Fabrication and study of organic and inorganic optoelectronics using a vapor phase deposition (VPD)

The majority of my research focuses on fabrication and characterization of organic and inorganic optoelectronics using a vapor phase deposition (VPD). The VPD, formerly called organic vapor phase deposition (OVPD), was developed to allow sublimation and deposition of organic, inorganic and metal molecules that exhibit high sublimation temperatures, ≤1000℃. Conversely, the OVPD was specifically developed to allow sublimation and deposition of organic‐inorganic compounds that display sublimation temperatures below 400℃. Chapter 1 provides an introduction to the original OVPD method and its accomplishments fabricating organic light emitting diodes (OLEDs), organic photovoltaics (OPVs) and organic thin film transistors (OTFTs). Further, chapter 1 also discusses how the OVPD evolved into the VPD, specifically to enable deposition of thin metal films for optoelectronic devices (discussed in chapter 3). ❧ In chapter 2 we present analysis and characterization of a novel molecule to fabricate OPVs using the OVPD. Specifically, we measured the exciton diffusion length (LD) of platinum tetra 1, 3‐di‐tert‐butylphenyl tetrabenzoporphyrin Pt(ᵗᵇᵘTPBP), by use of the spectrally resolved photoluminescence quenching (SR‐PLQ) method and the optical electric field distribution technique for optimization of OPVs. SR‐PLQ measurements were performed under vacuum (0.08torr), where an average LD of 12.1nm ± 2.4nm was found for Pt(ᵗᵇᵘTPBP). Concurrently, analysis of the films under nitrogen was also performed. However, the LD varied significantly with values ranging from 10nm to 120nm. Significant emission spectra intensity fluctuations were observed during the measurements under nitrogen suggesting oxygen quenching of the porphyrin. Clearly, oxygen quenching altered the LD thus revealing a weakness of the SR‐PLQ method. Likewise, using a program written in MATLAB language for the optical electric field distribution technique, an LD of 16.9nm ± 4nm and 14.8nm ± 3.5nm was obtained for Pt(ᵗᵇᵘTPBP) and C₆₀, respectively. Further, modeling of the optical electric field, photocurrent generation, exciton diffusion profile density, absorption, reflection, among other was also completed using the optical electric field distribution method. Both techniques proved to be great tools for the design and optimization of OPVs. Photovoltaic cells were fabricated in the OVPD and in a vacuum thermal evaporation (VTE) (reference). Fabricated OPVs had the following architecture: ITO/Pt(ᵗᵇᵘTPBP)¹⁰⁻⁶⁰ⁿᵐ/C₆₀¹⁰⁻⁴⁰ⁿᵐ/BCP¹⁰ⁿᵐ/Al¹⁰⁰ⁿᵐ. Open circuit voltages (Voc) in the order of 0.5V and 0.65V were obtained for devices made in the OVPD and VTE, respectively. The results suggest intermixing of the donor/acceptor layers during deposition of C₆₀ in the OVPD devices, thus leading to strong intermolecular interactions. ❧ Further, intermixing is common when films are annealed and strong intermolecular interactions are known to increase dark current, consequently, decreasing the Voc. The efficiencies obtained for the Pt(ᵗᵇᵘTPBP) devices were found to be 0.6% and 1.2% for the OVPD and VTE, respectively. The efficiency discrepancy is due to poor current contribution from the C₆₀ layer (10nm for OVPD vs 40nm for VTE). The external quantum efficiencies (EQE) from the OVPD devices clearly show weak contribution from the acceptor (C₆₀), whereas a stronger current contribution from C₆₀ can be observed on the VTE devices. The results suggest efficiency can be improved for devices made in the OVPD if a thicker C₆₀ film is employed. Similarly, the Voc can also be improved with proper substrate cooling to prevent intermixing of the donor/acceptor layers. Further, a similar Jsc trend was observed on the Pt(ᵗᵇᵘTPBP) versus the Pt(TPBP) devices, confirming a short LD in both molecules is accountable for the decrease in current when a thicker donor film is employed. ❧ In chapter 3 we introduce fully fabricated OPVs and OLEDs with the VPD, an improved version of the OVPD. The objective of the improved/modified VPD was to enable deposition of compounds with high sublimation temperatures e.g., metals and inorganics such as calcium, zinc, cadmium, magnesium, antimony, bismuth, indium, silver, aluminum, zinc sulfide, and manganese, among others. Of the above metals, magnesium (Mg) and zinc (Zn) were selected to fabricate metal films and devices, as they have sublimation enthalpies that are comparable to organic materials commonly used for OLEDs and OPVs. Magnesium and zinc films were deposited in the VPD to prepare optoelectronic devices under low vacuum conditions, i.e. 1 torr. The thin metal films, analyzed via scanning electron microscope (SEM), atomic force microscopy (AFM), X‐ray diffraction (XRD) and four‐point probe resistivity measurements, revealed comparable characteristics to metal films deposited in a VTE. Magnesium cathodes were fabricated for OLEDs and OPVs. OLEDs were fully made in either the VPD or VTE employing aluminum tris‐(8 hydroxyquinoline) [Alq₃] as the green fluorescent emitter or fac‐tris(2‐phenylpyridine)iridium [Ir(ppy)₃] as the green emitting phosphor. Analysis of the OLED devices made in the VPD showed external quantum efficiencies (EQE = 0.9 ±0.1%) and (EQE = 7.6 ±0.6%) at a luminance of 100 cd/m² for the fluorescent and phosphorescent devices, respectively. In addition, OPVs were fully fabricated by both methods employing copper phthalocyanine (CuPc) and C₆₀ as the donor/acceptor materials. Analysis of the OPV devices made in the VPD showed a power efficiency of 0.5 ±0.1%, an open circuit voltage of 0.45 ±0.05% and a fill factor of 0.50 ±0.05%. ❧ The final chapter introduces what we thought was the next logical subject, after studying organics (chapter 2) and metals (chapter 3), inorganic compounds. Chapter 4 presents hybrid organic‐inorganic lead based solar cells grown with the VPD, specifically, perovskite films i.e., methyl ammonium lead iodide (MAPbI₃) as the active layer. Perovskite films were fabricated with both the VPD and VTE and characterized via scanning electron microscopy (SEM), atomic force microscopy (AFM), X‐ray diffraction (XRD), ultraviolet‐visible spectroscopy (Uv‐Vis), X‐ray photoelectron spectroscopy (XPS), profilometry, spectroscopic ellipsometry, and inductively coupled plasma optical emission spectrometry (ICP‐OES). The films were found to be analogous to perovskite films previously reported in literature. One‐step and two‐step deposition methods were used to fabricate solar cells in the VPD. For the two‐step method, efficiencies of 3.3 ± 0.3% and 4.1 ± 0.4% and Voc of 0.6 ± 0.1V and 0.8 ± 0.1V were obtained for ITO/MAPbI₃(330nm)/C₆₀(40nm)/BCP(10nm)/Ag and ITO/MAPbI₃(330nm)/NPD‐MoO₃(140nm @ 50% doping concentration)/Ag, respectively. Further, we found that having excess of PbI₂ and/or CH₃NH₃I in the perovskite films considerably decreased device performance, i.e., current (Jsc) and fill factor (FF) were impacted greatly. Conversely, fully converted perovskite films proved to be critical to achieve diode like rectifying IV curves. Additionally, annealing also had a great impact on device performance were a decreased on Jsc was observed on films deposited on ITO electrodes, i.e., currents of 9.2 ± 0.6mA/cm² and 2.1 ± 0.4mA/cm² were observed for annealed and non‐annealed devices, respectively. Contrariwise, annealing marginally impacted films deposited on FTO/TiO₂ thus suggesting that a much rougher TiO₂ surface promotes crystallization of the perovskite film while a flatter electrode, i.e., ITO, is not conducive for film crystallization. ❧ However, regardless of the architecture used, we found that the devices fabricated in the VPD were unreliable and highly unreproducible, displaying poor‐rectifying IV curves and noticeably weaker Jsc and/or EQE response from the devices. Several possible explanations as to why this strange behavior happened were discussed. First, perovskite films were exposed to ambient air for its completion so it is conceivable that ambient air adversely affected the perovskite films, i.e., air and water diffused through the film creating problems at the ITO/perovskite interface. This possibility is validated with devices made in the VTE, which exhibited acceptable rectifying IV curves and comparable EQE and Jsc values. Second, unreacted (excess) PbI₂ and/or MAI influenced solar cells performance, delivering poor‐rectifying IV curves and noticeable discrepancies between Jsc and EQE values, whereas having a fully converted perovskite film lead to acceptable rectifying IV curves and comparable EQE and Jsc values. Third, it is also possible the perovskites films made in the VPD are not entirely crystalline, i.e., a blend of amorphous and crystalline phase exists in the film. The latter is less probable since amorphous materials tend to scatter X‐rays in many directions, thus leading to bumps on the collected spectra. Whereas, the periodicity of crystalline materials scatters X‐rays only in certain directions, consequently, leading to sharp peaks that are similar to those obtained in the collected XRD spectra. Regrettably, while it is most likely that perovskite film damage at the interfaces is what caused the previously mentioned problems, currently there are no methods/systems that can help us verify our theory.