ABSTRACT
This thesis focuses on developing earth abundant materials and compounds for solar energy applications. Thin films were grown and studied in search of improved p-type trans- parent contact layers in photovoltaic devices. Combinatorial co-sputtering was used to make thin films with compositional and temperature gradients across glass substrates. Materials were characterized using x-ray diffraction, x-ray fluorescence, Rutherford backscattering spectrometry, Kelvin probe measurements, four point probe measurements, Seebeck measurements, and optical transmission & reflection measurements.
The pseudo-ternary oxide system of Zn-Ni-Co-O was deposited on ambient temperature glass substrates to serve as a p-type selective contact. The properties of the ambient- temperature-deposited Zn-Ni-Co-O thin films were compared to films grown at higher substrate temperatures (350◦C) with the same compositions. It was found that the optical absorption and work function of the materials were similar in both cases.The optical absorption at 1.8 eV was large ( ∼105 cm-1) and the work functions were in the range of 5.1 to 5.8 eV.
However, the conductivity and crystallinity of the ambient-temperature-grown samples was poorer than the high temperature samples. The maximum conductivity was found to be 35 S/cm and corresponded to the region with the most crystalline disorder. A minimum grain size for the spinel films and disordered films were both estimated to be near 5 nm based on grain size analysis and neglecting strain effects. Overall, a factor of three decrease in the conductivity was the main difference observed when Zn-Ni-Co-O films were deposited on ambient temperature substrates when compared to 350◦C depositions.
A growth technique for sulfide and oxide-sulfide thin films is introduced. An RF solids single atom source (cracker) is used in conjunction with co-sputtering to deposit both sulfide and oxide sulfide compounds. Reactive sputtering from metallic and ceramic targets is demonstrated for binary sulfide growth. A range of oxygen to sulfur ratios is obtained by growing BiOxSy thin films to demonstrate the tunability of sulfur content using this technique. Lastly, independent tuning of two cation species and two anion species is demonstrated by the growth of the quaternary compound BiCuOS.
The properties of BiCuOS thin films are measured and compared to reported values for bulk crystals. The electrical conductivity and Seebeck coefficient are measured to show the electrical properties of the material and majority carrier type. The electrical conductivity was determined to be 0.059 +/- 0.005 S/cm and the Seebeck coefficient was positive indicating a p-type material. The optical transmission and reflection is used to determine the absorption and band gap. Tauc plots showed an indirect band gap at 1.03 +/- 0.01 eV with a first direct transition at 1.34 +/- 0.02 eV. The properties of the BiCuOS thin films are very similar to the bulk crystals further indicating successful growth of this material.
Bi-Cu-O-S films were grown using combinatorial sputtering and the properties were measured as the composition was varied. The films remained in the BiCuOS crystal structure as both the copper and bismuth contents were each individually increased in the films. The electrical conductivity increased as excess copper was added into the films and the conductivity decreased as excess bismuth was added. Seebeck measurements indicated that the materials remained p-type as the composition was varied.
Tauc plots indicated that the energy transitions remained relatively constant as both excess bismuth or excess copper were added into the material. These results provided a basis for a combinatorial study of the Bi-Cu-O-S material system that can be expanded with future experiments. These results also provided the foundation for combinatorial studies of oxide-sulfide compounds in search of p-type transparent contacts. The study of LaCuOS and La2O2SnS3 can now be attempted using these techniques and the insight gained from this study.
MOTIVATION TO STUDY AMBIENT TEMPERATURE DEPOSITED Zn-Ni-Co-O THIN FILMS
Figure 2.1: A figure showing donor and acceptor energy levels with respect to the energy bands for the four different doping types.
Figure 2.1: A figure showing donor and acceptor energy levels with respect to the energy bands for the four different doping types. The donor level is shown in blue and the acceptor level is shown in red. The image shows compensated (DT 1) with donor level above acceptor level and uncompensated (DT 2-4) with acceptor level above donor level. a) Doping type 1. b) Doping type 2. c) Doping type 3. d) Doping type 4. Figure adopted from Paudel et al.
Figure 2.3: Crystal field splitting for a d6 transition metal in an octahedral oxide environment.
The effective mass was large to the non-bonding nature of the states comprising the valence band maximum. A simplistic diagram of the crystal field splitting for a d6 transition metal in an octahedral oxide environment is shown in Fig. 2.3. The t2g level composes the valence band maximum but does not hybridize with the oxygen p level.
Zn-Ni-Co-O THIN FILMS DEPOSITED AT AMBIENT TEMPERATURE
Figure 3.3: XRD patterns from the ZnO-Co3O4 tie-line showing example patterns from the phases and mixed phases present along that tie-line
Figure 3.3: XRD patterns from the ZnO-Co3O4 tie-line showing example patterns from the phases and mixed phases present along that tie-line. The example XRD patterns are taken from the region outlined by an ellipse in the ternary diagram. Simulated reference patterns for Co3O4, NiO and ZnO are given at the bottom of the XRD patterns.
Figure 3.16: Semilog plot of the conductivity versus the absorption coefficient for the films.
Figure 3.16: Semilog plot of the conductivity versus the absorption coefficient for the films in the ambient temperature deposited Zn-Ni-Co-O material system. The conductivity is given on a logarithmic scale and the absorption coefficient for photon energies of 1.8 eV is given on a linear scale.
Figure 3.15: Optical absorption coefficient plotted against cation composition along the binary tie-lines from the ternary phase diagram.
Figure 3.15: Optical absorption coefficient plotted against cation composition along the binary tie-lines from the ternary phase diagram. a) Along the ZnO-Co3O4 tie-line. b) Along the NiO-Co3O4 tie-line. The blue lines in each figure represent the optical transitions or band gaps of the compounds and are labeled at the top of each line.
MOTIVATION TO STUDY OXIDE-SULPHIDE THIN FILMS
Figure 4.3: An image illustrating hybridization between transition metal cations and chalcogenide anions.
Figure 4.3: An image illustrating hybridization between transition metal cations and chalcogenide anions. The origin of the energy axis is set to the vacuum level. Adapted from calculation by Stephan Lany and Alex Zunger.
Figure 4.4: A plot of the LDA calculated band gap versus the calculated density of states effective mass for the candidate oxide-sulphide compounds.
Figure 4.4: A plot of the LDA calculated band gap versus the calculated density of states effective mass for the candidate oxide-sulphide compounds. The marker shape signifies if the anion atoms are segregated into layers or mixed throughout the compound. Calculations conducted by Kanber Lam and Giancarlo Trimarchi, plotted by Josh Ford.
OXIDE-SULHIDE THIN FILM GROWTH
Figure 5.2: Measured XRD pattern for the thin and thick Cu2S films grown from a metallic Cu target.
Figure 5.2: Measured XRD pattern for the thin and thick Cu2S films grown from a metallic Cu target. The thin (black) XRD pattern is for the thin sample that was used for RBS measurements of the film composition. The thick (grey) XRD pattern is for the sample with deposition conditions more suited for Cu2S phase formation. The colored patterns beneath the measured patterns are simulated reference patterns from the ICSD database for two common copper sulphide phases. ‘Thick’ sample grown and characterized by Adam Welch.
Figure 5.5: Oxygen to sulfur ratio plotted versus the sputter gun power and the k-cell temperature for the BiOxSy thin lms.
The carrier gas ow rate was 8 sccm through the RF solids cracker and the power on the RF coil was between 80 and 100 W. Nine samples were grown with different sulfur k-cell temperatures and sputter gun powers. RBS spectra were measured on each of the samples and analyzed to yield the composition of each film. The results are displayed as S/(O+S) for different combinations of sulfur uxes and gun powers in Fig. 5.5.
BiCuOS PROPERTIES
Figure 6.1: RBS spectra for the BiCuOS thin film sample.
Figure 6.1: RBS spectra for the BiCuOS thin film sample. Inset: the two low energy peaks of the spectrum on a smaller y axis to show the oxygen and sulfur peaks in more detail. The red lines indicate the high energy edge of the peak for each element.
Figure 6.6: XRD patterns for the films with excess bismuth and copper.
Figure 6.6: XRD patterns for the films with excess bismuth and copper along with a calculated reference pattern from the ICSD database. The copper cation percentage (Cu/(Bi+Cu)) is listed in blue for each measured XRD pattern.
Figure 6.10: Tauc plots used to determine the optical transitions in the Bi-Cu-O-S thin films.
Figure 6.10: Tauc plots used to determine the optical transitions in the Bi-Cu-O-S thin films. a) Indirect transition. b) and c) Direct transitions. d) Energy levels versus copper cation percentage for the three energy transitions determined using the Tauc plots.
CONCLUSION
he work in this thesis has served to explore earth abundant material systems that might be usable in thin film devices for solar energy harvesting. Through working as a member in the Center for Inverse Design (CID), a few material systems were chosen that have potential for application as specific component layers in photovoltaic devices. The material systems selected for research were Zn-Ni-Co-O thin films as well as oxide-sulfide compounds. The oxide-sulfide project focused on the growth of Bi2O2SnS3 and BiCuOS as initial research before growing the CID target compounds La2O2SnS3 and LaCuOS. Much of the research was spent on synthesizing the compounds and verifying the crystal structure and composition of the films.
The Zn-Ni-Co-O pseudo-ternary oxide system and the BiCuOS material system were both studied in pursuit of p-type contact layers in photovoltaic devices. The Zn-Ni-Co-O phase space was studied as more disorder was introduced through low temperature deposition and BiCuOS was studied as an analog compound to LaCuOS. The optoelectronic properties of both of these materials systems were studied as the composition was varied. A deposition technique was also developed in order to provide in-situ control over the sulfur content in sulfide and oxide-sulfide thin films. The use of this technique and the benefits that it provides are shown through the growth of sulfide and oxide-sulfide materials.
The study of Zn-Ni-Co-O thin films deposited on ambient temperature substrates found that the material properties are similar, albeit less desirable, to films grown at higher substrate temperature. The study was conducted to address the question of how the material properties of the system change when the substrate temperature is reduced to room temperature and disorder was added into the material. In order to make the comparison, data from a previously published study with high substrate temperatures (350◦C) was compared to the results obtained on ambient temperature substrates. All other deposition conditions besides the substrate temperature were the same providing a good comparison for the materials.
Source: University of Colorado
Authors: Joshua Cody Ford