† Corresponding author. E-mail:
A new transparent photovoltaic panel composed of a luminescent solar concentrator and Al/BaTiO3/ZnO/Pt ferroelectric solar cells is presented, in which a portion of the incoming solar illumination is converted by the fluorophores to ultraviolet (UV) light which is then absorbed by ZnO. Firstly, the solar cells are simulated using Atlas–Silvaco. Then, the panel is modelled based on the obtained solar cell characteristics. This panel would be of great importance for building integrated photovoltaics domain because of its high transparency.
Photovoltaic solar energy is the object of many scientific investigations. Researchers are constantly looking for cleaner, more efficient, and less costly systems. The pn junction solar cell (PNJSC) is the most used cell for solar energy harvesting,[1–3] and it is generally fabricated with silicon[1] containing a junction between two differently doped regions p and n; as a result, a build-in electric field appears therein, which separates the photogenerated charge carriers,[3–5] so the electrons and holes are drifted to the n and p regions, respectively, and then collected by the poles. Lastly, electrons will flow in the external circuit to provide energy and recombine with holes.[4,5]
Although PNJSCs are the most commercialized solar cells, many obstacles therein have to be surmounted in order to increase their efficiency again.[3–6] Among the solar light photons, only those with energy equal or higher than the semiconductor band gap energy (Eg) will be exploited; in addition, the excess energy coming from photons with the energy higher than Eg will raise the crystal temperature because of the photogenerated electrons and holes thermalization, which means that the output energy of the cell will be decreased.[4,5,7–9] In addition, the open circuit voltage (Voc) of a PNJSC is less than Eg/e (e is the elementary charge value)[3–5,10] and the attempt of increasing this voltage by adding dopants favors the non-radiative recombination; hence, the yield is reduced.[5]
Heterojunction solar cells are developed with the aim of absorbing the solar illumination more efficiently,[4–6,11] and their theoretical efficiency limit for an infinite number of junctions may reach 69%, 86%–86.8% under one sun, full concentration, respectively.[4,11,12] However, this type of cells is limited in material choice and has a high cost since it requires advanced growth techniques.[6]
Ferroelectric materials (FMs) under a specific temperature (Curie temperature) are characterised by an electric field which presents a hysteresis cycle, so it can be controlled by applying an electric field.[13] Since FMs present photovoltaic effect, many researchers have developed new solar cell structures based on these materials.[2,3,10,14–18]
At the beginning of ferroelectric solar cells (FSCs), the FM was used as a light absorber wherein the ferroelectric electric field is responsible for charge carrier separation. These cells present high values of Voc originated from the domain walls,[2,18] however, their short circuit current Isc is very low due to the weak electrical conductivity and the high Eg, which means that a small amount of charge carriers are generated using solar spectrum.[14,17,19]
In order to avoid these problems, FMs have been combined with a semiconductor light absorber. Fang Huang and Xiangxin Liu[15] fabricated a photovoltaic device in which the active layer is composed of CdTe absorber and CdS ferroelectric nanoparticles. Since the interface between CdS nanoparticles and CdTe is very important, the output current of this device is reduced due to surface recombination.[5]
A new approach to exploit FMs in photovoltaic cells was proposed by Wang et al.[19] in which FM was used to apply an electric field on a semiconductor absorber layer and, hence, to separate photogenerated carriers. In this new cell, photocurrent does not pass through the FM and photogeneration occurs at the semiconductor layer, which means that important Isc is accessible; nevertheless, the non-uniformity of the electric field and its weakness compared to the PNJSC limit the cell efficiency.
Luminescent solar concentrators (LSCs) have been studied for many reasons, such as their building compatibility and their economy since fewer solar cells are needed.[20–25] Down/up-conversion is a process in which particles absorb photons and then emit light at longer/shorter wavelength than the excitation wavelength. Many researchers are interested in down-conversion fluorophores embedded in LSCs[7,8,20,23,26–28] because silicon solar cells are the most used; however, up-conversion is also studied in order to benefit from longer wavelength radiations.[8,9,28]
In this theoretical work, we introduce a new transparent photovoltaic panel that is composed of ferroelectric solar cells and a luminescent solar concentrator, which can be realized using actual fabrication technologies. The panel schematics and its working principle are well explained in Section 2. Using ATLAS-Silvaco, we simulate and calculate the efficiency of Al/BaTiO3/ZnO/Pt ferroelectric solar cells; then, we calculate the transparent panel characteristics.
Figure
This panel structure is also suitable for other types of solar cells, such as silicon cell if the following conditions are met:
(i) The width of the cell must be comparable to the penetration depth.
(ii) The light absorber layer must be aligned with the electroluminescent layer.
(iii) The luminescent particles should be chosen in such a way that the emitted photon energy corresponds to the gap energy of the light absorbing material.
ZnO is a large band gap (3.4 eV)[29] semiconductor, hence, on the one hand, up-conversion luminescent elements must be embedded in the transparent matrix, with emission wavelengths less than 364 nm. Many up-conversion fluorophores are introduced in the literature. Table
![]() | Table 1.
UV emitting up-conversion fluorophores. . |
The fluorophores proposed here emit several wavelengths, but, in our case, only the ones less than 364 nm are utile for photogeneration.
Using Atlas–Silvaco software, two solar cells are simulated. These cells have vertical generation and lateral collection of the light generated carriers.[32] This type of cells suits the LSC perfectly. We remind that Atlas-Silvaco is a semiconductor device simulator, which is based on Poisson’s equation, the continuity equation, and the transport equation for electrons and holes.[33] In all our simulations, we consider the temperature T = 300 K and the wavelength of the monochromatic incident illumination
Firstly, we simulate cells composed of n-doped ZnO as an absorber, and Al and Pt as electrodes, as shown in Fig.
Platinum (Pt) has a work function of 5.7 eV[34] and zinc oxide (ZnO) has an electronic affinity of 2.088 eV,[35] hence, the n-ZnO/Pt forms a Schottky contact in which a spatial charge zone takes place near the junction. Photogenerated carriers in this region are diffused in different senses depending on their charge sign, so the separation is realized.
Secondly, we add a ferroelectric BaTiO3 (BTO) layer to our cell to form an Al/BTO/ZnO/Pt cell which is schematically represented in Fig.
Several cell dimensions (ZnO and BTO layer thickness, cell width) are compared in order to optimize the efficiency.
Al/ZnO/Pt solar cells with different dimensions x and y as shown in Fig.
We note that the short circuit current density is practically the same for all the dimensions, however, the open circuit voltage increases with the cell thickness unlike conventional solar cells,[36] and this is because the incidence of light is perpendicular to the cell thickness direction. Also, the voltage drops for a greater width x due to the diffusion of photogenerated carriers from the edges toward the middle. This diffusion occurs because of the photogeneration rate gradient as illustrated in Fig.
![]() | Fig. 10. Photogeneration rates. The photogeneration gradients are oriented from the center toward the edges. |
Figure
One can see that the short circuit current behavior is not so clear, but generally it increases with the thickness of zinc oxide and decreases slightly with the cell width; however, it is clear that the open circuit voltage varies depending on the cell parameters. Summarily, the power output is greater for x = 2 and y = 50.
We have simulated a number of Al/BTO/ZnO/Pt solar cells with the width value x = 2 and different ZnO and BTO thicknesses y and b, respectively, as shown in Fig.
It is evident from Figs.
Figure
We consider a transparent solar panel as designed in Fig.
![]() | Fig. 26. A schematic diagram of the simulated photovoltaic panel of 1 m2:((9 × 100 × 10) +(2 × 100 × 5)) cm2. |
Many cell connection methods are possible. In our work, the 10 cells are connected in series. Each cell has a (light collecting) surface
The LSC concentration in our case is the ratio of the incidence surface Sin to the emitting edges surface, and the latter is equal to the cell surface Scell multiplied by 2 (because two cells are connected to the LSC), so
By multiplying Eq. (
The solar panel contains 10 cells, so the total power can be written as
By considering ELSC=10% and an illumination of 1000 W/m2, the proposed panel current–voltage (I–V) and power–voltage (P–V) characteristics will be shown in Figs.
The panel characteristics are
In this work, it is shown that luminescent solar concentrator and ferroelectric solar cells can be combined together in order to form a new transparent photovoltaic panel. The ferroelectric solar cells are designed in such a way that the light incidence is lateral, which makes them adapted to luminescent solar concentrator.
Firstly, we commence the simulation using Atlas-Silvaco by considering a Schottky junction solar cell Al/ZnO/Pt. The results show that increasing the ZnO thickness and reducing the cell width lead to better efficiency values. The best simulated cell characteristics are Jsc=287.30 mA/cm2, Voc=0.92 V, Pmax=230.21 mW/cm2, fill factor = 0.87, and the efficiency is 11.51%.
Secondly, we simulate the structure Al/BTO/ZnO/Pt to investigate the effect of barium titanate layer on the cell efficiency. It is shown that thinner BTO layers give rise to better efficiency values; however, thicker ones reduce the fill factor, and this occurs due to the material’s electrical resistivity. The simulation of an optimized ferroelectric cell gives us Jsc=290.00 mA/cm2, Voc=2.00 V, Pmax=456.79 mW/cm2, fill factor = 0.78, and the efficiency = 22.84%.
Finally, a transparent photovoltaic panel is designed in which LSC absorbs a portion of the incident solar light and converts it to UV light which is guided into the ferroelectric solar cells. By considering the LSC efficiency as 10%, the estimated characteristics of a 1-m2 panel are
[31]
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] |