

and 5 cm long (a total active area of 1 cm2). And the flow channel was filled with graphite

felt. The Nafion 115 membrane was pretreated by boiling in 1 mol·L-1 H2SO4 and 3% H2O2

for 30 min, respectively. After that, the membrane was washed and soaked with DI water.

Besides, PTFE gaskets were placed on the circumference of the graphite flow-field plates to

prevent any leakage. The anolyte and catholyte in a 80 ℃ water bath were pumped to the

anode and cathode of the fuel cell with 10 mL·min-1, respectively. However, the cell was run

at ambient temperature in all the tests.

In order to quickly regenerate the catholyte, small amount of HNO3 (1 mL per 150 mL

catholyte) was added to (VO2)2SO4 solution at 80℃ with continuously O2 flow [34]. The

8
output current and output voltage were tested by the electrochemical working station.

2.6 Characterization

The gaseous productions in the reaction of glucose and FeCl3 were collected with a gas

collection bag and analyzed by the Agilent 490 Micro GC with a Pora Plat U column. The

components of the degradation solution was characterized by the GC-MS (Agilent 6890

with a 6890 autosampler) with an Agilent DB-5 capillary column. The Mass spectrometer

was a Micromass AutoSpec M. The electrochemical properties of the fuel cell were all

tested by the electrochemical working station (Princeton Applied Research). The absorbency

of Fe2+ in different solution was measured by UV-Vis spectrophotometer (Agilent

Technologies Inc.). The electrode potential was measured by mercurous chloride electrode

and platinum electrode.

3. Results and discussion

3.1 Decomposition of glucose oxidized by FeCl3

Since glucose is the typical structure unit and decomposition product of biomass, it was

used here as a model biomass for studying the mechanism and kinetic behavior of the

oxidation of biomass by FeCl3. Fig. 1 shows the decomposition products of glucose and the

generation rate of Fe2+ in anolyte which reflects the electron-transfer rate from glucose to

FeCl3. It was noted that the main gaseous product was CO2, which demonstrated that the

C-C bond of the glucose could be broken by FeCl3 (Fig. 1a). It should be noted that the gas

phase was collected by bubbling pure nitrogen to a sealed anode half-cell so all CO2 and

other carbon compounds in the gas phase were from glucose degradation but not from air.

And the formic acid, acetic acid and 4- hydroxybutyric acid, which are the typical

9
intermediate products during the decomposition of glucose, were detected in the reaction

solution (Fig. 1b). Besides, the Fe2+ concentration increased obviously with the reaction time

increasing. When the concentration of glucose was 0.2 mol·L-1, the generation rate of Fe2+

was improved greatly with the concentration of Fe3+ increasing (Fig. 1c). However, if the

amount of FeCl3 in the reaction solution was too high, crystallization could happen. Thus,

the preference mol ratio of C6H12O6/ FeCl3 is 1:10 to 1:12.

Fig.1 a) GC analysis of the gaseous products during the decomposition of glucose. b) GC-MS analysis of the liquid

products from the decomposition of glucose. c) Fe2+ concentration in anolyte with different initial contents of FeCl3 under

reflux. Note: the gas products were collected by bubbling N2 into the reaction system. And the ratios inserted in Fig.1c

were the mol ratios of C6H12O6/ FeCl3.

10
3.2 Principle of the anode half-cell

In principle, glucose can be decomposed to small molecules and even CO2 when it is

oxidized by FeCl3. And the electrons from glucose or its intermediates were stored in Fe 2+

and could be transferred to the anode for generating electricity when the anolyte was

pumped through the anode half-cell. The reaction in anolyte can be written as:

(2)

Then Fe2+ releases the electrons to the anode as shown in equation (3):

(3)

Therefore, the generation rate and concentration of Fe2+ in anolyte has a significant

influence on the output current and output voltage of the redox flow fuel cell.

3.3 Effects of glucose concentration and temperature on the generation rate of Fe2+ and

electrode potential in anolyte

Because the electricity generating efficiency of the redox flow fuel cell was directly

affected by the generation rate of Fe2+ and electrode potential of anolyte, the reaction

temperature and the initial molar ratio of glucose and FeCl3 were changed in order to

systematically investigate their influences. Fig. 2a and 2b show the changes of Fe2+

concentration and electrode potential in anolyte which has 0.2 mol·L-1 glucose and 2

mol·L-1 FeCl3 in the initial solution. It can be seen that the reaction temperature had a

significant influence on the generating rate of Fe2+. When the temperature was under 90 ℃,

the generation rate of Fe2+ was very slow, indicating that glucose was difficult to be oxidized

by FeCl3 at low temperature. However, with the temperature increasing, the redox reaction

rate became fast. On the other hand, the reaction temperature also has significant influence

11
on the electrode potentials of the anolyte (Fe3+/Fe2+). Clearly, the decreasing rate of

electrode potentials became faster with the temperature increasing (Fig. 2b). The results

were also reflected by the Nernst Equation as shown in equation (4):

0.059 aFe 3
e   0  lg （4）
n aFe 2

Where
 e is the electrode potential at the equilibrium state and 25℃,  0 is the standard

a Fe 3 a Fe 2
electrode potential, and are the chemical activity of Fe3+ and Fe2+ in the anolyte. n

is the number of moles of electrons transferred in the anode half-cell reaction Fe3+ + e = Fe2+.

It can be known from the equation (4) that the electrode potential
 e decreased when the

redox reaction between FeCl3 and glucose was carried out. That was because the

concentration of Fe3+ decreased while Fe2+ concentration increased with the redox reaction

a Fe 3
carrying on, resulting in the decrease of the ratio. Based on the same principle, the
a Fe 2 

electrode potential
 e at high temperature was relatively lower than at low temperature

with the same reaction time. That’s because the reaction extent and rate of equation (2) were

a Fe 3
both improved with the temperature increasing, resulting in the decrease of the ratio.
a Fe 2 

These results are consistent with the variation tendency of Fe2+ concentration in anolyte.

Furthermore, the initial content of glucose also had important influence on the generating

rate of Fe2+ and electrode potentials as shown in Fig. 2c and 2d. The concentration of Fe2+

increased obviously at the initial stage accompanied by the electrode potential decreasing.

Interestingly, the reaction process was almost linear with the reaction going on when the

initial glucose was lower than 0.1 mol·L-1. But when the initial glucose was above 0.2

mol·L-1, non-linear reaction occurred after reacting about 2 h. It should be pointed out that
12
the coking deposition was observed if the initial concentration of glucose was above 0.5

mol·L-1. Therefore, the preference temperature is from 100～110 ℃, and the preference

initial concentration of glucose is 0.2～0.5 mol·L-1 .

Fig.2 a) Fe2+ concentration and b) electrode potential in anolyte at different reaction temperature and time (Initial glucose

0.2 mol·L-1 and initial FeCl3 2.0 mol·L-1). c) Fe2+ concentration and d) electrode potential in anolyte with different initial

content of glucose when the temperature was 110℃ and initial FeCl3 was 2.0 mol·L-1.

3.4 Kinetics analysis of the redox reactions between glucose and FeCl3

Because Fe2+ ions are the charge carriers for transferring electrons to the anode of the

redox flow fuel cell, and we have found that the generation rate of Fe2+ significantly depend

on the kinetic parameters, including reaction temperature, time and content of ingredients.

Thus, detailed kinetic investigations with glucose as selected model compound were

performed. Here, the apparent kinetic equations were derived for investigating the reaction
13
rate affected by different parameters. The power law rate equation of the redox reaction

between glucose and FeCl3 was written as , which can be rewritten as:

（5）

According to Arrhenius equation:

（6）

（7）

Where k0 is the rate constant, E is the activation energy and R is the Boltzmann constant.

Therefore, equation (5) can be written as:

（8）

The reaction rate between glucose and FeCl3 can also be calculated by the expression:

（9）

When the initial concentration of glucose was 0.2 mol·L-1 and FeCl3 was 2.0 mol·L-1, the

experiment data in Fig. 2a was fitted by polynomial regression equations as written in

equation (10):

(10)

Where t is the reaction time. The regression results of the reactions under different

temperatures were shown in Table 1. Since the R2 of the polynomial models are all larger

than 0.97 in the apparent kinetic equation, it is proved that the equation fits the experimental

data well. Thus, when the reaction time was 1 h, the reaction rate at different

temperature could be obtained as shown in Table 1. Based on the data in Table 1, the linear

model was obtained which was determined by plotting versus of the reactions

between glucose and FeCl3 under different temperatures (Fig. 3). The R2 of the linear model

14
is also larger than 0.97, indicating that the data fits the theoretical equations well. Therefore,

it was obtained that apparent activation energy (E) was 127.85 kJ·mol-1.

Table 1 The polynomial regression results of the reactions under different temperatures.

Temp（℃） A B1 B2 R2
80 0.04001 0.01835 -0.00156 0.01523 0.97933
90 0.03227 0.02618 0.00149 0.02916 0.98706
100 0.03614 0.1262 0.00373 0.13366 0.99896
105 0.03013 0.30017 -0.01029 0.27959 0.99768
110 0.03831 0.35434 -0.01973 0.31488 0.99908
Note: was the reaction rate when the reaction time was 1h. Initial glucose was 0.2 mol·L-1 and initial FeCl3 was 2.0

mol·L-1.

Fig.3 Plot of versus of the reactions between glucose and FeCl 3 under different temperatures.

Based on the same principle, the polynomial regression results of the experiment data in

Fig. 1c and Fig. 2c were shown in Table 2. The R2 of the reactions with different initial

contents of glucose and FeCl3 at 110 ℃ were all larger than 0.99, indicating that the

experimental data fits the polynomial model well. Furthermore, when the reactions kept the

same temperature with same initial content of FeCl3, the plots of versus

fitted the line model well, and the slope 0.8929 in Fig. 4a was the value of the coefficient x

in equation (8). Similarly, a slope of 0.911 in Fig. 4b was obtained for the coefficient y in
15
equations (8). Therefore, the apparent rate equation of the reactions between glucose and

FeCl3 was as equation (11):

(11)

Table 2 The polynomial regression results of the reactions with different initial contents of glucose and FeCl3.

C[C6H12O6]
A B1 B2 R2
(mol L-1)
(mol L-1)
0.067 2 -0.00549 0.09362 0.00197 0.09756 0.99701
0.1 2 -0.00548 0.12898 -0.00005 0.12888 0.99698
0.2 2 -0.00087 0.38393 -0.02433 0.33527 0.99993
0.333 2 -0.0309 0.55274 -0.04985 0.45304 0.99341
0.5 2 -0.02294 0.64838 -0.06637 0.51564 0.99639
0.2 0.2 0.03078 0.04727 -0.00209 0.04309 0.9942
0.2 0.6 0.01868 0.09224 0.00284 0.09792 0.99352
0.2 1.0 0.04767 0.127 0.00055 0.12810 0.99591
0.2 2.0 0.03614 0.35707 -0.02024 0.31659 0.99944
0.2 2.4 0.03834 0.49942 -0.03115 0.43712 0.99969

Note: was the reaction rate when the reaction time was 1 h. Temperature was 110 ℃.

Fig.4 Plot of versus and b) of the reactions with different initial contents of glucose and

FeCl3 at 110 ℃.

3.5 Performance of the redox flow fuel cell fueled by glucose with different conditions

The performance of the redox flow fuel cell fueled by glucose with different redox

couples and reaction conditions were investigated as shown in Fig. 5. Firstly，two kinds of

catholytes were fabricated and compared. If the anolyte was the same used in the redox flow
16
fuel cell, the power density with (VO2)2SO4 solution as catholyte was about 2.5 times higher

than that with POM solution as catholyte (Fig. 5a). That is because the electrode potential of

(VO2)2SO4 solution (~1.18 V, versus the normal hydrogen electrode (NHE)) is higher than

that of POM solution (～1.08 V, versus NHE). Furthermore, the (VO2)2SO4 solution has

lower cost and simple process for fabrication. Thus, (VO2)2SO4 was employed as

cathode-active material to enhance the performance of the redox flow fuel cell. Fig.5b

shows the influence of Fe3+ reduction degree on the power density of the redox flow fuel

cell. It was found that the reduction degree of Fe3+ kept increasing with the reaction going

on. Clearly, the power density of the redox flow fuel cell was raised from 87.4 to 120.8

mW•cm-2 when the reduction degree of Fe3+ increased from 0.26 to 0.63 mole electron per

mole Fe3+. Fig. 5c shows that the initial content of glucose has great impact on the output

voltage and power density. Indeed, the maximum output voltage was improved from 0.52 to

0.64 V with the increase of glucose from 0.1 mol·L-1 to 0.333 mol·L-1 when the reaction

solution was heated 4 h at refluxing temperature. And accordingly the power density

increased from 87.4 to 125.7 mW•cm-2. Compared with the reported microbial fuel cell by

utilizing glucose as substrate, the power density of the redox flow fuel cell was improved

about 755 times [24]. However, if the content of glucose was further increased, the

performance of the redox flow fuel cell was not improved observably. That is because the

reaction solution with high content of glucose can generate coking deposition so that the

redox reaction was affected greatly. Fig. 5d shows the continuous discharging of the redox

flow fuel cell fueled by 0.2 mol·L-1 glucose, which was conducted under a discharge current

of about 250 mA•cm-2. It can be seen clearly that the redox flow fuel cell had a stable power

17
output at about 105 mW•cm-2.

Fig.5 a) Voltage-current density and power-current density plots with different kinds of catholyte. Plots for the glucose-

FeCl3 reaction system with b) different reaction times and c) different initial content of glucose at refluxing temperature. d)

Power density curve in continuous operation with the anolyte and catholyte at 80 ℃ water bath. Note: The cell was run at

ambient temperature. The reduction degree of FeCl3 was noted as r. The initial content of FeCl3 was 2.0 mol·L-1. The initial

content of glucose in Fig. 5a and 5b was 0.1 mol·L-1, and it was 0.2 mol·L-1 in Fig. 5d. The catholyte in Fig.5b-c was

(VO2)2SO4 solution.

3.6 Performance of the redox flow fuel cell fueled by various raw biomasses

Excitedly, the redox flow fuel cell reported in this research can also convert various

biomasses into electricity. Fig. 6 shows the generation rate of Fe2+ in various anolyte and the

performance of the redox flow fuel cell powered by bagasse, orange peel and corn stalks,

respectively. The concentration of Fe2+ reached 1.39 mol·L-1 for bagasse at the reaction time

of 5 h under refluxing with the initial concentration of FeCl3 1.5 mol·L-1, and the reduction

degree of Fe3+ reached 0.93 mole electron per mole Fe3+ (Fig. 6a). Accordingly, the output

18
power density reached 45.1 mW•cm-2 with bagasse as fuel when the fuel cell was run at

ambient temperature. It indicated that the bagasse could be oxidized by FeCl3 rapidly, and

the assembled redox flow fuel cell system had high performance due to the high diffusion

coefficient of the generated Fe2+ in anolyte and high potential difference between the two

redox couples. Besides, when orange peel and corn stalks were utilized as fuels, the

maximum power density were about 36.9 mW•cm-2 and 30.8 mW•cm-2, respectively. It was

about 1028 times higher than the reported microbial fuel cell with orange peel as substrate

[25]. Therefore, this approach can be a general and versatile approach for converting

different kinds of biomasses into electricity.

Fig.6 a) Fe2+ concentration in various anolyte; b) Performance of the redox flow fuel cell powered by different biomasses.

Note: 2.0 g biomasses and 27.0 g FeCl3•6H2O in 50 mL acid solution for heating 5 h.

4. Conclusion

In summary, we demonstrated a novel high-efficiency and low cost approach for directly

converting biomass into electricity at low temperature by assembling an environmentally

benign redox flow fuel cell without any noble metal catalyst , which utilized FeCl3 and

(VO2)2SO4 as the anode-active and cathode-active materials, respectively. The anode

oxidant FeCl3 has the unique properties of nontoxicity, low-cost and abundance. More

19
importantly, Fe3+ ions can be completely regenerated and are chemical stable because of its

inertness for all common chemical contaminants. The apparent reaction rate of biomass by

FeCl3 was greatly influenced by temperature and content of ingredients. At the initial stage

of the redox reaction, the reaction process was almost linear with reaction time, and the

apparent activation energy was about 127.85 kJ·mol-1 based on the linear relation equation.

Simultaneously, the chemical energy in biomass was successfully stored in Fe2+ during the

oxidation process, which could be transferred to the anode plate for outputting electricity

and regenerating Fe3+. When glucose was utilized as fuel, the power density of the redox

flow fuel cell can reach 125.7 mW•cm-2, Which was about 755 times higher than the

reported microbial fuel cell by utilizing glucose as substrate. The high performance of the

redox flow fuel cell was due to the fast generation and diffusion of Fe 2+ in anolyte and the

high potential difference between anolyte and catholyte. Furthermore, this approach was

very general and powerful for direct conversion of various crop straws, bagasse and other

biomasses without pretreatment into electricity efficiently.

Acknowledgements

This work was supported by the Pearl River S&T Nova Program of Guangzhou (No.

201710010144) and the Science and Technology Planning Project of Guangdong Province

(No. 2017B090915004).

Reference

1. Y. Ding, G. Yu, The promise of environmentally benign redox flow batteries by

molecular engineering, Angew. Chem. Int. Ed. 56 (2017) 8614 –8616.

20
2. T. Ibn-Mohammed, S.C.L. Koh, I.M. Reaney, A. Acquaye, G. Schileo, K.B. Mustapha,

R. Greenough, Perovskite solar cells: An integrated hybrid lifecycle assessment and

review in comparison with other photovoltaic technologies, Renew. Sustain. Energy Rev.

80 (2017) 1321-1344.

3. I.E.L. Stephens, J. Rossmeisl, I. Chorkendorff, Toward sustainable fuel cells, Science

354(6318) (2016) 1378-1379.

4. D. Guo, Y. Li, B. Cui, Z. Chen, S. Luo, B. Xiao, H. Zhu, M. Hu, Direct reduction of iron

ore/biomass composite pellets using simulated biomass-derived syngas: Experimental

analysis and kinetic modelling, Chem. Eng. J. 327 (2017) 822-830.

5. X. Zhao, J. Zhu, Efficient conversion of lignin to electricity using a novel direct biomass

fuel cell mediated by polyoxometalates at low temperatures, ChemSusChem 9 (2016)

197-207.

6. H. Li, Z. Xu, P. Yan, Z. Zhang, A catalytic aldol condensation system enables one pot

conversion of biomass saccharides to biofuel intermediates, Green Chem. 19 (2017)

1751-1756.

7. A. Jain, R. Balasubramanian, M. P. Srinivasan, Hydrothermal Conversion of Biomass

Waste to Activated Carbon with High Porosity: A Review, Chem. Eng. J. 283 (2016)

789-805.

8. H. Zhang, N. Li, X. Pan, S. Wu, J. Xie, Direct transformation of cellulose to gluconic

acid in a concentrated iron(III) chloride solution under mild conditions. ACS Sustainable

Chem. Eng. 5 (2017) 4066−4072.

9. S. Hossain, A.M. Abdalla, S.N.B. Jamain, J.H. Zaini, A.K. Azad, A review on proton

21
 conducting electrolytes for clean energy and intermediate temperature-solid oxide fuel

cells, Renew. Sustain. Energy Rev. 79 (2017) 750-764.

10. F. Santoni, D.M.S. Mosqueda, D. Pumiglia, E. Viceconti, B. Conti, C.B. Muñoz, B.

Bosio, S. Ulgiati, S.J. McPhail, In-situ study of the gas-phase composition and

temperature of an intermediate-temperature solid oxide fuel cell anode surface fed by

reformate natural gas, J. Power Sources 370 (2017) 36-44.

11. P. Choudhury, U.S.P. Uday, N. Mahata, O.N. Tiwari, R.N. Ray, T.K. Bandyopadhyay,

B. Bhunia, Performance improvement of microbial fuel cells for waste water

treatment along with value addition: A review on past achievements and recent

perspectives, Renew. Sustain. Energy Rev. 79 (2017) 372-389.

12. I. Lee, M.Y. Park, H.J. Kim, J.H. Lee, J.Y. Park, J. Hong, K. Kim, M. Park, J.Y. Yun,

K.J. Yoon, High-temperature current collection enabled by the in situ phase

transformation of cobalt−nickel foam for solid oxide fuel cells, ACS Appl. Mater.

Interfaces 9 (2017) 39407−39415.

13. Y. Zhu, W. Zhou, R. Ran, Y. Chen, Z. Shao, M. Liu. Promotion of oxygen reduction by

exsolved silver nanoparticles on a perovskite scaffold for low-temperature solid oxide

fuel cells, Nano Lett. 16 (2016) 512-518.

14. H. Park, X. Li, S.Y. Lai, D. Chen, K.S. Blinn, M. Liu, S. Choi, M. Liu, S. Park, L.A.

Bottomley, Electrostatic force microscopic characterization of early stage carbon

deposition on nickel anodes in solid oxide fuel cells, Nano Lett. 15 (2015) 6047-6050.

15. X. Tong, T. Luo, X. Meng, H. Wu, J. Li, X. Liu, X. Ji, J. Wang, C. Chen, Z. Zhan,

Shape-dependent activity of ceria for hydrogen electro-oxidation in reduced-temperature

22
 solid oxide fuel cells, Small 11(41) (2015) 5581-5588.

16. F. Curletti, M. Gandiglio, A. Lanzini, M. Santarelli, F. Maréchal, Large size biogas-fed

solid oxide fuel cell power plants with carbon dioxide management: Technical and

economic optimization, J. Power Sources 294 (2015) 669-690.

17. J. Yu, Y. Zhao, Y. Li, Utilization of corn cob biochar in a direct carbon fuel cell, J.

Power Sources 270 (2014) 312-317.

18. H. Jang, J.D. Ocon , S. Lee, J.K. Lee, J. Lee, Direct power generation from waste coffee

grounds in a biomass fuel cell, J. Power Sources 296 (2015) 433-439.

19. S.A.A. Naqvi, M.T. Mehran, R.H. Song, J.W. Lee, S.B. Lee, S.J. Park, D.R. Shin, T.H.

Lim, Performance evaluation of solid oxide carbon fuel cells operating on steam gasified

carbon fuels, Chem. Eng. J. 300 (2016) 384–393.

20. F. Ahmad, M. N. Atiyeh, B. Pereira, G.N. Stephanopoulos, A review of cellulosic

microbial fuel cells: Performance and challenges, Biomass Bioenerg. 56 (2013)

179-188.

21. W. Wang, S. You, X. Gong, D. Qi, B.K. Chandran, L. Bi, F. Cui, X. Chen, Bioinspired

nanosucker array for enhancing bioelectricity generation in microbial fuel cells, Adv.

Mater. 28 (2016) 270-275.

22. J. Ma, Z. Wang, J. Zhang, T.D. Waite, Z. Wu, Cost-effective chlorella biomass

production from dilute wastewater using a novel photosynthetic microbial fuel cell

(PMFC), Water Res. 108 (2017) 356-364.

23. Y. Lu, N. Zhu, F. Yin, T. Yang, P. Wu, Z. Dang, M. Liu, X. Wei, Biomass-derived

heteroatoms-doped mesoporous carbon for efficient oxygen reduction in microbial fuel

23
 cells, Biosens. Bioelectron. 98 (2017) 350-356.

24. S. You, X. Gong, W. Wang, D. Qi, X. Wang, X. Chen, N. Ren, Enhanced cathodic

oxygen reduction and power production of microbial fuel cell based on noble-metal-free

electrocatalyst derived from metal-organic frameworks, Adv. Energy Mater. 6 (2016)

1501497.

25. W. Miran, M. Nawaz, J. Jang, D.S. Lee, Conversion of orange peel waste biomass to

bioelectricity using a mediator-less microbial fuel cell, Sci. Total Environ. 547 (2016)

197-205.

26. M. Li, S. Zhou, Y. Xu, Z. Liu, F. Ma, L. Zhi, X. Zhou, Simultaneous Cr(VI) reduction

and bioelectricity generation in a dual chamber microbial fuel cell, Chem. Eng. J. 334

(2018) 1621–1629.

27. B. Zhang, W. Fan, T. Yao, S. Liao, A. Li, D. Li, M. Liu, J. Shi, S. Liao, C. Li, Design

and fabrication of a dual-photoelectrode fuel cell towards cost-effective electricity

production from biomass, ChemSusChem 10(1) (2017) 99-105.

28. B. Zhang, J. Shi, C. Ding, R. Chong, B. Zhang, Z. Wang, A. Li, Z. Liang, S. Liao, C. Li,

Conversion of biomass derivatives to electricity in photo fuel cells using undoped and

tungsten-doped bismuth vanadate photoanodes, ChemSusChem 8 (2015) 4049-4055.

29. J. Gong, W. Liu, X. Du, C. Liu, Z. Zhang, F. Sun, L. Yang, D. Xu, H. Guo, Y. Deng,

Direct conversion of wheat straw into electricity with a biomass flow fuel cell mediated

by two redox ion pairs, ChemSusChem 9 (2016) 1-9.

30. J.M. Gohil, D.G. Karamanev, Preparation and characterization of polyvinyl alcohol

polyelectrolyte-based membrane-anode assembly for hybrid Fe3+/H2 redox flow

24
 microbial fuel cell, Chem. Eng. J. 259 (2015) 25-33.

31. W. Liu, W. Mu, M. Liu, X. Zhang, H. Cai, Y. Deng, Solar-induced direct

biomass-to-electricity hybrid fuel cell using polyoxometalates as photocatalyst and

charge carrier, Nature Commun. 5 (2014) 3208.

32. W. Liu, W. Mu, Y. Deng, High-performance liquid-catalyst fuel cell for direct

biomass-into-electricity conversion, Angew. Chem. Int. Ed. 53 (2014) 13558-13562.

33. Y. Ding, B. Du, X. Zhao, J. Zhu, D. Liu, Phosphomolybdic acid and ferric iron as

efficient electron mediators for coupling biomass pretreatment to produce bioethanol

and electricity generation from wheat straw, Bioresour. Technol. 228 (2017) 279-289.

34. S.H. Bergens, C.B. Gorman, G.T.R. Palmore, G.M. Whitesides, A redox fuel cell that

operates with methane as fuel at 120℃, Science 265 (1994) 1418-1420.

25
 Highlights

 A redox flow fuel cell for directly converting biomass into electricity is reported.
 The redox flow fuel cell works at low temperature without any noble metal catalyst.
 The anode oxidant FeCl3 is low-cost, abundance, highly active and non-poisoning.
 The reaction rate in anolyte was significantly affected by the kinetic parameters.
 The power and discharge current density reached 125.7 mW•cm-2 and 314.2 mA•cm-2.

26
 Graphical abstract

Ferric ion pair mediated biomass redox flow fuel cell and related
chemical reaction kinetics study
Xihong Zua, Lele Suna, Jian Gongb, Xiaochun Liua,Yangxi Liua, Xu Dub, Wei Liub, Lifen
Chena, Guobin Yia,*, Weigang Zhanga, Wenjing Lina, Weizhao Lia, Yulin Dengb,*

A novel and low cost redox flow fuel cell that directly convert raw biomass to electricity
at low temperature without any noble metal catalyst is reported. FeCl3 and (VO2)2SO4 were
utilized as the anode-active and cathode-active materials, respectively. The power density of
the biomass redox flow fuel cell could reach 125.7 mW•cm-2.

27