Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/81475
Title: Study of photovoltaic effect in lead-free ferroelectric perovskite oxides
Authors: Ting, Ho Fung
Advisors: Kwok, K. W. (AP)
Keywords: Photovoltaic effect
Perovskite
Metallic oxides
Issue Date: 2019
Publisher: The Hong Kong Polytechnic University
Abstract: K₀.₅Na₀.₅NbO₃ (KNN) ceramics and KNN ceramics doped with 0.25 mol% MnO₂ (KNN-Mn) or 1 mol% CuO (KNN-Cu) have successfully been fabricated by a solid-state reaction method. All the samples are well densified into an orthorhombic perovskite structure. MnO2 and CuO act as sintering aids for enhancing the densification. CuO can also reduce the sintering temperature by 50°C. The grain size of the KNN ceramic is 1.4 μm on average, whereas the grain sizes of the KNN-Mn and KNN-Cu ceramics are reduced by 0.4 μm on average. The optical properties of the ceramics have been studied based on their transmittance and diffuse reflectance spectra, and their bandgaps have been estimated by Tauc's relation. Our results show that the bandgap of the KNN-Mn ceramic is slightly larger than that of the KNN ceramic. This should be attributed to the increase in the electronegativity difference between the B-site ions and oxygen arisen from the partial substitution of Mn3+ for Nb5+. Probably owing to the ineffective substitution of Cu2+ for the B-site Nb5+, the KNN-Cu ceramic exhibits a similar bandgap with KNN. The effects of the measurement methods, transmission or diffuse reflection, on the estimation of bandgap have been studied. It has been found that the bandgaps estimated by the transmission method are dependent on the sample thickness. Owing to the stepper change in transmittance, the estimated bandgap for thinner samples are larger. Moreover, the bandgaps estimated by the transmission method are generally lower than those obtained from the diffuse reflection method. The direct bandgaps of KNN and KNN-Mn estimated using the transmittance spectra are 3.16 eV and 3.21 eV, respectively, while their indirect bandgaps are 3.14 eV and 3.19 eV, respectively. On the other hand, the direct bandgaps of KNN, KNN-Mn and KNN-Cu estimated using the diffuse reflectance spectra are 3.34 eV, 3.41 eV and 3.35 eV, respectively, and their indirect bandgaps are 3.13 eV, 3.22 eV and 3.26 eV, respectively. The differences between the bandgaps estimated from the transmittance and diffuse reflectance spectra have been discussed based on their theoretical assumptions.
The ferroelectric photovoltaic properties of the samples have been studied. It has been shown that the direction of the short-circuit current is opposite to the net polarization or the poling field, while no short-circuit current is observed in unpoled samples under illumination. Also, the higher the poling field, the larger the net polarization, and then the stronger the depolarization field is. As a result, the observed Isc increases with the poling field because of the higher efficiency in separating and drifting the opposite charges. It is suggested that the driving force of the short-circuit current is attributed to the net polarization. As confirmed by a controlled experiment, the short-circuit current is originated from the photo-excited electron-hole pairs. It is note that the observed Isc increases rapidly to a maximum value once the sample is illuminated, but it decreases gradually with time and becomes saturated afterwards. The variations of the observed Isc is confirmed not contributing by the thermal effect. The repeatability of the observed Isc has also been evaluated. The observed Isc exhibits similar variations with time in each cycle that consists of 20-min illumination and 40-min darkness, but decreases slightly in the consecutive cycles. However, after keeping the sample short-circuited and in the dark for a long time (e.g., 85 hours), the Isc increases and reaches almost the same maximum level as in the first illumination cycle, suggesting that the reduction in Isc is reversible. Charged carriers may temporarily be trapped at the sample-electrode interfaces during the illumination, and the release of the trapped charges is a very slow process that requires a long period (85 hours) to complete. The photocurrent responses of the KNN-Mn and KNN-Cu ceramics have also been studied. Similar to KNN, the observed Isc for both the ceramic samples exhibit similar variations with time. However, their maximum values are much lower than that of KNN, only about 30%. It is interesting to note that, despite the difference in the maximum level they reached at the beginning of illumination, they all saturate at almost the same level. The greater Isc observed in KNN may due to its higher defect density. Charges may temporarily be trapped in the defects when the sample is stored short-circuited and in the dark. Under the illumination, the trapped charges are excited and released slowly, and thus contributing to the observed Isc. After the release of most of the trapped charges, the Isc becomes saturated and reaches a level similar to those for KNN-Mn and KNN-Cu. As all the ceramics have similar bandgaps, the saturated Isc should then reflect to the photovoltaic effect. More investigation is needed to further understand the phenomenon.
Description: xv, 94 pages : color illustrations
PolyU Library Call No.: [THS] LG51 .H577M AP 2019 Ting
URI: http://hdl.handle.net/10397/81475
Rights: All rights reserved.
Appears in Collections:Thesis

Files in This Item:
File Description SizeFormat 
991022270852503411_link.htmFor PolyU Users168 BHTMLView/Open
991022270852503411_pira.pdfFor All Users (Non-printable)5.44 MBAdobe PDFView/Open
Show full item record
PIRA download icon_1.1View/Download Contents

SCOPUSTM   
Citations

1
Citations as of Nov 15, 2019

WEB OF SCIENCETM
Citations

1
Citations as of Nov 10, 2019

Page view(s)

13
Citations as of Nov 13, 2019

Download(s)

5
Citations as of Nov 13, 2019

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.