Thursday at the 7th World Conference on Photovoltaic Energy Conversion saw a packed schedule of technical sessions and even ‘Battle Royale’ a la Game of Thrones.

The day kicked off with a morning session of oral presentations spanning Areas 1, 2, 3, 4, 6, 8, and 9 along with a special session of the IEA PVPS Workshop. The session chairs provided the following highlights:

Area 1 - Quantum-well, Wire, and Dot-Architectured Devices
Zhu introduced QD solar cell for IBSC applications. They validated the absence of phonon bottleneck by EQE and EL spectral. Using the absolute electroluminescence imaging system, Zhu is able to calculate the radiative limit J-V curve to fit with the actual experimental data. Tomic presented on InAs/GaAs QD solar cells. LO phonon bottleneck and Auger recombination were investigated. Auger is faster and causing imbalance between different transitions, meaning a large impact. We can reduce Auger by introducing type-II alignment to put CB-VB radiative time on the same timescale as IB-CB one. He shows the calculation and STM result of small 30A quantum dots and open up to 85meV of the IB-CB gap. A QD SC with 19.7% of power conversion efficiency at 5X concentration is demonstrated. Instead of using higher Eg barrier, they use quantum engineered QD to achieve 840mV Voc. Khim Kharel from Univ. of Houston talked about the dilute nitride solar cells. Resonant thermo tunneling effect was discussed and applied in the design of devices. Extract carrier escape time for reference 6.3ns, and the RTT solar cell is 0.2ns, which increases the extraction efficiencies from 0.94 to 0.98. They observed the unusual plateau of carrier temperature and the presence of hot carrier effect in the QE devices. Huang - UNSW: Carrier extraction in CQDs. Promising materials but suffer recombination losses. Here, the role of the hole transport was assessed, by doping the PbS CQDs with silver to improve the p-layer. Optimized conditions produced a PCE of 10%. Kim - UCL: Effects of delta doping in InAs/GaAs QDs. Modulation doping shown to improve performance over direct doping of the QDs. Large voltage improvements observed, but also a loss of current extraction. This was attributed to a reduction in the depletion width across the intrinsic region with higher doping. Shervin - U. Houston: Discussed novel resonant tunneling device to improve/enhance carrier extraction in dilute nitrides QW solar cells. 24 period superlattice shows unusually large Voc, which is close to radiative limit and currently under further investigation.

Area 3 - Low-Cost III-V Epitaxy and Processing
John Simon (NREL) put together progress on GaAs subcells, GaInP subcells, and tunnel junctions to show a 23.7% dual-junction device grown at high growth rate by HVPE, with a clear path towards higher efficiency. Robin Lang (ISE) showed improvements to material quality and utilization efficiency at high growth rates using MOVPE, and achieved 23.3% for a GaAs cell at 100 micron/hr. Ryuji Oshima (AIST) discussed methods to control interfaces during transfer between HVPE reactors, and achieved 22.1% for a GaAs cell in a horizontal HVPE. Alessandro Cavalli (NREL) showed GaAs subcell performance on spalled Ge substrates and identifies "arrest lines", which are related to nonradiative recombination, and discussed methods to reduce these defects. Kaitlyn VanSant (NREL) showed results of 4-terminal stacked GaAs//Si tandems using HVPE for the III-V cell and achieved 29% efficiency with a route towards over 30%. Kenneth Schmieder (NRL) showed the potential of regrowth on transfer printed GaAs chiplets for low cost substrates.

Area 4 - Interface Studies and Surface Passivation
Simone Bernadini from ASU presented an excellent talk about the injection level dependence of the surface passivation provided by remote PECVD silicon nitride on both n-type and p-type c-Si. He showed that the results could only be reproduced by including a J02 component which was assigned to a surface damage region. Bert Stegemann from HTW Berlin presented a detailed fundamental study about the properties of silicon oxide films prepared by various techniques. He showed that the interface defect densities were quite different and that these densities scaled with the density of sub-oxides, indicating that abrupt interfaces are preferred for low interface defect densities. Andreas Teppe from RCT solutions presented their metal catalyzed chemical etching. They have developed a high-volume solution with a cost of a bit over 1 cent per wafer which can be used to etch diamond wire sawn multicrystalline silicon wafers with a 0.4% efficiency gain compared to conventionally acid textured slurry sawn wafers. Catherine Collett (University of Oxford) reported on ionic passivation to boost the field effect component. Two techniques were reported: spray coating of KCl and corona charging. She reported effective lifetimes well beyond 6 ms with the corona charging approach. She claims the method to deliver stable performance on year scale. Chang-Yeh Lee (UNSW) presented an interesting study of non-destructive technique to characterize the silicon oxide that forms at the interface between c-Si and ALD AlOx. Angle dependent FTIR technique proves very sensitive to density variation as function of annealing temperature. With the help of modelling and corona charging technique a correlation is established between Dit and density of the film. Andre' Augusto (ASU) has reported on the impact of substrate thickness on surface passivation. First, a modelling framework is established reporting that a < 3 ms effective lifetime for cell precursor is sufficient to obtain record devices (eta > 26%) within thicknesses below 100 um. A cell precursor endowed with front and rear TCO and with 40-um thick bulk was reported to exhibit Voc = 760 mV, bringing down the Woc value to 0.4. This parameter indicates the band gap utilization. This study explores the possibility to enter the market of III-V cells competing for ultra-thin c-Si solar cells.

Area 6 - Methods for Scale-Up of Perovskite Devices
The rapid progress achieved in the scaling of perovskite solar cells was on display in the Area 6 session. Valerio Zardetto of Solliance presented the exciting development of a 144 cm 2 slot dye coated module with over 20 V and 14.5% PCE and showed their rapid progress in scaling more stable device architectures. Benjia Dou touched on the issues found in transferring spin-coating recipes to blade-coating processing and showed that the redesign of the perovskite solvent chemistry allowed for fully blade-coated perovskite solar cells to achieve nearly 20% PCE and actually out-compete comparable spin-coated devices. Tobias Abzieher of KIT presented the promise of all-evaporated perovskite solar cells which combined approximately 15% performance with good thermal and UV-stability. Tobias also showed an exciting image of an all-evaporated perovskite solar cell on textured silicon. Kai Zhu of NREL showed how most technologies decrease ~0.8% per decade increase in area, but perovskite solar cells still lag behind this general trend, there’s lots more work to do! Kai showed that NRELs recently developed solvent chemistry with a wide processing range can yield uniform films over large areas and obtain similar grain morphology and optoelectronic characteristics between spin-coating, blade-coating, and slot-die coating. Bert Stegemann of the University of Applied Sciences, Berlin discussed excess laser energy in the laser ablation of perovskite module interconnects can form PbI 2 along the scribe edges and showed how reducing laser pulse duration from ns to ps dramatically reduces PbI 2 formation and in turn reduces the cell to module losses. Jonas Schwenzer from the Karlsruhe Institute of Technology presented how temperature variations can introduce new degradation challenges in perovskite solar cells which need to be addressed by changes in device architecture, perovskite composition, and perovskite deposition methods.

Area 9 - Soiling 1
Lin Simpson presented effects of cleaning methodologies and fungus grow on solar glass at 6 different locations, where wet cleaning induces more fungus. Christian Hagendorf talked about reduction of soiling up to 65 % when dew formation can be prevented, due to heating of glass samples in Qatar.

IEA PVPS Workshop Highlights
AEO 2007 projected 3 GW of PV deployed through 2030 vs 122 GW in AEO 2017. A lesson learned is that modelling results are getting better if more realistic and recent data are used (e.g. NREL’s Annual Technology Baseline (ATB) provides a transparent set of electricity generation technology cost and performance data). The lessons learned from the Sunshine programme in Japan led to better forecasts in the New Sunshine Programme and later roadmap exercises. Perception about future technology mixtures still influence the choice of models used for forecasting and their outcome.

Following the morning session and a coffee pit stop, the final poster session with work presented across Areas 1, 2, 4, 5, 7, and 9. The following submissions were recipients of poster awards in their respective Areas:

Area 1
#890 Quantifying Parasitic Losses from Metal Scattering Structures in Solar Cells: How Uncertainty in Optical Constants Affects Simulation Results
P Pearce, A Mellor, N Ekins-Daukes         

Area 2 
#928  Boosting the efficiency of kesterite Cu2ZnSnS4 solar cells by optimizing the heterojunction interface quality
K Sun, C Yan, J Huang,  et al         
Area 4 - I
#971 Deposition of AlOx and SiNx for high efficient p-type PERC+ solar cells using high-throughput inline PECVD
M.Koenig, T. Grosse, H.P. Sperlich, D. Landgraf, G. Koehler
Area 4-II
#978 Electrothermal Simulation of Si Cells with Defects
M Nardone, H Lee         

Area 5
#999 Transmission Electron Microscopy Study on Degradation Mechanism of CdTe Thin-Film Solar Cells
J Liu, S Johnston, S Harvery,  et al         
Area 9 
#1039 Mars Soiling Sensor™
M Gostein, S Faullin, K Miller,  et al         
Area 7
#1032 Potential Analysis of a Rear-Side Passivation for Multi-Junction Space Solar Cells based on Germanium Substrates
C Weiss, J Schön, O Höhn,  et al         
During the lunch break, it was time for the Women in PV luncheon, where Sheila Bailey was the invited speaker. A long term advocate for women in STEM careers, Dr Bailey’s career in space photovoltaics and all of the interesting adventures and people that have made up that career give an inspiring message. The PV community is justifiably proud of the opportunities for women in the area and Sheila is a shining example of what is possible. A captivating speech from a major contributor to PV over more than three decades at an event that shows the continuing commitment to continuing and increasing the participation of women in PV. Sheila’s anecdotes and personal stories connected with and inspired the audience. 
Following lunch, attendees returned for a further two afternoon technical sessions. In the first set, from 1.30 to 3.00 pm in Areas 1, 4, 5, 6, 7, and 9, with the following highlights provided by the session chairs:

Area 6 - Examining Perovskite Stability
Rongrong Cheacharoen from Stanford presented encouraging results of perovskite solar cell stability, showing encapsulated perovskite solar cells passed the IEC standard tests (damp heat and thermal cycle test etc.). The mechanical strength of perovskite solar cell, which is another important aspect of stability, was well discussed by Nick Rolston from Stanford. Rongrong Cheacharoen presented promising results for perovskite stability, passing 1000 hrs of damp heat, 250 thermal cycles, and UV irradiation by using industry standard encapsulation. Professor Shengzhong Liu presented 21.5% efficient planar perovskite solar cells with organic molecule treatment of SnO2. The Duong showed how standard stability testing may be insufficient and specifically, metal induced degradation can be accelerated under heat and light. However, metal induced degradation can be mitigated by using transparent conducting oxides. Nicholas Rolston showed how mesoporous scaffolds can significantly improve the fracture energy of perovskite solar cells. Valerio Zardetto showed minimal loss in performance after >2500 hours at 85C with the introduction of a ZnO barrier layer deposited by spatial ALD.

Area 7 - Space PV Systems
Brian Spence of Deployable Space Systems gave a presentation on the recent successes with ROSA, or roll-out solar array, including the first ever flight experiment of ROSA on the International Space Station (ISS). ROSA is a boom/flexible blanket array geared towards improving W/kg metrics. Bao Hoang of SSL presented ROSA from a different perspective as SSL show the benefits of ROSA at the system level. ROSA can enable on orbit servicing, persistent platforms and high power interplanetary missions. SSL is presently qualifying a large scale ROSA product. John Gibb from Lockheed Martin gave an overview of "flex" solar arrays of the past, present, and the future. He announced the MMSA, multi- mission modular solar array that will be flown later this year, 2018. Christian Ruud of Penn State University makes the case for a novel space CPV design that leverages transfer printed microcells. These microcells are lightweight, very high efficiency, and when integrated into their molded reflective mirror array CPV design can lead to a module efficiency up to 32%, and 350 W/kg. Dave Scheimann of the Naval Research lab presented work on their Solar UAV program. The UAV needs about 70 watts to cruise and they have compared several wings populated with several different cell designs. Solar power has been shown that it can double flight duration of UAVs. Raymond Chan of Microlink Devices gave a talk detailing their work related to lightweight flexible sheets for solar flight. HALE, or High Altitude Long Endurance UAVs are a new class of aircraft that are able to stay aloft for weeks or potential months at a time. MLDs high specific power enable these aircraft to remain in flight. MLDs IMM solar sheets are currently operating at 1500 W/kg.

Area 1 - Quantum-well, Wire, and Dot-Architectured Devices / Advanced Light Management and Spectral Shaping
Stephen Bremner presented the effects of Sb in AlAs capped QDs that improves the transport for carriers photogenerated in the bulk. S.Asahi presented progress on improvements in the efficiency of 2-step photoabsorption in a heterojunction device. R.Vismara presented work on how decoupled pyramid textures deliver absorption enhancement beyond the Lambertian limit and simulation results presented on BaSi2 as well as nc-Si:H. S. Collin showed new calculations on multi-resonant absorption that can increase absorption by 4π n 2 higher than the traditional 4n 2 for more standard structures. Jose Ulloa working on type II strain-balance superlattices demonstrated that superlattice periods must be less than 6nm to obtain a 120% conversion efficiency enhancement over bulk material in GaAsN/GaAsSb devices. Haley Bauser presented the design of a high contrast metasurfaces (AlSb on glass array) to be included in tandem (colloidal quantum dots/InGaP and colloidal quantum dots/Si calculations give eff 24%) luminescent concentrator.

Area 5 - Module/Cell Characterization 2
Two excellent talks using statistical methods to extract influence of process variables in manufacturing for process optimization, by measuring large numbers of cells in process. Two excellent talks using imaging methods to extract physical quantities: IR imaging of water vapor transport; imaging of stress in Si wafers for determining effect of process variables on stress.

Area 4 - Passivated Contacts, Carrier Selective Contacts: Fundamentals
F. Feldmann presented convincing results showing that MIS theory can well explain some of the high performance passivated contacts, but also show that contacts including holes in the SiOx layer can perform well. Not all good contacts work the same! A. Morales demonstrated 2x2 cm 2 heterojunction cells with ZnO front and ZnO/Ag with 23.02% efficiency, an excellent result when using no Indium containing TCO. M.Koehler demonstrated a promising front heterojunction contact based on tunnel SiOx and CAT/CVD Sic. The QEE shows impressive current gain in the blue and implied Voc of 728 mV. ITO sputtering damages prevented good device fabrication. A. Kahle discussed in a detailed investigation of possible pinholes, in particular using EBIC measurements, he showed how thinned oxide location can lead to strong local current flow.1062. T. de Vrijer discussed the degradation in thin film silicon solar cells and pointed out that the degradation is caused by p-SiOx:H degradation. J. Bullock presented on passivated contacts based on TiOx/LiFx/Al applied locally on the rear-side with efficiencies exceeding 23%.

Area 9 - Adhesion and Corrosion
Nick Bosco, presenting on behalf of J. Tracy, reports on work to build a model for encapsulant to cell, and encapsulant to glass adhesion, using a combination of proposed chemical degradation mechanisms. In a second presentation, Bosco showed that temperature and RH during an extended adhesion test impacted the measured adhesion values of EVA/glass. Xin Hi described Finite Element Analysis using a cohesive zone model to simulate mixed mode delamination of EVA to glass, which showed the delamination mode shifts with temperature. Tadanori Tanahashi described AC impedance measurement technique used to localize source of degradation between the front contact and Si-emitter using acetic acid vapor. Yuji Ino presented an accelerated method using "pressure cooker" and extended damp heat tests to study the influence of Pb and non-PB solder and various solder fluxes on EL darkening that were responsible for approximately 20% series resistance increase. To close the session, Michael Kempe discussed using electric leakage currents to understand electrochemical and ionic species related corrosion.

After a quick re-charge the final oral technical sessions for the day covering work in Areas 2, 3, 4, 5, 8, 9, and 10 took place. Some of the highlights provided by the session chairs are given below:

Area 10 - Solar Variability and Hosting Capacity
Andrew Blakers presented on how off river pumped storage sites offer the option to move to 100% renewable power system over the next 15 years. An analysis of possible storage sites worldwide revealed a vast number of available sites which could be used. The associated costs of balancing PV and wind generation with such technology including HVDC transmission would add about 20 USD/MWh.

Area 3 - Hybrid Tandems: Battle Royale, Part 2
Tyler Grassman from Ohio State presented results on the development of a GaAsP/Si epitaxially grown tandem cell with a certified efficiency of 20%. Florent Sahli from EPFL presented work on monolithically integrated perovskite/Si tandem cells using fully textured Si cells, with a certified efficiency of over 25% and high quantum efficiency for the Si subcell. Markus Feifel from Fraunhofer ISE presented the former record (only broken 30 minutes earlier) for an epitaxially grown III-V on Si device with an efficiency of 19.7% from a GaInP/GaAs/Si multijunction cell. Subsequent growth improvements have enabled nucleation of GaP on Si with a threading dislocation density of 10^5 cm^-2, which has great potential to improve the efficiency of future devices. Valerio Zardetto showed the progress that Solliance has achieved in 4-terminal integration of high-transparency perovskite top cells with IBC and HIT Silicon base cells, reaching up to 26.3% efficiency. Adele Tamboli from NREL presented a transparent conductive bonding method to optically and electrically couple III-V top cells with IBC Si tandems. GaInP/Si cells reached 27.3% efficiency in 3T operation. Manoj Jaysankar from IMEC presented 26% and 23% efficient 4T tandems on Si and CIGS base cells respectively (0.13cm2 aperture area) and 20% from PVSK/Si at 16 cm 2 aperture area.

Area 2 - Industrial Aspects
Gang Xiong First Solar Invited presentation was a well presented, balanced presentation providing commercial status as well as technology and research direction for improvement of performance. Thrust is on Voc improvement with group V doping optimization. He underscored the move away from using Copper. Yoshiaki Hirai Solar Frontier: 22.9% efficiency CIGSSe Solar Cell. The speaker discussed how bandgap engineering and a treatment based on Cs lead to new record CIGSeS efficiency of 22.9%. Cs alkali addition, enhanced sulfurization and modified absorber contribute to the new record. Phillip Kratzert Solibro Hi-Tech: New world record for full size CIGS module18.7 % Aperture area efficiency full size module, new world record for thin film. Explained contributing factors to this record as optimized CIGS, Rb alkali addition and optimized metallization grid. Jinwoo Lee Global Solar Energy: Na interface accumulation in CIGS. The speaker discussed a loss mechanism (up to 2% in efficiency) in CIGS modules related to Na humps observed in CIGS films. The Na humps correlated with interfaces of the 3-stage process. Veronique Gevaerts Solliance (ECN and TNO) Backend Monolithic interconnect for electroplated CIGS. The speaker described a novel back end technique for laser scribing combined with inkjet printing for the monolithic interconnects.

Area 4 - Device Modeling and Power Loss Analysis
Andreas Fell gave an excellent comprehensive overview of the capabilities of an MIS (metal-insulator-semiconductor) model to describe passivated and even HJT contacts. This model is now included in Quokka3 for everyone to use. Olindo Isabella presented his implementation of passivated contacts in Synopsys Sentaurus and gave a detailed analysis of the behavior of different contact realizations in order to optimize devices and estimate their efficiency potential. In Sebastian Meier's comparison of the potential of bifacial p-PERT and PERC there was no clear winner. Only when turning to bifacial cells, p-PERT gained a significant advantage over PERC. Julian Weber used meta model device simulations to optimize the design of laser doped selective emitters.

Area 8 - PV Modules: Back Contact cells and Partial Shading
Two papers (by Ian Bennett from DSM and Bonna Newman from ECM) discussed the development of a new integrated conductive backsheet for back contact cells. Newman showed the application of this new capability in combination with integrated light management thinner cells to achieve a 15% module cost reduction.

That ended our technical program for Thursday. It was then time to get out your most tropical outfit and get along to the Luau for the Conference Banquet. After piling up their plates with delicious food attendees were treated to some traditional dancing from the South Pacific including Samoa and Hawai’i. Along with a thrilling fire dance there was also the challenge taken up by the conference Chair Alex Freundlich. We were even lucky to once again experience a fly-over of the International Space Station during our banquet. There was even the celebration of a sixtieth wedding anniversary, at the exact same spot! A great time was had by all as the conversations and fun flowed into the night.