Machine Learning in Nanoscience: Big Data at Small Scales.

Recent advances in machine learning (ML) offer new tools to extract new insights from large data sets and to acquire small data sets more effectively. Researchers in nanoscience are experimenting with these tools to tackle challenges in many fields. In addition to ML's advancement of nanoscience, nanoscience provides the foundation for neuromorphic computing hardware to expand the implementation of ML algorithms. In this Mini Review, we highlight some recent efforts to connect the ML and nanoscience communities by focusing on three types of interaction: (1) using ML to analyze and extract new insights from large nanoscience data sets, (2) applying ML to accelerate material discovery, including the use of active learning to guide experimental design, and (3) the nanoscience of memristive devices to realize hardware tailored for ML. We conclude with a discussion of challenges and opportunities for future interactions between nanoscience and ML researchers.

Perovskite Nanowire Extrusion.

The defect tolerance of halide perovskite materials has led to efficient optoelectronic devices based on thin-film geometries with unprecedented speed. Moreover, it has motivated research on perovskite nanowires because surface recombination continues to be a major obstacle in realizing efficient nanowire devices. Recently, ordered vertical arrays of perovskite nanowires have been realized, which can benefit from nanophotonic design strategies allowing precise control over light propagation, absorption, and emission. An anodized aluminum oxide template is used to confine the crystallization process, either in the solution or in the vapor phase. This approach, however, results in an unavoidable drawback: only nanowires embedded inside the AAO are obtainable, since the AAO cannot be etched selectively. The requirement for a support matrix originates from the intrinsic difficulty of controlling precise placement, sizes, and shapes of free-standing nanostructures during crystallization, especially in solution. Here we introduce a method to fabricate free-standing solution-based vertical nanowires with arbitrary dimensions. Our scheme also utilizes AAO; however, in contrast to embedding the perovskite inside the matrix, we apply a pressure gradient to extrude the solution from the free-standing templates. The exit profile of the template is subsequently translated into the final semiconductor geometry. The free-standing nanowires are single crystalline and show a PLQY up to ∼29%. In principle, this rapid method is not limited to nanowires but can be extended to uniform and ordered high PLQY single crystalline perovskite nanostructures of different shapes and sizes by fabricating additional masking layers or using specifically shaped nanopore endings.

Core-Shell CdS-Cu2S Nanorod Array Solar Cells.

As an earth-abundant p-type semiconductor, copper sulfide (Cu2S) is an attractive material for application in photovoltaic devices. However, it suffers from a minority carrier diffusion length that is less than the length required for complete light absorption. Core-shell nanowires and nanorods have the potential to alleviate this difficulty because they decouple the length scales of light absorption and charge collection. To achieve this geometry using Cu2S, cation exchange was applied to an array of CdS nanorods to produce well-defined CdS-Cu2S core-shell nanorods. Previous work has demonstrated single-nanowire photovoltaic devices from this material system, but in this work, the cation exchange chemistry has been applied to nanorod arrays to produce ensemble-level devices with microscale sizes. The core-shell nanorod array devices show power conversion efficiencies of up to 3.8%. In addition, these devices are stable when measured in air after nearly one month of storage in a desiccator. These results are a first step in the development of large-area nanostructured Cu2S-based photovoltaics that can be processed from solution.

Epitaxially aligned cuprous oxide nanowires for all-oxide, single-wire solar cells.

As a p-type semiconducting oxide that can absorb visible light, cuprous oxide (Cu2O) is an attractive material for solar energy conversion. This work introduces a high-temperature, vapor-phase synthesis that produces faceted Cu2O nanowires that grow epitaxially along the surface of a lattice-matched, single-crystal MgO substrate. Individual wires were then fabricated into single-wire, all-oxide diodes and solar cells using low-temperature atomic layer deposition (ALD) of TiO2 and ZnO films to form the heterojunction. The performance of devices made from pristine Cu2O wires and chlorine-exposed Cu2O wires was investigated under one-sun and laser illumination. These faceted wires allow the fabrication of well-controlled heterojunctions that can be used to investigate the interfacial properties of all-oxide solar cells.

Phase-selective cation-exchange chemistry in sulfide nanowire systems.

As a cation-deficient, p-type semiconductor, copper sulfide (Cu(2-x)S) shows promise for applications such as photovoltaics, memristors, and plasmonics. However, these applications demand precise tuning of the crystal phase as well as the stoichiometry of Cu(2-x)S, an ongoing challenge in the synthesis of Cu(2-x)S materials for a specific application. Here, a detailed transformation diagram of cation-exchange (CE) chemistry from cadmium sulfide (CdS) into Cu(2-x)S nanowires (NWs) is reported. By varying the reaction time and the reactants' concentration ratio, the progression of the CE process was captured, and tunable crystal phases of the Cu(2-x)S were achieved. It is proposed that the evolution of Cu(2-x)S phases in a NW system is dependent on both kinetic and thermodynamic factors. The reported data demonstrate that CE can be used to precisely control the structure, composition, and crystal phases of NWs, and such control may be generalized to other material systems for a variety of practical applications.

Absorption of light in a single-nanowire silicon solar cell decorated with an octahedral silver nanocrystal.

In recent photovoltaic research, nanomaterials have offered two new approaches for trapping light within solar cells to increase their absorption: nanostructuring the absorbing semiconductor and using metallic nanostructures to couple light into the absorbing layer. This work combines these two approaches by decorating a single-nanowire silicon solar cell with an octahedral silver nanocrystal. Wavelength-dependent photocurrent measurements and finite-difference time domain simulations show that increases in photocurrent arise at wavelengths corresponding to the nanocrystal's surface plasmon resonances, while decreases occur at wavelengths corresponding to optical resonances of the nanowire. Scanning photocurrent mapping with submicrometer spatial resolution experimentally confirms that changes in the device's photocurrent come from the silver nanocrystal. These results demonstrate that understanding the interactions between nanoscale absorbers and plasmonic nanostructures is essential to optimizing the efficiency of nanostructured solar cells.

Binary Superlattices of Infrared Plasmonic and Excitonic Nanocrystals.

Self-assembled superlattices of nanocrystals offer exceptional control over the coupling between nanocrystals, similar to how solid-state crystals tailor the bonding between atoms. By assembling nanocrystals of different properties (e.g., plasmonic, excitonic, dielectric, or magnetic), we can form a wealth of binary superlattice metamaterials with new functionalities. Here, we introduce infrared plasmonic Cu2-xS nanocrystals to the limited library of materials that have been successfully incorporated into binary superlattices. We are the first to create a variety of binary superlattices with large excitonic (PbS) nanocrystals and small plasmonic (Cu2-xS) nanocrystals, both resonant in the infrared. Then, by controlling the surface chemistry of large Cu2-xS nanocrystals, we produced structurally analogous superlattices of large Cu2-xS and small PbS nanocrystals. Transmission electron microscopy (TEM) and grazing-incidence small-angle X-ray scattering (GISAXS) were used to characterize both types of superlattices. Furthermore, our unique surface modification of the large Cu2-xS nanocrystals also prevented them from chemically quenching the photoluminescence of the PbS nanocrystals, which occurred when the PbS nanocrystals were mixed with unmodified Cu2-xS nanocrystals. These synthetic achievements create a set of binary superlattices that can be used to understand how infrared plasmonic and excitonic nanocrystals couple in a variety of symmetries and stoichiometries.

Effects of a Lead Chloride Shell on Lead Sulfide Quantum Dots.

The size of a quantum-confined nanocrystal determines the energies of its excitonic transitions. Previous work has correlated the diameters of PbS nanocrystals to their excitonic absorption; however, we observe that PbS quantum dots synthesized in saturated dispersions of PbCl2 can deviate from the previous 1Sh-1Se energy vs diameter curve by 0.8 nm. In addition, their surface differs chemically from that of PbS quantum dots produced via other syntheses. We find that these nanocrystals are coated in a shell that is measurable in transmission electron micrographs and contains lead and chlorine, beyond the monatomic chlorine termination previously proposed. This finding has implications for understanding the growth mechanism of this reaction, the line width of these quantum dots' photoluminescence, and electronic transport within films of these nanocrystals. Such fundamental knowledge is critical to applications of PbS quantum dots such as single-photon sources, photodetectors, solar cells, light-emitting diodes, lasers, and biological labels.

Vapor-liquid-solid and vapor-solid growth of phase-change Sb2Te3 nanowires and Sb2Te3/GeTe nanowire heterostructures.

We report the synthesis and characterization of radial heterostructures composed of an antimony telluride (Sb2Te3) core and a germanium telluride (GeTe) shell, as well as an improved synthesis of Sb2Te3 nanowires. The synthesis of the heterostructures employs Au-catalyst-assisted vapor-liquid-solid (VLS) and vapor-solid (VS) mechanisms. Energy-dispersive X-ray spectrometry indicates that Sb and Ge are localized in the Sb2Te3 and GeTe portions, respectively, confirming the alloy-free composition in the core/shell heterostructures. Transmission electron microscopy and diffraction studies show that Sb2Te3 and GeTe regions exhibit rhombohedral crystal structure. Both Sb2Te3 and GeTe grow along the [110] direction with an epitaxial interface between them. Electrical characterization of individual nanowires and nanowire heterostructures demonstrates that these nanostructures exhibit memory-switching behavior.

Understanding Detrimental and Beneficial Grain Boundary Effects in Halide Perovskites.

Grain boundaries play a key role in the performance of thin-film optoelectronic devices and yet their effect in halide perovskite materials is still not understood. The biggest factor limiting progress is the inability to identify grain boundaries. Noncrystallographic techniques can misidentify grain boundaries, leading to conflicting literature reports about their influence; however, the gold standard - electron backscatter diffraction (EBSD) - destroys halide perovskite thin films. Here, this problem is solved by using a solid-state EBSD detector with 6000 times higher sensitivity than the traditional phosphor screen and camera. Correlating true grain size with photoluminescence lifetime, carrier diffusion length, and mobility shows that grain boundaries are not benign but have a recombination velocity of 1670 cm s-1 , comparable to that of crystalline silicon. Amorphous grain boundaries are also observed that give rise to locally brighter photoluminescence intensity and longer lifetimes. This anomalous grain boundary character offers a possible explanation for the mysteriously long lifetime and record efficiency achieved in small grain halide perovskite thin films. It also suggests a new approach for passivating grain boundaries, independent of surface passivation, to lead to even better performance in optoelectronic devices.

Direct Observation of Halide Migration and its Effect on the Photoluminescence of Methylammonium Lead Bromide Perovskite Single Crystals.

Optoelectronic devices based on hybrid perovskites have demonstrated outstanding performance within a few years of intense study. However, commercialization of these devices requires barriers to their development to be overcome, such as their chemical instability under operating conditions. To investigate this instability and its consequences, the electric field applied to single crystals of methylammonium lead bromide (CH3 NH3 PbBr3 ) is varied, and changes are mapped in both their elemental composition and photoluminescence. Synchrotron-based nanoprobe X-ray fluorescence (nano-XRF) with 250 nm resolution reveals quasi-reversible field-assisted halide migration, with corresponding changes in photoluminescence. It is observed that higher local bromide concentration is correlated to superior optoelectronic performance in CH3 NH3 PbBr3 . A lower limit on the electromigration rate is calculated from these experiments and the motion is interpreted as vacancy-mediated migration based on nudged elastic band density functional theory (DFT) simulations. The XRF mapping data provide direct evidence of field-assisted ionic migration in a model hybrid-perovskite thin single crystal, while the link with photoluminescence proves that the halide stoichiometry plays a key role in the optoelectronic properties of the perovskite.

Preparation of Organometal Halide Perovskite Photonic Crystal Films for Potential Optoelectronic Applications.

Herein, a facile method for the preparation of organometal halide perovskite (OHP) thin films in photonic crystal morphology is presented. The OHP photonic crystal thin films with controllable porosity and thicknesses between 2 µm and 6 µm were prepared on glass, fluorine-doped tin oxide (FTO), and TiO2 substrates by using a colloidal crystal of polystyrene microspheres as a template to form an inverse opal structure. The composition of OHP could be straightforwardly tuned by varying the halide anions. The obtained OHP inverse opal films possess large ordered domains with a periodic change of the refractive index, which results in pronounced photonic stop bands in the visible light range. By changing the diameter of the polystyrene microspheres, the position of the photonic stop band can be tuned through the visible spectrum. This developed methodology can be used as blueprint for the synthesis of various OHP films that could eventually be used as more effective light harvesting materials for diverse applications.

Single-nanowire photoelectrochemistry.

Photoelectrochemistry is one of several promising approaches for the realization of efficient solar-to-fuel conversion. Recent work has shown that photoelectrodes made of semiconductor nano-/microwire arrays can have better photoelectrochemical performance than their planar counterparts because of their unique properties, such as high surface area. Although considerable research effort has focused on studying wire arrays, the inhomogeneity in the geometry, doping, defects and catalyst loading present in such arrays can obscure the link between these properties and the photoelectrochemical performance of the wires, and correlating performance with the specific properties of individual wires is difficult because of ensemble averaging. Here, we show that a single-nanowire-based photoelectrode platform can be used to reliably probe the current-voltage (I-V) characteristics of individual nanowires. We find that the photovoltage output of ensemble array samples can be limited by poorly performing individual wires, which highlights the importance of improving nanowire homogeneity within an array. Furthermore, the platform allows the flux of photogenerated electrons to be quantified as a function of the lengths and diameters of individual nanowires, and we find that the flux over the entire nanowire surface (7-30 electrons nm(-2) s(-1)) is significantly reduced as compared with that of a planar analogue (∼1,200 electrons nm(-2) s(-1)). Such characterization of the photogenerated carrier flux at the semiconductor/electrolyte interface is essential for designing nanowire photoelectrodes that match the activity of their loaded electrocatalysts.

25th anniversary article: semiconductor nanowires--synthesis, characterization, and applications.

Semiconductor nanowires (NWs) have been studied extensively for over two decades for their novel electronic, photonic, thermal, electrochemical and mechanical properties. This comprehensive review article summarizes major advances in the synthesis, characterization, and application of these materials in the past decade. Developments in the understanding of the fundamental principles of "bottom-up" growth mechanisms are presented, with an emphasis on rational control of the morphology, stoichiometry, and crystal structure of the materials. This is followed by a discussion of the application of nanowires in i) electronic, ii) sensor, iii) photonic, iv) thermoelectric, v) photovoltaic, vi) photoelectrochemical, vii) battery, viii) mechanical, and ix) biological applications. Throughout the discussion, a detailed explanation of the unique properties associated with the one-dimensional nanowire geometry will be presented, and the benefits of these properties for the various applications will be highlighted. The review concludes with a brief perspective on future research directions, and remaining barriers which must be overcome for the successful commercial application of these technologies.

Solution-processed core-shell nanowires for efficient photovoltaic cells.

Semiconductor nanowires are promising for photovoltaic applications, but, so far, nanowire-based solar cells have had lower efficiencies than planar cells made from the same materials, even allowing for the generally lower light absorption of nanowires. It is not clear, therefore, if the benefits of the nanowire structure, including better charge collection and transport and the possibility of enhanced absorption through light trapping, can outweigh the reductions in performance caused by recombination at the surface of the nanowires and at p-n junctions. Here, we fabricate core-shell nanowire solar cells with open-circuit voltage and fill factor values superior to those reported for equivalent planar cells, and an energy conversion efficiency of ∼5.4%, which is comparable to that of equivalent planar cells despite low light absorption levels. The device is made using a low-temperature solution-based cation exchange reaction that creates a heteroepitaxial junction between a single-crystalline CdS core and single-crystalline Cu2S shell. We integrate multiple cells on single nanowires in both series and parallel configurations for high output voltages and currents, respectively. The ability to produce efficient nanowire-based solar cells with a solution-based process and Earth-abundant elements could significantly reduce fabrication costs relative to existing high-temperature bulk material approaches.

Minimum voltage for threshold switching in nanoscale phase-change memory.

The size scaling of the threshold voltage required for the amorphous-to-crystalline transition in phase-change memory (PCM) is investigated using planar devices incorporating individual GeTe and Sb2Te3 nanowires. We show that the scaling law governing threshold switching changes from constant field to constant voltage scaling as the amorphous domain length falls below 10 nm. This crossover is a consequence of the energetic requirement for carrier multiplication through inelastic scattering processes and indicates that the size of PCM bits can be miniaturized to the true nanometer scale.