2014-Summer-REU-Program Project2-Energy Biweekly-Technical-Paper-3

PROJECT PROGRESS REPORT

Electrocatalytic Energy Conversion at the Interfaces of Hybrid Carbon-Bismuth Nanoparticle Assemblies

Submitted To

The 2014 Summer NSF REU Program

College of Engineering and Applied Science

University of Cincinnati

Cincinnati, Ohio

Prepared By

Trevor Yates, Chemical Engineering, University of Cincinnati

Adam McNeeley, Chemical Engineering, University of Cincinnati

Will Barrett, Chemical Engineering, University of Cincinnati

Report Reviewed By:

Dr. Anastasios Angelopoulos

REU Faculty Co-Mentor

Associate Professor

Department of Biomedical, Chemical and Environmental Engineering

University of Cincinnati

July 18th, 2014

Abstract.

Bismuth has potential as an electrocatalyst because of its conductive properties and low cost. Former studies have electrodeposited Bismuth nanoparticles on Graphite Felt electrodes. They have found that Bismuth improves the energy conversion efficiency of Vanadium Redox Flow Batteries (VRFBs). In this project, Bismuth nanoparticles will be further investigated as an electrocatalyst by implementing Standard Layer by Layer (sLBL) assemblies on a Glassy Carbon electrode.

1. Introduction.

Making solar energy more economical is one of the 14 National Academy of Engineering grand challenges presented to all engineers because of necessity to increase sustainability. Solar energy’s major flaw is its intermittence. When its its cloudy or at night there is no sun available, so no electricity can be generated to support a power grid. Having an effective way to store large amounts of energy to support large power grids would be a major step towards making solar energy viable. One method of storing energy that shows promise in this application is the VRFB. The VRFB is a flow battery that stores its charge in electrolyte tanks. The VRFB has tremendous potential for energy storage because the electrolyte solution is inert to the charging and discharging cycles; this means that they can run for indefinite periods being recharged without concern of degradation. The capacity and power output of the VRFB is also very customizable depending on the volume of electrolyte and number of cell stacks. The major issues with the VRFB include poor energy conversion, rate capability, and power density [2]. These issues keep the VRFB from becoming economically viable. Designing a more effective electrode for the VRFB may help solve some of these issues such as energy conversion and rate capability.

The first goal of this project is to investigate Bismuth as an electrocatalyst in electrodes to improve energy conversion and rate capability. Bismuth has great promise because it is cheap and serves as an excellent conductor. Bismuth as an electrocatalyst has not yet been extensively documented; however, previous research found that when bismuth is electrodeposited on graphite felt electrodes in VRFBs, there is an increase of energy efficiency by 11% at high current density (150 mA/cm^2) [2].

The next goal of this project is to test the performance of Bismuth and Carbon working together in Layer by Layer assembly. Carbon will be used in order to further stabilize the peaks generated by Bismuth.

2. Background Literature Review.

Background literature was reviewed to gain a better understanding of the research material. The first goal for the project was to gain a basic understanding of electrochemistry. The most important background reading assignments pertained to cyclic voltammetry because it has been used as the method of testing variations of bismuth electrocatalysts. Cyclic voltammetry is a basic method for characterizing electrochemical reactions. This involves measuring a current through a circuit as a function of the voltage potential. The voltage sweeps within a range of values linearly with respect to time, and current peaks will occur when a compound is oxidized or reduced.

A Cyclic Voltammeter is set up with three electrodes: the working, the reference, and the counter. The working electrode is connected to a potentiostat and is where the voltage is controlled. The reference electrode is used as a standard of reference to the voltage of the working electrode, and zero current passes through it. The counter electrode receives the current from the working electrode and its only purpose is to complete the circuit. All three electrodes are placed in a non-reactive electrolyte solution. The test is conducted by sweeping the voltage from low to high in a certain voltage range, and when the high limit is reached, the voltage will then sweep from high to low and repeating once the lower limit is reached. This is useful for electrochemistry because it generates redox reactions if swept in the right voltage ranges. A reactant loses electrons while being oxidized on the forward sweep, and gains electrons while being reduced on the reverse sweep. Based on Faraday’s Law of electrolysis, the accumulation of charge in the test is proportional to the mass of reactant altered at the electrode. Voltammogram graphs are set up with the voltage with respect to a reference on the x-axis and the current on the y-axis. Factors such as reaction rates, voltage sweep rates, concentration, and diffusion rates can all affect how a graph will turn out. In general as the voltage is swept higher or lower it will increase the reaction rate of oxidation or reduction. Peaks occur when the flux (diffusion rate) becomes a rate limiting factor in the redox reaction over the reaction rate. A higher potential pushes the equilibrium further to products, so the concentration gradient of the reactant begins to approach zero, and this causes a decrease in electron flow. By understanding further in depth how cyclic voltammetry works, it becomes possible to understand the ideal performance of layered assemblies.

Another part of the literature review was researching Bismuth nanoparticles as an electrocatalyst. This is a relatively new area, but two similar studies were found. The first study was conducted by Bin Li, Meng Gu, et al [2]. The second was conducted by David J. Suarez, Zoraida Gonzalez, et al [5]. Both of these studies involved the electrodeposition of bismuth on an electrode to use as an electrocatalyst in a Vanadium Redox Flow Battery (VRFB). Both studies also showed that Bismuth proved to be a useful electrocatalyst in the negative cell of the battery. The current research being done is different because the Bismuth is applied to the electrode using the Layer by Layer (LbL) technique. Also Bismuth and Carbon have never been used in a hybridized assembly before.

The last part of the literature review was on a study that used the Layer by Layer technique on an electrode only with Platinum instead of Bismuth, and for a different purpose [4]. This study was done by Samuel St. John, Indrajit Dutta, and Anastasios P. Angelopoulos and was helpful for learning the standard Layer by Layer (sLBL) technique.

3. Goals and Objectives of the Research Project.

The first objective of this project is to develop a basic understanding of electrochemistry. This will be achieved by reviewing literature. Most of this is to be done during the first couple weeks of the program. Once some background knowledge is obtained, the next objective is to test Bismuth as an electrocatalyst. This will be done by assembling an electrode with Bismuth nanoparticles using the Layer by Layer (LbL) technique and analyzing with cyclic voltammetry. This task should also be completed by the end of the second week. The third objective of this project is to create a hybrid Carbon-Bismuth nanoparticle assembly. It will be done by taking a Carbon electrode and implementing the Layer by Layer technique with the Bismuth. The theory behind this is that Carbon and the Bi-Sn nanoparticle surface charge are negative and can be stacked with a positive charged polymer in between. This test assembly will be tested using cyclic voltammetry in the standard electrolyte solution 0.5 M sulfuric acid as well as 2 M sulfuric acid in a Vanadium solution. This process will most likely take from week three to week eight. The final objective of this project is to efficiently complete all of the deliverables in the time restraints given at the beginning of the program.

4. Research Training Received.

Graduate Research Assistant (GRA), Abhinandh Sankar has demonstrated how to prepare Glassy Carbon for cyclic voltammetric scanning. The first major step is to make sure that the Glassy Carbon electrode is as clean as possible. Since there are some minor scratches on the surface, this makes it challenging to clear off the nanoparticles. In order to do that we must polish off the surface with alumina and rinse with deionized water. Then we have the Glassy Carbon sit in the ultrasonicator in a beaker of water for about five minutes before it is ready to soak in the acid solution. Abhinandh made a solution of Aqua Regia three parts hydrochloric acid and one part nitric acid. After the Glassy Carbon soaks in the acid it can be rinsed off and tested in the Cyclic Voltammogram. Once the peaks are no longer visible then the Glassy Carbon is ready for experimentation.

All experiments run with the same basic procedure, and the only things that change are the substances added to the glassy carbon. The group alternates between adding bismuth and polymer solutions to the carbon surface. Each layer requires at least two minutes to dry and then one minute of washing in a beaker of deionized water. This cycle repeats until all of the necessary layers are added for testing.

5. Methods.

Particle and Polymer Preparation

The preparation of Bismuth nanoparticles and the charged copolymer have been important to the testing, since they are the compounds used for the LbL assembly. The nanoparticles proved to be particularly difficult to deal with. The solution of nanoparticles is in a dark black solution, which has a tendency to separate. The solution must be thoroughly agitated before used in order to ensure it is well mixed. The Bi-Sn nanoparticles stability is dependent on the solution pH, which is ideal at 3. A higher pH and Bi and Sn would lose its bond. A lower pH and Bi would be oxidized.

The first Bi-Sn nanoparticles used were from a previous experiment and were old. These particles had a very thick consistency, and the solution oxidized becoming white. Fresh particles were prepared and required three days to set up. The new particles had a much thinner consistency and were easier to work with. After the old particles had been oxidized extra care was taken to prevent the new ones from being exposed to air. Parafilm was wrapped around the lid of the vial containing the particles when it was no longer being used.

The polymer was prepared prior to any testing. The polymer was labeled at 12 g/L, and was a clear solution. The same polymer solution was used for all of the tests. After initial test results showed the polymer having a poor effect with the Bi-Sn nanoparticles a portion of the existing polymer was used to create a polymer with a lower pH. The original polymer had a measured pH of 4.4. The sample polymer was stirred while 18 M HCl was added dropwise. The pH meter in the solution recorded the final pH of the solution being 2.67. The modified polymer was used for only one experiment.

The stability of the Bismuth nanoparticles has become a major concern in the experiment. After a few days of exposure to air the bismuth nanoparticles were observed to have a yellow white film that separated on top of the black solution. After an extended amount of time, usually a week later, the solution would turn white. Along with this noted change in physical appearance was a consistent decline in electroactivity of bismuth in older nanoparticle samples. This decline was noticed because of smaller bismuth peaks. It was hypothesized that the bismuth nanoparticles were being oxidized in air over time. To prevent this from happening measures were taken to keep the nanoparticles from being exposed to oxygen. When the particles were used for experimentation, a pure Nitrogen air stream was blown into the vial after uncapping it until bismuth was extracted using an automatic pipette. This Nitrogen stream was usually applied for about 5 seconds. Then after the bismuth was extracted the Nitrogen stream was reapplied for 30 seconds before recapping the vial. This procedure is repeated every time bismuth nanoparticles need to be applied to the electrode. After the bismuth nanoparticles are done being used Nitrogen is blown into the vial for 2 minutes before being capped and is then sealed with parafilm.

Cleaning Process

The glassy carbon electrode used for the assembly of the electrodes being tested, undergoes a vigorous cleaning after each test trial. It is important for accurate data that the glassy carbon is free of any impurities before using to assemble a test electrode. The glassy carbon electrode is tested for impurities by connecting to the cyclic voltammeter. The voltammeter is then run at a preset sweep rate of 250 mV/s; later changed to 50 mV/s to use as a base-line. The resulting graph is then analyzed for current peaks that would signify any electroactive impurities on the electrode. The electrode is scanned for 25 cycles to allow any minor impurities to be removed. When no current peaks are observed the glassy carbon electrode is ready for test assembly.

A standard cleaning procedure was first used to clean the electrode. This involved first scrubbing the electrode on a buffer pad using alumina paste, and Mili-Q deionized water. The electrode is scrubbed with the glassy carbon flat to the buffer pad applying light pressure in a figure eight motion. The electrode is scrubbed for a short length of time and is then rinsed with Mili-Q deionized water. The electrode is then placed in a 150 mL beaker half filled with Mili-Q deionized water. The beaker with the electrode is then placed in an ultrasonicator turned on to the max setting. The beaker is allowed sit in the ultrasonicator for at least 5 minutes to allow for any particulates to vibrate off of the electrode. The electrode is then removed from the 150 mL beaker and rinsed with Mili-Q deionized water. The electrode is then connected to the cyclic voltammeter to perform cleaning sweeps. After confirming the electrode is free of electroactive impurities from the cleaning scan, the electrode is removed from the cyclic voltammeter. The electrode is then rinsed with Mili-Q deionized water and placed on a stand to dry. Once dry the electrode is ready for test assembly. The Bi-Sn nanoparticles are difficult to remove from the electrode. Bismuth oxidation peaks were consistently observed in the cleaning scans. This increased cleaning time because all of the cleaning steps needed to be repeated until the cleaning scan showed the absence of a Bismuth peak.

An extra step was added to allow for a more efficient cleaning of the electrode. This step was done after removing the electrode from the ultrasonicator. Aqua regia was prepared in a vial by mixing 18M Hydrochloric acid with 18M Nitric acid in a 3:1 ratio. The vial of Aqua regia was stored with the vial capped inside a fume hood. After the electrode is removed from the ultrasonicator it is place inside the fume hood faced down inside the vial of Aqua regia so that just the surface of the glassy carbon comes in contact with the Aqua regia. The electrode is left to sit in the aqua regia between 15 minutes to an hour. The electrode is then removed from the vial and rinsed with Mili-Q deionized water. After rinsing the electrode it is placed in a 150 mL beaker half filled with Mili-Q deionized water and put in the ultrasonicator again. The electrode is removed from the beaker and rinsed with Mili-Q deionized water. The electrode is then connected to the cyclic voltammeter to check for electrochemical impurities.

It is important to note that for some of the early tests, the electrode showed very small Bismuth oxidation peaks after multiple intensive cleanings using the Aqua regia step. These small peaks were assumed negligible with the cleaning scan used as the base-line and assembly of the test electrode proceeded with.

Cyclic Voltammetry

When the electrodes were in Sulfuric acid, they were typically scanned between -650 mV and 650 mV with respect to an Ag/AgCl. When they were in a Vanadium Sulfate solution, they were scanned between 0.1 mV and 1.4 mV. These two values were chosen for these solutions because we want to avoid oxygen and hydrogen evolution and still be able to observe Bi oxidation. It is important to keep the window small because gas evolution would cause bubbles to form on the surface of the electrode and the electrode would have to be reconstructed. The electrodes were also scanned at different scan rates so we can find out more information about what is occurring. In general the scan rates were either 10, 25, 50, or 100mV/s. At the moment we are focusing on the initial peak current and the stability of the electrode. The initial peak current shows us how effective Bismuth’s electrochemical properties are. To investigate the stability, several cycles were run to see where the current would settle.

Standard Layer by Layer

Standard Layer by Layer (sLbL) is a method that allows for the controlled application of an electrocatalyst on and electrode surface by using electrostatic attractions. By taking advantage of the negative surface charge of the Glassy Carbon (GC), the positive surface charge of the cationic polymer, and the negative surface charge Bi-Sn nanoparticles they can be stacked forming layers. Both the cationic polymer and the Bi-Sn nanoparticles are in liquid form while the Glassy Carbon is the center of the electrode. Thus, the Glassy Carbon is just the base while the polymer and the Bi-Sn nanoparticles are applied using a pipet. The first layer applied to the Glassy Carbon (GC) is the cationic polymer. The positive surface charge should stick to the negative surface charge of the GC. Polymer is applied for two minutes to the GC. The electrode is washed in a beaker full of DI water to get rid of the excess polymer. The same process is then repeated for the Bi-Sn nanoparticles. Theoretically, the Sn shell with a negative surface charge should stick to the polymer. Once both are applied, a layer is complete. Finding how the number of layers affects the peak current is part of the investigation. This will be discussed in the results and discussion section. Several 4-layered and 8-layered electrodes were constructed and tested by using cyclic voltammetry.