⇦Chinese Physics Letters, 2019, Vol. 36, No. 5, Article code 057201 Absorption Enhancement of Silicon Solar Cell in a Positive-Intrinsic-Negative Junction * Gen Yue (乐艮)1,2, Zhen Deng (邓震)1, Sen Wang (王森)1,2, Ran Xu (徐然)1,2, Xinxin Li (李欣欣)1,2, Ziguang Ma (马紫光)1, Chunhua Du (杜春花)1, Lu Wang (王禄)1, Yang Jiang (江洋)1, Haiqiang Jia (贾海强)1, Wenxin Wang (王文新)1, Hong Chen (陈弘)1,2,3** Affiliations 1Key Laboratory for Renewable Energy, Beijing Key Laboratory for New Energy Materials and Devices, Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190 2Center of Materials and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing 100049 3Songshan Lake Materials Laboratory, Dongguan 523808 Received 21 March 2019, online 17 April 2019 *Supported by the National Natural Science Foundation of China under Grant No 11574362
**Corresponding author. Email: hchen@iphy.ac.cn Citation Text: Le G, Deng Z, Wang S, Xu R and Li X X et al 2019 Chin. Phys. Lett. 36 057201    Abstract Absorption coefficient is a physical parameter to describe electromagnetic energy absorption of materials, which is closely related to solar cells and photodetectors. We grow a series of positive-intrinsic-negative (PIN) structures on silicon wafer by a gas source molecule beam epitaxy system and the investigate the absorption coefficient through the photovoltaic processes in detail. It is found that the absorption coefficient is enhanced by one order and can be tuned greatly through the thickness of the intrinsic layer in the PIN structure, which is also demonstrated by the 730-nm-wavelength laser irradiation. These results cannot be explained by the traditional absorption theory. We speculate that there could be some uncovered mechanism in this system, which will inspire us to understand the absorption process further. DOI:10.1088/0256-307X/36/5/057201 PACS:72.80.Cw, 78.66.Db, 84.60.Jt © 2019 Chinese Physics Society Article Text Solar photovoltaics (PV), one kind of renewable energy, have attracted a great deal of attention since they can offer a practical and sustainable solution to the challenge of meeting the global energy and environment requirements.[1–6] Now, demands for photovoltaics keep growing, the balance-of-system cost has become more important for PV systems. Therefore, it has become urgent to consider how the PV system cost can be reduced. A greater-efficiency solar cell would be one promising solution.[7–10] Up to date, solar cells are mainly based on silicon materials due to their advantages of natural abundance, thermal stability, and nontoxic properties. Since Si is an indirect bandgap near 1.1 eV and a direct bandgap at about 3.4 eV, the optical absorption process is accompanied by the absorption or emission of phonons.[11,12] In the recent decade, research of thin film Si solar cell has become a rich area of searching for high-efficiency and low-cost solar cell solutions.[13–16] Thus, optical characteristics of Si are of great importance and have been extensively studied with particular interests in the photon energy region from 1.0 eV to 4 eV (0.3–1.2 $\,µ$m). Generally, the absorption coefficient is a constant for a certain material, which is dependent only on the incident light wavelength. As mentioned above, silicon is the mostly common used matter for solar cells, and it has an indirect band. It is important and hopeful to increase its absorption coefficient. Recently, researchers experimentally found that the absorption coefficient is not a constant and becomes larger in low-dimensional GaAs systems compared to the bulk material.[17] However, there have been a few reports on silicon systems with low-dimensional structures. In this Letter, a series of Si-based PIN junctions with different intrinsic-layer thicknesses are investigated. We demonstrate an anomalous enhancement of the absorption of silicon solar cells of the PIN junctions with thin intrinsic layers. In this work, four samples with different intrinsic-layer thicknesses were grown on 2-inch boron-doped p-Si wafers with resistivity of 1 $\Omega$$\cdot$cm by our gas source molecule beam epitaxy system (GSMBE, VG V80 S) and the schematic diagram of the four samples is shown in Fig. 1(a). The used silicon source is Si$_{2}$H$_{6}$. The sources for n-type and p-type dopings were 1% diluted PH$_{3}$ and 1% diluted B$_{2}$H$_{6}$, respectively. The intrinsic region was sandwiched between the 500-nm p-Si layer with doping density of $5\times 10^{17}$ cm$^{-3}$ and the 250-nm n-Si layer with doping density of $5\times 10^{17}$ cm$^{-3}$. The thicknesses of the intrinsic layers of the four samples were 100 nm (sample A), 300 nm (sample B), 1000 nm (sample C) and 3000 nm (sample D), respectively. Lastly, an n$^{+}$ layer was heavily doped to form the ohmic contact. During the growth, the doping densities were controlled by the flow ratios of Si$_{2}$H$_{6}$ to the n-type and p-type dopings. The epitaxial growth temperature was 750$^{\circ}\!$C for all the samples, and the growth rate controlled by the gas flow was 1$\,µ$m/h. After the growth, the solar cell chips with sizes of 0.5$\times$0.5 cm$^{2}$ were fabricated by photolithography and etching, as shown in Fig. 1(b). In our experiment, 30 nm Ti and 300 nm Al were deposited in sequence on the face of n$^{+}$ layers with electron-beam evaporation for n-type ohmic contact. Then, a 300 nm Al film was deposited on the face of a p-Si substrate for p-type ohmic contact. The devices were annealed at 420$^{\circ}\!$C for 30 s to form good ohmic contact.
cpl-36-5-057201-fig1.png
Fig. 1. (a) Schematic diagram of the PIN junction with different intrinsic-layer thicknesses of 100 nm (sample A), 300 nm (sample B), 1000 nm (sample C), 3000 nm (sample D). (b) The chips of silicon PIN solar cells fabricated by the standard technology.
cpl-36-5-057201-fig2.png
Fig. 2. The built-in electric field distribution of different intrinsic region thicknesses: (a) 100 nm, (b) 300 nm, (c) 1000 nm, (d) 3000 nm, calculated by the diffusion-drift model.
To investigate the relationship between the intrinsic layer and the built-in electric field, we calculate the built-in electric field intensity using the diffusion-drift model, as shown in Fig. 2. The simulation results suggest that although the electric field intensities in the four samples are different, the intrinsic layer is fully covered by the built-in electric field. Moreover, although the carriers exist everywhere in the samples during light irradiation, only the carriers generated from the intrinsic layer can be extracted to the electrodes on both sides. Because the drift velocity of carriers is much faster than the diffusion velocity, the current in the external circuit mainly originates from the carriers generated and transported from the intrinsic layer. Therefore, the intrinsic layer thickness can be regarded as the absorption thickness, which suggests that only the contribution of photon-generated electrons in the intrinsic layer is taken into consideration in later discussion. To obtain the absorption coefficient, the photocurrent was measured with varied wavelengths from 400 nm to 1100 nm in steps of 10 nm. Then we calculate the absorption coefficient using the method introduced in our previous report,[17] where the calculation process was described in detail. Without considering the scattering effect during the carrier transport process, the calculated absorption coefficient dependence on the incident wavelength is plotted in Fig. 3. The absorption coefficients for samples A–D are $8.9\times 10^{4}$, $1.1\times 10^{4}$, $1.0\times 10^{4}$, $3\times 10^{3}$ cm$^{-1} $, respectively, showing that the absorption coefficient of sample D near the silicon bandgap is much smaller than that of sample A by 30 folds. In addition, the absorption coefficient of sample A is 1–2 orders of magnitude higher than those given by early works.[18–21] According to the calculation process, the actual absorption coefficient should be larger than the calculation due to the reflection loss and carrier combination. Therefore, absorption coefficients for a given material under different conditions are different. The possible reason for the increased absorption coefficient of sample A is the existence of a PIN junction.
cpl-36-5-057201-fig3.png
Fig. 3. Absorption coefficient of samples dependent on the excitation wavelength. The excitation wavelength ranges from 400 nm to 1100 nm in steps of 10 nm.
To confirm the enhancement of absorption coefficient further, a laser with wavelength 730 nm was used to explore the change of absorption coefficient with different excitation powers of incident light. Figure 4 shows that the absorption coefficient is almost invariable with the increasing excitation power. Moreover, the absorption coefficients for all the samples agree with the data shown in Fig. 3. As a result, the absorption coefficients for all the samples are enhanced in silicon of the PIN solar cells. Generally, dopings for solar cells with conventional p–n structures can decrease the absorption coefficients. However, absorption coefficients of Si-based PIN structures can be as high as the direct bandgap materials such as GaAs. The physical mechanism for the absorption enhancement in a PIN junction is unknown and it needs more detailed studies on the mechanisms of optical transition and photon-generated carrier transport in such a structure. The enhanced light absorption with a thinner intrinsic layer may have potential applications in solar cells and photodetectors. Because of the absorption enhancement, the conversion efficiency can be higher, which can reduce consumption of materials without killing the photoelectric conversion efficiency. In addition, signal-to-noise ratios of photodetectors could be much higher due to the larger absorption coefficient, which will be very helpful to increase the working temperature for some kind of photodetectors.
cpl-36-5-057201-fig4.png
Fig. 4. Absorption coefficients of the four samples versus the laser power at 730 nm.
In summary, silicon-based solar cells in a PIN junction of different intrinsic-layer thicknesses have been grown by our GSMBE system and the absorption coefficient is measured for four samples. From the photoelectric conversion theory, we find that the absorption coefficient of silicon can be modulated by a PIN junction and the absorption coefficient can be increased specially in the range from 1.5 eV to 2.5 eV. To verify the result, we measure the absorption coefficient under 730 nm laser irradiation with changing the exciting power. The absorption coefficient is enhanced and almost a constant under different power excitations. Although these experimental results cannot be explained by the existing absorption theory, they can provide us a new guide for higher efficiency solar cells. References Photovoltaics: The Semiconductor Revolution Comes to SolarOrganic photovoltaics: technology and marketOrganic photovoltaicsPoly(diketopyrrolopyrrole−terthiophene) for Ambipolar Logic and PhotovoltaicsSpectrum-Splitting Diffractive Optical Element of High Concentration Factor and High Optical Efficiency for Three-Junction PhotovoltaicsEfficient planar heterojunction perovskite solar cells by vapour depositionOptimum Nanoporous TiO 2 Film and Its Application to Dye-sensitized Solar CellsDesign Rules for Donors in Bulk-Heterojunction Solar Cells—Towards 10 % Energy-Conversion EfficiencyEnhanced solar to electrical energy conversion of titania nanoparticles and nanotubes-based combined photoanodes for dye-sensitized solar cellsOptimized Spatial Correlations for Broadband Light Trapping Nanopatterns in High Efficiency Ultrathin Film a-Si:H Solar CellsThree dimensional a-Si:H thin-film solar cells with silver nano-rod back electrodesSuperior optical response of size-controlled silicon nano-crystals in a-Si:H/nc-Si:H superlattice films for multi-junction solar cellsDirect imaging of dopant distributions across the Si-metallization interfaces in solar cells: Correlative nano-analytics by electron microscopy and NanoSIMSAnomalous enhancement of the absorption coefficient of GaAs in a p-n junctionDielectric functions and optical parameters of Si, Ge, GaP, GaAs, GaSb, InP, InAs, and InSb from 1.5 to 6.0 eVEllipsometric determination of optical constants for silicon and thermally grown silicon dioxide via a multi-sample, multi-wavelength, multi-angle investigationModeling Radiative Properties of Silicon with Coatings and Comparison with Reflectance MeasurementsAbsorption Coefficients of Crystalline Silicon at Wavelengths from 500 nm to 1000 nm
[1]Bequerel E 1839 Acad. Sci. 9 145
[2] Hammond L 1977 Science 197 445
[3] Brabec C J 2004 Sol. Energy Mater. Sol. Cells 83 273
[4] Kippelen B and Bredas J L 2009 Energy Environ. Sci. 2 251
[5] Bijleveld J C, Zoombelt A P, Mathijssen S G J et al 2009 J. Am. Chem. Soc. 131 16616
[6] Lin D F, Quan B G, Zhang Q L et al 2016 Chin. Phys. Lett. 33 094207
[7] Liu M Z, Johnston M B and Snaith H J 2013 Nature 501 395
[8] Dai S Y and Wang K J 2003 Chin. Phys. Lett. 20 953
[9] Scharber M C, Wuhlbacher D, Koppe M et al 2006 Adv. Mater. 18 789
[10] Jeyaraman A R, Balasingam S K, Lee C et al 2019 Mater. Lett. 243 180
[11]Cohen M L and Chelikowsky J R 1988 Electronic Structure, Optical Properties of Semiconductor (Berlin: Springer)
[12]Zhang Z M 2007 Nano/Microscale Heat Transfer (New York: McGraw-Hill)
[13] Ferry V E and Verschuuren M A 2011 Nano Lett. 11 4239
[14] Jang J, Kim M and Kim Y 2014 Curr. Appl. Phys. 14 637
[15] Kar D and Das D 2015 RSC Adv. 5 61118
[16] Kumar P, Pfeffer M, Willsch B et al 2017 Sol. Energy Mater. Sol. Cells 160 398
[17] Sun L, Wang L, Liu J et al 2018 Superlattices Microstruct. 122 80
[18] Aspnes D E and Studna A A 1983 Phys. Rev. B 27 985
[19] Herzinger C M, Johs B, McGahan W A, Woollam J A et al 1998 J. Appl. Phys. 83 3323
[20] Lee B J, Zhang Z M, Early E A et al 2005 J. Thermophys. Heat Transfer 19 558
[21] Wang H and Liu X L 2013 Int. J. Thermophys. 34 213