A Highly Selective and Efficient Electrocatalytic...

Running head: ELECTROCATALYTIC CONVERSION OF CARBON DIOXIDE

A Highly Selective and Efficient Electrocatalytic

Conversion of Carbon Dioxide using

Nanostructured Electrodes

Nikhil Murthy

The Catlin Gabel School

8825 SW Barnes Rd, Portland, OR 97225

Correspondence concerning this article should be addressed to Nikhil Murthy

Email: [email protected]

ELECTROCATALYTIC CONVERSION

2

TABLE OF CONTENTS

I. Abstract ................................................................................................................................3

II. Purpose of the Study ............................................................................................................4

III. Introduction ..........................................................................................................................5

IV. Experimental Details ..........................................................................................................14

V. Results and Discussion ......................................................................................................19

VI. Conclusions ........................................................................................................................31

VII. Future Directions ...............................................................................................................32

VIII. Acknowledgements ............................................................................................................32

IX. References ..........................................................................................................................33


3

I. Abstract

An efficient and economically profitable process for the electrochemical conversion of carbon

dioxide to valuable chemicals such as methanol and formic acid was investigated to address the

steady increase of atmospheric CO2. Undoped, Lithium (p-type) and Antimony (n-type) doped

nanoparticles were synthesized over a wide range of particle sizes using a sol-gel precipitation

method, to act as electro-catalysts in the reduction process. Scanning Electron Microscopy was

used to characterize the morphology and nanoparticle size. The CO2 reduction was carried out by

cyclic voltammetry using a home-made three-electrode system based on Indium Tin Oxide (ITO)

working electrode coated with SnO2 nanoparticles, Pt counter electrode and Ag/AgCl reference

electrode. Nuclear Magnetic Resonance (NMR) was used to determine the concentration of

products produced. NMR studies show a 30-fold increase in the conversion of CO2 when p-type

SnO2 particles were used. For the stand-alone ITO electrodes (control), a Faradaic efficiency of

0.01% was obtained as compared to an efficiency of 47.4% for ITO with p-type SnO2

nanoparticles to form formic acid. For the reduction to methanol, a control Faradaic efficiency of

5.09% was obtained as compared to an efficiency of 84.4% for p-type SnO2 electrodes. The

application of novel p-type SnO2 highlights the originality of this method as a significant

improvement in electrochemical efficiency was achieved. The improved reduction efficiency

allows for the process to be economically profitable as the energy cost is less than the revenue

generated when selling the primary products, formic acid and methanol. These highly

encouraging results open up the viability of electrochemical reduction of CO2 as a way to answer

global challenges such as global warming.


4

II. Purpose of the Study

The primary purpose of this study was to explore ways to improve Faradaic efficiency and

minimize over-potential in the electrochemical reduction of CO2 to methanol and formic acid in

order to make the process economically profitable in terms of the energy cost and profits made

from selling the products. This was accomplished through tin dioxide nanoparticles which act as

a catalyst in this electrochemical reduction process. The scope of the study included synthesis

and characterization of doped and undoped SnO2 nanoparticles, scanning electron microscopy to

determine the particle size, cyclical voltammetry with a 3-electrode system to generate CV plots

for the reduction reactions and nuclear magnetic resonance spectroscopy to determine the

concentration of methanol and formic acid.


5

III. Introduction

The atmospheric carbon dioxide (CO2) has continuously increased due to the extensive use of

fossil fuels (Centi & Perathoner, 2009; Ganesh, 2014; Lim et al., 2014). The level of CO2 was

about 278ppm during the Industrial Revolution, and has since reached 400ppm mark (Lim et al.,

2014). The present CO2 levels in atmosphere are higher than at any time during the last 650,000

years for which reliable data exists from ice cores (Ganesh, 2014).

Fig. 1 Atmospheric CO2 levels over the last 10,000 years (Ganesh 2014).

Multiple climate modeling studies show a strong correlation of the increasing greenhouse gases

to adverse impact on climate (Centi & Perathoner, 2009; Chen & Kanan, 2012; Ganesh, 2014;

Lim et al., 2014; Lv, Zhang, Gao, & Lei, 2014; Zhang, Kang, & Meyer, 2014).


6

Many methods have been proposed to countering the rising carbon dioxide levels. The limited

capacity of enhanced photosynthesis (a natural photochemical approach to reduce CO2) taking

place during the spring-summer period, in northern hemisphere, can reduce atmospheric CO2 but

by only less than 1%. Simply planting more trees is not sufficient to overcome the current

atmospheric CO2 burden. There is not enough land-area to house additional plants on the scale

that is needed for remediation. Even with the proposed crossover to electric cars, the CO2

emissions are unlikely to budge, since electric car battery charging needs increase the use of

electric power, which is mainly generated from fossil fuels. Therefore, a solution that effectively

can reduce carbon dioxide from its primary sources is desirable. Below is a chart highlighting the

carbon dioxide emissions by sector:

Fig. 2. Global carbon dioxide emission by sector (US EPA, n.d.)

As shown above in Fig.2, the majority of the carbon dioxide emissions stems from fossil fuel

processing, transportation fuels, industrial processes, and power stations. This project focuses on

Transportation Fuels

Fossil fuel processing

Agricultural Byproducts

Residential and commercial

Land use and biomass burning

Waste treatment

Power stations

Industrial processes


7

removal of carbon dioxide from its major (and concentrated) source of emission, the flue gas

released from fossil fuel burning power plants, rather than sequestering CO2 from the

atmosphere, far from emission sources. In the chart above, the sectors that can be targeted from

such a solution (in which concentrated carbon dioxide emissions are removed) make up over two

thirds (66.8%) of all emissions (in orange).

Although a large percentage of carbon dioxide emissions can be targeted with such a system, a

major hurdle that prevents the implementation of such systems is the economic cost associated

with the process itself. If the reduction process requires the input of money (to pay for the energy

input), companies would not willingly install such systems. In that case, the response to a

potentially novel solution would be low, especially considering the sectors that produce the

majority of carbon dioxide emissions (industry). Therefore, a system that is economically

profitable, one that produces products that can be sold for more than the input energy cost, is

much more viable. If such a method for reducing carbon dioxide can be developed, producers of

concentrated carbon dioxide such as manufacturing plants or factories would adopt reducing

their carbon footprint, ultimately addressing the rising levels of carbon dioxide in our

atmosphere.

Reduction of CO2 is thus of considerable interest, and especially so, if it can be turned into

useable fuels. The opportunities and prospects in the chemical recycling of CO2 to fuels have

been explored extensively in the recent past (Centi & Perathoner, 2009; Lim et al., 2014). The

role of nanostructured catalysts was recently studied by Zhang et al (Zhang et al., 2014) and

found that particle size affects the efficiency of electrocatalytic reduction of CO2. In addition,

other recent studies have shown an increased Faradaic efficiency with nanostructured electrodes

(Lee et al., 2013; Wu, Harris, Sharma, & Zhou, 2013). Tin oxide dependence on the CO2


8

reduction efficiency was also recently investigated by Chen and Kanan (Chen & Kanan, 2012),

and observed a four-fold improvement in Faradaic efficiency over stand-alone tin foil electrodes.

The reduction of carbon dioxide using bulk tin dioxide was also investigated recently

(Bumroongsakulsawat & Kelsall, 2014; Floriano, Scalvi, Sambrano, & Geraldo, 2010). While

improved Faradaic efficiency with SnO2 is reported in a few recent studies (Kwon & Lee, 2001;

Lv et al., 2014; Prakash, Viva, & Olah, 2013; Russell, Kovac, Srinivasan, & Steinberg, 1977;

Wu, Risalvato, Ke, Pellechia, & Zhou, 2012), the role of grain size, microstructural homogeneity

and doping effect is not as understood.

In the current project, a comprehensive study was carried out to synthesize undoped and doped

(Antimony and Lithium) SnO2 nanoparticles, over the desirable size range of 10-30nm to

maximize the surface adsorption and catalysis. Cyclical voltammetry with a three electrode

system was extensively used to study and characterize CO2 electrochemical reduction for a

number of working electrodes, including Indium Tin Oxide (ITO) and ITO with synthesized

SnO2 nanoparticles. Nuclear Magnetic Resonance (NMR) spectroscopy was employed to study

and characterize concentration of formate produced during electrochemical reduction of CO2

with varying working electrodes. The overall goal of the project was thus to determine and

optimize working electrode characteristics to achieve high Faradic efficiency and minimal

overpotential.

The overall electrochemical CO2 reduction reaction and equilibrium potentials (at pH=7) may be

described as follows (Wu et al., 2012).

CO2 + 2H+ + 2e− ↔ HCOOH Eo = −0.61 V vs. SHE

CO2 + 2H+ + 2e− ↔ CO + H2O Eo = −0.52 V vs. SHE


9

The reaction pathways involve initial adsorption followed by a one electron reduction to form the

intermediate CO2 *- species (Zhang et al., 2014). The weakly adsorbed CO2

*- species goes

through a protonation reaction and a second electron transfer to form formate.

CO2(solution) → CO2(ads)

CO2(ads) + e− → CO2∗−(ads)

CO2∗−(ads) + HCO3

− + e− → HCO2−(ads) + CO3

2−(ads)

HCO2−(ads) → HCO2

−(solution)

CO32− + CO2 + H2O → 2HCO3

−

For the reduction of carbon dioxide to methanol, the CH2(OH)2 species is adsorbed onto the

cathode, which is then converted to formaldehyde. This is then reduced to methanol to a double

electron transfer.

CO2(solution) → CO2(ads)

CO2(ads) + e− → CO2∗−(ads)


− + e− → HCO2−(ads) + CO3

2−(ads)

HCO2−(ads) → HCO2

−(solution)

HCO2− + 2H+ + OH− → HCOOH2−

HCOOH2− + 2H+ → CH2(OH)2(ads)

CH2(OH)2(ads) → HCHO + 𝐻2O


10

HCHO + 2e− + 2H+ → CH3OH

In this study, the electrocatalytic reactions were carried out by cyclical voltammetry based on a

homebuilt 3-electrode system. A potentiostat (model 273, EG & G Princeton Applied Research)

was used in conjunction with the electrode system. Voltammograms were obtained for varying

scan rates from +2V to -2.5V at scan rates, ranging from 5mV/s to 50mV/s. For the working

electrode, Indium Tin Oxide (ITO) coated with doped or undoped SnO2 nanopowders were used.

Sol-gel precipitation method was used for the synthesis of the nanoparticles. P-doping of SnO2

was achieved with Lithium and N-doping using Antimony as the dopant species. Various process

parameters for the hydrolysis of SnO2 powders were studied to understand the reaction chemistry

and the mechanisms behind the formation of nanoparticles. Nuclear Magnetic Resonance (NMR)

spectroscopy was utilized to study and characterize the concentration of formate produced during

electrochemical reduction of CO2 with varying working electrodes.

The Nernst equation derived from Gibbs free energy is given by,

QlnFn

RT E E cellcell

where Ecell = nonstandard cell potential in volts

E˚cell = standard cell potential in volts

R = constant, 8.314 J/mol·K

T = temperature in Kelvin

n = moles of electrons transferred

F = Faraday’s constant = 96485 C/mol eˉ and

Q = the reaction quotient expression

Alternative representation of the Nernst Equation at standard temperature (298K) is


11

Qlogn

V 0.0592 E Eor Qln

n

V 0.0257 E E cellcellcellcell

In a reversible redox reaction, the peak potential, 𝐸𝑝= |𝐸𝑝, 𝑎𝑛𝑜𝑑𝑖𝑐 − 𝐸𝑝, 𝑐𝑎𝑡ℎ𝑜𝑑𝑖𝑐| = 59

𝑛 𝑚𝑉 for all

scan rates at 298K. Thus, the peak separation can be used to determine the number of electrons

transferred, and as a criterion for a Nernstian behavior. Accordingly, a fast one-electron process

exhibits a Ep about 59 mV. Both the cathodic and anodic peak potentials are independent of the

scan rate.

The peak current 𝑖𝑝 in a reversible system at 298 K is given by the Randles–Sevcik equation

(Ref.9) , 𝑖𝑝 = (2.69×105) 𝑛3

2 𝐴 𝐷1

2 𝜈1

2 𝐶 where n is the number of electron equivalent

exchanged during CO2 reduction process, 𝐴 (cm2) the active area of the working electrode, 𝐷

(cm2 s–1) the diffusion coefficient, C (mol cm–3) the bulk concentration of the electroactive

species and v is the voltage scan rate (V s–1) . Accordingly, the current is directly proportional to

concentration and increases with the square root of the scan rate.

The role of the nanomaterials in increasing the efficiency of the reduction process is one of a

catalyst: Tin dioxide decreases the activation energy associated with charge transfer from the

nanomaterial cathode to adsorbed species. The rate determining steps for both the formation of

formic acid and methanol both consist of a charge transfer. Therefore, the addition of tin dioxide

nanoparticles directly impacts this step by decreasing Ea, the activation energy. The rate

determining steps for both formic acid and methanol production is reproduced below:

Formic acid rate determining step:


− + e− → HCO2−(ads) + CO3

2−(ads)


12

Methanol rate determining step:

HCHO + 2e− + 2H+ → CH3OH

According to the Arrhenious equation, as the activation energy decreases, the rate constant k

increases:

𝑘 = 𝐴𝑒−𝐸𝑎𝑅𝑇

where k is the rate constant

A is the Arrhenius constant

Ea is the activation energy

R is the universal constant and

T is the temperature in Kelvin.

The effect of the nanomaterial catalyst is illustrated in Fig.3, an energy diagram that shows the

reaction pathway with and without a catalyst. While the change in Gibb’s free energy remains

the same so the reaction maintains the same level of spontaneity, the activation energy decreases

allowing smaller initial input of energy to complete the reaction. This therefore increases the rate

of the reaction. This is also exemplified by the graph in Fig.3, which illustrates the Maxwell-

Boltzmann distribution of kinetic energy and that a greater percentage of molecules will react if

the activation energy decreases.


13

Fig. 3 A schematic representation of activation energy with and without a catalyst. Graph on the right illustrates the

Maxwell-Boltzmann distribution of kinetic energy and that a greater percentage of molecules will react if the

activation energy decreases.

Ultimately, the usage of tin dioxide nanoparticles allow the rate determining step, a charge

transfer, to occur with a lower activation energy, increasing the rate of reaction and the

efficiency of the process.


14

IV. Experimental Details

For the synthesis of SnO2 nanoparticles, 10 g of SnCl4.5H2O in a 1:1 ratio with H2O was used. A

few drops of concentrated HCl were added next in order to decrease the pH to 0.4. This solution

was magnetically stirred until it reached homogeneity. The pH of this solution was then raised to

8 using around 14 ml of NH4OH, and was added in small increments. The precipitate

comprising Tin Dioxide was washed and filtered in H2O, and then dried at 25°C for 24 hours.

Following the drying, the precipitate underwent calcinations at 700°C for 3 hours in a furnace to

obtain the final Tin Dioxide nanopowders (Chaparadza, Rananavare, & Shutthanandan, 2007;

Tran & Rananavare, 2011; Verma, Kumar, & Verma, 2012).

In order to obtain p and n doped SnO2 nanoparticles, LiCl was used at 5-20% by wt., and SbCl3

at 1-5% by wt. were added to the SnCl4 solution. For both lithium doped and antimony doped

SnO2 nanopowders, the precipitate comprising of SnO2 underwent calcinations of 150° for 2

hours and 450° for 2 hours to obtain the nanopowders.

Fig. 4 Filtering and washing of n-type SnO2 in H2O

Photos taken by: Nikhil Murthy

Fig. 5 Collection of various nano-particle powders, including

Li and Sb doped SnO2


15

Fig. 6 A p-type lithium doped tin dioxide

nanoparticle pellet created using 500kPa

compression

Fig. 7 Calcination furnace and setup for resistance measurement

of pellets under 4000 ppm H2 in N2

All photos taken by: Nikhil Murthy

In order to form the nanoparticle pellets, 2g of the nanoparticles were ground up into finer

particles, and then mixed with 400 µL of tetraethyl orthosilicate in ethanol (2%). This was

mixed and the resulting paste was added to compressor cylinder. 100 µL of tetraethyl

orthosilicate in ethanol (2%) was added, and then compressed at 5,500 kPa of pressure, creating

the pellet. Fig. 6 shows an image of a representative P-type SnO2 pellet.

X-ray diffraction of the nanopowders was carried out on a Philips XRG-300 diffractometer with

Cu Kα (λ = 1.5418Å) at 40kV and 2mA. The diffraction patterns were collected on a computer

interfaced to the analog output over 2θ from 10o to 90o. The diffraction pattern was analyzed for

the presence of any amorphous phase. Indexing of the diffraction peaks was carried out and

compared to against expected planes for the known SnO2 Cassiterite (tetragonal) phase.


16

Electrochemical reduction of CO2 using 3-electrode setup and cyclic voltammetry

Fig. 8 3-electrode setup for cyclic voltammetry. A potentiostat (model 273, EG & G Princeton Applied Research)

was used to carry out voltammetry.

CO2 Reduction carried out with a home-made 3-electrode system based on Working Electrode of

ITO glass coated with 0.5g SnO2 nanoparticles and Nafion (0.5 ml of 0.05g/ml NR50) solution,

Counter Electrode comprising Pt and Reference Electrode based on Ag/AgCl. An electrolyte

solution of 1M KCl +1M KHCO3 saturated with dry ice (CO2 source) was used for all

experimental runs.


17

Cyclic voltammetry was carried out using Potentiostat (Model 273, EG & G Princeton Applied

Research) at scan rates from 5mV/s to 50mV/s. Product solution post electrochemical reaction

was further analyzed by NMR.

Fig.9 Cyclic voltammetry setup: The working electrode consisted of ITO glass coated with nanoparticles and

Nafion.

The Faradic efficiency is defined as: 𝜂𝐹𝑜𝑟𝑚𝑎𝑡𝑒 = 2 𝑛𝐹𝑜𝑟𝑚𝑎𝑡𝑒 𝐹

𝑄

where F = Faraday’s constant = 96485 C/mol eˉ and

Q = the reaction quotient expression

Counter Electrode: PtWorking Electrode: ITO Glass coated with doped

SnO2 + Nafion NR50 Reference ElectrodeAg/AgCl

Electrolyte: 1M KCl + 1M KHCO3

Dry Ice (CO2 saturated)


18

𝑄 = 𝑗 ×𝐴 ×𝑡,

where Q = charged passed

j = current density from cyclic voltammetry

A = the area of working electrode = 7.0 cm2

t = reaction time

n = number of electrons required to reduce species, in this case 2 for the

electrochemical reduction of CO2 to formate (Lv et al., 2014)

The current density and reaction time are deduced from the voltammograms.

Faradaic efficiency was used to compare different samples from one another. While the NMR

product concentration provides some insight into the relative effectiveness of each working

electrode, different energy inputs may have been used to achieve such a result. Therefore, the

Faradaic efficiency combines the concentration of product, the number of electrons needed to

reduce CO2 and the total charge used to do so (input energy). This therefore provides a

quantitative measure that allows one to not only compare samples from one another but compare

samples with the industry.

NMR samples were prepared by dissolving product solution from electrolysis with 10 μL of

D2O. NMR spectroscopy was carried out using Bruker ER300 EPR. Formate concentration, n

(for 10 μL solution) was determined from area under curve for HCOO- at 8.22ppm normalized to

CH3OH.


19

V. Results and Discussion

Representative X-ray diffraction pattern obtained from undoped SnO2 powder is shown in Figure

10. From the diffraction pattern, it is evident that excellent crystallinity is observed in the

synthesized nanopowders with well defined peaks. The peaks correspond to Cassiterite structure

of SnO2. No additional phases are evident from the diffraction pattern. The non-existence of any

diffuse/broad peak is also indicative of absence of amorphous phase.

Fig. 10 Powder X-ray diffraction of undoped SnO2 nanopowders. The peaks are sharp and very well defined.

Indexed peaks correspond to the Cassiterite structure. Crystal structure and (110) surface structure of SnO2 are

shown on the right. Equilibrium structure is Rutile type tetragonal (also called Cassiterite).

Scanning Electron Microscopy was carried out on all three classes of samples. SEM images in

the secondary electron mode were collected at 5keV. Figure 11showa a secondary electron

images collected for p-doped SnO2 powder. The image shows a uniform distribution of SnO2

nanoparticles. The average grain size was measured to be in the range of 10-30nm. Similar

Angle (2 )

10 20 30 40 50 60 70 80 90

Arb

. In

ten

sity

-0.020

0.000

0.020

0.040

0.060

0.080

0.100


20

microstructure was also obtained for undoped SnO2 nanoparticles, as shown in Fig. 12. For

optimal sensor performance, a grain size of ~10nm is desirable to maximize the surface

adsorption and catalysis during the electrochemical reduction reaction.

Fig. 11 Scanning Electron Microscopy (SEM) image of p-doped SnO2 nanopowders synthesized by sol-gel

precipitation. The average grain size is in the range 10-30nm.


21

Fig. 12 Scanning Electron Microscopy (SEM) image of undoped (pure SnO2) and p-type (Li doped

SnO2) synthesized by sol-gel precipitation. Grain size ranges from 10 to 20 nm, which is optimal for

catalysis.

Cyclic Voltammetry was carried out in order to both reduce carbon dioxide and characterize

the electrochemical reaction. CV measurements allowed for the charge transfer, reaction time,

and average current density to be accurately measured. Charge transfer is greatly enhanced

with the addition of nanoparticles and is also a key component in calculating Faradic

Efficiency. The plots below show many different trials of the three electrode system in order to

test for these components:


22

1- Scan Rate – 20 mV/sec

2- Scan Rate -30 mV/sec

3- Scan rate – 50 mV/sec

Fig. 13 Voltammogram 1M NaHCO3; WE-Pt, CE- Pt, RE-Calomel) over a Range of 1 to -1 V

-1.1 -0.6 -0.1 0.4 0.9-0.0150

-0.0125

-0.0100

-0.0075

-0.0050

-0.0025

0

0.0025

E (Volts)

I (A

mp

s/c

m2)

SBC, 1M, 1-1.corSBC, 1M, 1-2.corSBC, 1M, 1-3.cor


23




Fig. 14 Voltammogram 0.5M NaHCO3; WE-Pt, CE- Pt, RE-Calomel) over a Range of 1 to -1 V

The plot above shows a standard cell in order to test the three electrode system. With a working

and counter electrode of platinum and a reference electrode of silver-silver chloride, the CV plots

were well known and easily comparable. The plot above matches literature and therefore

confirms that the homebuilt system works reliably.

-1.0 -0.5 0 0.5 1.0-0.0100

-0.0075

-0.0050

-0.0025

0

0.0025

E (Volts)

I (A

mp

s/c

m2)

SBC, 0.5M, 1-1.corSBC, 0.5M, 1-2.corSBC, 0.5M, 1-3.cor


24

Fig. 15 Voltammogram 0.5M KHCO3 + 1M KCl + Dry ice ( WE-SnO2 nanoparticles (undoped) with Nafion binder

and on ITO glass, CE-Pt, Ref-Calomel),Range –(-0.5 to -2.0V)

The above voltammogram consists of the nanoparticles (SnO2, undoped) with the Nafion binder

and on ITO glass. According to the reaction below, the Randles-Sevcik equation, the scan rate

influences the shape of the plot:

𝑖𝑝 = (2.69×105) 𝑛3

2 𝐴 𝐷1

2 𝜈1

2 𝐶

where n = the number of electron equivalent exchanged during CO2

reduction process

A = the active area of the working electrode

-2.2 -1.7 -1.2 -0.7-0.10

-0.05

0

E (Volts)

I (A

mp

s/c

m2)

PBC red,0.5,1, Sn-1.corPBC red,0.5,1, Sn-2.corPBC red,0.5,1, Sn-3.corPBC red,0.5,1, Sn-4.cor






25

D = Diffusion Coefficient

C = The bulk concentration of the electroactive species

𝜈 = The voltage scan rate

By the equation above, the peak current density is related to the square root of the scan rate when

all other variables are held constant. Therefore, as the scan rate increases, the peak current

increases and the graph shifts downwards, as shown in the graph above.

Fig. 16 Voltammogram 0.5M KHCO3 + 1M KCl + Dry ice ( WE-Db Pt, CE-Pt, Ref-Calomel),Range –(-0.5 -2.0V)

-2.2 -1.7 -1.2 -0.7-0.0100

-0.0075

-0.0050

-0.0025

0

0.0025

E (Volts)

I (A

mp

s/c

m2)

PBC red,0.5,1-1.corPBC red,0.5,1-2.corPBC red,0.5,1-3.cor





26

Fig. 17 Voltammogram for varying scan rates from +2V to -2.5V at scan rates, ranging from 5mV/s to 50mV/s. (a)

control sample comprising stand-alone ITO electrode (b) the ITO working electrode coated with SnO2 (n-type)

nanoparticles and NR50 Nafion

Fig. 18 1H NMR spectra at 400MHz at E = -2V vs. SCE , 1M KCl +1M KHCO3 with CO2 bubbling for ITO coated

with nanoparticles and Nafion.

Formate: 1.685 molChemical shift: 8.22 ppm

MethanolChemical shift: 3.3 ppm

D2O solventTemp: 298.2KAcquisition time: 3.4s

NMR Spectra for 1HCO2 reduction on ITO with SnO2 Nanoparticles and Nafion


27

Fig. 19 1H NMR spectra at 400MHz at E = -2V vs. SCE , 1M KCl +1M KHCO3 with CO2 bubbling for the control

ITO sample.

Nuclear Magnetic Resonance (NMR) spectroscopy was carried out using Bruker ER300 EPR

with D2O as the solvent. Figs. 18 and 19 illustrate the 1H NMR spectra obtained for the formate

solution dissolved in D2O. The integrated concentration of formate at a chemical shift 8.22ppm

was observed to be 0.0585 mol (normalized to 1 mol of CH3OH which had a chemical shift of

3.3ppm) for the control ITO sample, while the formate concentration was 1.685 mol for ITO

coated with nanoparticles and Nafion.

Chemical shift:3.3 ppm

Formate: 0.0585 molChemical shift: 8.22 ppm

NMR Spectra for 1HCO2 Reduction on ITO (Control)

D2O solventTemp: 298.2KAcquisition time: 3.4s


28

Working

Electrode

NMR Spectroscopy Cyclic Voltammetry

Faradic

Efficiency Concentration of

Formate (mol/μL)

Average

current

density

(A/cm2)

Electrode

Area

(cm2)

Reaction

time (s)

Charge

transferred

(C)

Control: ITO

glass

(formate)

0.00555 0.000506 7.10 300 1.08 * 108 0.0994%

Control: ITO

glass

(methanol)

0.0949 0.000506 7.10 300 1.08 * 108 5.09%

ITO with

doped SnO2

Nanoparticles

and Nafion

(formate)

1.705 0.000421 3.67 450 6.94 * 108 47.4%

ITO with

doped SnO2

Nanoparticles

and Nafion

(methanol)

1.012 0.000421 3.67 450 6.94 * 108 84.4%

Table I. Summary of Results from Cyclic Voltammetry and NMR Spectroscopy

The results tabulated in Table I compare the concentration of formate produced, the average

current density during the reduction reaction, respective electrode area, the charge transferred

and the Faradaic efficiency for the two types of electrodes.

The Faradic efficiency is calculated using, 𝜂𝐹𝑜𝑟𝑚𝑎𝑡𝑒 = 2 𝑛𝐹𝑜𝑟𝑚𝑎𝑡𝑒 𝐹

𝑄 , where F = 96,485C/mol

and Q, charge passed.

𝑄 = 𝑗 ×𝐴 ×𝑡, where j is current density from cyclic voltammetry, A, the area of working

electrode = 7cm2 and t, reaction time (Lv et al., 2014). In the current electrochemical reduction

of CO2 to formate, n = 2, as 2 electrons are involved. The current density and reaction time are

deduced from the voltammograms. The average current density for the control electrode, which

had an area of 7cm2, was 0.0005A/cm2. For a reaction time of 300s, the total charge transferred


29

was 1.05C. For the ITO electrode coated with SnO2 nanoparticles, which had a larger area of 11

cm2, the total transferred charge over 1500s was 4.13C.

The respective Faradaic efficiencies for the stand-alone ITO and ITO coated with nanoparticles

and Nafion were 11% and 79%. The reaction mechanism resulting in such higher efficiencies for

the electrode with nanoparticles is discussed next.

Fig. 20 Representation of a possible pathway for the generation of formate from CO2 during electrochemical

reduction.

Fig. 21 Representation of the distribution of SnO2 nanoparticles and Nafion ionomer on the cathodic surface of ITO

working electrode.


30

Fig 20 shows a representation of formate formation. Highly covered H+ accepts an electron to

generate Had (Tran & Rananavare, 2011). The adsorbed Had species react with CO2 diffused from

the electrolyte solution to the electrode surface to form HCOOad, which subsequently forms

HCOO- on acceptance of an electron. The reduction reaction can proceed with reduced

overpotential when aided by a catalyst. Nafion isomer and SnO2 nanoparticles are both known to

electrocatalysts (Wu, Sharma, Harris, & Zhou, 2014). Fig. 21 shows an illustration of SnO2 and

Nafion distributed on the cathodic surface of ITO. At the nano-scale of 10-30nm, the surface

adsorption sites of these electro-catalysts are maximized. The improved efficiencies are evident

from the significantly reduced overpotential, as well as the high Faradaic efficiency of 79%

compared to 11% for the stand-alone ITO electrode.


31

VI. Conclusions

In this study, undoped, N-doped and P-doped tin dioxide (SnO2) nanocrystals were successfully

synthesized using sol-gel precipitation method. N- and P-doping in SnO2 were achieved with

Antimony and Lithium respectively. SEM imaging of the nanocrystals shows uniform

microstructure in the desirable size range of 10-30nm for optimal catalysis. X-ray diffraction

shows excellent crystallinity of the synthesized nanoparticles. Electrochemical reduction of CO2

to formic acid and methanol was successfully achieved by cyclic voltammetry using a home-

built 3-electrode system and a potentiostat. NMR studies show > 30x higher concentration of

formate when reaction was catalytically aided by SnO2 nanoparticles and Nafion. For stand-

alone ITO electrodes, a Faradaic efficiency of 5.09% was obtained for the reduction to methanol.

For ITO with SnO2 nanoparticles and Nafion, Faradaic efficiency of 84.4%, a 17X improvement

was achieved. With ITO and SnO2 nanoparticles, a significant reduction in over-potential of

~0.7V was demonstrated. Analysis of economic impact indicates that energy costs would be

significantly reduced for the electrochemical reduction of CO2 at the Faradaic efficiencies

achieved in the current study. These results offer potential viability of electrochemical reduction

of carbon dioxide to formic acid and methanol as a way to answer global challenges such as

depletion of fossil fuels and reducing greenhouse gases, as well as economically produce formic

acid and methanol.


32

VII. Future directions

Having successfully developed synthesis method for Antimony and Lithium-doped tin dioxide

nano-crystalline powders, and electrochemical reduction of CO2 to formate, electrodes based on

these nanocrystals may be used to fabricate a novel urea synthesis diode cell to sequester Carbon

and reduce CO2, a greenhouse gas. The extension of the current project would essentially entail

concomitant photo-electrochemical (PE) reduction of CO2 (at the cathode) and oxidation of NH3

(at the anode) to recycle CO2. The steps would involve:

Formate Synthesis: Optimize p-type nano-particulate (Li doped SnO2) electrode for

electrochemical CO2 reduction

Ammonia Oxidation: Optimize n-type nano-particulate (Sb doped SnO2) electrode for

photoelectrical NH3 oxidation

Photo-electrochemistry of Carbamates: Optimize electrolyte composition and photo-

electrochemical cell operational conditions.

VIII. Acknowledgements

The author wishes to acknowledge Prof. S. Rananavare for valuable guidance and the use of

nanofabrication facilities.


33

IX. References

Bumroongsakulsawat, P., & Kelsall, G. H. (2014). Effect of solution pH on CO: formate

formation rates during electrochemical reduction of aqueous CO2 at Sn cathodes.

Electrochimica Acta, 141, 216–225. https://doi.org/10.1016/j.electacta.2014.07.057

Centi, G., & Perathoner, S. (2009). Opportunities and prospects in the chemical recycling of

carbon dioxide to fuels. Catalysis Today, 148(3–4), 191–205.

https://doi.org/10.1016/j.cattod.2009.07.075

Chaparadza, A., Rananavare, S. B., & Shutthanandan, V. (2007). Synthesis and characterization

of lithium-doped tin dioxide nanocrystalline powders. Materials Chemistry and Physics,

102(2–3), 176–180. https://doi.org/10.1016/j.matchemphys.2006.11.022

Chen, Y., & Kanan, M. W. (2012). Tin Oxide Dependence of the CO2 Reduction Efficiency on

Tin Electrodes and Enhanced Activity for Tin/Tin Oxide Thin-Film Catalysts. Journal of

the American Chemical Society, 134(4), 1986–1989. https://doi.org/10.1021/ja2108799

Floriano, E. A., Scalvi, L. V. de A., Sambrano, J. R., & Geraldo, V. (2010). Evaluation of bulk

and surfaces absorption edge energy of sol-gel-dip-coating SnO2 thin films.

ResearchGate, 13(4), 437–443. https://doi.org/10.1590/S1516-14392010000400004

Ganesh, I. (2014). Conversion of carbon dioxide into methanol – a potential liquid fuel:

Fundamental challenges and opportunities (a review). Renewable and Sustainable Energy

Reviews, 31, 221–257. https://doi.org/10.1016/j.rser.2013.11.045

Kwon, Y., & Lee, J. (2001). Formic Acid from Carbon Dioxide on Nanolayered Electrocatalyst.

Electrocatalysis, 1(2–3), 108–115. https://doi.org/10.1007/s12678-010-0017-y


34

Lee, S., Ju, H., Jeon, H., Machunda, R. L., Kim, D., Lee, J. K., & Lee, J. (2013). Electrocatalytic

Reduction of Gas-Phased CO2 on Nano-Sized Sn Electrode Surface. ECS Transactions,

53(29), 41–47. https://doi.org/10.1149/05329.0041ecst

Lim, R. J., Xie, M., Sk, M. A., Lee, J.-M., Fisher, A., Wang, X., & Lim, K. H. (2014). A review

on the electrochemical reduction of CO2 in fuel cells, metal electrodes and molecular

catalysts. Catalysis Today, 233, 169–180. https://doi.org/10.1016/j.cattod.2013.11.037

Lv, W., Zhang, R., Gao, P., & Lei, L. (2014). Studies on the faradaic efficiency for

electrochemical reduction of carbon dioxide to formate on tin electrode. Journal of

Power Sources, 253, 276–281. https://doi.org/10.1016/j.jpowsour.2013.12.063

Prakash, G. K. S., Viva, F. A., & Olah, G. A. (2013). Electrochemical reduction of CO2 over Sn-

Nafion® coated electrode for a fuel-cell-like device. Journal of Power Sources, 223, 68–

73. https://doi.org/10.1016/j.jpowsour.2012.09.036

Russell, P. G., Kovac, N., Srinivasan, S., & Steinberg, M. (1977). The Electrochemical

Reduction of Carbon Dioxide, Formic Acid, and Formaldehyde. Journal of The

Electrochemical Society, 124(9), 1329–1338. https://doi.org/10.1149/1.2133624

Tran, H. A., & Rananavare, S. B. (2011). Synthesis and characterization of N- and P- doped tin

oxide nanowires (pp. 144–149). Presented at the IEEE Conference on Nanotechnology.

Retrieved from https://www.infona.pl//resource/bwmeta1.element.ieee-art-000006144603

US EPA, O. (n.d.). Global Greenhouse Gas Emissions Data [Overviews and Factsheets].

Retrieved January 2, 2017, from https://www.epa.gov/ghgemissions/global-greenhouse-

gas-emissions-data


35

Verma, I., Kumar, R., & Verma, N. (2012). Synthesis and characterization of Sb doping on thick

film SnO2 gas sensor by sol-gel method. International Journal of Scientific Research

Engineering & Technology (IJSRET), 1(2), 49–51.

Wu, J., Harris, B., Sharma, P. P., & Zhou, X.-D. (2013). Morphological Stability of Sn Electrode

for Electrochemical Conversion of CO2. ECS Transactions, 58(2), 71–80.

https://doi.org/10.1149/05802.0071ecst

Wu, J., Risalvato, F. G., Ke, F.-S., Pellechia, P. J., & Zhou, X.-D. (2012). Electrochemical

Reduction of Carbon Dioxide I. Effects of the Electrolyte on the Selectivity and Activity

with Sn Electrode. Journal of The Electrochemical Society, 159(7), F353–F359.

https://doi.org/10.1149/2.049207jes

Wu, J., Sharma, P. P., Harris, B. H., & Zhou, X.-D. (2014). Electrochemical reduction of carbon

dioxide: IV dependence of the Faradaic efficiency and current density on the

microstructure and thickness of tin electrode. Journal of Power Sources, 258, 189–194.

https://doi.org/10.1016/j.jpowsour.2014.02.014

Zhang, S., Kang, P., & Meyer, T. J. (2014). Nanostructured tin catalysts for selective

electrochemical reduction of carbon dioxide to formate. Journal of the American

Chemical Society, 136(5), 1734–1737. https://doi.org/10.1021/ja4113885

A Highly Selective and Efficient Electrocatalytic...

Documents

Transcript of A Highly Selective and Efficient Electrocatalytic...