Symposium F.EN02—Silicon for Photovoltaics

Thin films, based on amorphous and microcrystalline silicon or silicon oxides, are the key players in the future silicon photovoltaics. They very well passivate the surface of crystalline silicon, enabling high open circuit voltages and fill factors of finalized cells. Two most promising strategies are Silicon Heterojunction solar cell (SHJ) and Tunel Oxide Passivated Contacts (TOPCon) based cells.
Interdigitated back contacted SHJ solar cell (IBC SHJ), is the current silicon efficiency world record holder with 26.7%. The record IBC SHJ cell is fabricated by many photolithography steps, not fully compatible with mass production. Thus straightforward shadow masks technology has been developed. In order to check the profiles of ~10 nm thin passivating/selective contacts prepared by shadow mask process on rough substrates, Raman spectroscopy based thickness measurement with resolution bellow 0.2 nm was introduced [1]. Alternative, industrially relevant way, to image the silicon film thickness by photoluminescence imaging will be introduced.
In the same time it is important to probe the structural and local electrical properties of these films. For this end conductive atomic force microscope (C-AFM) in the tomography regime is suitable for characterisation of mixed phase amorphous/nanocrystalline ultrathin films as well as the tunnelling oxide contacts. 3D visualisations of high temperature passivated back contacts (TOPCon) will be used as an example.
But since the silicon wafer based technology approaches the practical efficiency limit (the theoretical is at ~29%), the only way how to go beyond 30% is combination of two cells in a so called tandem. By doing that the theoretical limit is shifted to over 40%. Suitable candidates for the top cell accompanying well mastered silicon bottom cell are halide perovskites. We will discuss the extraordinary properties of this material family [2]. Finally, we introduce a simple semiconductor property dictating the maximal open circuit voltage obtainable with the given material [3]. In this way we can rapidly preselect suitable candidates for top tandem cell.

References
[1] M. Ledinský, B. Paviet-Salomon, A. Vetushka, J. Geissbühler, A. Tomasi, M. Despeisse, S. De Wolf, C. Ballif and A. Fejfar: Sci. Rep. 6, 37859 (2016).
[2] De Wolf, S.; Holovsky, J.; Moon, S.-J.; Loeper, P.; Niesen, B.; Ledinsky, M.; Haug, F.-J.; Yum, J.-H.; Ballif, C. J. Phys. Chem. Lett., 5, 1035–1039 (2014).
[3] Ledinsky, M.; Schönfeldová, T.; Holovský, J.; Aydin, E.; Hájková, Z.; Landová, L.; Neyková, N.; Fejfar, A.; De Wolf, S. Temperature Dependence of the Urbach Energy in Lead Iodide Perovskites. J. Phys. Chem. Lett., 10, 1368–1373 (2019).

Acknowledgments
This work was supported by the Czech Science Foundation Project No. 17-26041Y, Operational Programme Research, Development, and Education financed by the European Structural and Investment Funds and the Czech Ministry of Education, Youth and Sports (Project No. CZ.02.1.01/0.0/0.0/16_019/0000760 - SOLID21). This work is also supported by Operational Programme Research, Development and Education financed by EU Structural and Investment Funds and the MEYS CR Project No. CZ.02.1.01/0.0/0.0/16_026/0008382 (CARAT).

Light-beam induced current (LBIC) measurements with a bias voltage or current are used to obtain photogenerated charge collection profiles at different points in the I-V curve of a solar cell. Such measurements can be analysed to obtain local electrical parameters such as dark saturation current and series/shunt resistance and identify aspects for improving solar cell performance.

In the present work, such an analysis technique was developed for solar cells made from 14 µm liquid-phase-crystallized silicon. In particular, it compares a reference cell with the current best cell exhibiting 657 mV open-circuit voltage (V_OC) and 32.7 mA/cm² short-circuit current density (J_SC) (with an anti-reflection foil on the illumination side). Since the cell architecture used is ‘silicon heterojunction interdigitated back contact’ (SHJ-IBC), the implemented circuit model is designed for an IBC architecture. The circuit is fit to the measured data on a MATLAB script, which will be available for downloading and adapting to other solar cell architectures and materials. The LBIC setup was built in-house at the University of Ljubljana. It uses an alternating current (AC) measurement with a direct current (DC) bias and produces full cell images at different bias points. A 40 µm step size was used. The setup offers a minimum step size of 1 µm and an illumination spot diameter <= 10 µm.

Each pixel is treated as an independent one-diode model, surrounded by 2 one-diode model representing the remaining (dark) cell. With this approximation, full cell maps of the following parameters of a one diode model were obtained: dark saturation current (I0), ideality factor (n), effective series resistance, shunt resistance and photo-generated current. While treating each pixel as an independent, illuminated solar cell, local I-V curves were simulated, giving full cell maps of V_OC, J_SC, maximum power points (V_MPP and I_MPP) and efficiency (η). The following key observations were made:

1. Top 50 percentile of 40 µm x 40 µm pixels show a J_SC between 32-33.5 mA/cm², V_OC between 600-640 mV, fill-factor (FF) between 73-76 % and η 14-15.1% in the reference cell and J_SC between 33.6-34 mA/cm², V_OC between 658-700 mV, FF between 71-76% and η between 15.7 %-16.5% in the best cell. Due to the presence of crystal defects, local shunts and lower charge collection under the electron contact fingers, the resultant full cell parameters for the reference cell are J_SC: 28.5 mA/cm², V_OC: 617 mV, FF 65.4 % and η 11.5 % and for the best cell are J_SC: 30.3 mA/cm², V_OC: 652 mV, FF 64.6 % and η 12.8 % (as measured under a AAA-rated sun simulator).
2. Grain boundaries exhibit J_SC of approximately 7-9 mA/cm², V_OC of approximately 340-400 mV, 50%-70% FF and 1.5-5 % η , subject to the size and recombination activity. Similarly, stacking faults in the reference cell exhibit average values of 26 mA/cm² J_SC, 515 mV V_OC, 70 % FF and 9.5 % η.
3. Laser firing used to lower contact resistance on the electron contact fingers does not lower the over-all V_OC of a cell because it is primarily limited by grain boundaries present in the cell.
4. Electroluminescence imaging shows a gradient towards the direction of the electron contact busbar primarily due to high series resistance in the electron contact fingers. This gradient is also visible in the LBIC images taken at voltage bias > 0.6 V and is reflected in the local series resistance fitted by the model.

A comparison with the best efficiency cell (15.1 % with anti-reflection foil) is ongoing in which similar trends are expected, but with higher power conversion efficiency. In conclusion, this method successful models current paths of photogenerated charge carriers within the cell and to external electrodes in an IBC architecture. The high resolution maps of electrical parameters quantify the excellent performance of intra-grain regions and limitations primarily due to crystal defects and minority carrier (hole) diffusion length under the electron contact.

Although 26.7% efficiency has been attained with hetero-junction and back contact (HBC) structure and by using bulk minority-carrier lifetime of around 10 msec, high efficiency of more than 29% is expected to be realized as a result of decreasing non-radiative recombination loss. This paper presents analytical results for high efficiency potential of crystalline Si solar cells, analyzed by using knowledge (C_i-O_i center, B_i-O_i center, E’ center and Pb center), in radiation-induced defects in Si.
One of problems to attain the higher efficiency crystalline Si solar cells is to reduce non-radiative recombination loss. The open-circuit voltage drop compared to bandgap energy (Eg/q-Voc) is dependent upon non-radiative voltage loss (Voc, nrad) that is expressed by external radiative efficiency (ERE). Open-circuit voltage is expressed by
V_oc = V_oc,rad + (kT/q)ln(ERE), (1)
where the second term shows non-radiative voltage loss, and V_{oc, rad} is radiative open-circuit voltage and 0.87V was used in this study. Correlation between Voc values for HIT, HBC, PERL and multi-crystalline Si cells and ERE values in the references estimated by eq. (1) suggests that crystalline Si solar cells have still non-radiative loss and further improvements in efficiency are thought to be possible by improving minority-carrier lifetime [1]. Calculated values of minority-carrier lifetime is expressed by the following equation
1/τ = BN + CN², (2)
where B is radiative recombination probability and 2x10^-15 cm³/s for Si, N is carrier concentration and C is the Auger coefficient and is 1.6x10^-30 cm⁶/s for Si. Correlation between calculated values and obtained values of minority-carrier lifetime suggests that further improvement in minority-carrier lifetime is expected to be realized. Really, high bulk lifetime of around 100 msec has been obtained by the MCZ (Magnetic-field-applied Czochralski) method and decreasing carbon concentration to less than 10¹⁴ cm^-3 in Si. C_i-O_i center induced by 1-MeV electron irradiations have been used for calibration of carbon concentration in Si. Higher minority-carrier lifetime of around 100msec by decreasing carbon and oxygen precipitate concentrations will provide lower voltage drop of around 0.3V and higher Voc of around 800mV and thus higher efficiency Si solar cells of more than 28%.
Regarding light-induced degradation of Si solar cells, although B_i-O_i center in Si was previously thought to act as defect center for light illumination degradation, B_i-O_i center doesn’t act as a non-radiative recombination center according to our previous studies [2] for radiation-induced defects in Si. Defect introduction rates of B_i-O_2i center induced by light illumination are compared with those of B_i-O_i center induced by 1-MeV electron irradiations in this study.
Surface recombination degradation of Si solar cells under light illumination is also another important issue for realizing highly reliable Si solar cells. In this study, surface recombination degradation of Si solar cells due to 1-MeV electron irradiations is compared with surface degradation of Si solar cells under light illumination by considering Pb center generation due to 1-MeV electron and ultraviolet light exposures.
In summary, understanding defects in Si solar cells by learning radiation-induced defects in Si is thought to useful for further development of high-efficiency and highly reliable Si solar cells.

References
[1] M. Yamaguchi et al., J. Mater. Res. 33, 2621 (2018).
[2] M. Yamaguchi et al., J. Appl. Phys. 86, 217 (1999).

Interfaces such as grain boundaries (GB) that break the transitional symmetry of a crystal play a dominant role in the properties of materials. The role of GBs is substantial in multi- and polycrystalline optoelectronic and photovoltaic materials: Multi-crystalline Si (mc-Si), which nowadays dominates the Si based photovoltaics industry has a density of 10⁴/cm GBs. These defects, as well as their interaction with impurities constitute a major limiting factor in the efficiency of the aforementioned devices. More specifically it has been shown that recombination at GBs constitute a major channel for electrical losses in high performance mc-Si solar cells. Σ3 tilt GBs in Si is a family of interfaces with special fundamental and technological interest. The Σ3{111} twin boundaries (TB) have negligible boundary energy and are electrically inactive. They are therefore preferred over other electrically active boundaries. However, Σ3{112} GBs exist in two stable/metastable states, the symmetric (S-Σ3) and the asymmetric (A-Σ3), with the former having deep states into the fundamental bandgap. Furthermore, facet junctions of the TB with the A-Σ3 show a unique anisotropic segregation mechanism, that goes beyond the classical picture of planar McLean type segregation [1]. Nevertheless, a prerequisite to get full control on the properties of these interfaces is to gather a deep understanding of the corresponding faceting mechanisms.

In the present work we employ density functional theory (DFT) and large scale modified embedded atom method (MEAM) potential calculations to investigate the structure, energetics and strain associated with these boundaries. Based on the our DFT calculations we parametrize a MEAM potential suitable to describe GB energetics and long range elastic interactions in Si. Our calculations revealed that Σ3 tilt GBs with the [110] rotation axis are intrinsically unstable against faceting towards {111} and {112} facets. Furthermore, based on these calculations we derive a facet and line junction phase diagram that describes and predicts the reconstructions of these extended defects as function of facet length and boundary inclination angle. This diagram reveals that large (small) inclination angles with respect to {111} plane and/or large (small) facet lengths favor junctions with A-Σ3 (S-Σ3) facets, respectively. This is explained in terms of (i) the large anisotropy in the energies of S- and A-Σ3 and (ii) the presence (lack) of extended defects at the {111} and A- (S-) {112} line junctions, respectively: While the low energy of the A-Σ3 facets favors these boundaries over their symmetric counterparts, line junctions of TBs with A-{112} facets are accommodated by edge type dislocations. The long range repulsive interactions make these facets highly unfavorable at small facet lengths. On the other hand, S-Σ3 GBs are commensurate to the Σ3 coincidence site lattice and the energy of the corresponding faceted junction is independent of the facet length. These findings shed light upon GB faceting phenomena and have important implications on the GB mobility mechanisms, the electronic properties as well as on selective impurity segregation at line junctions and facets in mc-Si: At large inclination angles the A-Σ3 GBs will adopt the maximum length allowed by the microstructure and grain size. Furthermore, the highly strained core region and its topology at the line junctions will strongly influence impurity segregation at these regions. On the other hand, at small inclination angles, the deep states introduced by the S-Σ3 facets will act as a recombination centers and will be detrimental for the device efficiency. Based on the aforementioned results will further discuss the implications on impurity segregation at the flat and facet junctions Σ3GBs as well as on their mobility.

[1] C. H. Liebscher, A. Stoffers,M. Alam, L. Lymperakis et al., Phys. Rev. Lett. 121, 015702 (2018).

Light-induced degradation (LID) is a decrease in the minority carrier lifetime in boron-doped Czochralski (Cz) silicon solar cells due to injection of excess carriers. It is well accepted that LID occurs due to a defect involving a substitutional boron atom and an oxygen dimer. There are three characteristic stages of LID: an annealed state where defects are metastable and non-recombination-active, a degraded state where defects are recombination active, and a stabilized state where defects are permanently non-recombination-active. While a technological solution to mitigate LID is slowly being implemented into industrial manufacturing lines, the fundamental mechanism for lifetime stabilization and the structure of the LID defect remain speculative. Empirical studies relate B and O to the defect structure, but atomistic evidence is lacking. If the defect structure and mechanism of LID is discovered, mitigation strategies could be quickly studied and implemented improving the long-term reliability of high-efficiency, B-doped Si solar cells.
In this work we utilize electron paramagnetic resonance (EPR) spectroscopy, magnetic susceptibility, and lifetime spectroscopy to study the structure of the LID defect and the degradation mechanism. We report on the EPR spectra of Cz and float-zone (FZ) Si in different stages of LID. Firstly, we show that the paramagnetic signal at low temperature (6 K) is dependent on the state of LID and on the sample doping. The light-degraded state in p-type Cz Si shows a strong, sharp paramagnetic defect signature (g-value = 2.003) compared to samples in the annealed state, where the EPR spectra is relatively featureless. In contrast, the broad EPR active signature related to boron interaction with lattice, is present in the annealed state but diminishes after light exposure. Both EPR signals change under indoor light exposure at room temperature and therefore are sensitive to initial stages of LID. We see a smaller decline in the wide-spectrum boron/lattice interaction EPR signal in p-type FZ Si and none in n-type Cz Si. We hypothesize the decrease in the oxygen concentration in the p-type FZ sample results in incomplete quenching of the boron acceptor spins. We suggest that these changes in the EPR spectra are related to the BO defect that causes LID. Additionally, we utilized p-type FZ with an order of magnitude higher B doping to investigate the effect of the B/O concentration ratio on extent of quenching of B_a spins after indoor light exposure. We observe that in the p-type FZ Si sample that is more heavily doped with B (B/O ~ 10:1) the B_a spin signal is almost unaffected by LID compared to the more lightly doped B-doped p-type FZ Si sample (B/O ~ 1:1). Therefore, if the B atoms are in excess, there are simply not enough O atoms to quench all the B_a spins and the EPR signal remains. Therefore, O is involved in this paramagnetic to diamagnetic transition of the B_a in the Si bulk and for this transition to occur, BO complexes are required. Finally, we show that magnetic susceptibility measurements confirm that in the annealed state of LID, the p-type Cz Si sample is paramagnetic, and in the degraded state, diamagnetic, which is consistent with our EPR findings. Thus, for the first time, EPR and magnetic susceptibility have been used to show manipulation of the LID defect after dark-annealing and light exposure.

Crystalline silicon (Si) solar cells currently occupy an important place in the solar cell market. However, potential-induced degradation (PID) of crystalline Si photovoltaic (PV) modules has recently been observed in large systems comprising significant numbers of PV modules. In these PV systems, numerous such modules are interconnected in series, such that potential stress is imparted to some of the modules with consequent high power losses. PID in Si PV modules based on p-type crystalline Si solar cells has been previously investigated in detail [1]-[5].
About estimated changes in solar cell during PID, for example, it has been reported that, due to the penetration of Na into the solar cell, a deep level is formed in the depletion layer or the base layer (p layer) and, as a result, carrier recombination through the level occurs [4]. Na is known to create a level in Si at a position 0.35 eV from the valence band. It also has been suggested that current shunt path is formed by intrusion of Na into stacking faults penetrating the pn junction (or that Na invading Si forms stacking faults) [5].
Even so, there is still a lack of understanding of the PID mechanism in solar cells, and obtaining more information regarding this phenomenon will require measurements of carrier distributions on a microscopic scale. One of the authors previously succeeded in quantitatively analyzing such distributions in monocrystalline Si solar cells using scanning nonlinear dielectric microscopy (SNDM) [6][7].
In the present study, the microscopic carrier distributions in monocrystalline Si solar cells exhibiting PID were investigated using this same technique. In addition, an enhanced version of this technique referred to as super-higher-order (SHO)-SNDM was employed to visualize depletion layer distributions in such cells both with and without PID.
As a result, reductions in carrier density in accordance with the extent of PID were clearly observed, and quantitative measurements demonstrated that the electron concentration was reduced by serval order of magnitude due to PID. Depth profile measurements showed that the solar cells were affected by PID to a significant depth of approximately 90 μm, equal to almost half the cell thickness, suggesting that Na⁺ migration proceeded easily upon the application of a high voltage. The SHO-SNDM data showed that the depletion layer of the non-PID sample was thinner than that of a PID sample, indicating that the carrier concentration in the former was greater.
This work demonstrates that SNDM is a very useful means of investigating the PID effect through the measurement of carrier distributions in monocrystalline silicon solar cells.

References:
[1] S. Jonai and A. Masuda, AIP Adv. 8, 115311 (2018).
[2] Berghold, O. Frank, H. Hoehne, S. Pingel, B. Richardson and M. Winkler, Proceedings of the 25th European Photovoltaic Solar Energy Conference and Exhibition, Valencia, Spain, 3753 (2010).
[3] P. Hacke, M. Kempe, K. Terwilliger, S. Glick, N. Call, S. Johnston, S. Kurtz, I. Bennett and M. Kloos, Proceedings of the 25th European Photovoltaic Solar Energy Conference and Exhibition, Valencia, Spain, 3760 (2010).
[4] W. Luo, Y. Sheng Khoo, P. Hacke, V. Naumann, D. Lausch, S. P. Harvey, J. Prakash Singh, J. Chai, Y. Wang, A. G. Aberle and S. Ramakrishna, Energy Environ. Sci. 10, 43 (2017).
[5] V. Naumann, D. Lausch, A. Hähnel, J. Bauer, O. Breitenstein, A. Graff, M. Werner, S. Swatek, S. Großer and J. Bagdahn, and C. Hagendorf, Sol. Energy Mater. Sol. Cells 120, 383 (2014).
[6] K. Hirose, K. Tanahashi, H. Takato, and Y. Cho, Appl. Phys. Lett. 111, 032101 (2017).
[7] Y. Cho, A. Kirihara, and T. Saeki, Rev. Sci. Instrum. 67, 2297 (1996).

In the present work, the effect of surface states and bulk opto-electronic properties of Al doped c-Si(p)/a-SiC:H(p) heterojunction photocathodes were evaluated on the photocatalytic performance for photoelectrochemical water splitting, before and after thermal treatments. The absorber a-SiC:H(p) thin film of around 320 nm was deposited onto p-type Si and fused silica substrates by reactive Radio frequency magnetron co-sputtering using high purity SiC and Al targets under Ar-H₂ atmosphere. a-SiC:H(p) thin films on Si substrates were chosen to perform morphological, structural and photoelectrochemical characterization. Whereas the samples prepared onto fused silica substrates, were used to perform optical characterization. The absorption coefficient and optical bandgap energy were determined since they play an important role in the device performance. Optical properties of the a-SiC:H(p) films were assessed systematically by means of UV-Vis transmittance measurements. Fourier transform infrared spectroscopy measurements were performed to track the hydrogen content, molecular bond-densities and changes in the structural properties of the a-SiC:H(p) layer for the distinct annealing temperatures. Van der Pauw electrical measurements revealed an improvement in the electrical behavior of the photocathode device after performing the annealing step at 600^oC. Ohmic contact, which is a key requirement for an optimal photoelectrochemical cell performance, was achieved after the annealing step at this temperature.

Annealing treatments induced changes in electrical, structural and optical properties of the material. These changes were reflected in the achieved good photocathode performance. Thus, an abrupt enhancement of the photocurrent was observed during the device photoelectrochemical test in 1 M H₂SO₄ solution under chopped light illumination. Reaching a saturation value of 17 mA/cm² at -1.75 V vs. Ag/AgCl (3.5 M KCl). In this research we propose that this enhancement is associated to a charge transfer kinetic mechanism influenced by surface states and the p-type Si substrate. The latter most likely due to the space charge region extension beyond the absorber layer until reaching the substrate, as it was demonstrated by means of electrochemical impedance spectroscopy measurements and some of our estimations. Current density-potential and electrochemical impedance spectroscopy measurements in dark, revealed a reduction of the SiO₂ native layer at cathodic potentials higher than -1 V vs. Ag/AgCl (3.5 M KCl). This last effect might contribute to the high charge transfer kinetic of the system. In that sense, our findings so far conduct to believe that SiO₂ layer reduction triggers a favorable semiconductor-electrolyte interface, possessing less overpotential barriers that facilitates photocurrent generation. We believe that these results will help to the developing of Si/SiC junction devices in tandem for solar-to-hydrogen energy conversion.

Transparent solar cells, which combine visible transparency and solar energy conversion, are being developed for applications in which conventional opaque solar cells are unlikely to be feasible, such as windows of buildings or vehicles. However, transparent solar cells developed to date have limitations in efficiency and stability. Crystalline silicon (c-Si) is one of the best candidates to develop transparent solar cells with high efficiency and stability, because conventional c-Si solar cells are known to exhibit high efficiency and long-term stability compared with other solar cells. However, the opaque characteristic of the c-Si wafer hinders the development of transparent solar cells using c-Si. In this presentation, we introduce a novel approach to develop neutral-colored transparent c-Si solar cells that exhibit the highest efficiency among neutral-colored transparent solar cells developed to date. Also, we review recent progress in transparent solar cells along with strategies that enable the transparency of conventional solar cells, including thin film technology, selective light transmission technology, and luminescent solar concentrator technology. From fundamental research to commercialization of the transparent solar cells, three main perspectives should be considered: (1) high power conversion efficiency at the same average visible transmittance; (2) aesthetic factors, which should not detract from applications such as buildings and vehicles; and (3) feasibility for real-world applications, including modularization and stability evaluation. We present the distinct analysis criteria for these main perspectives and discuss their importance.

Texturing the front surface of thin film photovoltaic cells with ordered or disordered arrangements of subwavelength structures is beneficial in terms of efficient light harvesting as well as efficient carrier extraction. Previous studies demonstrated efficient broadband absorption of solar radiation with surface arrays of subwavelength inverted cones (light funnels – LFs). In the current work, we use threedimensional finite-difference time-domain electromagnetic calculations as well as three-dimensional device calculations to examine carrier extraction from photovoltaic cells that are composed of LF arrays on top of underlying substrates. For the selected geometry under examination, we show a broadband absorption enhancement of 14% for the LF photovoltaic cell compared with a cell based on the respective optically optimized nanopillar arrays. However, we show that the nominal power conversion efficiency is 60% higher in the LF cell which is due to the enhancement of both open-circuit voltage and short-circuit current. The higher open-circuit voltage in the LF cell is due to the higher injection of photocarriers, and the higher short-circuit current is a result of the unique LF geometry that supports efficient carrier extraction due to the naturally occurring gradients of the quasi-Fermi levels and minority carrier conductivity that allow for enhanced contact selectivity. We believe that this work paves the way towards a new approach for carrier collection in photonic devices for energy applications.

Silicon solar cells among different types of solar energy harvesters have entered the commercial market owing to their high power conversion efficiency and stability. By replacing the electrode and the p-type layer by a single layer of carbon nanotubes, the device can be further simplified. This greatly augments the attractiveness of silicon solar cells in the light of raw material shortages and the solar payback period, as well as lowering the fabrication costs. However, carbon nanotube-based silicon solar cells still lack device efficiency and stability. These can be improved by chemical doping, antireflection coating, and encapsulation. In this work, the multifunctional effects of p-doping, antireflection, and encapsulation are observed simultaneously, by applying a polymeric acid. This method increases the power conversion efficiency of single-walled carbon nanotube-based silicon solar cells from 9.5% to 14.4% and leads to unprecedented device stability of more than 120 days under severe conditions. In addition, the polymeric acid-applied carbon nanotube-based silicon solar cells show excellent chemical and mechanical robustness. The obtained stable efficiency stands the highest among the reported carbon nanotube-based silicon solar cells.

We present calculations of performance characteristics of Indium Gallium Nitride-Silicon Heterojunction Schottky barrier solar cells. The effect of growth axis and spontaneous and piezoelectric effects in the Indium Gallium Nitride are taken into account. We consider both wurtzite Indium Gallium Nitride layers on 111 silicon and cubic indium gallium nitride layers on 100 silicon. The short-circuit current generation as a function of depletion-layer thickness is studied along with the effect of Indium Gallium Nitride composition on the Schottky barrier dark current. The absorption of high energy solar photons in a wider band-gap indium gallium nitride layer with the absorption of lower energy photons in the silicon layer provides the best efficiencies for this device as the conduction and valence band discontinuities are of the right sign to provide unimpeded electron and hole flow for collection in a high-efficiency cell. We consider the effect of composition grading on solar cell characteristics.

Common forms of elemental silicon (mono-, multi-crystalline and amorphous) materials have been playing a key role in photovoltaic technologies for almost seventy years and are studied thoroughly. Exotic, low-density forms of silicon have recently regained interest due to alluring semiconducting and optical properties as films in regard to the conventional ones. In this work, the focus is put on an allotrope of silicon: silicon clathrate films. Inorganic clathrates are intermetallic structures defined by a 3D porous framework (Si or group IV elements) of polyhedrons, which can enclose guest-atoms (Na or alkali metals). Due to this specific organization, some silicon clathrates can exhibit an interesting direct band-gap in the range of 1.8-2 eV^[1], at low guest-atom concentration, which is specifically beneficial for photovoltaic applications.
In this work, we will introduce the preparation and characterization of silicon clathrate films using different silicon reagents (p-doped and intrinsic c-Si (100) and (111) wafer, a-Si on sapphire...). The structural, electrical, optical, surface properties are analyzed and the fabrication conditions optimized to give the best properties and mechanical integrity. Furthermore, various post-synthesis processes are investigated in order to maximize the photovoltaic efficiency of silicon clathrates in regard to the first silicon clathrates basic solar-cell obtained^[2]. From Ion implantation experiments and other doping techniques to annealing under press, we will discuss how these treatments could, not only help to strengthen the properties of silicon clathrates, but also to bring some enlightenment about the underlying mechanisms that are still unclear.
^[1] Valeri I. Smelyansky, John S. Tse. (1997). The electronic structure of metallo-silicon clathrates NaxSi136 (x = 0, 4, 8, 16 and 24). Chemical Physics Letters. 264 (5).
^[2] Tetsuji. Kume et al. (2016). Thin film of guest-free type-II silicon clathrate on Si (111) wafer. Thin Solid Films. 609.

Hyperdoped semiconductors are promising material candidates for intermediate band photovoltaic applications with theoretical efficiency exceeding 50% [1]. Despite of considerable amount of literature reports on the optical and structural characterizations, few studies focus on crucial parameters for device designs such as charge carrier lifetime and mobility. In recent years, ion implantation followed by non-equilibrium pulse laser melting has shown to produce high quality hyperdoped material. The resulting hyperdoped layer is highly substitutional, containing up to 1 atomic % of dopant exceeding the typical solid solubility limit [2]. Previous studies have reported Si hyperdoped with S, Se, Te and Au as promising materials to enhance light absorption to the order of ~10⁴ cm^-1 in sub-band gap wavelength range [3-5]. Furthermore, Te dopant exhibits higher thermal stability than S and Se, and less dopant segregation than Au [6]. This study aims as understanding charge charrier dynamics such as carrier lifetime and mobility of Te-hyperoped silicon (Si:Te).
We study Si:Te on both p-type and n-type substrates with varying laser processing conditions. The optical absorption spectroscopy and secondary ion-mass spectrometry were employed to investigate the sub band-gap absorption and concentration profile. The results show that the absorption values are comparable with the other well studied materials like Au-hyperdoped Si (Si:Au), where sub bandgap absorption increases with increasing dopant dose. Time resolved terahertz spectroscopy (TRTS) is used to study the charge carrier lifetime. Optical pump THz probe is a non-contact photoconductivity measurement for probing carrier lifetime with picosecond time resolution. We found that Si:Te lifetimes are much longer, consisting of a rapid decay in the order of ps followed by a longer, ns-lifetime component. To explain the observed results, we study two different optical pump wavelengths at 400nm and 266nm to investigate surface recombination, carrier mobility and diffusion properties. The carrier lifetime and mobility results together with optical absorption and structural properties will be discussed to better understand and optimize Si:Te as a material candidate for intermediate band photovoltaics.

[1] A. Luque, et al., Nature Photonics 6, 146 (2012).
[2] J. P. Mailoa, et al., Nature Communications 5, 3011 (2014).
[3] T. G. Kim, et al., Appl. Phys. Lett. 88, 241902 (2006).
[4] M. Tabbal et al., Appl. Phys. A 98, 589 (2010).
[5] W. Yang, et al., Phys. Rev. Mater. 1, 074602 (2017).
[6] M. Wang, et al., Phys. Rev. Applied 10, 024054 (2018).

Development of new materials with novel optical, electronic, and spin properties is critical to next generation electronic and photonic technologies. Although Si has been the most widely used element in this area, the most stable Si phase, diamond Si, offers an indirect band gap and weak absorption. A cage-like, crystalline Si allotrope, Si clathrate, has been reported to have a direct or nearly direct bandgap with potentially exciting optoelectronic properties. Room temperature and atmospheric pressure metastable clathrates are synthesized in the presence of alkali guest atoms, such as Na, which occupy the interstitial sites in the cages. Low Na concentration Type II Si clathrate Na_xSi₁₃₆ in the powder form has been identified to be a heavily doped semiconductor with Na guest(s) as a shallow donor. Compared with clathrate powder, the thin-film clathrates are more interesting due to their potential for practical application. However, the synthesis of high-quality clathrate films with well-defined properties for studies of band structure, defects, dopant diffusion, and electronic transport has been difficult. A detailed understanding of tunable and controlled thin-film synthesis and its optical properties is critical to eventually realizing the potential of these semiconductors for electronic and optoelectronic devices.
Here, we present a protocol for silicon clathrate thin film synthesis that produces thin, uniform, well adhered type II clathrate films with low Na concentration and a low concentration of competing phases. The films have been formed using a two-step process: formation of the precursor NaSi film and thermal decomposition of the precursor into the clathrate phase. By adding an additional annealing process between the two steps above, Na content was reduced before the thermal decomposition in an effort to reduce the voids and holes in the final film due to Na removal. XRD in combination with Rietveld refinement was used to characterize film crystallinity, phase purity, and Na content. The effect of varying the time and temperature of each of the processing steps was explored. Also, films were formed directly on Si wafers and Si films on alternative substrates, such as sapphire, to simplify electrical and optical characterization. Post growth treatment procedures to remove residual Na contamination were also developed. The films are expected to grow randomly and SEM images confirm morphological variation, but also show regions with sufficient uniformity to allow further measurement with optical and electronic techniques that sample small areas (e.g. photoluminescence (PL) and Time of Flight SIMS). We also obtained the correlation between the thin film thickness and the annealing time of the first step which makes controllable synthesis possible.
Characterization of the synthesized thin film samples was performed and compared to low Na content powder samples. The PL spectrum shows a wide emission peak centered around 2 eV, with a full width at half maximum of about 0.54 meV. This peak is near the reported direct bandgap of Si clathrates but extends to higher energy, suggesting a degenerately doped semiconductor with an impurity band moving into the conduction band. Moreover, similar to the Electron Paramagnetic Resonance spectra of the clathrate powder at low temperature, we observe a feature at g~2.002 associated with free carriers and four hyperfine lines with the splitting of 133mT in the thin film due to the interaction of the Na valence electron with the nuclear spin (I = 3/2) of Na²³. The observation of the hyperfine lines suggests that the Na content in the films is below the metal-insulator transition. The findings in this work reinforce the notion that this Si allotrope could be a next-generation material for semiconductor technology.
This work was supported by the National Science Foundation Award #1810463.

Deposition of TMOs for PV purposes on industrial techniques (e.g. Sputtering, ALD, etc) is a work that has been repetedly reported by relevant photovoltaic research groups[1][2], however so far literature results indicate that the hole selective properties of evaporated TMOs can not be successfully reproduced by other deposition methods Oposing to this thesis we report the deposition of vanadium oxide (V2O5) films by ALD
effectively performing as hole-selective contact for n-type crystalline silicon solar cells. Reasonable specific contact resistance of 100 mΩcm2 was measured by transfer length method. In addition, capacitance-voltage measurements confirms the presence of a strong inversion layer at the c-Si/V2O5 interface with a build in voltage of 683 mV. Furthermore, soft deposition of ALD achieved high effective lifetime values up to 800
μs as a result of passivation properties of the film. Finally, a proof-of-concept solar cells fabricated exhibiting efficiencies beyond 18% and with similar if not better photovoltaic parameters to evaporated TMOs without the
use of any additional passivation layer in the front contact. The technolgical implementation of ALD as the deposition technique to state of the art TMO solar cells could be the key for the fabrication of full wafer high efficiency dopant free c-Si solar cells.

[1] Sputter-deposited WOx and MoOx for hole selective contacts, Martin Bivour et Al, 10.1016/j.egypro.2017.09.259, Fraunhofer Institute for Solar Energy Systems ISE.
[2] Atomic layer deposited molybdenum oxide for the hole-selective contact of silicon solar cells, Martin
Bivour et Al, 10.1016/j.egypro.2016.07.125, Fraunhofer Institute for Solar Energy Systems ISE

Silicon nanowires (SiNWs) have gained great attention in various emerging applications that include solar cells, Li- & Na- ion batteries, electronic memory devices, supercapacitors, and thermoelectric generators [1-3]. Although the semiconductor industry is dictated by silicon, there are many limitations posed by the bulk material limiting its use in emerging electronic devices. Interestingly, these limitations are solved by its micro to nano-sized silicon. Smaller dimensions of SiNWs have provided opportunities in solar cells due to increased absorption by trapping of light, anti-reflection effect, effective charge separation, decoupling the path of light absorption and carrier separation [1, 2]. The use of SiNWs requires a fraction amount of silicon as compared with the traditional Si wafer-based devices, which minimises the cost of production. With the Use of Plasma Enhanced Chemical Vapor Deposition process (PECVD) it is possible to lower the energy consumption for the growth of crystalline SiNWs by efficiently dissociating precursors gases, controlling the process with plasma and growth rate is an order of magnitude faster as compared with the Chemical Vapour Deposition and Molecular Beam Epitaxy. Also, enables the use of relatively cheaper substrates such as glass and plastic. However, currently SiNWs are doped with diffusion processes and/or gaseous precursors (diborane and phosphine). However, our work proposes doping from a metal catalyst [2-3] that is already used as a seed for the growth of nanowires. This method has added advantages including the elimination of toxic gases, avoiding the complexity of counter-doping caused by the catalyst, and reducing costs. Thereby moving towards green manufacturing processes.

This research focuses on understanding and optimisation of controlled growth of self-doped SiNWs (both n-doped and p-doped) by a PECVD process at relatively low temperatures (<300°C). Metals used as a catalyst in the VLS growth process provides some additional advantages along with the nucleation and precipitation if carefully selected. Hence, the selection of catalyst (metal) with ionisation energy level closer to the valence and/or conduction band of Si will contribute to self-doping. In this study, various metals (such as gallium, indium as p-type dopant, and Bismuth as n-type dopant) are used to facilitate the growth and to dope the nanowires. This work also introduces and discusses the novel process to grow simultaneously and selectively in situ n- and p-type silicon nanowires in a single pump-down PECVD process without the use of dopant gases. To the best of our knowledge, such a process is not reported. Such nanowires are possible to be grown on plastic and glass substrates thus reducing cost and open up a number of opportunities. Optical and electrical analyses including Seebeck and resistivity measurements will be discussed to understand the properties of catalyst doped SiNWs. Morphology and crystallinity by various characterisation techniques such as transmission and scanning electron microscopy, X-Ray diffraction, and Raman spectroscopy are performed. This work is not just limited to the growth process, but also adapting SiNWs to fabricate solar cells. For example, fabricated SiNWs photovoltaic solar cells have shown an efficiency of 6% and there is still room for improvement as the study is at the exploratory stage.

[1] Manjunatha, Krishna Nama, et al. "Comparative Study of Silicon Nanowires Grown From Ga, In, Sn, and Bi for Energy Harvesting." IEEE Journal of Photovoltaics (2020).
[2] Manjunatha, Krishna Nama, and Shashi Paul. "In-situ catalyst mediated growth and self-doped silicon nanowires for use in nanowire solar cells." Vacuum 139 (2017): 178-184.
[3] Manjunatha, Krishna Nama, and Shashi Paul. "Carrier selective metal-oxides for self-doped silicon nanowire solar cells." Applied Surface Science 492 (2019): 856-861.

Photovoltaic technology is considered one of the leading fields for sustainable solar energy production. In recent years, considerable efforts have been done in the development of functional materials for the fabrication of increasingly sophisticated and advanced photovoltaic devices (PV) that can ensure an efficient exploitation of solar spectrum. In this context, lanthanide (Ln)-doped fluoride materials are considered particularly promising for energy conversion applications due to their unique luminescent properties under light irradiation. A proper combination of doping elements allows an efficient luminescent pathway in a large spectral range through up-conversion (UC) or down-conversion (DC) mechanisms. UC and DC processes are the main mechanisms which enhance the efficiency of photovoltaic devices since they make possible to collect the radiation energy outside the absorption range of the photoactive material (usually silicon) by shifting the energy to a more suitable optical region. Among the fluoride-based systems, BaF₂ is considered a suitable host for the UC and DC luminescence of Ln³⁺ ions because of its low phonon energy (similar or even lower than the extensively used hosts such as NaYF₄), high chemical stability, wide transparency (i.e. BaF₂ is transparent from 150 nm to 15 μm) and versatile synthetic strategies. It should be mentioned that host materials require lattice matches with Ln³⁺ dopant ions. Trivalent lanthanide ions in particular can be well incorporated into crystal structure of alkaline-earth fluorides due to the similar coordination sphere. However, a charge compensation mechanism is required to balance the additional charge introduced after the replacement of divalent alkaline-earth ions with trivalent Ln dopants.
In the present study, we report the synthesis of Ln³⁺ co-doped BaF₂ thin films (Ln³⁺ = Yb³⁺, Pr³⁺) on silicon and quartz substrates, using metalorganic chemical vapor deposition technique (MOCVD). The advantages of this approach include: simplified instrumental apparatus, high deposition rate, high purity and easy industrial scale-up. For the fabrication of BaF₂ thin films co-doped with Yb³⁺, Pr³⁺, a multicomponent mixture of metalorganic adducts in an appropriate molar ratio is used as a single molten source. It is possible to easily tune the chemical compositions of the films, to tailor the energy transfer processes, by changing the composition of the precursor mixture, that is the Ba:Yb:Pr ratio in the multicomponent source. X-ray diffraction analysis (XRD), field-emission scanning electron microscopy (FE-SEM) and energy dispersive X-ray analysis (EDX) allowed a structural, morphological and compositional characterization of the films. Finally, the film functional properties have been assessed through luminescence spectroscopy.

Atomic layer deposited (ALD) niobium oxide (NbO_x) has shown potential for use as a passivating layer or as a selective contact for silicon-based photovoltaics (PVs). However, there is large gap in understanding of the synthesis-structure-electronic property relationships. It has been shown by Macco et. al. that ALD niobium oxide can provide effective passivation of silicon substrates with an interfacial oxide down to 1 nanometer layer thickness.¹ In a separate work it was also show that ALD niobium oxide can serve as an electron selective contact.² In this work, the ALD growth of niobium oxide is optimized for temperature and co-reactants of either ozone (O₃) or water (H₂O). The oxidation states are then determined via X-ray photoelectron spectroscopy (XPS) to determine the exact stoichiometry of the films. The surface recombination velocity of niobium oxide contacts to silicon is measured and related to deposition conditions, film thickness, and post-deposition processing. Some samples are annealed in both inert and oxygen atmospheres and the stoichiometry is determined as well as the crystal structures measured via X-ray diffraction (XRD). This work provides new insight into the applicability of niobium oxides for passivation and selective contact layers for Si photovoltaics.

1. Macco, B. et al. Effective passivation of silicon surfaces by ultrathin atomic-layer deposited niobium oxide. Appl. Phys. Lett. 112, (2018).
2. Macco, B. et al. Atomic-layer deposited Nb2O5 as transparent passivating electron contact for c-Si solar cells. Sol. Energy Mater. Sol. Cells 184, 98–104 (2018).

Carrier selective contact enhances the efficiency of silicon (Si) solar cell by providing surface passivation and an effective majority carrier extraction. Recently, ultrathin silicon oxide (SiO_x) passivated n⁺ polysilicon (poly-Si) contact has drawn attention as a promising electron selective contact because of its excellent surface passivation quality, low contact resistivity and high thermal stability compared to amorphous Si (a-Si:H) layer. Record efficiency of 25.7 % has been reported for n-type silicon solar cell fabricated using this as a full area rear contact [1].
In this work, we present a tunnel oxide passivated phosphorus doped poly-Si electron selective contact for n-type silicon solar cell where the in-situ doped poly-Si layer (around 100 nm) is deposited using an in-line high-throughput atmospheric pressure chemical vapor deposition process (APCVD). This in-line APCVD system can perform single-sided deposition in atmospheric environment and is well-suited for industrial mass production. In this process, thermal dissociation of silane (SiH₄) and phosphene (PH₃) gases deposits the phosphorus doped Si layer on top of the ultrathin SiO_x (<2nm) layer which is grown in a wet chemical process and acts as the tunnel oxide. Later, the as-deposited Si layer goes through different post deposition treatments such as annealing, firing, hydrogenation, etc. for solid phase crystallization and passivation quality improvement. We have studied the effect of the deposition temperature and different post deposition treatments on Si surface passivation quality and contact resistivity (ρ_c) of this poly-Si contact. The change in microstructure, dopant profile and crystallinity of the APCVD Si layer due to change in process parameters are observed to understand the process-structure-property relationship of this electron selective contact.

1. Richter, Armin, Jan Benick, Frank Feldmann, Andreas Fell, Martin Hermle, and Stefan W. Glunz. "n-Type Si solar cells with passivating electron contact: Identifying sources for efficiency limitations by wafer thickness and resistivity variation." Solar Energy Materials and Solar Cells 173 (2017): 96-105.

Amorphous/Crystalline Silicon heterojunction (SHJ), which has been invented in 1990 and has been industrialized in 1997 by Sanyo, is regarded as one of the ideal junction technologies for silicon-based solar cells. The conversion efficiencies are exceeding >24% for “standard” bifacial SHJ solar cells, and >25% for back contact SHJ solar cells, respectively, using industrially feasible processes. We are aiming to improve power generation efficiency and reduce power generation cost by combining thinner and larger wafering, reducing optical losses reduction, innovating texturing, hetero-junction and metallization technologies. We are also developing perovskite/SHJ tandem solar cells in order to break the "Silicon Limit." I will review our recent R&D activities.

We are reporting the findings of the 18 months project on impact of undoped substrates on high performance silicon solar cells supported by the U.S. Department of Energy. The aim of the project was to develop fundamental understanding and experimentally demonstrate the mechanisms in play when silicon solar cells are manufactured using different bulk resistivities (1 ohm.cm to over 15000 ohm.cm). The project focus was understanding how the performance of the solar cells, with different bulk resistivities, is impacted by in-field temperatures (up to 80 ^oC) and light conditions (0.1-1 suns). Today's highest-efficiency solar cells typically operate near the threshold between low and high injection. It is not well understood whether pushing the cell operation point further into a high injection regime throughout the day is beneficial for the solar cell performance. In this study, we present a comprehensive assessment, both experimental and by simulation, of how bulk resistivity, defects, light conditions, and operation temperature impact the performance of the silicon solar cell. Analytical simulations indicate that high bulk resistivity wafers (10 ohm.cm to over 15000 ohm.cm) require bulk Shockley-Read-Hall lifetimes in the millisecond range to outperform wafers with standard bulk resistivities (below 10 ohm.cm). Additionally, for bulk resistivities beyond 10 ohm.cm, the cell efficiency is shown to be weakly dependent of bulk resistivity. This study is particularly relevant today, as solar cell architectures with better surface passivation and milliseconds lifetimes wafers are commercially available, leveraging potential benefits of using higher bulk resistivities. Previous studies on high bulk resistivities are limited, inconclusive and lacked experimental data above 100 ohm.cm. In this work, we fabricate silicon heterojunction solar cells and samples to study lifetimes on n-type wafers with bulk resistivities from 1 ohm.cm to over 15000 ohm.cm. We measured their performance at different operation temperatures between 30 ^oC and 80 ^oC and for light intensities between 0.1-1 suns. The temperature coefficients are calculated from voltages at open circuit and maximum power, and from fill factors as a function of temperature. They are shown to be independent of the bulk resistivity and exhibit similar values as previously reported on standard bulk resistivities wafers. The surface saturation current density increases by several orders of magnitude for all bulk resistivities and show a cubic dependence of temperature. This temperature dependence was previously reported on standard bulk resistivities wafers. The significant increase of the surface saturation current density with temperature can be partially explained with a similar increase of the effective intrinsic carrier concentration. We are currently in the last quarter of the project and finalizing the measurements at different light conditions, the conclusions of which will be reported at the conference.

Lifetime prediction is a critical first step as the basis of a warranty or risk assessment. Research that identifies a PV module’s primary degradation or failure mechanism, quantifies the associated acceleration factors, and applies those to operating conditions is a straightforward approach to lifetime prediction. The value of this process is obvious and researchers around the world are investigating degradation and failure mechanisms of PV modules with this goal in mind. Making a lifetime prediction is actually quite easy: we can predict 25 years ±25 years for pretty much all PV products on the market, but such a prediction is useless. Making a lifetime prediction with a small uncertainty that leads to a useful prediction is much harder. This talk will describe why lifetime prediction is difficult, when it is essential, when it is a distraction, and how we expect lifetime prediction may evolve in the future.

Vehicle integrated photovoltaics has recently seen a rise in interest due to the increased employment of full or hybrid electric vehicles. The topic itself is not new, already in the 80ies first cars have been equipped with solar cells, such as for the “Tour de Sol” in 1985. However, only within the last few years there has been a peak in announcements and news about this topic. New car manufactures such as SonoMotors or Lightyear as well as established car manufacturers such as Toyota or Hyundai have announced sizeable plans to introduce cars with solar energy support. The current spike in interest is not odd, solar technologies have become historically low priced in recent years and electric cars are more and more available for which an addition of a solar support is most reasonable.
After a brief introduction of the topic with support of a historical review, we provide an overview on the current approach of vehicle-integrated photovoltaics (VIPV) at Fraunhofer ISE. While VIPV for cars focusses currently on the embedding of silicon solar cells within curved laminated glass, the approach for utility vehicles such as trucks and their trailers is vastly different. Here, lightweight and glass-free setups are preferred due to the crash behavior and less intense requirements regarding surface finish. As promising solar cell and interconnection technology for VIPV, we introduce the concept of shingled solar cells, offering benefits regarding efficiency, voltage level, curvature and the overall aesthetics.
To precisely calculate the potential benefits of VIPV we provide insights into the project “PV2Go”, where we compile a solar register of roads in Germany, which provides a detailed overview of the shading on roads in Germany. The modelling approach using spatial, geographic as well as digital terrain models is supported by a measurement campaign, where 100 cars are equipped with irradiation sensors.
We round off the topic with an overview of the future potential prospects of VIPV, such as equipping further vehicle parts like motor hood, side doors or even the windows with solar cells. It is concluded that by utilizing next generation solar cells such as tandem cells and further advances in increasing the efficiency of electric vehicles could lead to affordable vehicles with solar as the main source for propulsion.

With the lifespans of photovoltaic (PV) modules expected to last >20 years, rapid and spatially-resolved characterization methods are essential for understanding module reliability and degradation processes. Luminescence and thermal imaging have been gaining popularity both for fast, outdoor-compatible module inspection and for more detailed post-mortem analysis of failure mechanisms. Importantly, many failure mechanisms are not evident by visual inspection, and characterization of electrical properties alone is insufficient to identify the weakest features of cell or module design. This is especially true as PV technologies and module packaging schemes become more advanced and degradation becomes more subtle. Here, we present various examples of PV module degradation mechanisms in field-weathered modules, where we (1) locate problem areas with luminescence and thermal imaging, (2) core these areas out of the module, and (3) use various microscopy methods to identify chemical and structural causes of power loss. We particularly highlight causes of voltage loss or series resistance increases that vary depending on the technology, and we show studies including conventional Al back surface field (Al-BSF), silicon heterojunction with intrinsic thin layer (HIT), and passivated emitter and rear contact (PERC) modules. These results highlight the utility of imaging methods for improving module reliability.

Silicon solar cells have been interconnected by soldering and laminated into modules using essentially the same process that was used in the 1970’s. When Suntech commenced the mass manufacture of silicon photovoltaic modules in China in 2006, soldering was performed manually due to the low-cost of labour. However, automated stringing and tabbing machines quickly became the norm in Chinese manufacturing for the commonly produced two-busbar cells. Since then we have seen numerous advances in module engineering as silicon photovoltaics has become a low-cost form of electricity generation across the world. Starting from 3 busbar patents to large-sized modules fabricated from cut-cells designed to increase the module area efficiency especially for utility scale photovoltaics. It is now not just the module power that matters, it is module specific yield or how much electricity (kWh) will be generated per unit of purchased power (kW). This presentation will trace the advances in silicon photovoltaic module manufacturing over the years and discuss the current challenges of rapidly changing module architectures to service both rooftop and utility-scale applications.

Despite being labelled as “first generation” by some of the most eminent experts, crystalline-silicon technology has been dominating the photovoltaics industry for several decades and shows no sign of giving up its leading position. There are many reasons for that and the fast ramp up of production volumes combined with an aggressive learning curve makes it hard for challengers to compete. But another element should not be underestimated when considering the dominance of crystalline silicon: the constant improvement of the base material, the crystalline silicon wafer.
In this presentation, we will identify the main trends the silicon wafers have experienced along the years: the secular battle between mono and multi, the reduction of thickness incentivized in periods of raw material shortage, the increase of grain size to reduce effect of grain boundaries followed by a reduction of the grain size to avoid intra-grain defects, etc. We will also discuss the market rationale as well as the technical implications of the most recent trends in this domain: rapid increase of the wafer size, cutting of wafers into smaller pieces and emerging interest for n-type technologies.
The discussion will rely 10-years of in-factory experience, and data collection (module bill-of-materials and performance metrics) that STS is enjoying to address the impact of each crystalline silicon wafer trend on all aspects of the module performance.

The Durable Module Materials Consortium (DuraMAT) launched in November of 2016 with five years of funding as part of the Energy Materials Network. DuraMAT is a multi-lab consortium, led by NREL and Sandia with SLAC and LBL as core research labs. The overarching goal of DuraMAT is to discover, develop, de-risk, and enable the rapid commercialization of improved materials, designs, predictive tests and models for photovoltaic modules that increase performance, extend lifetime, and enable new applications. Our core objectives are focused on establishing and applying critical capabilities to address big challenges for PV module durability. These challenges include the community’s inability to make meaningful service life and durability predictions for PV modules with our current testing protocols and knowledge. It is incredibly difficult to make meaningful predictions for 30+ year outdoor durability for cutting edge products with a six month product development cycle! Improving outdoor durability and our ability to predict it requires more durable materials and designs, better tests to screen for weaknesses, modeling, materials and module characterization data, and a way to combine all of this data with historical performance data to extract meaningful results.

With the increasing global demand for silicon photovoltaics (PV), new cell and module technologies are quickly emerging in the market. The implementation of these technologies requires investigation of the long-term performance in order to maintain PV module reliability standards. In this work, we perform multiscale characterization to gain insight into the different failure modes and degradation mechanisms commonly observed in PV modules. Multiscale characterization methods can be used to link observed degradation to specific loss mechanisms (i.e., optical, recombination, resistive) and to the root causes (i.e., changes in chemistry and/or microstructure). We investigate a range of module types with different cell technologies like traditional Al-BSF (Aluminum Back Surface Field) and newly emerging mono/multi-crystalline PERC (Passivated Emitter and Rear Cell). Both field exposed and damp-heat aged modules are investigated to track the long-term degradation. Module and cell level current-voltage (I-V), Suns-V_OC, photoluminescence (PL) and electroluminescence (EL) imaging have been performed on aged and control modules. For the aged modules, cell darkening has been observed in the EL images which can be related to high resistance and fill factor (FF) losses. For the older modules, this darkening happens across the entire cell area, whereas for newer modules with new generation silver (Ag) paste, the darkening is mainly observed to start from the busbars. Our analysis suggests this degradation is related to metallization corrosion. To gain more insight into this failure mode, samples have been cored out from the modules followed by the removal of encapsulation to perform materials characterization. Top-down and cross-sectional SEM and EDS analysis have been performed on the degraded Ag fingers. Furthermore, XPS analysis on the top of the bulk Ag finger and on the glass-layer (after etching away the Ag) has been performed. We observe that the degradation results in delamination of the Ag finger from the cell with voids and gaps forming within the metal-Si interface. EDS and XPS analysis show tin (Sn) in the busbar region and oxidation of the Ag finger. Further investigations are being performed to understand the effect of Sn migration and oxidation on the corrosion and delamination process.

In silicon and perovskite solar cells, the presence of surface states at their electrical contacts is increasingly recognized as a key factor limiting device performance [1,2]. In this presentation, I will first discuss how such surface states result in Fermi-level pinning and surface recombination. Passivating-contact technology aims at minimizing the interface-gap state density, resulting in improved operating voltages and low contact resistances. I will focus on recent advancements in this topic, both for crystalline silicon and perovskite solar cells. In a second part of my presentation, I will discuss the fabrication of efficient monolithic tandem solar cells, combining silicon and perovskite solar cells, aimed at very high-efficient photovoltaics at affordable cost. Specific emphasis will be laid on different possible strategies to build efficient tandems from fully-textured silicon bottom cells; the use of such macro-scale textured bottom cells brings optical advantages, but also brings perovskite/silicon tandems close to industrial implementation [3]. In a final part, I discuss the field performance of such devices and draw conclusions from this for future device development.

[1] E. Aydin, M. De Bastiani, and S. De Wolf, Defect and Contact Passivation for Perovskite Solar Cells, Adv. Mater. 31, 1900428 (2019).
[2] T.G. Allen, J. Bullock, X. Yang, A. Javey, and S. De Wolf, Passivating contacts for silicon solar cells, Nature Energy 4 (2019).
[3] Y. Hou. E Aydin, M. De Bastiani, C. Xiao, F.H. Isikgor, D.-J. Xue, B. Chen, H. Chen, B. Bahrami, A.H. Chowdury, A. Johnston, S.-W. Baek, Z. Huang, M. Wei, Y. Dong, J. Troughton, R. Jalmood, A.J. Mirabelli, T. Allen, E. Van Kerschaver, M.I. Saidaminov, D. Baran, Q. Qiao, K. Zhu, S. De Wolf, and E.H. Sargent, Efficient Tandem Solar Cells with Solution-Processed Perovskite on Textured Crystalline Silicon, Science 367, 1135 (2020).

Silicon PV modules have become a worldwide commodity nowadays. New module types and production technologies are rapidly introduced and the consumption of costly or valuable materials is minimized with increasing conversion efficiencies of sunlight to electricity. The latest approaches utilize, for example, advanced silicon and perovskite silicon tandem cells to increase the conversion efficiencies to more than 30% of the incident sunlight. These 2 examples are selected to assess the current and forthcoming challenges in the reuse and waste treatment sector of end-of-life modules in this presentation. The five topics covered are:

Regulations and standards
Economic and environmental aspects
Decommissioning of PV plants
Preparation for re-use
Recycling and end-of-waste criteria

The PV modules have a long lifetime of beyond 25 years, but with increasing installations sound end-of-life treatment solutions have to be introduced and any potential environmental burdens have to be reduced as far as possible. Even though the waste amounts are rather small compared to other waste streams like electronics, a sound circular economy should be developed with a good eye for the market demands, resources, costs and environment. This approach requires a detailed analysis of the components and materials used in a PV module and an environmental risk and value analysis.

Usually end-of-live PV modules are considered as electronic waste. The approaches to establish a sound waste treatment economy are manifold. E.g., mandatory end-of-life collection and waste treatment systems are already in force in all EU countries and in some areas of USA. At the end of the life of a PV module the collection and treatment is frequently covered by general waste, waste electronic and electrical equipment (WEEE) laws. International and local regulations exist in trans-boundary shipments of used modules and wastes. The waste regulations are applied in combination with the globally harmonized system (GHS). Examples of current practices of waste treatment with recycling and/or landfill disposal are presented. A brief overview about a possible classification of the wastes according to the European list of wastes is presented for both example module types and their possible output materials. End-of-waste criteria for intact modules having entered the waste regime are discussed and, additionally, the major recycling outputs like glass, metals and semiconductor materials are included.

New standards like ANSI/NSF 457-2019 “Sustainability Leadership Standard for Photovoltaic Modules and Photovoltaic Inverters, or CENELEC EN50625-1-4 (2018) on collection, logistics & treatment requirements for WEEE, and other have been introduced. Like in the automotive and the electronics industry stewardship systems and product databases including the individual components are proposed and producer responsibility principles are realized. This is closely linked to the introduction of new assessment methodologies like the European Environmental Footprint (PEF). Recommendations for the cases of a second life of the modules including reuse and repair are under development. The presentation includes an overview about these topics.

Perovskite/silicon tandem solar cells promise power conversion efficiencies (PCE) beyond the Shockley–Queisser limit of single-junction devices, however, their actual outdoor performance is yet to be studied and understood. Here we fabricate two-terminal (2T) monolithic perovskite/silicon tandem solar cells with certified PCE values of 25% and current densities of 19.8 mA/cm² and test them outdoors at elevated temperatures in a hot and sunny climate. With these data, we then calibrated a device-temperature model. By combining this model with relevant meteorological data, we simulated the tandem temperature over a complete year (2016), giving a realistic framework for tandem-performance analysis. Later, we experimentally characterized our tandems in a controlled laboratory environment under various temperatures and illumination spectra, with specific attention paid to their impact on current matching. Finally, to gain further insight, we developed a temperature-dependent radiative-efficiency model to support our findings. We find that the temperature dependence of the bandgaps (E_g) of both crystalline silicon and perovskite – which follow opposing trends – shifts the devices away from current matching for 2T tandems optimized at standard test conditions (STC, i.e., AM1.5G spectrum, 1000 W/m², 25 °C). Consequently, we argue that the optimal perovskite bandgap at STC is lower compared to earlier findings. Through temperature-dependent radiative-efficiency modeling, we show that the ideal bandgap of the perovskite in perovskite/silicon tandems is E_g <1.68 eV (at STC) for optimal field performance at operational temperatures >55 °C. This implies that Br-lean perovskites with narrower bandgaps at STC – and, therefore, better phase stability – hold great promise for the commercialization of perovskite/silicon tandem solar cells.

I will present different strategies to improve the efficiency and power yield of solar cells, both by reducing optical losses of the cell itself as well as by optical design of the surroundings.
As solar researchers during the past decade, we have been privileged to witness solar energy transitioning from a niche technology into the most promising and affordable way to meet our demand for electricity. In many regions of the world, solar cells have reached grid parity with conventional fossil fuel based electricity sources and the cost of solar energy is further declining. Finally, the transition to renewable energy sources is no longer driven by environmental concerns only, but also offers competitive return of investment.
Simultaneously to this rapid cost decline, the efficiency of solar cells has been increasing steadily. Thanks to very careful interface design and reduction of surface recombination, silicon solar cells are on a strong track to approach the practical Auger limit.
Nevertheless, creative approaches from many directions still enable new record efficiencies and power per cost enhancements. Here, I will present some of our strategies to improve the power output of conventional photovoltaic panels by controlling and managing the photon influx. First, I will talk about effectively transparent contacts (ETCs), the only front contact technology eliminating close to all shading losses while providing high conductivity. ETCs are triangular microscale silver contacts featuring high aspect ratio (>1:2.5) and specular side walls which redirect incoming sunlight towards the active area of the solar cell instead of reflecting it away. We have integrated ETCs with different types of solar cells, and observed up to 7% power output enhancement. Upscaling and commercialization of the ETC technology by the start-up ETC Solar is underway.
As the second topic, I will present our experimental study of spectro-angular solar irradiance and the influence on the yield of solar power plants in different locations and configurations. In particular, I will present comparisons between monofacial and bifacial solar cells with respect to the measured spectrum and angle dependent irradiance. Bifacial cells show to be superior both under cloudy as well as under sunny conditions. Under sunny conditions, the benefit mainly arises from ground reflection. The ability of surface to reflect light, also known as albedo, heavily influences the output of bifacial solar cells. The output is strongly dependent on the spectral reflection properties of the albedo material and it can be shown that an ideal spectral albedo material exists. We are designing and fabricating ideal albedo materials with the goal to achieve affordable low x concentration in bifacial solar cells.

A bottom-up approach to creating silicon solar cells has the potential to not only avoid kerflosses incurring in wafer slicing, but also the energy involved in kerfless slicing techniques. Liquid-phase-crystallized silicon (LPC-Si) is one such approach in which amorphous/nanocrystalline silicon is deposited on glass up to a desired thickness (5-40 µm) and crystallized with a line-shaped energy source, which is a laser in this work. LPC-Si of ~14 μm has previously demonstrated power conversion efficiency up to 14.2 % despite illumination through glass, using a heterojunction interdigitated back contact cell architecture (HJ-IBC) [1]. The first part of this work reports efforts to create thicker heterojunction interfaces for better surface passivation. A ‘multilayer’ a-Si surface passivation is also shown which consists of three cycles of a-Si(i) deposition, each one followed by a short hydrogen plasma treatment but with different hydrogen dilution levels. The second part covers laser firing on the electron contact. It enables a controllable trade-off between charge collection and fill-factor by creating a low resistance Ag-Si electron contact, while preserving a-Si:H (i) passivation in other areas [2,3]. Short-circuit current density (J_SC) was observed to be up to 33.1 mA/cm², surpassing all previously recorded values for this technology. Open-circuit voltage (V_OC) of up to 658 mV also exceeded every previous value published at a low bulk doping concentration of 1 x 10^16 cm^-3. Laser firing reduced J_SC by 0.6 mA/cm² on average but improved fill-factor (FF) by 22.5 % absolute on average, without any significant effect on V_OC. Collectively, these efforts have helped achieve a new in-house record efficiency for LPC-Si of 15.1 % with 31.4 mA/cm² J_SC, 645 mV V_OC and 74.7 % FF. The best cell in terms of J_SC (32.7 mA/cm²) and V_OC (657 mV) lacked only in series resistance and shows potential to surpass the current best cell performance in the near future. While excluding the effects of series resistance by measuring a Suns-V_OC curve, this cell shows a pseudo-efficiency of 17 %. Therefore, current efforts are focused on creating larger electron contact widths and heterojunction layers optimised for lower contact resistance.

[1] Trinh, Cham Thi, Natalie Preissler, Paul Sonntag, Martin Muske, Klaus Jäger, Martina Trahms, Rutger Schlatmann, Bernd Rech, and Daniel Amkreutz. "Potential of interdigitated back-contact silicon heterojunction solar cells for liquid phase crystallized silicon on glass with efficiency above 14%." Solar Energy Materials and Solar Cells 174 (2018): 187-195.
[2] Garud, Siddhartha, Cham Thi Trinh, Holger Rhein, Sven Kuehnapfel, Stefan Gall, Rutger Schlatmann, Bernd Rech, and Daniel Amkreutz. "Laser firing in silicon heterojunction interdigitated back contact architecture for low contact resistance." Solar Energy Materials and Solar Cells 203 (2019): 110201.
[3] Garud, Siddhartha, Matevz Bokalič, Cham thi Trinh, Bernd Rech, Daniel Amkreutz, and Marko Topič. "Analysis of Surface Passivation and Laser Firing on thin-film silicon solar cells via Light-Beam Induced Current." IEEE Journal of Photovoltaics (2020) (Accepted for publication)

To be able to continue its steady reduction of solar module cost, the photovoltaic industry must keep increasing the usage efficiency of all the materials, particularly that of silicon (Si). For the latter, the three main levers are increasing cell efficiency, reducing cell thickness and reducing all the losses along the wafer manufacturing chain. To reach below today’s average consumption of 4 g/Wp, incremental improvements can be still be envisaged. However, going far below that value, down to 1 g/Wp, requires disruptive technologies. The 2019 ITRPV roadmap is thus predicting the increase in appearance of kerfless multi- and monocrystalline wafers on the market.

A wide range of kerfless wafering methods have been developed as alternatives to the traditional Czochralski wafer manufacturing process. These disruptive concepts can supply wafers thinner than 100 µm and increase the Si utilization rate by shortcutting the Cz process chain. These various methods can be classified according to the process step at which they introduce the shortcut, and at which silicon is available in different states: solid/ingot, liquid/melt or gas. At the end of the chain, are kerfless wafering methods, such as spalling, that propose alternatives to wafer sawing. One step upstream are methods that enable growing multi- or monocrystalline wafers directly from the melt, without ingot growth, such as ribbon growth techniques or Epilift. And further upstream, skipping the energy-intensive poly-Si production step, are methods that grow wafers from the gas (silane) phase, such as epi-Si lift-off. Except for the ribbon-growth methods, all rely on the lift-off of a thin silicon wafer or seed layer with manifold re-uses of the parent substrate (wafer or ingot). The first reported lift-off silicon foil, by Milnes and Feucht, dates back to 1975. Forty-five years later, the concept has been explored in countless labs and has resulted in dozens of creative ways of producing thin and high-quality silicon layers. Only very few have the potential however to make it from lab to fab.

We have recently reviewed this rich field of silicon lift-off techniques for photovoltaics and propose to give an overview of all the different innovative ways that have been imagined and realized to fabricate thin silicon slabs with minimal silicon wastage, and to try and understand what has determined their fade-out or survival. And what may determine their survival or revival in the future. Today, even though the drive for the entry of kerfless lift-off methods in the PV market is still rather low and hindered by different technological hurdles, they may, with their lower carbon dioxide footprint, be better suited to the world of tomorrow.

Introducing thin, light-weight and high efficiency photovoltaics (PV) will make solar cells more suitable to be integrated in urban landscapes or even small gadgets and would largely contribute to solving the global warming threat that we are facing today. Stacking of solar cells with different characteristic bandgaps is the most common strategy to increase efficiency to surpass the Shockley-Queisser limit, but such tandem devices are typically heavy weight, rigid and costly. An effective approach to tackle this problem is nanopatterning the absorbing layer.

In this work we introduce various hyperuniform designs incorporated on a 1-um thin silicon film as an effective way to control scattered light into particular range of angles to be efficiently absorbed in silicon. These structures are first designed by a 6-fold optimization to achieve 2D designs, and later Meep simulations to optimize absorption spectra in 3D structures. We experimentally show the redirection of light into the engineered range of angles as a ring in the k-space, which matches greatly with the expected structure factors. Further, using an integrating sphere we measure 65% sunlight absorption in the free-standing 1 um c-Si with fabricated hyperuniform designs. The equivalent photocurrent derived from our measurement is far above the highest found in literature, and the estimated photovoltaic efficiency for a state-of-the-art ultra-thin solar cell incorporating our designed nanotexturing is above 20%. This is a highly remarkable efficiency for such a thin indirect-bandgap material and establishes a new breakthrough in thin light-weight and flexible solar cells.

Silicon heterojunction (SHJ) solar cells coupled with hydrogenated intrinsic amorphous silicon (a-Si:H) layers enable high open-circuit voltages (V_oc) of up to 750 mV, but suffer from short-circuit current density (J_sc) losses due to parasitic absorption in a-Si:H, transparent conductive oxide (TCO) layers and shadowing of the metal fingers. Significant improvement in V_oc is expected only for thinner wafer. Increasing J_sc is a more promising route to further improve cell efficiency. Besides the use of high bandgap materials such as nanocrystalline silicon oxide or low carrier density TCO layers to reduce parasitic absorption, an improved light trapping that enables more light to be absorbed within the silicon wafer represents another important opportunity for J_sc improvement, which becomes even more significant when thinner wafers are used. In this work, a thin intrinsic a-Si:H layer was introduced after metallization of SHJ solar cell, formed a TCO/a-Si:H back reflector. Unlike sophisticated structures such as diffraction gratings, black silicon or other kind of nanophotonics, TCO/a-Si:H back reflector is much easier to be fabricated. Compared with the very often used silver back reflector, a-Si:H is less expensive and can still keep some of the bifaciality of the cells.

The intrinsic a-Si:H layer was deposited by plasma enhanced chemical vapor deposition (PECVD). It has a refractive index of around 4 at infrared (IR) wavelengths extracted from ellipsometry measurement. Sentaurus TCAD simulation was used for the rear optics in terms of optimizing the thickness of TCO/a-Si:H stack. We systematically investigated how the TCO/a-Si:H back reflector increases rear internal reflectance by measuring reflectance and external quantum efficiency before and after deposition of the a-Si:H layer. As predicted from simulation, due to increased IR reflection, J_sc was 0.4 mA/cm² higher with TCO/a-Si:H back reflector when applied on SHJ solar cells using 110 mm thickness wafer, while fill factor (FF) and V_oc remained constant. Combined with an optimized front MgF₂ double antireflection coating, we could demonstrate a 0.9 mA/cm² improvement in J_sc, which boosts the cell efficiency by 0.5%. Ultimately, a M2 size total area cell efficiency of 23.7% has been achieved with V_oc of 745 mV, J_sc of 39.4 mA/cm² and FF of 80.8% on 110 um thin wafer. It’s worth noting that the cell still kept a bifaciality of more than 50%. The reflector concept introduced here has a possibility to be used in advanced double-glass module encapsulation.

Hydrogenated amorphous silicon (a-Si:H) is a key material in realizing efficient passivating contacts for Si solar cells known as the silicon heterojunction (SHJ) technology. However, one drawback of the SHJ solar cells is that the a-Si:H, particularly when doped with boron, gives rise to a parasitic optical absorption loss due to its high absorption coefficients in the visible wavelengths. This adverse effect is pronounced in bifacial solar cells, as this type of device allows illumination from both the front and the rear surfaces. Here we pursued alternative materials focusing on hole-selective passivating contacts: one is the hydrogenated nanocrystalline Si (nc-Si:H) and the other is the novel metal-oxide material. For the nc-Si:H material, it is demonstrated that the replacement of (p)a-Si:H by (p)nc-Si:H provides not only an increase of J_sc, but also improvements in V_oc and FF. This results in an efficiency increase by 0.6% absolute compared to our reference cell with (p)a-Si:H, and a best cell efficiency of 23.5%. The factors that alter the optimum hole contact layer thickness depending on the material and deposition condition will be discussed. For the metal-oxide material, we have developed an atomic-layer-deposited TiO_x nanolayer that functions as an efficient and transparent hole-selective passivating contact, leading to proof-of-concept solar cells with efficiencies above 20%. This new function of TiO_x is opposite from the previous understanding that it acts solely as electron contact.

The use of metal-insulator-semiconductor (MIS) tunnel diodes as photovoltaics (PVs) relies on the Schottky barrier between the metal and semiconductor, which induces band bending in the semiconductor absorber allowing for efficient separation of electrons and holes. In the years following early MIS PV research in the 1970s and 1980s, the advent of atomic layer deposition (ALD) has provided a tool for enhanced control of ultrathin film deposition. One method to potentially improve MIS PV performance is the use of ALD oxide bilayers as tunnel insulators. Oxides with different oxygen areal densities have been demonstrated to form interfacial dipoles, manipulating the Schottky barrier. Previous studies have only explored the impact of dipoles between a single ALD oxide and interfacial SiO_x. The present work combines two ALD oxides, one (AlO_x) with oxygen areal density higher than that of SiO_x and another (LaO_x) with oxygen areal density lower than that of SiO_x according to literature. It has been hypothesized that the greater oxygen areal density difference would result in a greater impact on the Schottky barrier. It is demonstrated that, compared to AlO_x alone, an ALD LaO_x/AlO_x stack increases the average n-Si/Ni Schottky barrier height from 0.63 to 0.70 eV, and the average p-Si/Al barrier height from 0.81 to 0.90 eV. LaO_x was the oxide in direct contact with the Si substrate. The observation that the barrier height increases for both substrate types suggests that the effect is due to the ability of LaO_x to depin the Fermi level rather than an interfacial dipole.

Carrier selective contacts have become one of the leading advancements in photovoltaics with the most efficient structures exceeding 26% conversion efficiency. Selective carrier conduction implies that one type of charge carrier is preferentially collected at an electrical contact. Carrier selectivity increases the efficiency of solar cells by reducing recombination at metal contacts and avoiding highly doped emitter layers. Popular selective contact materials are often transition metal oxides (TMOs) due to high optical transparencies and large work functions that are useful for hole selective contacts to silicon. In contrast to previous articles, recent studies have indicated that the interface of p-type silicon and the molybdenum oxide (MoO_x) exhibits a significant Schottky barrier that decreases the efficiency of hole-selective contacts by impeding majority carrier hole collection. To alleviate this issue, atomic layer deposited (ALD) alumina (Al₂O₃) might be added between Si wafer and MoO_x with the expectation to generate a negative interface fixed charge (N_f) after annealing, decreasing band bending. ALD Al₂O₃ also has the ability to passivate interface traps with diffusion of precursor hydrogen during annealing. In this work, we hypothesize that insertion of a tunneling ALD Al₂O₃ layer between MoO_x and p-type Si will enable Schottky barrier height (Φ_bh) minimization that is tunable with N_f, enabling a high efficiency hole-selective contact.
Since there should not be quantifiable N_f in tunneling Al₂O₃ using capacitance-voltage (C-V) analysis, non-tunneling n-Si|Al₂O₃|Al MOSCAP structures were used to determine ideal processing for maximum negative N_f. The ALD growth of Al₂O₃ utilized trimethylaluminum and H₂O at 4 deposition temperatures. Half of the samples were annealed at 425°C in a nitrogen atmosphere prior to the application of Al contacts. After N_f quantification, p-Si|Al₂O₃|MoO_x|Al tunnel diodes with thin (~1 nm) as-deposited and annealed Al₂O₃ layers were fabricated to determine Φ_bh. Current density-voltage-temperature (J-V-T) and Mott-Schottky 1/C²-V measurements were used to establish a range for Φ_bh. Both methods showed that the as-deposited tunneling Al₂O₃ diodes had a higher Φ_bh compared to the diodes with annealed tunneling Al₂O₃, and therefore also supported the existence of a Φ_bh at the Si|MoO_x interface. Future work will involve using larger barrier height diodes using n-Si to more accurately calculate Φ_bh and confirm or reject the phenomenon observed.

Numerous alternative materials and solar cell device structures have been widely explored in order to reduce the cost of photovoltaic (PV) manufacture. Organic-inorganic component solar cells based on nanostructured semiconductor have built up in past two decades, which may promise the potentially low cost and high performance. However, the device performances are generally lower than its pristine all-inorganic PV devices, resulting from the numerous surface defect and improper organic-inorganic phase segregation. Here, we demonstrate that hybrid PVs based on organic conjugated molecular and nanostructured silicon can achieve a high power conversion efficiency by controlling the interface contact as well as surface passiviation.
An advantage of this type heterojunction devices presents the excellent light harvest capability of nanostructured as well as benefiting likely simple fabrication process. Especially, the composite of conjugated organic materials and nanostructured inorganic materials are potential candidates for cost-effective and efficient solar-energy-harvesting devices¹. This device can be operated by a light-modulated field effect solar cell, which can improve the device performance under accumulation of light illumination². In addition, this type device can be easily integrated with other type energy device, for example, an energy harvesting structure that integrates an organic-silicon heterojunction solar cell and a triboelectric nanogenerator (TENG) device is built to realize power generation from both sunlight and raindrop³. In addition, a self-charging power unit based on an organic-Si heterojunction solar cell and an organic supercapacitor, which simultaneously achieved both photoelectric conversion and energy storage⁴. Finally, we will demonstrate that low temperature processed conductive silver-based ink has wide application usage for low temperature processed solar cell and related plastic electronics.

1. Sun, B.; Shao, M.; Lee, S.-T. Advanced Materials 2016, 28, (47), 10539-10547.
2. Wang, Y.; Xia, Z.; Liu, L.; Xu, W.; Yuan, Z.; Zhang, Y.; Sirringhaus, H.; Lifshitz, Y.; Lee, S. T.; Bao, Q.; Sun, B. Advanced Materials 2017, 29, (18), 1606370.
3. Liu, Y.; Sun, N.; Liu, J.; Wen, Z.; Sun, X.; Lee, S.-T.; Sun, B. ACS Nano 2018, ACS Nano 2018, 12 (3), 2893-2899.
4. Liu, R.; Wang, J.; Sun, T.; Wang, M.; Wu, C.; Zou, H.; Song, T.; Zhang, X.; Lee, S.-T.; Wang, Z. L.; Sun, B. Nano Letters 2017, 17, (7), 4240-4247.

Photo-corrosion of anode participating the oxygen evolution reaction (OER) in photoelectrochemical cell (PEC) for water splitting is one of the obstacles for the long term stability and efficient solar to hydrogen conversion. In order to prevent the phot-corrosion, “electrically leaky” thick TiO₂ film was deposited onto a n-Si photoanode surface. However, the carrier transport mechanism through the thick dielectric layer and the influence of oxygen vacancy-related midgap states on the mechanism are still unclear. In order to verify the carrier transport through the midgap states of thick TiO₂ film, we only modulated the defect density of the protective TiO₂ film with no significant change of optical, physical, and chemical properties. It has been demonstrated that the defect density of the TiO₂ film is proportional to water-splitting activity. This result enables to explain the hole transport mechanism of the previously reported electrically leaky TiO₂ protection layer in n-Si photoanode. For the defect-level optimization, titanium sub-oxide (sub-TiO_x) protection layer with higher defect density compared to the TiO₂ layer was also investigated. The controlled incorporation of defects into TiO_x dramatically enhances the hole transport from the photoanode surface to electrolyte solution. The relationship between band structure and PEC activity of the photoanode system was explored depending on the defect density of the protection layer. Mott-Schottky analysis of this system suggests that the defect level of TiO_x films influences the band bending of n-Si, which governs the accessible density of defect states and the carrier recombination.

We report on improvement in the passivation quality of titanium oxide (TiO_x) by doping with tantalum (Ta). TiO_x has been considered one of the most promising materials for electron-selective contacts (ESCs) for crystalline silicon (c-Si) solar cells. There has been a high efficiency of 23.1% successfully achieved for a c-Si solar cell with a TiO_x rear-side ESC [1]. To realize higher efficiencies, it is necessary to further improve the passivation quality. One possible candidate for realizing such improvements is Ta doping. Ta doping reportedly causes a negative shift of the flat band voltage of TiO_x films [2], and the passivation quality of TiO₂ ESCs can therefore be improved by Ta doping. In addition, tantalum oxide (TaO_x) has long-term stability against UV light irradiation [3,4], and it is possible that Ta doping may also improve the long-term stability. In this work, we focus on an improvement in the passivation quality of TiO_x ESCs by doping with Ta.

We deposited TiO_x:Ta films by atomic-layer deposition on the both-sides of n-type c-Si substrates which were in advance oxidized in 30% hydrogen peroxide solution at 80 °C. Ta doping was realized by inserting a unitcycle of TaO_x monolayer into unitcycles of TiO_x. In this study, Ta doping concentration is characterized by a ratio of the number of TaO_x unitcycles to that of TiO_x unitcycles. The thickness of TiO_x:Ta passivation films was basically set to 3.5 nm. After the deposition, the TiO_x:Ta-passivated c-Si substrates were annealed at 130 °C in air. The effective minority-carrier lifetimes (τ_effs) were acquired from quasi steady-state photoconductance measurements, which were used as an indicator for the passivation qualities of the TiO_x:Ta ESCs.

The All the samples passivated with undoped TiO_x films and with TiO_x:Ta films with unitcycle ratios of 0.05, 0.09, 0.17, and 0.50 showed improvements in the τ_eff by annealing at 130 °C. However, the τ_eff of samples with TiO_x:Ta after annealing was, in most cases, higher than that of samples with undoped TiO_x. This indicates that doping with Ta improves the passivation quality of TiO_x ECSs. In addition, TiO_x:Ta with a relatively low amount of Ta doping tended to have high passivation qualities. The TiO_x:Ta with a unitcycle ratio of 0.05 showed the highest τ_eff of 784 μs in this study, which corresponded to an effective surface recombination quality of 24.9 cm/s. The TiO_x:Ta with a unitcycle ratio of 0.50 exhibited a rather lower passivation quality than that of the undoped TiO_x, which indicates excess Ta doping rather degrades the passivation quality.

The above results imply that TiO_x:Ta is one of the promising materials for ESCs. The results of experiments to elucidate the mechanism of the improvements in the passivation quality will be presented.

[1] J. Bullock et al., Adv. Energy Mater. 9 (2019) 1803367.
[2] X. Feng et al., Angew. Chem. Int. Ed. 48 (2009) 8095.
[3] Y. Wan et al., ACS Energy Lett. 3 (2018) 125.
[4] T. Wang et al., Ind. Eng. Chem. Res. 58 (2019) 5510.

Multijunction all-perovskite solar cells offer a route towards efficiencies of III-V materials at low cost by combining the advantages of low thermalization loss in multijunction architectures with the beneficial properties of perovskites - namely low processing cost, high throughput fabrication, and compatibility with flexible substrates. Here, I will discuss why flexible all-perovskite tandems are a very attractive iteration of the perovskite technology. I will focus on what is needed to make a good perovskite-perovskite tandem on a flexible substrate, and the advances that we have made towards this goal.
I will cover advances including a) the development and understanding of wide bandgap materials that leverage a novel bandgap-tuning strategy (without increasing bromine, hence avoiding significant phase segregation) to reach the ideal bandgaps for all-perovskite tandems, resulting in efficient devices that are stable to the combined stresses of heat and light; b) advances that we have made in interlayer design, working towards more ideal material properties for reducing shunt resistance; and c) the integration of these approaches to develop >23% efficient monolithic all-perovskite tandems on rigid substrates and 21% efficient devices on flexible plastic substrates.

Perovskite solar absorbers have the potential to disrupt the photovoltaic industry predominantly with their simultaneous promise of high efficiency and low cost. Whilst these two properties ensure a steady stream of research and development opportunities, commercial scaling requires assuring that the devices are sufficiently characterized to be confident of their long-term performance. Oxford PV has been addressing the challenges associated with commercially scaling its perovskite on silicon tandem solar cell design. Key practices that have been adopted during the development of highly efficient, reliable, and cost-competitive solar cells are the use of rigorous characterization tools and techniques, and the development of necessary models. This talk will highlight some of the challenges and considerations associated with bringing perovskite on silicon tandem solar cells from a laboratory prototype to a commercial product ready for manufacturing and how characterization has shaped and continues to inform that process.

Single-junction crystalline silicon solar cells dominate the photovoltaics market today, but their power conversion efficiencies (PCEs) are approaching the practical limit for this technology. The emergence of hybrid organic-inorganic metal halide perovskite semiconductors over the past decade has opened a window of opportunity to advance the performance of the market-dominating photovoltaic technologies through the development of tandem solar cells. These tandem solar cells promise high PCEs (>33%). To reach this landmark, a number of scientific and technological challenges related to high-efficiency perovskite-based tandem solar cell architectures, stable and high-quality perovskite semiconductor materials, and light management concepts need to be addressed. In this presentation, our progress on passivation strategies and novel interlayers to reduce the voltage losses in wide-bandgap and low-bandgap perovskite solar cells (PSCs), and in turn to increase the PCE of perovskite-based tandem photovoltaics, will be presented.
The search for efficient passivation interlayers for wide-bandgap perovskite top solar cells in n-i-p architecture directed our research in the past towards solution-processed 2D/3D perovskite heterostructures. Making use of these heterostructures for interface passivation, wide-bandgap PSCs (E_g ~1.72 eV) with record open-circuit voltage for this bandgap and very high PCEs of up to 19.4% were demonstrated [1]. The reduced voltage deficit is a result of the decreased non-radiative recombination losses at the perovskite/hole-transport layer (HTL) interface. The solution-processed 2D/3D perovskite heterostructure is compatible with 3D perovskite absorber layers of a wide range of bandgaps (1.65 eV ≤ E_g ≤ 1.85 eV) [2]. Using this strategy in 4-terminal perovskite/silicon and perovskite/CIGS tandem solar cells, high PCEs > 27% [3], and > 25% [2] were demonstrated, respectively, thereby significantly exceeding the record PCESs reported for the respective single-junction solar cell technologies.
Beyond these past results, at the MRS fall meeting we will further present our latest work on novel interlayers and passivation strategies to minimize the voltage loss in wide-bandgap (1.65 eV ≤ E_g ≤ 1.75 eV) as well as low-bandgap (E_g ~ 1.26 eV) perovskite solar cells in p-i-n architecture. The p-i-n architecture is a key to realize high-performance 2-terminal perovskite/silicon as well as perovskite/perovskite tandem solar cells. Moreover, we will present a novel route to deposit uniform ultra-thin polymer passivation layers onto perovskite thin films by chemical vapor deposition (CVD) [4]. The CVD polymerization promises reliable, scalable, and homogeneous deposition for future large-area processing of perovskite-based tandem solar cells.

References:
[1] Gharibzadeh, S. et al. Record Open-Circuit Voltage Wide-Bandgap Perovskite Solar Cells Utilizing 2D/3D Perovskite Heterostructure. Adv. Energy Mater. 9, 1–10 (2019).
[2] Gharibzadeh, S.; Hossain I. M. et al. 2D/3D Heterostructure for Semitransparent Perovskite Solar Cells with Engineered Bandgap Enables Efficiencies Exceeding 25% in Four-Terminal Tandems with Silicon and CIGS. Adv. Funct. Mater. 30, (2020).
[3] Duong, T. et al. High Efficiency Perovskite-Silicon Tandem Solar Cells: Effect of Surface Coating versus Bulk Incorporation of 2D Perovskite. Adv. Energy Mater. 10, 1903553 (2020).
[4] Malekshahi Byranvand M.; Behboodi-Sadabad F. et al. Chemical Vapor Deposited Polymer Layer for Efficient Passivation of Planar Perovskite Solar Cells. submitted

Integrating metal halide perovskite top cells with crystalline silicon (c Si) bottom cells into monolithic tandem devices has recently attracted increased attention due to the high efficiency potential of these cell architectures. To further increase the tandem device performance to a level well above the best single junctions, optical and electrical optimizations as well as a detailed device understanding for such architectures need to be developed [1]. Here we present our recent results on such monolithic tandem cells in the common 2-terminal (2T) as well as in a novel 3-terminal (3T) contacting scheme.
In the 2T design, we recently demonstrated a certified power conversion efficiency (PCE) of 29.15% [2,3]. I will discuss some key elements to this result: using a silicon heterojunction cell electron contact on the rear side, enabling the integration of a perovskite top cell in the p-i-n configuration; employing self-assembled monolayers (SAMs) as hole selective contacts for further reduction of non-radiative recombination losses, etc. With the latter, efficiencies of up to ~ 21 % were demonstrated for p-i-n perovskite single junctions [3].
In 3T tandems, one of the three contacts is shared by both subcells, allowing for the individual operation of both subcells in a monolithic device with a single substrate, thus combining the characteristic advantages of 2T and 4-terminal (4T) devices: high combined photogeneration currents as in state-of-the-art 2T tandems, and ideally independent electrical operation similar to 4T tandems, making the cell more robust against varying spectra. I will discuss our recent proof-of-concept study [4], demonstrating a 3T perovskite/silicon tandem cell by manufacturing an n-i-p perovskite top cell on an interdigitated back contact (IBC) silicon heterojunction solar cell. In our proof-of-concept device, the combined PCE of ~ 17% is still limited both optically and electrically, and it will be discussed how optimizations of the cell stack can bring this architecture to a PCE level comparable to 2T tandems. Furthermore, I will demonstrate that the subcells can indeed be operated largely independently at their individual maximum power points, allowing to extract the maximum power output for varying spectral conditions similar to 4T cells. A minimized series resistance at the shared electron contact on the cell’s rear side is identified as the crucial parameter for independent operation of the subcells, and I will discuss pathways to approach the simulated efficiency potential by an optimized device design.

References:
[1] Jost, Kegelmann, Korte, Albrecht, Adv. Energy Materials, 2020, doi: 10.1002/aenm.201904102
[2] NREL Best Research-Cell Efficiency Chart, https://www.nrel.gov/pv/cell-efficiency.html
[3] Köhnen, Jost, Stannowski, Albrecht et al., Sustainable Energy and Fuels, doi: 10.1039/C9SE00120D
[3] Al-Ashouri, Magomedov, Korte, Albrecht et al., En. Env. Sci, 2019, doi: 10.1039/C9EE02268F
[4] Tockhorn, Wagner, Korte et al., ACS Appl. En. Mat., 2020, doi: 10.1021/acsaem.9b01800

Among all commercialization routes for perovskite photovoltaics (PV), the perovskite/Si tandem is of particular interest as it capitalizes the mature silicon industry with decades of experience. Here we report low-temperature solution process for efficient two-terminal perovskite/Si tandem solar cells, wherein electrical and optical properties of the tunneling junction are carefully modulated. Moreover, we improve materials stability by introducing Eu²⁺/Eu³⁺ “redox shuttle” to suppress the Pb⁰/I⁰ impurities which inevitably occurred during device operation. In addition, we carefully analyze the cost of tandem devices in terms of levelized cost of energy (LCOE) by adopting the “bottom-up cost model”. It reveals perovskite PV exhibit low materials cost, which reduces the LCOE substantially in tandem devices. We conclude that further efforts are required to simultaneously improve the efficiency and lifetime of perovskite modules, which ultimately endows the competitiveness of perovskite tandem solar cells.

Lifetime prediction is a critical first step as the basis of a warranty or risk assessment. Research that identifies a PV module’s primary degradation or failure mechanism, quantifies the associated acceleration factors, and applies those to operating conditions is a straightforward approach to lifetime prediction. The value of this process is obvious and researchers around the world are investigating degradation and failure mechanisms of PV modules with this goal in mind. Making a lifetime prediction is actually quite easy: we can predict 25 years ±25 years for pretty much all PV products on the market, but such a prediction is useless. Making a lifetime prediction with a small uncertainty that leads to a useful prediction is much harder. This talk will describe why lifetime prediction is difficult, when it is essential, when it is a distraction, and how we expect lifetime prediction may evolve in the future.

With the lifespans of photovoltaic (PV) modules expected to last >20 years, rapid and spatially-resolved characterization methods are essential for understanding module reliability and degradation processes. Luminescence and thermal imaging have been gaining popularity both for fast, outdoor-compatible module inspection and for more detailed post-mortem analysis of failure mechanisms. Importantly, many failure mechanisms are not evident by visual inspection, and characterization of electrical properties alone is insufficient to identify the weakest features of cell or module design. This is especially true as PV technologies and module packaging schemes become more advanced and degradation becomes more subtle. Here, we present various examples of PV module degradation mechanisms in field-weathered modules, where we (1) locate problem areas with luminescence and thermal imaging, (2) core these areas out of the module, and (3) use various microscopy methods to identify chemical and structural causes of power loss. We particularly highlight causes of voltage loss or series resistance increases that vary depending on the technology, and we show studies including conventional Al back surface field (Al-BSF), silicon heterojunction with intrinsic thin layer (HIT), and passivated emitter and rear contact (PERC) modules. These results highlight the utility of imaging methods for improving module reliability.

Perovskite solar absorbers have the potential to disrupt the photovoltaic industry predominantly with their simultaneous promise of high efficiency and low cost. Whilst these two properties ensure a steady stream of research and development opportunities, commercial scaling requires assuring that the devices are sufficiently characterized to be confident of their long-term performance. Oxford PV has been addressing the challenges associated with commercially scaling its perovskite on silicon tandem solar cell design. Key practices that have been adopted during the development of highly efficient, reliable, and cost-competitive solar cells are the use of rigorous characterization tools and techniques, and the development of necessary models. This talk will highlight some of the challenges and considerations associated with bringing perovskite on silicon tandem solar cells from a laboratory prototype to a commercial product ready for manufacturing and how characterization has shaped and continues to inform that process.

Single-junction crystalline silicon solar cells dominate the photovoltaics market today, but their power conversion efficiencies (PCEs) are approaching the practical limit for this technology. The emergence of hybrid organic-inorganic metal halide perovskite semiconductors over the past decade has opened a window of opportunity to advance the performance of the market-dominating photovoltaic technologies through the development of tandem solar cells. These tandem solar cells promise high PCEs (>33%). To reach this landmark, a number of scientific and technological challenges related to high-efficiency perovskite-based tandem solar cell architectures, stable and high-quality perovskite semiconductor materials, and light management concepts need to be addressed. In this presentation, our progress on passivation strategies and novel interlayers to reduce the voltage losses in wide-bandgap and low-bandgap perovskite solar cells (PSCs), and in turn to increase the PCE of perovskite-based tandem photovoltaics, will be presented.
The search for efficient passivation interlayers for wide-bandgap perovskite top solar cells in n-i-p architecture directed our research in the past towards solution-processed 2D/3D perovskite heterostructures. Making use of these heterostructures for interface passivation, wide-bandgap PSCs (E_g ~1.72 eV) with record open-circuit voltage for this bandgap and very high PCEs of up to 19.4% were demonstrated [1]. The reduced voltage deficit is a result of the decreased non-radiative recombination losses at the perovskite/hole-transport layer (HTL) interface. The solution-processed 2D/3D perovskite heterostructure is compatible with 3D perovskite absorber layers of a wide range of bandgaps (1.65 eV ≤ E_g ≤ 1.85 eV) [2]. Using this strategy in 4-terminal perovskite/silicon and perovskite/CIGS tandem solar cells, high PCEs > 27% [3], and > 25% [2] were demonstrated, respectively, thereby significantly exceeding the record PCESs reported for the respective single-junction solar cell technologies.
Beyond these past results, at the MRS fall meeting we will further present our latest work on novel interlayers and passivation strategies to minimize the voltage loss in wide-bandgap (1.65 eV ≤ E_g ≤ 1.75 eV) as well as low-bandgap (E_g ~ 1.26 eV) perovskite solar cells in p-i-n architecture. The p-i-n architecture is a key to realize high-performance 2-terminal perovskite/silicon as well as perovskite/perovskite tandem solar cells. Moreover, we will present a novel route to deposit uniform ultra-thin polymer passivation layers onto perovskite thin films by chemical vapor deposition (CVD) [4]. The CVD polymerization promises reliable, scalable, and homogeneous deposition for future large-area processing of perovskite-based tandem solar cells.

References:
[1] Gharibzadeh, S. et al. Record Open-Circuit Voltage Wide-Bandgap Perovskite Solar Cells Utilizing 2D/3D Perovskite Heterostructure. Adv. Energy Mater. 9, 1–10 (2019).
[2] Gharibzadeh, S.; Hossain I. M. et al. 2D/3D Heterostructure for Semitransparent Perovskite Solar Cells with Engineered Bandgap Enables Efficiencies Exceeding 25% in Four-Terminal Tandems with Silicon and CIGS. Adv. Funct. Mater. 30, (2020).
[3] Duong, T. et al. High Efficiency Perovskite-Silicon Tandem Solar Cells: Effect of Surface Coating versus Bulk Incorporation of 2D Perovskite. Adv. Energy Mater. 10, 1903553 (2020).
[4] Malekshahi Byranvand M.; Behboodi-Sadabad F. et al. Chemical Vapor Deposited Polymer Layer for Efficient Passivation of Planar Perovskite Solar Cells. submitted

Integrating metal halide perovskite top cells with crystalline silicon (c Si) bottom cells into monolithic tandem devices has recently attracted increased attention due to the high efficiency potential of these cell architectures. To further increase the tandem device performance to a level well above the best single junctions, optical and electrical optimizations as well as a detailed device understanding for such architectures need to be developed [1]. Here we present our recent results on such monolithic tandem cells in the common 2-terminal (2T) as well as in a novel 3-terminal (3T) contacting scheme.
In the 2T design, we recently demonstrated a certified power conversion efficiency (PCE) of 29.15% [2,3]. I will discuss some key elements to this result: using a silicon heterojunction cell electron contact on the rear side, enabling the integration of a perovskite top cell in the p-i-n configuration; employing self-assembled monolayers (SAMs) as hole selective contacts for further reduction of non-radiative recombination losses, etc. With the latter, efficiencies of up to ~ 21 % were demonstrated for p-i-n perovskite single junctions [3].
In 3T tandems, one of the three contacts is shared by both subcells, allowing for the individual operation of both subcells in a monolithic device with a single substrate, thus combining the characteristic advantages of 2T and 4-terminal (4T) devices: high combined photogeneration currents as in state-of-the-art 2T tandems, and ideally independent electrical operation similar to 4T tandems, making the cell more robust against varying spectra. I will discuss our recent proof-of-concept study [4], demonstrating a 3T perovskite/silicon tandem cell by manufacturing an n-i-p perovskite top cell on an interdigitated back contact (IBC) silicon heterojunction solar cell. In our proof-of-concept device, the combined PCE of ~ 17% is still limited both optically and electrically, and it will be discussed how optimizations of the cell stack can bring this architecture to a PCE level comparable to 2T tandems. Furthermore, I will demonstrate that the subcells can indeed be operated largely independently at their individual maximum power points, allowing to extract the maximum power output for varying spectral conditions similar to 4T cells. A minimized series resistance at the shared electron contact on the cell’s rear side is identified as the crucial parameter for independent operation of the subcells, and I will discuss pathways to approach the simulated efficiency potential by an optimized device design.

References:
[1] Jost, Kegelmann, Korte, Albrecht, Adv. Energy Materials, 2020, doi: 10.1002/aenm.201904102
[2] NREL Best Research-Cell Efficiency Chart, https://www.nrel.gov/pv/cell-efficiency.html
[3] Köhnen, Jost, Stannowski, Albrecht et al., Sustainable Energy and Fuels, doi: 10.1039/C9SE00120D
[3] Al-Ashouri, Magomedov, Korte, Albrecht et al., En. Env. Sci, 2019, doi: 10.1039/C9EE02268F
[4] Tockhorn, Wagner, Korte et al., ACS Appl. En. Mat., 2020, doi: 10.1021/acsaem.9b01800

Multijunction all-perovskite solar cells offer a route towards efficiencies of III-V materials at low cost by combining the advantages of low thermalization loss in multijunction architectures with the beneficial properties of perovskites - namely low processing cost, high throughput fabrication, and compatibility with flexible substrates. Here, I will discuss why flexible all-perovskite tandems are a very attractive iteration of the perovskite technology. I will focus on what is needed to make a good perovskite-perovskite tandem on a flexible substrate, and the advances that we have made towards this goal.
I will cover advances including a) the development and understanding of wide bandgap materials that leverage a novel bandgap-tuning strategy (without increasing bromine, hence avoiding significant phase segregation) to reach the ideal bandgaps for all-perovskite tandems, resulting in efficient devices that are stable to the combined stresses of heat and light; b) advances that we have made in interlayer design, working towards more ideal material properties for reducing shunt resistance; and c) the integration of these approaches to develop >23% efficient monolithic all-perovskite tandems on rigid substrates and 21% efficient devices on flexible plastic substrates.

Among all commercialization routes for perovskite photovoltaics (PV), the perovskite/Si tandem is of particular interest as it capitalizes the mature silicon industry with decades of experience. Here we report low-temperature solution process for efficient two-terminal perovskite/Si tandem solar cells, wherein electrical and optical properties of the tunneling junction are carefully modulated. Moreover, we improve materials stability by introducing Eu²⁺/Eu³⁺ “redox shuttle” to suppress the Pb⁰/I⁰ impurities which inevitably occurred during device operation. In addition, we carefully analyze the cost of tandem devices in terms of levelized cost of energy (LCOE) by adopting the “bottom-up cost model”. It reveals perovskite PV exhibit low materials cost, which reduces the LCOE substantially in tandem devices. We conclude that further efforts are required to simultaneously improve the efficiency and lifetime of perovskite modules, which ultimately endows the competitiveness of perovskite tandem solar cells.