It has been proven that Pb-halide perovskite solar cells (PSCs) are 25.5% to 21.6% efficient, which makes them competitive with artificial solar cells made of multicrystalline silicon, cadmium telluride, and copper indium gallium selenide. Pb-halide PSC research is now mainly focused on making big modules and making them more stable over time so that they can be sold. But worries about Pb’s harmful effects on humans and the world still exist. Halide perovskites based on Sn have bandgaps of 1.2 to 1.4 eV and are great for single-junction solar cells. They are getting more attention as a Pb-free option to Pb-halide PSCs. It is likely that these perovskites with narrow bandgaps will be good choices for the bottom cell in perovskite–perovskite tandem solar cells. This part talks about the latest progress, problems, and future hopes for Sn-halide PSCs.
Sn–Pb Perovskite Solar Cells
Background
Halide perovskites, which are often written as ABX3, are a type of solar material that has a bandgap between 1.2 and 1.4 eV. Because they can collect light well and have a narrow bandgap, like single-crystal Si semiconductors, they are perfect for solar cells. The best power conversion efficiency (PCE) for Sn-based halide perovskite solar cells is 33%. On the other hand, PSCs made from Pb-halide perovskites can only get as high as 30%.
The Sn-Pb mixed-cation perovskite has a bandgap between 1.1 and 1.2 eV and a light absorption edge between 1000 and 1100 nm, which is similar to how single-crystal Si semiconductors absorb light. This Sn-Pb mixed PSC with a small bandgap is needed for the bottom cell of all-perovskite tandem cells that want to have more than 30% efficiency.
The efficiency trend of Sn-Pb mixed PSCs, which includes the cells we made in the lab, goes from 20.4% to 23.3% at the start of the study. The PCE wasn’t as high as that of Pb PSCs, though, and the effectiveness has been going up year after year and has now passed 20%. Table 10.3 shows a summary of recent studies that use Sn-Pb-halide perovskites (PVK) and report yields of over 20%.
Stabilization of Sn(II) Ions
Problem with Sn-based perovskites is that Sn2+ changes into the more stable Sn4+, which is a place where charges can come back together. Adding Sn0 (metal) to the precursor solution, lowering the concentration of Sn4+ through a comproportionation reaction, and improving the balance of energy levels at the contact are some of the ways that have been tried to lower its concentration.
Lin et al. found that metallic tin could be used to reduce Sn gaps in mixed Pb–Sn perovskites, which led to a 21.1% efficiency. A combined solar cell with a perovskite top cell and a perovskite bottom cell that uses this Sn–Pb PSC as the top cell has been shown to work 24.8% of the time. Jiang et al. said that adding Sn metal slowed down non-radiative recombination, which made the charge carrier lifetime of the perovskite film go from 115 ns to 701 ns.
Tong et al. discovered that adding guanidinium ion and an antioxidant like formamidine sulfinic acid also lowered the amount of Sn4+ and made the Sn–Pb PSCs work better. They also made two-terminal and four-terminal all-perovskite tandem solar cells that worked 25% and 23.1% of the time, respectively.
Yang et al. found that cadmium ions decreased the number of electron traps and increased the length of electron diffusion. This made the perovskite films thicker, reaching 1000 nm, and produced 20.2% and 22.7% PCEs for Sn–Pb PSCs and perovskite–perovskite tandem cells, respectively.
Gas cooling is a different method that has been looked into for making Sn–Pb PSCs that work very well. Instead of PEDOT-PSS, the authors used polybis(4-phenyl)(2,4,6-trimethylphenyl)amine as the hole transport material (HTM). For 0.06 cm2 and 1 cm2 active areas, the PCEs were 20% and 17.5%, respectively.
Efficiency Enhancement
PSCs made of Sn-Pb perovskite show an increase in efficiency over time based on how much energy they lose during conversion. For Pb PSCs, the Voc loss (eV) and Jsc loss (mA/cm2) are found to be between 0.45 and 0.32 eV and 8.8 and 6.5 mA/cm2, respectively. The FF value is around 0.8. In 2019, the effectiveness was stated to be 20.4%. Table 10.1 shows the makeup of the perovskite and the shape of the solar cell. Not long ago, the effectiveness reached 23.3%.
In 2014, Fan Zuo et al. reported the first reversed structure Sn-Pb PSC (MAPb0.85Sn0.15I3) that worked 10.1% of the time. Instead of just SnF2, the efficiency went up to 15.93% when SnF2 DMSO complex was added. Adding a thin layer of PCBM at the junction between the perovskite layer and the C60 layer made the spike conduction band (CB) structure, which increased the efficiency to 17.59%. The efficiency went up to 18.88% when the F-doped tin oxide (FTO) substrate was used instead of the indium tin oxide (ITO) substrate (No. 5). This is because the FTO substrate has better optical transmittance in the near infrared (800–1100 nm) range, which is where the Sn–Pb perovskite layer collects light. Finally, a Cs ion was added to the A site to reduce the local lattice strain (disordering) (No. 6), which made the efficiency go up to 20.4%.
This is a summary of the improvements that can be made to the device using all of these methods:
- When the reversed structure (No. 2) was used instead of the regular structure (No. 1), FF and Voc got better, but Jsc stayed the same. Adding the SnF2–DMSO complex (No. 3) made FF and Voc much better, while also slightly improving Jsc. The other steps from No. 4 to No. 6 helped improve Jsc while keeping the FF and Voc numbers the same. This finally led to a 23.3% PCE (No. 7).
The normal-shaped PSC had a big Voc loss of 0.78 eV, which is about the same as what amorphous Si solar cells have. The Voc loss was lowered to 0.71 eV by creating an upside-down structure, and it was lowered even more to 0.39 eV by using the way in No. 6. In the beginning, the JSC loss was 20 mA/cm2, but it slowly went down until it was only 6.9 mA/cm2.
Interfacial Engineering and Device Architecture
The study looks at how well Sn and Sn-Pb PSCs work when they are made in shapes that are backwards (p–i–n) instead of straight (n–i–p). The low efficiency of the regular structure was found to be caused by the surface where mesoporous TiO2 and MASnPbI3 meet. A quartz crystal microbalance device was used in situ to track the uptake of PbI2 or SnI2 on the TiO2 layer for the study. Scientists looked at the surface of TiO2 and found Sn–O– or Pb–O– ions. The weight gain didn’t go down when liquids were used to wash the TiO2 layer. This suggests that PbI2 or SnI2 is tightly attached to the TiO2 surface.
It had a single layer of Pb–I or Sn–I molecules with Ti—O—Sn—I or Ti—O—Pb—I bonds on top of it. The surface’s Pb—O—Ti bond was quickly replaced by Sn–O–Ti when it was treated again with SnI2 solution. This suggests that the Sn–O–Ti linkage covered most of the TiO2 surface, even when SnI2 and PbI2 solution were mixed together. Because SnI2 prefers to stick to porous TiO2 layers, there were a lot of Sn ions in the porous TiO2 layer.
The Pb/Sn ratio changed from the top to the bottom of the Sn–Pb perovskite layer that was made on top of a thin TiO2 layer. The Pb/Sn ratio would be one across the whole thickness of the film if MAPb0.5Sn0.5I3 was made. There were, however, times when the Pb/Sn ratio was higher than 1.0 and times when it was less than 1.0. This means that the layer at the bottom had a lot of Sn and the layer on top had a lot of Pb.
The graded structure on the flat substrate stood out more than on the flat substrate. Because the perovskite predecessors don’t all dissolve in water, the graded structure on the flat base makes sense. The energy levels in the conduction band and valence band get lower as the Sn content of Sn–Pb perovskite rises. This makes the layer more p-type. The top layer of the graded structure needs to be in touch with an n-type layer (an electron capture layer) like fullerenes for it to work better. In the end, the arrangement that is turned upside down (p–i–n) is better for efficiency.
We measured thermally stimulated currents (TSC) in a bare TiO2 layer, a TiO2 layer treated with SnI2, and a TiO2 layer treated with PbI2 to see how the carrier traps were spread out in each layer. It was cooled down to the temperature of liquid N2 and then put under deep UV light with a small bias to fill the traps. When the Pb rich(n) Sn rich(p) TCO substrate side was put under deep UV light with a small bias to fill the traps, TSC showed up.
The TiO2 layer’s TSC went down after being treated with PbI2 solution, but it went up after being treated with SnI2 solution. These results show that treating with PbI2 lowers the number of carrier traps and treating with SnI2 raises them. The drop in Voc was caused by the traps that were made at the border between the Sn–Pb perovskite layer and the TiO2 layer. Voc went from 0.2 V to 0.4 V when Y2O3 or C60–COOH was used to passivate the TiO2 layer and the Sn–Pb perovskite layer was kept from coming into direct touch with it.
A regular MAPbI3 light-harvesting layer was used to see how well the device worked with and without the Ti–O–Sn layer. The MAPbI3 solar cells with the FTO/(TiO2)/MAPbI3/P3HT/Ag/Au structure worked much better than the FTO/(TiO2 passivated with SnI2)/MAPbI3/P3HT/Ag/Au structure, where SnI2 was used to passivate the TiO2 layer. This strongly supports the idea that the I–Sn–O–Ti connections make a place for charges to mix at the differentiating layer between the Sn–Pb perovskite and TiO2 layers, which lowers the efficiency of the solar cell.
The efficiency went up from No. 3 to No. 4 in Figure 10.2 when the spike band structure was used at the border between the Sn–Pb perovskite layer and the C60. To stop charge recombination, the spike structure made with the added PCBM layer is suggested. This is because it slows down charge recombination between holes in the Sn–Pb perovskite layer and electrons in C60.
The research is mostly about making Sn-Pb perovskite solar cells (PSCs) that can increase PCE by more than 20%. Sn-Pb PSCs work better when they reduce charge exchange at the interface, make a graded band structure, control light, and fix local lattice disordering. It was discovered that the flow of electrons from the Sn-Pb perovskite layer to C60 was the same in the spike structure with PCBM and the cliff structure without it. This stops charge from recombining at the spike layer, which leads to better fill factor (FF) and short-circuit photocurrent density (Jsc).
The efficiency went up from No. 4 to No. 5 when FTO glass substrate was used instead of ITO substrate. This is because FTO glass substrate has a high optical transmittance in the near infrared range. The rise in JSC helped make things run more smoothly. The most efficient mixture was the one with 2.5% Cs and the least amount of local lattice strain (disordering). The average lifetime found by photoluminescence (PL) got longer as the relative local lattice strain (disordering) went down.
Voc loss and Jsc loss in Sn-Pb PSCs are about the same as those in Pb-PSCs, which are the most efficient at this point in time. But there are still ways to cut down on these costs. Also, the FF value, which is lower (0.80) than the best Pb-PSC (0.85), should be raised to make the cell even more efficient overall. To make Sn-Pb PSCs work better, more tests need to be done to find ways to reduce the charge recombination site, create a light confinement structure, and lower the series resistance while keeping the shunt resistance high.
Pb-free Sn Perovskite Solar Cells
Background
Most likely, Sn, a metal ion in the same group as Pb on the periodic table, can be used instead of Pb in halide perovskites. It is known that Sn-halide perovskites have a smaller bandgap and can absorb light up to 900 nm. This is because atomic orbitals meet and band-levels form. The valence bands of Sn-iodide and Pb-iodide perovskites are made up of hybrid orbitals that are antibonding and are made up of Sn 5s–I 5p and Pb 6s–I 5p, respectively. Sn5s has a lower energy level than Pb6s, which means that its valence band maximum (VBM) and conduction band minimum (CBM) are also lower.
The bandgap of MA perovskites MA(Sn1−xPbx)I3 that are mixed with Sn and Pb changes in a way that is not linear with respect to x. This is known as bandgap bending. The energy difference between the s and p orbitals of Pb and Sn is what makes the bandgap bow. MAPbI3 has a valence band of Pb(s)I(p) and a conduction band of Pb(p)I(p). MASnI3 has a valence band of Sn(p)I(p), and MASn1−xPbxI3 has a conduction band of Pb(p)I(p).
The Sn perovskite’s electric structures are very similar to those of the Pb perovskite. This means that the Sn PSCs should work very well, if not better. At first, yields went from about 5% for regular structures to 13–14% for structures that were turned upside down. Sn-PSCs don’t work very well because they have flaws like Sn2+ vacancies and Sn4+ that is made when Sn2+ oxidises. Sn2+ vacancies are easy to make because they don’t need much heat to form.
To sum up, Sn is a good metal ion to use instead of Pb in halide perovskites because it has a similar electronic structure and a low efficiency. But problems like Sn2+ vacancies and Sn4+ made by oxidation of Sn2+ stay close to the problem, which makes these solar cells less effective.
The main goal of this study is to make Sn-Pb perovskite solar cells (Sn-PSCs) more efficient by lowering or getting rid of the Sn4+ content. The inverted structure of Sn-Pb PSCs is better than the normal structure. The first attempt to make Sn-PSCs with an inverted design led to cells that worked with a PCE of 5.05%. The efficiency was raised to 7.9% by adding a small amount of GeI2. Adding ethylenediamine diiodide (EDAI) to the A site, according to Diau et al., makes it work better. Surface passivation with diaminoethane (DAE) can also help improve efficiency, leading to a 10.65% rise in PCE. Ge ion doping, surface passivation by DAE, and band energy alignment optimisation can all be put together to get a PCE of 13.24%.
There are several types of recent studies that report Sn-PSCs that work better than 10% of the time. These include surface passivation, adding reducing agents to lower the content of Sn4+, improving the bulk structure to include defect-less crystallisation, and finding the best band energy level. Almost every method is meant to lower the quantity of Sn4+. Chen et al. used a divalent organic cation of 4-(aminomethyl)-piperidinium to protect the grain borders of Sn perovskite and found that it worked 10.9% of the time. Li et al. reported an efficiency of 10.17 percent and better stability by passivating the grain border of Sn perovskite with ortho-fluoroaniline, which made Sn perovskite without any flaws.
It has been found that additives can change the direction of crystals, which leads to better performance. Pentafluorophen-oxyethylammonium (FOE) iodide was added to the precursor solution by Meng et al. to control the position of Sn perovskite crystal faces on the substrate. This led to a 10.81% efficiency. It was discovered by Wang et al. that phenylhydrazine hydrochloride could stop the formation of Sn4+. Cells that use this addition have a PCE of 11.4%. Nakamura et al. discovered that a dihydropyrazine product lowers SnF2 to make Sn metal, which removes Sn4+ and makes cells work better.
Increasing the productivity has also been helped by better bulk qualities like crystallinity. Wu et al. described a way to make crystals with no flaws by adding a π-conjugated Lewis base to the predecessors of Sn perovskite and slowing down the crystallisation process. The team led by Liu created an amorphous-polycrystalline structure made up of a tin triple-halide amorphous layer and a Cs-FA tin iodide polycrystals. This structure achieved a PCE of 10.4% and stayed stable for longer.
Ge-Doped Sn Perovskites
A big reason for the poor performance is that pure Sn perovskites have a lot of carriers, which makes them act like metals. They have tried to cut down on this by reducing it. The use of SnF2 has been a well-known way to improve the effectiveness of Sn PSCs because it helps lower the number of carriers. In the beginning, the amount of Sn perovskite carriers dropped from 4.1 × 1020/cm3 to 9.11 × 1017/cm3. A later study found that adding Ge2+ made the carrier density in the Sn perovskite layer go down even more, from 9.11 x 1017/cm3 to 1.69 x 1015/cm3. This is a little higher than the density in MAPbI3 and multicrystalline Si.
By adding SnF2 and GeI2, the carrier mobility of Sn perovskite went from 1.05 to 12.49 and 98.27 cm2/V/s, respectively. When both SnF2 and GeI2 are treated, the best result is seen. Around the edges of PEDOT-PSS and Sn perovskite and between Sn perovskite and C60, there was more Ge than in the main Sn perovskite layer. When 5 mol% of GeI2 was added to the surface, the ratio of Sn4+ to Sn2+ dropped from 1.26 to 0.89. This suggests that the Ge ion slows down the process of Sn2+ turning into Sn4+.
GeI2 lowers the amount of Sn4+ in the system because oxidants like oxygen and liquids can easily change it to Sn2+. Thin layers of GexOy are made on the surface and along the grain boundaries. These layers protect the bulk Sn perovskite layer from oxygen and water, which makes the device more stable. GexOy passivation on the grain border might lower the flaws connected to I− and Sn2+ holes.
Sn2+ gap is another possible charge recombination centre that can be lowered by adding extra SnI2. The Ge2+ ion might help reduce the Sn2+ gap in the lattice, according to a study that is still going on.
Efficiency Enhancement by Grain Boundary Passivation
There are Sn2+ vacancies and Sn4+ sites in the grain border that affect how well Sn-PCSs work. Passivation with diaminoethane (DAE) as the passivation agent has been shown to make these solar cells work better. We used a PSC structure made of ITO, PEDOT:PSS, perovskite, C60, BCP, Ag, and Au (with a perovskite formula of FA0.98EDA0.01SnI3) to check how well the passivation worked.
The tests showed that Jsc, Voc, FF, and efficiency were 24.37 mA/cm2, 0.49 V, 0.61%, and 7.28%, respectively, when the surface wasn’t passivated. It was passivated with 0.05 mM chlorobenzene, which made Jsc, Voc, FF, and PCE better to 23.09 mA/cm2, 0.60 V, 0.73%, and 10.18%, in that order. The passivation effect worked best at 0.05 to 0.10 mM amounts.
Through intensity-modulated photocurrent spectroscopy (IMPS), the transport lifetime did not change before or after DAE passivation. However, the charge recombination lifetime got longer, going from about 6 μs before passivation to about 15 μs after it. Also, the charge collection efficiency went up from about 30% to about 72%, and the charge diffusion length went up from 170 to 220 μm after the passivation.
A makeup high in iodine works against Sn2+ at the grain border to help Sn4+ form. In our testing, the Sn2+-rich FAxSnyIx+2y mixture worked better, and the SnI2/FAI ratio could be changed. These results may back up the results of the math.
Improving the effectiveness of Sn-PSCs can also be done by changing the cation makeup at the A spot. For example, adding (CH2-NH3+I−)2, or ethylenediammonium diiodide, to the A site of FASnI3 made it work better. The results were even better when Ge2+ doping was combined with DAE passivation for cells that had the structure ITO/PEDOT:PSS/perovskite(FA0.98EDA0.01SnI3)/C60/BC. They showed a Jsc of 23.04 mA/cm2, Voc of 0.625 V, FF of 0.74, and PCE of 10.65%.
The study looks into how well solar cells work with halide perovskite (Sn) in the A site, which is written as Q in Q0.1(FA0.75MA0.25)0.9SnI3. The researchers found a strong link between the tolerance factor and the disorder in the lattice, which they called “relative local lattice strain.” The relative local lattice strain went down when the tolerance factor got close to 1.0 by switching part of the Q site from Na+ to EA+. The 3D structure stayed the same. But when 10% of BA+ was added, the tolerance factor went up to 1.11, the relative local lattice strain went down, and a two-dimensional structure was made.
As the lattice strain was loosened, the phase of FAPbI3 became more stable. The solar cell’s PCE was strongly linked to the relative local lattice strain, and it got better as the relative local lattice strain went down. When EA+ was added to Q, it caused the least relative local lattice strain and the most efficiency. The entry of Na+ caused the most relative local lattice strain and the least effectiveness.
The same trend was seen in carrier mobility: low local lattice strain (disordering) led to high carrier mobility. It’s likely that the local grid disordering spreads out the carriers, making them less mobile and lowering the efficiency. PCE was about 2.5% without adding EA+ ((FA0.75MA0.25)0.9 SnI3), but it went up to 5.4% when 10% EA was added to the A site (Q0.1(FA0.75MA0.25)0.9 SnI3 composition).
This increase in efficiency can be talked about from the point of view of aligning band energies. Adding Na+, K+, Cs+, and EA+ to Q in the form of Q0.1(FA0.75MA0.25)0.9 SnI3 made the perovskite layer’s conduction band go deeper, with Na+, K+, Cs+, and EA+ coming first. So, the difference in band shift between the perovskite’s conduction band and that of C60 got smaller. This is why the EA0.1(FA0.75MA0.25)0.9 SnI3 had the highest Voc.
The efficiency was raised even more to over 13% by adding Ge ions and EA+ replacement to the mix of DAE passivation and doping. We saw that the Voc loss went down to 0.54 eV for a cell with 13.2% PCE. This is a little higher than the 0.37 eV loss for 25.2% Pb PSC and the 0.39 eV loss for 20.4% Sn–Pb PSC in our lab.
In the end, the research shows that adding EA+ replacement and DAE passivation makes solar cells based on halide perovskite work better. When you add EA+ to the Q site of Q0.1(FA0.75MA0.25)0.9, the valence band and conduction band levels go down. The conduction band level is almost the same as that of regular Pb perovskite.
Conclusion
This study talks about how to make Sn-Pb mixed perovskite solar cells (Sn PSCs) more efficient by lowering the quantity of Sn4+, reducing the void of Sn2+, and making the best use of band energy alignment. Several studies have shown that lowering the quantity of Sn4+ defects can increase performance by more than 20%. This study improved the performance of Sn PSCs by using an upside-down structure to avoid carrier traps, making the crystals more dense with a SnF2−DMSO complex, adding a spike structure between the perovskite layer and fullerene layer, and lowering the strain in the local lattice. It became 20.4% and 22.3% more efficient.
It’s possible for the Sn-Pb perovskite to be more efficient than Pb PSCs because it has the best bandgap for low power loss. It is likely that the narrow-bandgap solar cells will be one of the options for all perovskite tandem solar cells.
Several groups have raised the effectiveness of Sn PSC to over 10% by focussing on lowering the amount of Sn4+ defects. Adding Ge ions, making a GexOy layer, passing through DAE on FASnI3, lowering the Fermi level, and changing the make-up of A-site cations all improved the efficiency of Sn PSCs. Adding EA+ to the A site of FASnI3 made it more efficient, and it got as efficient as 13.2%.
Researchers have found Sn PSCs that are more than 20% efficient. These are likely to be used as the top layer of Pb-free perovskite/Si tandem, perovskite/CIGS tandem, and all-perovskite tandem solar cells.
Leave a Reply