Design and fabrication of a down-conversion solar cell

Wang LiangXing

Supervisor: Ming Lu

SQ limit was first calculated by William Shockley and Hans Queisser in 1961.[1] It assumes that each photon with energy greater than Eg generates one photon. Based on this theory, silicon solar cell has a maximum efficiency of 29% (practical efficiency is 25%). [1,2] The mainly losses are reflection of the front surface and recombination of interfaces.[3]

In order to break the SQ limit many approaches are suggested. The keys are: (1) upconversion - to combine sub-Eg photons into one photon than can be absorbed, (2) downconversion - to split high energy photons into multiple photons that each have energy greater than Eg. [4,5]

In this paper we will be concerned primarily with downconversion. New approaches that Al/SiO2 structure and texturisation of front surface used for solar downconversion are described. Detailed design results and fabrication process are presented.

There are two main aspects in this work. One is focused on the research of the preparation of the downconversion material, especially Al/SiO₂ structure is explored. This structure can be grown on Si wafer without high temperature which could decrease the performance of si solar cell. The other aspect is focused on the texturisation of front surface, and the formation of a pyramidal structure on the surface of <100> c-silicon wafers is an effective method to reduce reflection loss. The approach discussed here is different from other texturisation [6,7,8] which has shown the lowest reflectivity that has been reported. These results suggest that this new method for solar downconversion can be beneficial for si solar cell.

References

[1] William Shockley and Hans J. Queisser, Journal of Applied Physics, Volume 32 (March 1961), pp. 510-519;

[2] Zhao J, Wang A, Green MA, Ferrazza F. Applied Physics Letters 1998; 73: 1991–1993.

[3] A. Richter, M. Hermle, S.W. Glunz (Oct 2013). IEEE Journal of Photovoltaics 3 (4): 1184–1191.

[4] Mark B.Spitzer, HansP. Jenssen, Arlete Cassanho, Solar Energy Materials & Solar Cells108 (2013) 241–245

[5] B.M.vander Ende ,L.Aarts, A.Meijerink, Physical Chemistry Chemical Physics 11 (2009) 11081–1109

[6] M. Moreno, D. Murias, J. Martı´nez, C. Reyes-Betanzo, A. Torres, R. Ambrosio, P. Rosales, P. Roca i Cabarrocas, M. Escobar，Solar Energy 101 (2014) 182–191

[7] S. Belhadj Mohamed, M. Ben Rabha, B. Bessais, Solar Energy 94 (2013) 277–282

[8] Rocı´o Barrio, Nieves Gonza´lez, Julio Ca´rabe, Jose Javier Gandı´a, Solar Energy 86 (2012) 845–854

The mechanism of perpendicular exchanging bias in [Co/Ni]_N/TbCo composite structure

Minghong Tang

Supervisor：Zongzhi Zhang

       Since the exchange-bias (EB) effect was firstly observed in CoO/Co ferromagnetic (FM) /antiferromagnetic (AFM) bilayer system by Meiklejohn and Bean¹ in 1956, the biasing effect has become a significant part of spintronics for basic research and practical applications of giant magnetoresistance or tunnel magnetoresistance devices in hard disk drives and spin-transfer-torque magnetic random access memories (STT-MRAM)². This phenomenon is typically characterized by a shift of the magnetic hysteresis loop towards a non-zero field defined as the exchange bias field (H_E),when the system is below a blocking temperature (T_B). Commonly, the magnetic interface coupling in the heterostructures with uncompensated spins³ of the antiferromagnetic (AFM) layer is considered as the origin of the EB effect. However, other exchange-coupled bilayer systems such as FM/FI⁴, FI/FI⁵ have also shown EB effect. However the origin of EB effect with a ferrimagnetic layer is still not yet fully understood.

In order to thoroughly understand the PEB mechanism of the heterostructures, in this work, we have fabricated various series samples including Tb_xCo_100-x alloy and [Co(0.28)/Ni(0.58)]_N/Tb_xCo_100-x composite structures, where both the ferrimagnetic TbCo alloy and ferromagnetic[Co/Ni]_N show strong PMA. The choice of ferromagnetic [Co/Ni]_N pinned layer is due to its relatively high spin polarization⁶ and small Gilbert damping factor⁷, which are valued in the MRAMs for high STT efficiency. Similar to the conventional ferrimagnetic alloy⁸, a ferrimagnetic-like coupling is clearly presented in the heterostructures, but the Tb content corresponding to the composition compensation (at which the effective total magnetic moments become zero) was pushed to a higher value due to the existence of Co/Ni multilayer. By tuning the deposition conditions and structure parameters, large room temperature H_E as high as 10 kOe is peculiarly observed in a perpendicular coupled [Co(0.28)/Ni(0.58)]₅/Tb₃₀Co₇₀(12.0) heterostructure (in unit of nanometer). Different from the FM/AFM system in which the FM layer is pinned by the AFM layer, we found that the pined layer in this composite structure can be set as TbCo or [Co/Ni]_N, depending on their effective magnetic moments. In other words, the one owning smaller moments can be biased. In order to distinguish the biased layer between the ferromagnetic[Co/Ni]_N and ferrimagnetic TbCo layers, temperature dependent magnetization measurements have been carried out. Furthermore, we choose TbCo as the biased layer to investigate the PMA effect of the FM pinning layer on the H_E. The effective magnetic anisotropy (K_eff) of the FM pinning layer was tuned by inserting a Co interlayer between the FM and FI and by changing the repetition number N of the Co/Ni MLs. In all, we demonstrate that the PEB can be established and controlled by designing the magnetic moments of FI or FM part and the K_eff of pinning layer, which are of great significance for applications in spintronics.

References

1. Meiklejohn, W. & Bean, C. P. Phys. Rev. 102, 1413–1414 (1956).

2. S. Mangin, et al. Nat. Mater. 5, 210–215 (2006).

3. K. Takano. et al. Phys. Rev. Lett. 79, 1130 (1997).

4. William C. Cain. et al. J. Appl. Phys. 67, 5722 (1990).

5. F. Radu. et al. Nat. Commun. 3,715 (2012).

6. S. Mangin. et al. Nat. Mater. 5, 210–215(2006).

7. H.S. Song. et al. Appl. Phys. Lett. 102, 102401 (2013).

8. Alebrand, S. et al. Appl. Phys. Lett. 101, 162408 (2012).

The photoluminescence and photoelectrochemical properties of II-IV semiconductor shell/ZnO corenanorod arrays

Xu Yang

Supervisor: Jiada Wu

With the increasing demand for energy, the limited fossil fuel reserves and the increasingly serious environmental pollution, searching for reproducible resources is becoming an imperious task. Solar energy is an ideal new energy to meet the target because of its abundant, clean and inexhaustible characteristic. Solar cell is an important application for solar energy. Zinc oxide (ZnO) is an important wide band gap (3.37 eV) semiconductor material with a large exciton binding energy (60 meV) [1]. ZnO has been demonstrated as a promising candidate for photoanode due to its appropriate energy band position and good corrosion resistance in aqueous solution. ZnO also have amazing electronic transmission characteristics, good morphology control, simple preparation process and low cost [2]. However, ZnO itself cannot absorb and utilize visible light because of its wide band-gap. An efficient approach to extend the spectral region of photoresponse and enhance the photocatalytic or photovoltaic efficiency of ZnO-based devices is to sensitize the surface of nanostructured ZnO with photosensitizers such as narrower band-gap semiconductors forming type-II core/shell nanostructure. Depositing a thin coating of a narrower band-gap semiconductor (e.g. CdS, ZnSe) outside nanosizedZnO provides an approach to permanently sensitize ZnO [3]. Meanwhile, the inorganic semiconductor coating can protect nanosizedZnO from surface degradation.

In the previous work, we studied the influences of shell thickness and core-shell structure on the optical properties of ZnO/CdS core/shell nanorods (NRs) for an elucidation of the mechanisms responsible for extended photo-response. Well aligned ZnO/CdS core/shell NRs were fabricated on indium tin oxide substrates using hydrothermally grown ZnO NRs as the cores and pulsed laser deposited CdS coatings as the shells. The sample structure was characterized by X-ray diffraction and Raman backscattering spectroscopy, revealing the wurtzite structure of both the ZnO cores and CdS shells, and the improvement in the structure after annealing. The optical properties were studied through optical transmittance, absorbance and photoluminescence measurements, showing the optical properties featured with type-II heterogeneous nanostructures constructed from ZnO and CdS. The results provide a support that the optical properties of the CdS covered ZnO NRs are attributed to the suppressed radiative recombination of photo generated carriers due to the efficient spatial separation of electrons and holes in the nanosized ZnO-CdS heterostructures.

After that, we grew the ZnSe coated ZnO nanorod arrays. We found this nano-hetero structure has the same photoluminescence properties with the CdS coated ZnO nanostructure. Next, we will mainly study the photoelectrochemical properties of the fabricated CdS/ZnO and ZnSe/ZnOnanorod arrays used as the photoelectrodes under visible illumination. In the photoelectrochemical (PEC) process, the solar conversion efficiency is mainly determined by the photoelectrodes. Morphological and structural control of photoelectrode play an important role for enhancement of PEC properties. The development of photoelectrodes with high utilization of solar energy, high energy conversion efficiency and excellent stability is the key issue for PEC technology [4].

References

[1] S.Khanchandani, S.Kundu, A.Patra, A. K.Ganguli,J. Phys. Chem. C 116 (2012) 23653-23662.

[2] Jinwen Ma, Shi Su, Wuyou Fu, Haibin Yang, Xiaoming Zhou etc. CrystEngComm 16 (2014) 2910-2916.

[3]P. Reiss, M.Protiere, L. Li,Small5 (2009) 154-168.

[4] Juan Wang, Wei-De Zhang etc. ElectrochicaActa. 71 (2012) 10-16.

Band Structure of Chalcogen-Hyperdoped Silicon

Jiamin Xiao

Supervisor: Jun Zhuang

       Nowadays, chalcogen-hyperdoped silicon has drawn an increasing interest because of the unique optical properties (a near-unity broadband absorption)[1,2] and electricalcharacteristics(insulator-to-metal transition(IMT))[3,4] which render chalcogen-hyperdopedsilicon a promising application in silicon-based infrared photodetectors and photovoltaics with increased efficiency[5,6]. To understand the origins of these novel features, the study of band structures of chalcogen-hyperdopedsilicon is indispensable.

       The density functional theory (DFT) has become the method of choice for studying the electronic structure of heavily doped systems.And there are different treatments of theelectronic band structures from DFT calculations. For example, Sa´nchez et al. have employed a scissor operator to separate the conduction band(CB) from the impurity band(IB), whereas Ertekin et al. have made no correction and left unchanged the relative positions of IB and CB. Then Sa´nchez et al. have shown that the hyperdopedchalcogen could introduce IBs located in the middle of the silicon gap when the chalcogen impurities are doped at a concentration of 0.463% (or 1/216).[7]

       Ertekin et al., however, have performed DFT calculations for Se-hyperdoped silicon at various concentration of dopants and shown that when the doped Se reached 0.4% (or 1/250), the IB would merge with the CB, and they also proposed that the IMT may be from the merging of IB and CB.[4]Two results are obviously inconsistent. So the acquisition of the accurate band structures for chalcogen-supersaturated silicon is urgent and crucial.

       We have employed the hybrid functional proposed by Heyd et al. (HSE06)and the many-body perturbation theory based on the GW approximation to investigate band structures of hyperdoped silicon. Furthermore, considering the adopted models in conformity with the actual heavily doped silicon material, super cell is made corresponding modificationin order to study the influence of the disorder of impurity position on the band structure of sulfur-hyperdoped silicon in depth.

References

[1] Wu C., Crouch C. H., Zhao L., Carey J. E., Younkin R., Levinson J. A., Mazur E., Farrell R. M., Gothoskar P. and Karger A., Appl. Phys. Lett., 78(2001) 1850.

[2] Kim T. G., Warrender J. M. and Aziz M. J., Appl. Phys. Lett., 88(2006) 241902.

[3] Winkler M. T., Recht D., Sher M. J., Said A. J., Mazur E. and Aziz M. J., Phys. Rev. Lett., 106(2011)178701.

[4] Ertekin E., Winkler M. T., Recht D., Said A. J., Aziz M. J., Buonassisi T. and Grossman J. C., Phys.

Rev. Lett. , 108(2012) 026401.

[5] Carey J. E., Crouch C. H., Shen M. Y. and Mazur E., Opt. Lett., 30(2005) 1773.

[6]Mahmood A. S., Sivakumar M., Venkatakrishnan K. and Tan B., Appl. Phys. Lett., 95(2009) 034107.

[7] K. Sa´nchez, I. Aguilera, P. Palacios, and P. Wahno´n, Phys. Rev. B 82(2010) 165201.

Time: 6:30 pm, Thursday, 2014.12.25

Location:Optical Building. Room 525