Oxygen Reduction Catalysts in Alkaline Electrolyte

(1)

KA103X - Degree project in Engineering Chemistry, undergraduate level

Department for Applied Physical Chemistry, KTH and Department of Process Technology, KTH

2020-05-22

Authors:

Anders Abrahamsson Avital Cherednik Bjarne Falk

Supervisors:

Inna Soraka, Department of Applied Physical Chemistry Yi Yang, Department of Applied Physical Chemistry Yohannes Kiros, Department of Process Technology

(2)

Abstract

Alkaline fuel cells are a promising technology, with their sturdy design and many applications they are held back mostly by their cost. By introducing a catalyst, the activation energy of the cell can be reduced to an overcomable amount. Unfortunately, due to the high cost and sparse availability of the most used catalyst metal today, platinum, it has become apparent that a new suitable catalyst must be found in order to make the fuel cells economically feasible. Silver and palladium have been proposed as promising alternatives, sharing a majority of the traits but with a fraction of the cost. The original aim of this project was to study the performance of electrodes in an alkaline electrolyte loaded with different ratios of palladium and silver. However, due to the COVID-19 situation the project was not able to be completed and the aim of the project changed. The new aim was divided into two parts. The first one being to study how the initial concentration of silver ions affects the size of the obtained particles. This was achieved by a radiolysis-based method of synthesis in an aqueous solution. The second aim was to study the performance of the electrodes loaded with different amounts of silver and different average particle size. However, this part was not possible to conduct either. Therefore, results from a previous study performed by I. L.

Soroka et al. was used for discussion. The results point towards a lower initial concentration achieving a smaller average particle size and a lower loading of catalyst on the electrode can be compensated by a smaller average particle size of the catalyst.

(3)

Introduction

The original aim of this project was to investigate how the palladium to silver ratio affects the performance of an electrode in an effort to lower the cost of a fuel cell. This would have been achieved by synthesizing carbon based electrodes with different ratios of palladium and silver as catalyst , via radiolysis, and to study their performance in an alkaline electrolyte. However, the experiments were not completed due to the COVID-19 situation.

Thus, the aim of this project changed to study how the initial silver ion concentration affects the obtained particle size after gamma irradiation. This was done by preparing four samples of different silver ion concentration and irradiating the samples with gamma radiation while keeping other parameters constant. The support used for the obtained nanoparticles was carbon powder, XC-72. To discuss how the silver particle size affects the performance of an electrode, the results from a previous study performed by I. L. Soroka et al. was borrowed and analyzed. This project therefore serves as a pre-study to following research for further lowering of the cost of fuel cells.

Fuel cells

A fuel cell converts chemical energy to electrical energy. The main construction of a fuel cell consists of two connected electrodes and an electrolyte, see Figure 1. The fuel, or the reactants, react indirectly with one another via the electrodes. The most common fuel used is hydrogen gas which reacts spontaneously with the oxygen in the air as seen in Equation 1.

The reason why hydrogen and oxygen is utilized is due to the high energy release of the reaction and the only product being water. This means, depending on the source of hydrogen gas, the reaction is a green source of energy.

H H O

2 ₂+ O₂ → 2 ₂ + Δ (1)

This reaction is often represented as two half-cell reactions consisting of a reductive and an oxidative reaction:

(5)

Oxidative: H₂→ 2 H ⁺+ 2 e ⁻ (2) Reductive: O₂+ 4 e ⁻+ 4 H → 2 H O⁺ ₂ (3)

As previously mentioned, the reaction is indirect. The hydrogen oxidizes at the anode and releases its electrons. The electrons travel to the cathode where oxygen reduces together with hydrogen ions, formed at the anode, producing water. The electrons traveling, i.e the current, can be used as power. Through this fashion, the reaction can be controlled and reach high efficiency.¹

Figure 1. A schematic illustration of a simple fuel cell.¹

(6)

Types of fuel cells

There are different types of fuel cells, the main differences between them are the type of electrolyte, electrode material, catalyst and type of fuel. However, all of them operate on the same principle of harnessing the energy released in an indirect chemical reaction. Fuel cells can be categorized into either having a low or high operating temperature. The operating temperature naturally affects the practical applications of a fuel cell. Alkaline fuel cells (AFCs) are the main focus of this project. Alkaline fuel cells use an alkaline electrolyte consisting of either NaOH or KOH.² AFCs have a relatively low operating temperature of 60-80 °C, which means they are applicable in mobile devices. ² This is one reason why AFCs were used in the Apollo program as a source of energy and water onboard the spacecrafts.

Historically, the disadvantages of AFCs have been their corrosive electrolyte and susceptibility to CO2 poisoning, which decreases the performance of the cell. However, many of the issues related to the earliest AFCs are no longer present in modern AFCs.³

Since AFCs use an alkaline electrolyte, the oxidation and reduction reaction changes, these changes can be seen in Equation 4 and Equation 5. The overall net reaction stays the same.³

Oxidative: H₂+ 2 OH →2 H O ⁻ ₂ + 2 e ⁻ (4) Reductive: O₂+ 4 e⁻+ 2 H O → 4 OH ₂ ⁻ (5)

Alternatives to AFCs are polymer electrolyte membrane fuel cells (PEMFC) and solid-oxide fuel cells (SOFC). PEMFCs utilize a membrane that allows ions to pass through which separates the electrodes, in this type of fuel cell the reaction usually takes place over a catalyst consisting of platinum or a platinum alloy. Because of their high cost and the difficulties of managing them they are not suitable or attractive as consumer products. ^4,5 SOFCs on the other hand are one of the most efficient types of fuel cells available but due to their high operating temperature of +600℃ they are not suitable as consumer products either.⁶

(7)

Oxygen reduction reaction - ORR

The ORR is a reaction in aqueous solutions where oxygen and hydrogen combine to form water with the help of a catalyst via one of two paths. Which path is the dominant one is determined by the reaction conditions and catalyst. The available paths are the direct four-electron path seen in Equation 6.

H O e OH

O₂+ 2 ₂ + 4 ⁻ → 4 ⁻ (6)

As well as the two part two-electron path seen in Equation 7 and Equation 8.⁷

O e O H

O₂+ H₂ + 2 ⁻ → H ₂⁻+ O ⁻ (7)

O O e OH

H ₂⁻+ H₂ + 2 ⁻ → 3 ⁻ (8)

The ORR is the limiting reaction within a fuel cell since the standard reduction potential of the oxidative half-cell shown in Equation 4 is -0.828 V, ⁸ while the standard reduction potential of the reductive half-cell Equation 5 is +0.40 V. ⁹ The reduction potentials shows that equation 4 is a stronger reducing agent than equation 5 is an oxidizing agent.

Therefore, catalyzing the ORR is beneficial for the cell performance.

(8)

Catalyst

Due to the slow nature of the ORR it is desirable to use a catalyst with a high activity. The catalysts often consist of rare or noble metals. Because of their cost and rarity they are a deciding factor in the price of fuel cells. To counteract this cost as well as increase the overall efficiency of the catalyst nano-dispersed metal particles on a high surface area support like carbon is often used. A carbon based electrode will therefore be cheaper than an electrode constructed by noble metal solely. In this project, the carbon powder XC-72 was used as support for the catalyst. By decreasing the size of the particles and increasing the total surface area it is possible to keep the use of these expensive metals to a minimum while keeping the catalytic activity high enough to sustainably run the ORR.

The most common catalyst today is platinum but because of its rarity it has proven to be too expensive for use in consumer products. It was therefore this project's intention to examine alternatives to the noble metal but due to the COVID-19 situation only one alternative was examined. Silver has previously proven to display similar properties and has a fraction of the price making it a good candidate for replacing platinum as the standard in fuel cell catalysts.¹⁰

Nanoparticles

However, the surface area to volume ratio is a key factor in this project. When an object is divided into smaller portions, the volume remains constant as the total surface area increases, this is shown in Figure 2. If the particles are approximated to have a spherical shape, the formula for the surface area and volume of a sphere can be used to graph the ratio between the two as the particle size differs. The surface area over volume ratio increases exponentially as the particle diameter decreases as seen in Figure 3. Therefore, smaller particles are desired to minimise the amount of material, i.e. catalyst, used.

(9)

Figure 2. Illustration of the increase of total surface area as an object is divided

Figure 3. Graph of the surface area to volume ratio of a sphere as a function of particle diameter.

(10)

Radiolysis

Introduction to radiation chemistry

Ionizing radiation are particles or electromagnetic waves with enough energy to cause multiple ionizations to matter when absorbed. The unit of how much energy is absorbed per unit mass is called the absorbed dose and is expressed in the SI-unit Gray or Gy (J/kg) for short. The absorbed dose per unit of time is known as the dose rate and these units play a vital role in radiation chemistry. The G-value is another important factor, also known as the radiation chemical yield and defined as the number of moles produced of a given particle per joule of energy (mol/J). Chemical changes caused by radiation depend on both the absorbing chemical, dose and dose rate as well as traditional reaction conditions like temperature and concentration.

Metal nanoparticle synthesis in water

Sample preparation by radiation has plenty of benefits, low energy consumption coupled with lower risk of contamination and simpler synthesis makes it an attractive method for the creation of metal nanoparticles. Nanoparticles that can be synthesized in water bring additional advantages as they require little to no other solvents and allow for very green synthesis paths. When water is exposed to ionizing radiation, in the case of this study gamma radiation, it absorbs the energy and enters either an excited state or an ionized state. From here the excited or ionized molecules start fragmenting in a set way visualised in Figure 4.

(11)

Figure 4. Schematic view of the fragmentation process for water. Source: Dispenza et al.¹¹

The result of this process is two radical species, H ॱ and HO ॱ .¹² In order to reliably reduce metal ions to form the desired product the oxidizing species must be removed, this can be done with the addition of a radical scavenger. Addition of isopropanol as a radical scavenger to the solution before radiation converts any unwanted oxidizing radicals, OHॱ , in the solution to the reducing radical CH ) COH( _{3 2} ॱ through Equation 9.¹¹ Thus, obtaining completely reducing conditions.

Hॱ CH ) CHOH O CH ) COHॱ

O + ( _{3 2} → H₂ + ( _{3 2} (9)

The reducing radicals are capable of reducing metal ions to metal atoms through Equation 10 and Equation 11. Were the Hॱ can also be CH ) COH( _{3 2} ॱ in Equation 11. These evenly distributed particles then act as seeds for the formation of clusters. Metal ions are more likely to form bonds with the reduced metal than retain their bonds with the solvent making them gather and form larger charged particles. These are then reduced by the reducing radicals in the solution. Presence of any oxidizing radical would heavily counteract this process by oxidizing the formed metal particles back to metal ions.¹¹

e e e

M ⁿ⁺+ n ⁻ → M ⁰ (10)

e Hॱ e H

M ⁿ⁺+ n → M ⁰+ n ⁺ (11)

(12)

Analytical methods

XRD

The two main analytical methods for this experiment were X-ray diffraction (XRD) and inductively coupled plasma atomic emission Spectroscopy (ICP-AES). XRD is a method to analyze the structure of crystals. It is used within a wide variety of fields of science since it is both a qualitative- and quantitative analyzing method. In this project, XRD is used to confirm that the catalyst is present in the electrode material and to determine the particle size of the catalyst. The diffractometer has three main components; an x-ray tube, sample holder and a detector. The most common wavelength used in XRD is λ=1.5419 Å due to the use of copper which in turn produces CuK𝝰 radiation. The X-rays are directly illuminating the sample and the detector records the intensities of the diffracted rays. XRD utilizes constructive interference of light and because of the ordered structure in the atoms, the wavelength of XRD and the distance between the atoms in crystals creates the diffraction effect. The incident rays excite electrons of an atom and X-ray of equal wavelength as the incident rays are scattered and an illustration of it can be seen in Figure 5. The scattered X-rays interfere with one another and the angles where constructive interference occurs can be determined. These angles satisfy the Bragg Equation resulting in a stronger intensity detected.¹³

Figure 5. A schematic illustration of elastic scattering. λ represents the X-rays wavelength, is the distanced between atoms in the crystal and 2θ is twice the Bragg angle.

(13)

position, 2θ, wavelength, λ, and Full Width at Half Maximum (FWHM), β, are obtained from one of the peaks in the spectrum and used in Equation 12.¹⁴

Dp = _{β×cos (θ)}^0,94×λ (12)

ICP

ICP-AES is a common method of analyzing a sample both qualitative and quantitative. The basic principle behind ICP-AES, like all spectroscopy methods, is that when excited atoms return to their ground state a photon is released. The wavelength of the emitted photon corresponds to a difference in energy between the excited state and the ground state of the atom, which is characteristic for all atoms. By letting argon flow through an inductively coupled coil with an alternating electric field, plasma is created. The plasma causes the excitation of the incoming atoms from the sample. The photons emitted are processed by a monochromator and a signal amplifier, depending on the setup. The signal is then interpreted by a computer and a spectra is obtained. The wavelength and the intensity indicates what and how much analyte was in the analyzed sample.¹⁵

BET

BET stands for Brunauer- Emmet- Teller which are the developers of the BET theory. The theory builds upon the Langmuir theory where the adsorption of gas to solid surfaces is modelled. The BET method is therefore a useful way of determining the surface area of a sample. The most common way of doing this is by the use of nitrogen gas due to its strong interactions with surfaces and its availability on high purity. The sample is placed in a vacuum where nitrogen gas is released into stepwise. By measuring the amount of nitrogen flowing in and out of the sample chamber at different relative pressures of nitrogen, the surface area can be determined.¹⁶

(14)

Complementing research

As previously mentioned, the project was not able to be fully completed. Therefore, the results from a study performed by Soroka et. al. was used to further discuss how nanoparticles of silver affects the electrode performance. These results are presented below inTable 1 and Figure 6. ¹⁰The different electrodes containing different amounts of Ag ratios will from now on be denoted as follows;

Electrode A: 1.8 wt% Ag on carbon Electrode B: 3.8 wt% Ag on carbon Electrode C: 6.7 wt% Ag on carbon

Table 1. Data regarding the synthesis of the different electrodes and electrode material. Data acquired by I.L.

Soroka et al.¹⁰

Electrode Ag

concentration in Ag-C electrode material, (wt. %)

AgNO3

concentratio n

(mM)

Amount of PVA in solution (g)

Total irradiation dose (kGy)

Ag yield, measured with ICP-OES, (%)

Crystallite size of Ag particles calculated from XRD patterns, (nm)

A 1.8 5.5 0.3 31.2 44 ± 3 11 ± 6

B 3.8 5.5 0.15 26.7 82 ± 5 41 ± 5

C 6.7 11 0.3 41.3 91 ± 6 19 ± 3

(15)

Figure 6 . Current-potential characteristics of Ag/C electrodes. (a) and (b) display the correlation between current and potential per area of the working electrode at room temperature, (a), and 60℃, (b). (c) and (d) display the correlation between potential and current normalized by weight Ag at room temperature, (c), and 60℃. Data acquired by I. L. Soroka et.al.¹⁰

Methods

Sample preparation

A standard solution of AgNO 3 was prepared by adding 0.1689 g of AgNO 3 in 10 ml of MilliQ water, resulting in a 0.1 M AgNO 3 solution. The amount of carbon for each sample was calculated from the goal of 10 wt% of catalyst on carbon and assuming complete reduction of the metal ions. Each amount of carbon powder was weighed and put in four different vials. To vials 1, 2, 3 and 4 the amount of carbon powder, XC-72, added was 0.0262 g, 0.1311 g, 0.2631 g and 0.5242 g respectively. The volume of MilliQ water added to each vial was 22.601 ml, 21.521 ml, 20.171 ml and 17.471 ml respectively. To each vial 4.129 ml of isopropanol was added. The samples were placed in an ultrasonic bath for 15 minutes to solve the carbon powder. Then the different volumes, 0.27 ml, 1.35 ml, 2.7 ml and 5.4 ml of the standard solution (SS) were added to each sample respectively and were placed in the

(16)

ultrasonic bath for another 15 minutes. The composition of each sample can also be seen in Table 2. Two 1 ml samples were collected for pH and ICP analysis. A magnetic stirrer was added and the vials were capped with rubber caps and wrapped with parafilm and aluminium foil. The vials were placed on stirring overnight at 800 rpm. After stirring each sample was bubbled with N2 gas for 30 min and then placed into the gamma cell. After 20 hours in the gamma cell, the samples were taken out of the gamma cell and prepared for centrifugation. The samples were then centrifuged for 20 minutes at 8000 rpm. From each sample two 1 ml samples were taken for pH and ICP analysis. The solution was removed and isopropanol was added to each sample to aid the drying process. The samples were placed in an oven overnight to dry. To analyze how much catalyst was incorporated onto the carbon powder, 1.8 mg, 2.8 mg, 5.4 mg and 4.1 mg was taken from sample 1, 2, 3 and 4 respectively and placed into clean vials. 10 ml of 5 mole% HNO 3was added to each vial and the vials were placed in an ultrasonic bath for 20 minutes. 1 ml samples were taken from each vial.

Table 2. Table over the composition of each sample before irradiation

Sample # [Ag⁺] (mM) XC-72 (g) VSS (ml) VIsopropanol (ml) VMilli-Q water (ml)

1 1 0.0262 0.27 4.129 22.601

2 5 0.1311 1.35 4.129 21.521

3 10 0.2631 2.7 4.129 20.171

4 20 0.5242 5.4 4.129 17.471

Sample analysis

As previously mentioned two 1 ml samples were taken from each sample before and after irradiation for pH and ICP analysis. The pH-meter was calibrated in advance and the electrode was washed thoroughly between samples. To determine the Ag content in the samples collected before and after irradiation was analyzed with ICP. The dried precipitate, i. e the electrode, from each sample was grinded and analyzed with XRD. The electrode material was dissolved in 5 % HNO3 for ICP analysis of the Ag loaded on the carbon.

(17)

instruction for future investigations in the subject.

Electrode preparation

The obtained catalytic materials were mixed with 60 wt% polytetrafluoroethylene and hydrocarbons solvent was added and slowly homogenized until a rollable paste was obtained. To receive a thin layer of 0.7 mm thickness, a heavy 180 kg roller on a flat layer was used. After pressing at 150 kg cm ^-2the electrodes were dried in a 60 ℃ oven overnight and thereafter sintered at 320 ℃ for 1 hour in N 2. Under the same condition a blank sample of Vulcan XC-72 was made.

Electrode measurement

For electrode measurements a three-electrode configuration was made, a counter electrode of nickel wire mesh and a reference electrode of Hg/HgO. Galvanostatic measurements were carried out using power supply (PS3005) on the 4 cm ² half-cells after the electrodes reached a steady state. A constant flow of oxygen (99.9 % purity) of 5 ml/min was used.

After approximately 15 minutes the electrodes reached steady-state, the galvanostatic measurements were made. For controlling the temperature a thermostated bath was used and kept at 65±1 ℃ and the temperature was registered by thermocouple.

(18)

Results

Table 3. The measured pH of each sample before and after irradiation

Sample no. pH before irradiation pH after irradiation

1 5.361 4.692

2 3.911 2.647

3 3.338 2.504

4 3.069 2.428

Table 4. Concentration of Ag⁺ before and after irradiation for each sample from ICP together with the calculated yield.

Sample no. Before irradiation (mM)

After irradiation (mM) Yield (%)

1 0.2101 0.0009639 99.5

2 2.5813 0.2621 89.8

3 3.4465 0.5444 84.2

4 14.2370 7.1248 49.9

(19)

Figure 7. Results from the X-ray diffraction, initial sample concentration shown in corresponding graph

Table 5. The particle size of each sample calculated from X-ray diffraction patterns using the Scherrer formula.

Sample no. Particle size (nm)

1 36.15

2 111.56

3 30.99

4 111.51

(20)

Figure 8 . Particle size of each sample plotted against the respective Ag ⁺concentration before irradiation and a plotted fitted linear relationship.

Table 6 . Ag⁺concentration of the Ag particle loaded carbon, mass carbon analysed and the calculated loading of the carbon.

Sample no. Silver on carbon (mM)

Silver on carbon (mg)

Mass electrode material in sample (mg)

mAg/mC Total mass Ag on carbon

(mg)

1 0.09018 0.09818 1.8 0.0577 1.51

2 0.1522 0.16418 2.8 0.0623 8.17

3 0.2290 0.24702 5.4 0.0479 12.61

4 0.06384 0.06886 4.1 0.0171 8.95

Table 7. Ag⁺ concentration of Ag particle loaded carbon, percentage yield calculated

Sample no. Before irradiation (mg) Yield (%)

1 0.5666 268

2 6.9610 117

3 9.2942 136

4 38.3929 23

(21)

Discussion

The ICP analyses of the Ag⁺ concentration before irradiation shows lower than expected results. The four samples were prepared to have a Ag ⁺ concentration of 1 mM, 5 mM, 10 mM and 20 mM, although the ICP results show about 0.2 mM, 2.6 mM, 3.4 mM and 14.2 mM in respective samples as seen in Table 4. This indicates a systematic error in either the ICP analysis or in the sample preparation. Comparing the total amount of loaded Ag on carbon with the amount of Ag ⁺ before irradiation, the yield for sample 1 through 4 is about 268 %, 117 %, 136 % and 23 % respectively, seen in Table 7. The calculations for the total amount of loaded Ag on carbon assumes homogeneous distribution of Ag. These yields are not possible and are a strong indicator of a faulty preparation. Since the ICP was calibrated before the analysis using stock solutions, the most likely source of this error is the preparation of the standard solution of 0.1 M AgNO 3. The scale used to weigh the AgNO 3salt was assumed to be calibrated, although this may have not been the case. The uncalibrated scale was also used to weigh both the carbon powder loaded with Ag and the non-loaded carbon powder. This introduces errors in several steps of the synthesis.

The yield of Ag ⁺(aq) to Ag(s) was also calculated using the concentrations of Ag ⁺before and after irradiation. Sample 1, 2 and 3 all show a yield of over 84 % and sample 4 shows about 50% yield, seen in Table 4. The irradiation time was calculated for full conversion of sample 2, which was 20 hours. In theory, this would result in an approximate yield for sample 1, 3 and 4 of 100 %, 50 % and 25 % respectively. Sample 1 and 2 show a lower than expected conversion, while sample 3 and 4 show a significantly higher conversion than expected. The results for sample 3 and 4 could be explained by the lower than expected Ag ⁺concentration before irradiation. These yields are more likely to be correct than the yields calculated from the Ag loaded on the carbon powder since the source of error from the scale is cancelled out. The yield calculated from the loaded Ag shows over 100 % yield, see Table 7, which is impossible. This further indicates that the scale is the source of the error since it was used to weigh the loaded carbon for analysis.

(22)

The XRD results show varying particle sizes as seen in Table 5. These results were plotted against the Ag⁺ concentration before irradiation of each sample respectively, see Figure 8. A regression analysis was performed and the resulting linear function indicates a positive trend between ion concentration and particle size, also seen in Figure 8. However, the correlation is not strong. The varying results could be because of the sample preparation for the XRD analysis. Each sample was grinded and prepared by a different person by hand, which introduces inconsistencies in the preparation of the XRD samples.

The pH measurements after the irradiation, also seen in Table 3, show a decrease in pH for every sample. An explanation for this could be the use of the radical scavenger. Since the OH^• is removed by the isopropanol, as shown in Equation 9, it cannot react further to produce OH^-. Meanwhile, H^• still reacts and produces H⁺. This results in a higher concentration of H⁺ and consequently a lower pH. As for the pH measurements of the unirradiated samples the same trend can be observed in Table 3. This can be explained by the acidity of the used scavenger, isopropanol. With a pKa of 17.1 the alcohol will exist in a deprotonated form even in a neutral solution, addition of it will therefore lower the pH.

Another contributing factor could be the carbon. Activated carbon can have either an alkaline or acidic effect on the solution depending on the manufacturing process. Unlike the isopropanol, which had a constant amount added, the carbon was added in a fixed ratio to the metal ion concentration. This would explain the downward trend seen. In Table 3 a clear trend is seen where the pH decreases as the initial Ag ⁺ concentration increases. Together with the results for the amount of remaining Ag ⁺ in the solution from the ICP analysis in Table 4 . It can be concluded that in solution with higher initial concentrations of Ag ⁺ more ions are reduced to form metal atoms. As seen in Equation 11 hydrogen ions are formed when a hydrogen radical reduces another ion. This makes the pH a measurement of how much reductant has been consumed and further a measurement of how many metal atoms has been formed.

As seen in graph (a) and graph (b) in Figure 6, electrode A performs similarly to electrode C at 24 °C and 60 °C. The performance of electrode B is somewhere in between the electrode without any catalyst and the other two electrodes. However, when these results are normalized against the mass of the loaded catalyst of each respective electrode as shown in

(23)

results show that all the electrodes have similar surface area within 200 ± 20 m ²/g. ⁹ The XRD shows an average crystallite size of 11 nm, 41 nm and 19 nm for electrodes A, B and C respectively, as seen in Table 1. Assuming identical surface area of the electrodes, the results in graphs (a) through (d) could be explained using the average crystallite size and the mass Ag loaded on the electrode. Electrode B has the largest Ag crystallite size of the electrodes, which means the smallest active area of the catalyst. This reduces the electrode’s performance compared to electrode A and electrode C. As mentioned, electrode A and electrode C show similar performance to one another. Electrode A has the smallest average Ag crystallite size of the electrodes and therefore the largest active area of catalyst.Although, electrode C has almost the double average Ag crystallite size than electrode A, the performance is compensated by the loading of Ag. The loading of Ag on electrode C is almost four times greater than electrode A. This explains why the performance of electrode C decreases in graph (c) and graph (d).

Conclusion

To conclude, silver nanoparticles on carbon support were created in a water solution by radiolysis. The pH- and ICP-analysis of the solutions and XRD-analysis of the particles expressed two observable trends. The pH of the solutions was affected by too many variables such as acidic residue on the carbon support, acid by-products from radiolysis and addition of alcohols to draw any clear conclusions from on its own. The other observed trend was that of lower initial concentrations and lower average particle size, with the lowest initial silver ion concentration yielding the lowest average particle size. This trend is backed by data previously acquired by I.L Soroka but further investigation is required to confirm the trend as a result of the high variance in results. The results from the electrode measurements indicate an increase in performance as the Ag loading of the electrode increased and as the average Ag crystallite size decreased. This means that a lower degree of silver loading on the electrode can be compensated by decreasing the average particle size of the silver.

(24)

Bibliography

(1) Köster, V. How Do Fuel Cells Work? chemistryviews.org February 7, 2012.

(2) Gülzow, E. Alkaline Fuel Cells: A Critical View. J. Power Sources ^1996,61 (1), 99–104.

https://doi.org/10.1016/S0378-7753(96)02344-0.

(3) U.S Department of Energy. Types of Fuel Cells

https://www.energy.gov/eere/fuelcells/types-fuel-cells.

(4) Chapter 41 - Engineered Nanomaterials for Energy Applications. In Handbook of Nanomaterials for Industrial Applications; Mustansar Hussain, C., Ed.; Micro and Nano Technologies; Elsevier, 2018; pp 751–767. https://doi.org/10.1016/B978-0-12-813351-4.00043-2.

(5) Ren, H.; Chae, J. 2 - Fuel Cells Technologies for Wireless MEMS. In Wireless MEMS Networks and Applications; Uttamchandani, D., Ed.; Woodhead Publishing Series in Electronic and Optical Materials; Woodhead Publishing, 2017; pp 35–51.

https://doi.org/10.1016/B978-0-08-100449-4.00002-6.

(6) Esposito, V.; Garbayo, I.; Linderoth, S.; Pryds, N. 15 - Solid-Oxide Fuel Cells. In Epitaxial Growth of Complex Metal Oxides; Koster, G., Huijben, M., Rijnders, G., Eds.; Woodhead Publishing, 2015; pp 443–478. https://doi.org/10.1016/B978-1-78242-245-7.00015-4.

(7) Yang, Y.; Zhou, Y. Particle Size Effects for Oxygen Reduction on Dispersed Silver + Carbon Electrodes in Alkaline Solution. J. Electroanal. Chem. ^1995,397 (1–2), 271–278.

https://doi.org/10.1016/0022-0728(95)04178-7.

(8) Harris, D. C. Quantitative Chemical Analysis, 7. ed..; New York : Freeman: New York, 2007.

(9) Burrows, A. Chemistry3 : Introducing Inorganic, Organic, and Physical Chemistry, Third edition.;

Chemistry 3; Oxford : Oxford University Press, 2017.

(10) Soroka, I. L.; Bjervås, J.; Ceder, J.; Wallnerström, G.; Connan, M.; Tarakina, N. V.; Maier, A. C.;

Kiros, Y. Particle Size Effect of Ag-Nanocatalysts Deposited on Carbon as Prepared by γ-Radiation Induced Synthesis. Radiat. Phys. Chem. ^2019,169^.

https://doi.org/10.1016/j.radphyschem.2019.108370.

(11) Grimaldi, N.; Sabatino, M. A.; Soroka, I.; Jonsson, M. Radiation-Engineered Functional Nanoparticles in Aqueous Systems. J. Nanosci. Nanotechnol. ^2015,15^.

https://doi.org/10.1166/jnn.2015.9865.

(12) Soroka, I. Radiation Induced Synthesis of Nanomaterials. Appl. Phys. Chem. R. Inst. Technol.

KTH^36.

(13) Dutrow, B. L.; Clark, C. M. X-Ray Powder Diffraction (XRD); Science Education Resource Center Carleton College.

(14) Patterson, A. L. The Scherrer Formula for X-Ray Particle Size Determination. Phys. Rev. ^1939,56 (10), 978–982. https://doi.org/10.1103/PhysRev.56.978.

(15) Raja, P. M. V.; Barron, A. R. 1.5: ICP-AES Analysis of Nanoparticles

https://chem.libretexts.org/Bookshelves/Analytical_Chemistry/Book%3A_Physical_Methods_i n_Chemistry_and_Nano_Science_(Barron)/01%3A_Elemental_Analysis/1.05%3A_ICP-AES_Anal ysis_of_Nanoparticles (accessed Apr 29, 2020).

(16) Raja, P. M. V.; Barron, A. R. 2.3: BET Surface Area Analysis of Nanoparticles

https://chem.libretexts.org/Bookshelves/Analytical_Chemistry/Book%3A_Physical_Methods_i n_Chemistry_and_Nano_Science_(Barron)/02%3A_Physical_and_Thermal_Analysis/2.03%3A_

BET_Surface_Area_Analysis_of_Nanoparticles (accessed May 22, 2020).

Oxygen Reduction Catalysts in Alkaline Electrolyte

Abstract

Contents

Introduction

Methods

Results

Discussion

Conclusion

Bibliography