Electron Transfer Dynamics in Photocatalytic CO₂ Conversion

Coal is the workhorse of our power industry, responsible for approximately half of the electricity consumed by
Americans. Managing carbon dioxide (CO₂) emissions from coal utilization is one of the most challenging issues facing the fossil energy industry today. To cost-effectively capture and manage CO₂, new and flexible photocatalytic technologies are being developed that can be used at large storage facilities (geological, oil fields, etc. ) to slowly convert stored CO₂ into more useful products such as methane and methanol.

Photocatalysis is the acceleration of a light-induced reaction in the presence of a catalyst. During this process an electron-hole pair is created as a result of exposure to ultraviolet radiation or visible light and the resulting free-radicals are very efficient oxidizers of organic matter. The most common photocatalyst for CO₂ conversion is titanium dioxide (TiO₂), which suffers from two fundamental inefficiencies preventing industrial scale deployment:
1) a large UV spectral band gap that only utilizes 1-5% of the sunlight reaching the earth's surface, and
2) rapid charge carrier recombination that causes electrons and holes to recombine before they can initiate the photoreduction of CO₂.

To overcome these inefficiencies, one approach uses semiconductor nanocrystals such as cadmium selenide (CdSe) or lead sulfide (PbS) to photosensitize TiO₂ and improve its optical activity in the visible and near infrared regions of the solar spectrum. The advantage of using these nanocrystals is that they are easily tunable to absorb light at any wavelength, are thermally stable, and are photochemically robust. When combined with TiO₂, these semiconductor nanocrystals form a type II band alignment, meaning that photoexcited electrons in the conduction band of the nanocrystal can be injected into the conduction band of TiO₂. Photoinjections of electrons are ultra-fast events, causing the electron and hole in the nanocrystal to separate, reducing charge recombination. But the mechanisms by which these events happen are not well understood and they are the key processes in photocatalysis. To simulate electron transfer at interfaces composed of two dissimilar materials such as CdSe and TiO₂ is technically challenging.

NETL researchers are investigating the dynamics of these photoinduced electrons at the interface of a CdSe nanoparticle with a TiO₂ nanoparticle. The goal is to generate valuable insights into the transfer mechanisms of photoinduced electrons and to provide guidelines for system design and improvement. Simulations are performed using the supercomputers at the National Energy Research Scientific Center (NERSC), where NETL researcher De Nyago Tafen was awarded a total of 600,000 MPP Hours (hours X nodes used X 24 cores per node) for a period of two years starting in January 2012. To tackle this problem, Dr. Tafen uses a variety of techniques and a set of tools in collaboration with a researcher in the Chemistry Department at the University of Rochester. The present work combines nonadiabatic molecular dynamics (NAMD) with time-domain density functional theory (TDDFT) to help understand the mechanisms responsible for the movement of charge through the heterostructure, identify the vibrational motions that promote charge transfer, and provide a better understanding of the electron transfer dynamics.

In molecular models, approximations assume low-energy density and near equilibrium (adiabatic) situations. NAMD accounts for conditions or subsystems that are quickly driven out of equilibrium by an external perturbation. Time Dependent Density Functional Theory provides a framework to describe electron dynamics out of the electronic ground state. Recent findings at NETL have shown that uniform distribution and close/direct contact of the semi-conductor nanoparticle with the TiO₂ nanoparticles are needed for efficient carrier separation (for fast electron transfer to occur) across the semiconductor nanocrystal and TiO₂ junction. Additionally the charge injection rate from the semiconductor nanocrystal into the TiO₂ nanoparticle depends on the size of the nanocrystal.

Figure 1. Schematic of the photoinduced electron injection process in CdSe/TiO2 nanocrystal heterostructure. An absorbed photon promotes an electron from the ground state of the CdSe nanocrystal located inside the TiO2 band gap into an excited state (photoexcited state). Then, the excited electron is transferred into the TiO2 conduction band (CB). VB represents the valence band of TiO2 and ET the electron transfer from the photoexcited state to the TiO2 CB.

NETL Studies High Throughput Membrane Screening

Membranes offer a potential low-maintenance and economical method for gas separations from power plant flue gas streams. Polymer membranes and supported liquid membranes show great promise to solve problems in the area of clean energy production. Carbon dioxide, a greenhouse gas, is a principal by-product of energy production from fossil fuels. Capturing CO₂ from power plant flue gas streams is critical to the goal of reducing the nation’s carbon footprint and preserving the environment. Currently, there is no technology that can meet the goals for carbon capture as set forth by the U.S. Department of Energy. These goals are 90% capture of the CO₂ with a less than 35% increase in the cost of energy.

Dr. David Luebke at work in the lab.

The National Energy Technology Laboratory (NETL) is pursuing the development of both polymeric and supported ionic liquid membranes for CO₂ capture. Development of adequate membrane technology requires equipment capable of rapidly measuring membrane performance. Typical membrane testing equipment operates under either constant pressure or constant volume conditions. Constant pressure instruments pass feed gas over one side of the membrane and a sweep gas over the other side of the membrane. The feed gas is comprised of the gases which are to be separated while the sweep gas is inert and serves the purpose of carrying away the gas that passes through the membrane (i.e. , the separated gas). By carrying away the separated gas, the sweep gas allows for increased efficiency of the separation. Constant volume instruments are set up with a membrane separating a pressurized vessel and an evacuated vessel. The pressurized vessel contains the gases which are being separated. As the gases permeate through the membrane, the pressure in the evacuated vessel will increase. The rate of pressure increase permits a determination of the ability of the membrane to separate the gases.

At NETL, membrane development is guided by a thorough understanding of the fundamental membrane transport mechanisms as well as a thorough understanding of materials properties. However, even after carefully selecting promising materials, there are still an abundance of potential membrane material choices. To greatly increase the speed of screening materials for gas separation membranes, a high throughput membrane screening system was developed at NETL. This system is unique, and a patent application covering its design has been submitted. This system was developed by assembling in parallel multiple single cell constant pressure test systems. The instrument consists of 16 membrane cells and, therefore, can test up to 16 membranes simultaneously. The feed gas flows to the 16 cells such that the top side of all 16 membranes has a constant flow of gas passing over the top of the membrane. Similarly, the sweep gas is fed such that the bottom side of all 16 membranes has a constant flow of gas passing over the bottom of each membrane. Special selector valves are used to select one of the gas streams flowing away from the membrane for analysis. The selector valves are then rotated through all 16 cells so that each membrane is sampled. This process permits the testing and analysis of 16 membranes in one instrument.

NETL's membrane screening system accelerates membrane testing by a factor of 12.

To illustrate the value of this instrument, consider one type of membrane technology – mixed matrix membranes. These are traditional polymer membranes with filler particles embedded in them. If the particles have favorable gas transport properties, the overall performance of the membrane can be improved. For this type of membrane, there are many variables to be considered including polymer type, filler type, filler loading, and filler size. Simply by selecting one polymer and two fillers, while testing 5 loadings and 5 filler sizes, 50 membranes would need to be tested. Using standard testing equipment, 1 membrane takes 3 days to test. Using the high throughput system, 16 membranes take only 4 days to test. As a result, NETL has greatly enhanced its capability to solve important gas separation problems associated with clean energy production.

Computational Study of Ionic Liquids Illuminates Detailed CO₂ Interactions

Ionic liquids (ILs), which can be thought of as salts that are molten at room temperature, are being studied for use as part of CO₂ adsorption and/or separation technologies. These applications depend on having strong interactions between the CO₂ and the ions of the IL. In order for significant advances to occur in this area of research, the interaction between the CO₂ and each IL must be understood and described with accuracy. Computational methods are used to describe these interactions on a molecular level.

NETL scientist Jan Steckel has used a variety of methods to elucidate the complex nature of the interactions between CO₂ and acetate ion. The results of this study were published recently in the Journal of Physical Chemistry A. The acetate ion was chosen because it is representative of the anions used in many ILs currently under investigation as CO₂ sorbents or as part of a separation technology.

Dr. Steckel has shown that the acetate-CO₂ potential energy surface is very complex. Eight energy minima, representing the most stable configurations, were located and characterized using computational methods that apply first-principles molecular orbital calculations to obtain an accurate description of these interactions at the molecular level.

The most stable structure is denoted η2, (eta 2, pictured below on the left) where the CO₂ interacts with both oxygen atoms of the acetate. This complex structure is predicted to have a binding energy of -10.6 kcal/mol, a measure of stability. There are several other complexes with binding energies close to -8.5 kcal/mol, but of these, the η1-CT complex (eta 1, pictured below on the right) is unique. This complex is notable because the CO₂ is bent to about 140°, the C atoms of the CO₂ are only 1.54 Å away from the O of acetate, and there is evidence of charge being transferred from the acetate to the CO₂ upon complexation.

The η2 and η1-CT structures of the acetate-CO2 complex.

Using these interaction energies as benchmarks, it was possible to investigate the degree to which more affordable methods can describe these complexes. Unfortunately, many popular and affordable computational methods do not succeed in describing the η1-CT complex accurately.

This study helps to provide a clear understanding of the acetate-CO₂ interaction and supplies previously missing energetic and structural benchmark data. However, another important contribution made by this work is the revelation that widely-used but less accurate methods fail to accurately describe this interaction.