Photovoltaic (PV) solar cells play a central role among renewable energy sources, which currently account for 33% of electric power generation worldwide. PV cell efficiency is known to be limited by their incapability of converting the portion of the solar spectrum whose energy is lower than the PV material energy gap. However, such low-frequency part of the solar spectrum can be converted into heat and then, in turn, transformed into electric power by thermoelectric generators (TEGs). Considering that 90% of PV modules are based on Si, either single-crystalline or polycrystalline, the development of Si-based hybrid solar harvesters with enhanced conversion efficiencies would provide a remarkable contribution to the overall renewable power generation. This review will analyse recent advancements in the PV-TEG hybridisation focusing on Si-based PV cells at low solar concentration. Two well-established approaches will be mainly discussed, namely tandem hybridization, where TEGs are in a thermal series with the PV cell, and spectral-split hybridization, where instead suitable optics splits the solar spectrum, conveying its high-frequency part to the PV cell and the remaining part toward the TEG. Account will be given to both energetic and economic profitability, elaborating on the strategies to evade physical and economic constraints that are still limiting the exploitation of such technologies in the most relevant and diffused class of solar harvesters.

Thermoelectric generation via radiative cooling holds promise as a renewable electricity source. However, the governing mechanisms have not been fully explained, and the optimal operating conditions are yet to be identified. To address this issue, detailed parametric analyses of the most basic, passive configuration are presented here, where the cold side of the thermoelectric generator is coupled to a radiative cooling surface and the hot side collects heat from the ambient surroundings. We note that the operation deviates significantly from the constant temperature conditions because the radiative cooling power under realistic conditions is not large enough to act as a cold reservoir. The system is thus considered as imperfectly coupled to the reservoirs with a finite thermal conductance. In this regime, the contributions of both electrical and thermal conduction within the thermoelectric generator, along with their interdependence, must be carefully treated to describe the steady-state behaviour. An analytical solution is derived, as an explicit function of system parameters such as the radiative cooling surface area, the degree of thermal insulation of the radiative cooling surface, and the heat transfer coefficient at the hot side. As a result, electrical and thermal impedance matching conditions are presented for the generation of maximum electrical power in terms of these parameters. Our findings serve as a design guide for renewable thermoelectric generation based on the temperature difference between the ambient and the coldness of the sky.

Hybrid thermoelectric-photovoltaic (HTEPV) systems offer a promising route to enhance solar energy conversion by utilizing waste heat from photovoltaic cells. Suppressing radiative heat loss, which is a key challenge to reach high efficiency, necessitates implementing a transparent heat mirror that transmits solar radiation to the photovoltaic cell while reflecting thermal infrared radiation back to the system. Such spectrally selective heat mirrors are also highly relevant for hybrid photovoltaic–thermal (PV-T) and thermal solar harvesting systems, where suppressing radiative heat losses can directly enhance thermal efficiency and heat recovery. This work reports the fabrication and optimization of a flexible, multilayer heat mirror on cyclic olefin polymer (COP) substrates, compatible for low-temperature processing (<100 °C). The heat mirror is designed to combine high solar transmittance (300–2500 nm) with strong mid-infrared reflectance (>2500 nm), specifically targeting the blackbody emission peak at operating temperature around 300–400 K, typical of hybrid PV-thermal and HTEPV systems. The core of the structure is an aluminum-doped zinc oxide (AZO) film deposited by pulsed DC magnetron sputtering. The AZO layer was optimized by tuning its thickness (220 nm) and sputtering conditions (1% oxygen at a base pressure of 700 µPa). Performance was considerably enhanced through multilayer engineering, first by incorporating a symmetric Al2O3/AZO/Al2O3 sandwich structure deposited via plasma-enhanced atomic layer deposition (PE-ALD), and then depositing a top 100 nm MgF2 anti-reflection coating. Remarkable values of weighted transmittance and reflectance of 92.2% and 65.8% were obtained, establishing a viable low-temperature pathway for fabricating efficient, flexible, and spectrally selective heat mirrors and radiative-loss management across a broad class of hybrid solar harvesters.

Silicon nanopillars (SiNPs) are quasi-1D Si nanostructures extending normally to the substrate. As Si nanowires, SiNPs display large surface-to-volume ratios and a remarkable reduction of their thermal conductivity by almost a factor one-hundred compared to bulk Si. SiNPs have found applications in energy-related fields, including thermoelectrics and photovoltaics, and in chemical sensing and biosensing. Among the available preparation techniques, Metal-Assisted Chemical Etching (MACE) has been extensively used to prepare high density SiNP forests, as the method is facile and scalable. However, while MACE lets easily obtain low-to-medium doped SiNPs, p-type SiNPs with doping level ≥ 1020 cm-3, essential for energetic applications, could never be obtained. In this paper we provide evidence that this is related to the competition between metal-assisted and noncatalyzed etching. Thus, guided by a model of the silicon-solution and silicon-metal-solution interfaces, we could develop a strategy that enabled for the first time the preparation by MACE of p-type SiNPs with doping levels as large as 1020 cm-3, using Au as the catalyst and Na2S2O8 as the oxidizing agent. The possibility of preparing heavily doped SiNPs largely extends MACE application to critical technological fields spanning from thermoelectrics and photovoltaics to batteries and sensing.

Thermoelectric generators are devices capable to convert heat into electric power with no moving part. However, and despite a tremendous research effort on materials, their conversion efficiency is still limited, especially in the low temperature range where most of the discarded heat is available. We show that the exact solution of the time-dependent Domenicali's equation predicts that, when the temperature difference across the thermoelectric legs is modulated in time, efficiency at maximum power (eta(MP)) improves by up to 50% compared to the stationary case - with a power output equivalent to that attainable by doubling the material figure of merit. Building on this evidence, we additionally show how, even for sources delivering heat at a constant rate, simple heat flux pre-processing leads to a comparable eta(MP) improvement. Since the operational mode we propose is material-agnostic and does not require changes of the device layout, it could find prompt application.

Silicon nanopillars (SiNPs) are quasi-1D Si nanostructures extending normally to the substrate. As Si nanowires, SiNPs display large surface-to-volume ratios and a remarkable reduction of their thermal conductivity by almost a factor one-hundred compared to bulk Si. SiNPs have found applications in energy-related fields, including thermoelectrics and photovoltaics, and in chemical sensing and biosensing. Among the available preparation techniques, Metal-Assisted Chemical Etching (MACE) has been extensively used to prepare high density SiNP forests, as the method is facile and scalable. However, while MACE lets easily obtain low-to-medium doped SiNPs, p-type SiNPs with doping level ≥ 1020 cm-3, essential for energetic applications, could never be obtained. In this paper we provide evidence that this is related to the competition between metal-assisted and noncatalyzed etching. Thus, guided by a model of the silicon-solution and silicon-metal-solution interfaces, we could develop a strategy that enabled for the first time the preparation by MACE of p-type SiNPs with doping levels as large as 1020 cm-3, using Au as the catalyst and Na2S2O8 as the oxidizing agent. The possibility of preparing heavily doped SiNPs largely extends MACE application to critical technological fields spanning from thermoelectrics and photovoltaics to batteries and sensing.

Metal-Assisted Chemical Etching (MACE) using Ag as the catalyst lets prepare vertically aligned crystalline silicon nanopillars (SiNPs), a highly promising system for thermoelectric applications, with high aspect ratios in a wide doping range. MACE may be implemented either by using Ag both as the catalyst and the oxidant (so-called one-pot MACE) or by using another chemical (typically H2O2) as the oxidant (two-pot MACE). This study investigates how the localized etching rate depends upon Si doping in both MACE implementations, accounting for the concurrent non-catalyzed etching. The latter, which shortens SiNPs, is found to become more significant in p-type Si at higher doping levels due to the narrower space-charge regions at the bare Si-solution interface. We demonstrated that in both one- and two-pot MACE the etching rate is controlled by the band bending at silicon-silver interface. In p-type silicon, it decreases with doping due to faster hole diffusion, while the Schottky barrier at the interface hinders hole injection in n-type silicon at any doping level. Overall, we highlight that MACE may be effectively implemented in its one-pot version, facilitating MACE scale-up toward SiNP large-scale manufacturing.