Lead halide perovskite semiconductors: the role of morphological control for high performance solar cells

Tri-dimensional lead halide perovskite semiconductors are an emerging solution-processed class of materials that has led to great advances in the performances of solar cells since 2013, and has demonstrated good potential when embodied in light emitting devices and photodetectors. The exceptional opto-electronic properties of these materials, such as the tunable direct band-gap, the extremely high absorption coefficient, the low exciton binding energy, and the balanced ambipolar carrier transport have been extensively studied in the last few years. At the same time, many of the finer details of the device physics remain under debate. In particular, the understanding of the role of structural and chemical defects, and their effects on the properties of devices, has become imperative. Overall, the three-year work highlights the primary properties of the same class of material obtained by various synthetic routes, combining structural and morphological characterizations to the photophysical ones. The thesis work shows a direct comparison between two different approaches to the realization of a perovskite-based thin film. The first one consists in the direct crystallization of the semiconductive material onto a substrate, starting from precursors solutions, while the second one relies on the deposition of pre-formed high-quality colloidal perovskite-based nanocrystals. The first approach is the most widely used in literature. It consists in the self-organization of ionized precursors dispersed in high boiling point organic solvents, which undergo a crystallization upon thermal treatment. Chemical purity of the precursors, presence of additives, relative humidity, temperature and surface properties of the substrates employed are just few of the parameters that can deeply affect the organization of the perovskite crystalline lattice. In this case, a non-negligible level of unintentional structural and chemical defects is expected, within the bulk and at the interfaces of the crystals. Such defects can introduce localized energy levels that are confined in the band-gap of the semiconductor, creating non-radiative recombination pathways for the photo-generated charge carriers. A reproducible and reliable protocol to fabricate polycrystalline thin film, with different average-crystallite dimensions, is reported. Varying the crystallites mean size, from tens of nanometers to a few micrometers, it is possible to tune the opto-electronic behaviour of the material, directly influencing the number and the type of electrically-active defect sites. From these results, the existence of a tight tolerance window for the optimal processing parameters of these self-assembled materials clearly emerges. A novel approach is then proposed, consisting in the realization of the solar cell active layer by the deposition of an ink composed by suspended nanoparticles. The ink formulation can be tailored prior to the film formation, meaning that the co-optimization of large-area uniform coverage with a high-quality crystal growth is no longer needed. The results obtained by following this approach are presented in the second part of the thesis. A fully inorganic cesium lead bromide (CsPbBr3) perovskite has been employed in the study. Since bromide-based compounds exhibit a wide band-gap, around 2.32 eV, these materials can achieve much higher open-circuit values, extending the interest for the applications to multijunction solar cells, visible light-emitting devices and photo-electrochemical cells. A new, fast and one-step-injection synthesis is proposed, leading to the formation of stable colloidal nanocrystalline perovskite nanoparticles, passivated by organic ligands. The innovative synthesis has the advantages of being a room-temperature and moisture-insensitive process, that employs only low boiling point and less toxic solvents, compared to the ones required in the standard approach. The use of ligands, composed by short alkyl chains, circumvents post-synthesis treatments typical of the longer and bulkier chains usually employed in this field. The passivating nature of the nanocrystal’s ligands gives rise to extremely good photo-luminescence properties. The photo-luminescence quantum yield for the colloidal suspension is higher than 75%, and it drops to 35% after the particles have been deposited. The thin film obtained shows an amplified spontaneous emission (ASE) threshold as low as 1.5 J/cm2, which sets a record for not-passivated inorganic nanostructures. The optical quality observed is close to those of nanocrystals made with high-temperature hot-injection syntheses previously reported in literature. Interestingly, the quantum yield from the films prepared with bulkier ligands are comparable to the ones typically observed from the films here reported, but they remain non-conductive. Both the quantum yield and the ASE threshold values are indicative of a lower defect density compared to perovskite thin film obtained from standard direct-crystallization approaches. The CsPbBr3 perovskite inks are then used to fabricate a halide perovskite nanocrystalline-based photovoltaic prototype device. The deposition of highly uniform and complete films has been uniquely facilitated by the use of low-boiling point solvents for the preparation of the colloidal suspension. Sequential deposition cycles enable a fine control over the final active-material thickness. In particular, a layer of 550 ± 50nm, corresponding to an optical thickness of more than 1.5 OD, exhibited a power conversion efficiency of 5.4%, which is comparable to the best-performing and fully optimized devices reported so far. The solar cell exhibits short circuit current (JSC) and an open circuit voltage (VOC) values of 5.6 mA/cm2 and 1.5 V, as well as a fill factor of 0.62. The value for VOC is among the highest reported for perovskite halides, underlying the high quality of the perovskite crystals. For a better understanding of the reference, the calculated ideal maximum VOC and JSC are 2.05 V and 7.78 mA/cm2 under AM1.5 illumination, considering a bandgap of about 2.38 eV. Finally, it has to be highlight that this process is fully carried out in ambient air. Moreover, the technique provides electrically stable devices. Finally, a fine analysis over the ligands’ role in the device performances has been carried out. At the same time, the correlation between the nanocrystal size and the figures of merit of the devices has been investigated, highlighting the nanocrystal size-dependence of the current density in the device. The validity of this novel approach has been further demonstrated by successfully applied it to large-area deposition techniques (bar-coating), obtaining working devices with performances comparable to the previously reported ones.

Semiconduttori con struttura a perovskite basati su composti di alogenuri di piombo sono stati utilizzati negli ultimissimi anni come materiali attivi, processabili per soluzione, in dispositive optoelettronici, specialmente nel campo della conversione fotovoltaica. Inoltre, hanno dimostrato promettenti risultati quando impiegati in rivelatori e in dispositivi per emissione di luce. Tra le proprietà interessanti che questa classe di materiali presenta, vanno ricordati il band-gap di tipo diretto e facilmente modificabile attraverso la composizione chimica dei materiali; il coefficiente di assorbimento estremamente elevato, l’energia di legame degli eccitoni particolarmente bassa e una ottima conducibilità di carica ambipolare. Tuttavia, molti dei dettagli più fini dei comportamenti chimico-fisici che regolano questi sistemi rimangono in fase di dibattito. Il lavoro di tesi mette in luce la stretta correlazione tra le proprietà optoelettroniche di questa classe di materiali auto-assemblanti e il loro abito cristallino, ovvero la morfologia dei cristalli stessi. Nel lavoro vengono presentati materiali identici processati attraverso vie sintetiche differenti, combinando caratterizzazioni morfologiche e strutturali a quelle fotofisiche. I materiali investigati sono stati ottenuti, nel primo caso, per cristallizzazione diretta dei componenti sul substrato scelto partendo da precursori solubilizzati. Nel secondo caso, invece, il film viene ottenuto tramite deposizione di nanocristalli di alta qualità ottenuti per sintesi colloidale. Il primo approccio mostra come la qualità strutturale del semiconduttore sia fortemente influenzata da molteplici fattori ambientali, spesso non direttamente o facilmente controllabili. In questo caso, un numero non trascurabile di difetti viene introdotto nel reticolo cristallino, che si traduce nell’inserimento di livelli energetici localizzati all’interno del band-gap del materiale, che aggiungono canali di ricombinazione non radiativi per le cariche fotogenerate. Il secondo approccio prevede la realizzazione del film attivo nel dispositivo fotovoltaico attraverso la deposizione di nanocristalli pre-formati, dispersi soluzione. Le proprietà optoelettroniche di questi nanocristalli possono essere modificate in fase di sintesi, e vengono mantenute una volta che i cristalli vengono deposti. In particolare, si dimostra la realizzazione di celle solari a base di cesio-piombo-bromuro, un particolare materiale inorganico che non sarebbe possibile processare attraverso il metodo presentato precedentemente. Le celle solari mostrano un’efficienza di conversione maggiore del 5.4%, con un voltaggio di circuito aperto superiore a 1.5 V. Questo valore è il più alto presentato in letteratura, e dimostra la possibilità di utilizzare efficientemente questo materiale per la conversione luce-energia elettrica o per applicazioni di water-splitting. Nell’ultima parte della tesi vengono illustrati i risultati ottenuti durate l’ottimizzazione della sintesi delle nanoparticelle, dimostrando la compatibilità dell’approccio con i metodi di deposizione per larga area (bar-coating).