This thesis is a culmination of efforts to study and identify optoelectronic effects in cadmium telluride (CdTe)-based photovoltaic (PV) absorbers. Specifically, it elaborates on characterization of films alloyed with selenium (Se) in Chapter 3 or doped with antimony (Sb) in Chapter 4. The goal of this research is to investigate fundamental impacts of processing conditions on absorber properties from alloying and doping using non-contact optical probes. Polycrystalline thin film PV are desirable for their short time and low energy intensity to manufacture compared to single-crystal silicon PV. Thus far, CdTe-based absorbers are the only successful thin film technology on the global PV market, comprising ~4% of global module production. The power conversion efficiency from the detailed balance limit for CdTe, is ~31%, however CdTe-based technologies to date have only reached 23.1% efficiency. Despite successful commercialization, CdTe shows a large deficit in open circuit voltage (V_[OC]), which represents a viable direction for technological improvement. Research efforts to support CdTe manufacturing have improved efficiency to 23.1% at lab scale and ~19.5% at the commercial scale in the last 10 years. Improvements are largely the result of incorporation and optimization of: 1) selenium alloy gradients and 2) the transition from copper doping to arsenic doping. Both changes have been implemented commercially; however, the full extent of microstructural and electronic effects of the processing is not fully understood. Both Se-alloying and As-doping were critical to enhance performance. However, there are known problems and knowledge gaps with both Se and As in CdTe, and both can be correlated to harmful defect chemistry. Selenium alloying improved performance greatly, but it is induces defects in quantities that can limit V_[OC]. The identities of Se-related defects are not known precisely, and their behaviors have only recently been studied. Films containing selenium show also show native n-type conductivity, which reduces p-type doping ability and can result in compensation. Arsenic-doped CdSeTe are the current state-of-the-art devices; however, dopants show low activation, leading to inactive impurities. Inactive impurities can form new defects of varying impact, in the worst cases causing extreme electronic disorder, but in general reducing or limiting performance in devices. The exact contributions of each impurity or defect complex on performance limitations require detailed study. Optical spectroscopies are non-destructive and non-contact methods that can be applied to absorber films or full devices. Photoluminescence (PL) and Time Resolved Photoluminescence (TRPL), as well as Optical Pump Terahertz-Probe (OPTP) and terahertz Time Domain Scans (TDS) are valuable and complementary experimental that are used in this work to inform on band and defect structures, as well as carrier transport and recombination, with regard to processing. In general, developing a deeper understanding of how chemistry and processing affect thin film electronic behavior can help improve performance. Se-alloying appears to increase carrier lifetimes, yet estimates from TPRL do not match performance expectations from device models. In Chapter 3, we report the results extracted from fitting a photocarrier transport and recombination model to TRPL and OPTP transients. We quantify effects of Se concentration on carrier dynamics and recombination in CdSe_[x]Te_[1-x]. Bulk carrier and surface lifetimes are improved with Se alloying, however, carrier mobilities are reduced. We show that carrier trapping in shallow defects can complicate interpretations of carrier lifetime from TRPL decays. We show that a model incorporating trapping physics results in good description of dynamics of Se alloyed CdTe films. These results demonstrate the need for both careful parameter estimation for devices and controlled processing with Se. Additionally, using terahertz TDS analysis, we show that the incorporation of Se does not fundamentally change the carrier mobility, and there are instead other causes for mobility reduction. Incorporation of group V dopants in place of traditional copper led to a boost in device V_[OC] and efficiencies; however, low activation and defect chemistry of state-of-the-art dopants such as As or P can lead to unintended V_[OC] deficits. Sb was proposed for its size match with the Te-anion site and demonstrated high activation in CdTe. In Chapter 4, we show that Sb-doped films achieve high carrier concentrations (>2x10¹⁶ cm⁻³) in graded CdSeTe devices with aggressive CdCl₂, at the cost of inducing defect luminescence ~350 meV below the gap energy, which can limit V_[OC]. In samples with mild CdCl₂, Sb doping has no deleterious effects on carrier dynamics or emission spectrum in CdTe or CdSeTe up to 3x10¹⁷ cm⁻³. High dopant heater temperature during processing results in complex chemistry and transport physics in newly identified radiative defects. While it has yet to be demonstrated, this work suggests the combination of high doping and sufficient optoelectronic performance is possible. Ultimately, many material properties of relevance to device performance can be probed with optical spectroscopies, which can be used to enhance understanding of processing-structure-property-performance relationships. The methods used in this thesis are well suited for initial analyses and screening of experimental thin films as well as deeper correlation between processing variables and their potential impacts on device physics and performance.
Metrics
9 Record Views
Details
Title
Investigating carrier dynamics and recombination in cadmium telluride photovoltaics
Creators
Gregory Andrew Manoukian
Contributors
Jason B. Baxter (Advisor)
Awarding Institution
Drexel University
Degree Awarded
Doctor of Philosophy (Ph.D.)
Publisher
Drexel University
Number of pages
xv, 96 pages
Resource Type
Dissertation
Language
English
Academic Unit
Chemical (and Biological) Engineering [Historical]; College of Engineering (1970-2026); Drexel University