Contents lists available at ScienceDirect

Solar Energy

journal homepage: www.elsevier.com/locate/solener

Broadband omnidirectional anti-reflection property of V-groove texturedsilicon

Yan Zhaoa,c, Yaoping Liua,b,⁎, Quansheng Chena,c, Wei Chena,b, Juntao Wua,c, Yan Wanga,b,Xiaolong Dua,b,c,⁎

a Key Laboratory for Renewable Energy, Beijing Key Laboratory for New Energy Materials and Devices, Beijing National Laboratory for Condensed Matter Physics, Instituteof Physics, Chinese Academy of Sciences, Beijing 100190, Chinab Songshan Lake Materials Laboratory, Dongguan, Guangdong 523808, Chinac School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100049, China

A R T I C L E I N F O

Keywords:OmnidirectionalAnti-reflectionV-groove textured siliconCopper-assisted chemical etching

A B S T R A C T

Omnidirectional silicon solar cells with an outstanding anti-reflection property over a broad angular range ofincident light show great potential for application in the photovoltaic industry. Many studies have been carriedout to investigate the omnidirectional anti-reflection properties of different silicon textured structures. However,the omnidirectional anti-reflection property of the V-groove textured silicon has not been reported, even thoughthe V-groove structure is also one of most frequently used anti-reflective structures for silicon solar cells. In thisresearch, ray tracing simulation is utilized to analyze and validate the omnidirectionality of a micron-sizedsilicon V-groove structure. In addition, in consideration of the previous complicated fabrication methods of V-groove structure, a brand new fabrication method for the V-groove texturization onto diamond-wire-sawnmonocrystalline silicon wafers through copper-assisted chemical etching is presented. Consistent with the si-mulations, the experimental results similarly indicate that the V-groove textured silicon not only possesses lightabsorption ability that is comparable to the common random pyramid structure but also exhibits an unexpectedexcellent omnidirectional anti-reflection property over a wide range of incident angles, demonstrating the greatpotential for the fabrication of omnidirectional silicon solar cells.

1. Introduction

Texturization is an effective way to enhance the light absorptionand conversion efficiency for silicon solar cells. Monocrystalline silicon(c-Si) solar cells with different textured structures have been fabricated,and the optical properties of these structures have been investigated inprevious studies (Campbell and Green, 1987; Smith and Rohatgi, 1993;Wang et al., 2015; Yang et al., 2017a; Zhao et al., 1999). However, theoptical properties of these structures are always measured under thestandard conditions in which light is incident perpendicularly to the Siwafers despite these test conditions being inconsistent with the actualapplicable situations of Si solar cells. Actually, the angle of incidence(AOI) between light and Si wafers changes all the time, since the sunmoves from east to the west during the daytime. For this reason, theomnidirectional anti-reflection ability has been presented to evaluatethe optical properties for Si wafers and Si solar cells under differentlight incident conditions, and the omnidirectional performances ofvarious Si textured structures, including the upright and inverted

pyramid structures, some nanostructures and hierarchical structures,have been studied (Chhajed et al., 2008; Haiyuan et al., 2018; Huanget al., 2007; Kuang et al., 2016; Lin et al., 2014a; 2014b; Savin et al.,2015; Spinelli et al., 2012; van Dam et al., 2016; Wei et al., 2013; Zhonget al., 2017). It has been demonstrated in relevant studies that thetextured c-Si wafers with random nano inverted pyramid structure ornanopyramid structure possess quasi-omnidirectional anti-reflectionproperties since the light absorption performance will not be dramati-cally degraded over certain angular ranges (Haiyuan et al., 2018; Zhonget al., 2017). In addition, the excellent omnidirectional anti-reflectionabilities of the Si nanowires (Huang et al., 2007; van Dam et al., 2016),Si nanopillars (Savin et al., 2015; Spinelli et al., 2012) and some otherSi nanostructures (Kuang et al., 2016; Lin et al., 2014a, 2014b; Weiet al., 2013) have also been verified. In previous study, a new parameterhas been brought up to further evaluate the omnidirectionality of dif-ferent samples, which is defined as (Chhajed et al., 2008)

https://doi.org/10.1016/j.solener.2019.09.048Received 23 May 2019; Received in revised form 15 August 2019; Accepted 12 September 2019

⁎ Corresponding author.E-mail addresses: ypliu@iphy.ac.cn (Y. Liu), xldu@iphy.ac.cn (X. Du).

Solar Energy 193 (2019) 132–138

0038-092X/ © 2019 Published by Elsevier Ltd on behalf of International Solar Energy Society.

T

∫ ∫=Rλ θ

Rdθdλ1Δ

1Δ λ

λ

θ

θ

min

max

min

max

(1)

where θ represents the zenith angle during the whole day, while λ andR represent the wavelength of the incident light and the correspondingreflectance respectively. Nevertheless, omnidirectional research for a SiV-groove structure has not been proposed, even though the V-groovestructure is a quite frequently-used textured structure and the lighttrapping ability of the structure has been emphasized in many studies(Backlund and Rosengren, 1992; Borojevic et al., 2014; Campbell andGreen, 1987; Dobrzański and Drygała, 2008; Nussbaumer et al., 1994;Scheibe and Obermeier, 1995; Untila et al., 2013; Vangbo andBäcklund, 1996; Wilbers et al., 2018; Willeke et al., 1992; Winderbaumet al., 1997; Yagi et al., 2006; Yang et al., 2017b; Zhang et al., 2019).

Numerous fabrication techniques for the texturization of the V-groove structure on Si wafers have been put forward. Mechanical dicingused to be considered as a promising texturization for multi-crystallinesilicon wafers as there is no preference for the crystal orientations, anddifferent types of V-groove structures can be obtained by using variousbeveled blades (Nussbaumer et al., 1994; Willeke et al., 1992; Yanget al., 2017b). However, this mechanical technique with low accuracy isnot suitable at present with the large decrease of wafer thickness andthe request for high conversion efficiencies of solar cells. Methods usingmasks combined with anisotropic etching have been widely adopted, assuch masks can be prepared through photolithography (Backlund andRosengren, 1992; Wilbers et al., 2018), thermal oxidation (Vangbo andBäcklund, 1996), chemical vapor deposition (CVD) (Scheibe andObermeier, 1995) and so on. In addition, laser and reactive-ion etching(RIE) have been applied for the production of V-groove structures aswell (Dobrzański and Drygała, 2008; Untila et al., 2013; Winderbaumet al., 1997; Zhang et al., 2019). However, the preparation processes ofthese techniques are rather complicated, and the costs are extremelyhigh when applied in the large-scale production.

In this study, ray tracing simulation has been applied for the ana-lysis and verification of the omnidirectionality on the V-groove texturedSi. The simulation results show varying anti-reflection performance aslight is incident along different directions. For further confirmation ofthe simulation, a one-step Cu-assisted chemical etching process wasutilized for the fabrication of a V-groove structure on diamond-wiresawn (DWS) c-Si wafers. The micron-sized V-grooves formed throughthis all-solution-processed method are arranged in parallel on the wa-fers. In addition, we have observed and analyzed the formation processfor the V-grooves and proposed a possible formation mechanism for theV-groove textured structure. More importantly, systematic explorationand quantitative analyses on the omnidirectionality of this V-groovetextured Si were also carried out, in which the random pyramidstructure was studied for comparison. Consistent with the simulationresults, the experimental results similarly indicate that the V-groovetextured Si has an excellent quasi-omnidirectional characteristic over abroad range of AOI. In consequence, the V-grooves texture Si which hasoutstanding omnidirectional optical advantages can be utilized for thefabrication of the omnidirectional silicon solar cells.

2. Material and methods

The analytical model for the ray tracing simulation in this paper is aSi V-groove array in which the V-shaped grooves are arranged in par-allel on a Si substrate. The opening size of the grooves is 2 µm, while theincluded angles for the grooves are 90°. The light beams in the simu-lation are incident in a line that is perpendicular to the extending di-rection of the V-grooves, considering the uniformity of the structurealong the extending direction. The wavelength of the incident light is650 nm, and the quantity of incident rays in each groove is 200. Thetotal power of the incident rays is 1W. The incident direction for thebeams changes, and in the simulation, the AOI between the incidentlight and the normal of the substrate lies in the range of 0–80°.

(1 0 0)-oriented Czochralski (CZ) monocrystalline silicon wafers (p-type boron-doped, 156 µm×156 µm, 200 ± 10 µm thick, resistance1–3Ω·cm) sliced using diamond-wire sawing technology were used forthe texturization in this study. The DWS c-Si wafers were textured in amixed solution containing 100mmol L−1 Cu(NO3)2, 4.5 mol L−1 HFand 1.2mol L−1 H2O2 for 3 s, 30 s, 1 min, 3min and 5min, respectivelyat 30 °C to obtain the V-groove structure and analyze the formationprocess. After that, the wafers were then immersed in an aqueous HNO3

solution to remove the Cu-NPs deposited onto the surface of the Siwafers during etching. Additionally, the samples were immediatelyrinsed in deionized water after removal from the mixed texturing so-lution as well as the HNO3 solution. A random pyramid structure on a Sisubstrate was obtained through etching in a solution containing 2 wt%KOH and 10 vol% IPA for 25min at 80 °C.

Scanning electron microscopy (SEM, Zeiss Sigma 300) was em-ployed to investigate the surface and cross-section morphologies of theV-groove samples. The reflectance spectra at varying AOIs for thesamples were measured using a UV–Vis-NIR spectrophotometer(Agilent Cary 7000) with an integrating sphere. The reflectance spectrawere measured in the wavelength range from 300 to 1000 nm, and theAOI between the incident light and normal of the Si wafers varied in therange of 0–80°.

3. Results and discussion

3.1. Ray tracing simulation on the V-groove textured silicon

To evaluate the omnidirectional anti-reflection ability of the V-groove structure on the Si substrate, ray tracing simulation was used tocalculate the reflectance of the V-groove structure for changes in theAOI ranging from 0° to 80°. In the simulation, two simplified circum-stances in which the incident rays change along the different directionswere analyzed and discussed. Further investigation on reflectance forthree-dimensional radiation have also been performed, as can be seenin the Fig. S1. The incident directions and AOIs for these two conditionsare shown in Fig. 1a. In the first case, the direction of the incident lightchanges in the plane xoz that is parallel to extending direction of the V-grooves and perpendicular to the surface of the Si substrate at the sametime, which is called the parallel for short in this paper. Meanwhile, theperpendicular direction represents the direction of incident light thatchanges in the plane yoz perpendicular to the V-shaped grooves as wellas the surface of the sample. The simulated reflectances under varyingAOIs in these two cases are provided in Fig. 1a.

It can be seen clearly that the reflectance gradually decreases as theAOI increases from 0 to 70° in the parallel direction, accompanied withan evident rise when the AOI is in the range of 70–80°. Despite theincrease at high AOI, the reflectance over the wide AOI range of 0–80°still remains lower than 15% in this condition, indicating an excellentomnidirectional anti-reflection property. However, a different changingtendency is revealed when light is incident perpendicular to the groovesin the figure, as the calculated reflectance will first increase to morethan 30% and then remain unchanged for the AOI larger than ap-proximately 45°. Reflection paths and the relative intensity of the lightin these two cases are shown in Fig. 1b and c (Supplementary Video 1and 2). When light irradiates in parallel, the light will always undergo asecondary reflection regardless of the AOI, see Fig. 1b. However, thereflection paths in Fig. 1c show that the light will only be reflected onceat high AOI values when the incident light is perpendicular to thegrooves, resulting in a decrease in the light absorption.

Comparing with the other commonly used surface structures in-vestigated in previous studies, the simulated results show that the Si V-groove structure possesses an outstanding omnidirectional antireflec-tion ability in the AOI range of 0-80° when light is incident along the V-grooves. The reflectances of the random micron pyramid structure,nano pyramid structure as well as nano inverted pyramid structureraised slowly when AOI is in the range of 0–50° but then increased

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

133

rapidly, resulting in the great loss of light absorption (Haiyuan et al.,2018, p.; Zhong et al., 2017). In addition, the other nanostructures,which has also been frequently used for Si solar cells, hardly exhibitedsatisfying anti-reflection properties when light was incident obliquelyat large angles, especially when the AOI was larger than 60° (Savinet al., 2015; Spinelli et al., 2012; van Dam et al., 2016). In consequence,the V-groove structure is quite advantageous for the fabrication ofomnidirectional Si solar cells.

3.2. Fabrication of the V-groove textured silicon

To further confirm the omnidirectionality of the V-groove structureon the Si substrate, a V-groove textured structure was fabricated ontothe surface of DWS c-Si wafers through a brand new fabricationmethod. Cu-assisted chemical etching, which has been generally usedfor the preparation of inverted pyramid structure and some nanos-tructures, was utilized for preparing the V-grooves in this work. TheDWS c-Si wafers were etched in a mixed solution to obtain the V-groovestructure in this study; the surface morphologies of the textured struc-ture are shown in Fig. 2a and b in which some tightly arranged parallelV-shaped grooves can be clearly seen. The opening sizes and depths forthe grooves are in the range of 1–5 µm; such wide openings will makethe structure suitable for adequate passivation in subsequent fabrica-tion processes for solar cells. The cross section of the V-groove structurecan be seen in Fig. 2b, and measurements prove that the included an-gles for the grooves are 90°. Additionally, it is worth noting that thesurfaces of these deep grooves on the textured samples are quitesmooth, which will ensure low recombination of carriers for solar cells.

The etching process and formation mechanism for the V-groovestructure were also analyzed. The initial surface appearance for a rawDWS c-Si wafer is shown in Fig. 3a. The surface of the raw wafer is not

very rough, while some smooth grooves and cracks resulting from thewafer dicing can be found in the figure. During the etching process,obvious deposition of Cu-nanoparticles (Cu-NPs) on the substrate canbe observed when the Si wafer is immersed in the etching solutioncontaining Cu2+, F− and H2O2 for only 3 s, as shown in Fig. 3b, Thedistribution of the Cu-NPs on the surface is not uniform, and it seemsthat the Cu-NPs prefer to nucleate at the bottom of the deep grooves orat cracks at previous etching stage, while fewer Cu-NPs are found onshallow grooves and smooth areas. An approximate line-shaped dis-tribution for the Cu-NPs can be observed when the etching time extendsto 60 s in Fig. 3d. As the etching time is prolonged from 60 s to 300 s,the distribution of Cu-NPs stays consistent in terms of line shape, andthe V-groove structure gradually forms as the Si is etched away, seeFig. 3c–f.

The main cause for this selective deposition of Cu-NPs should be theheterogeneous surface of the DWS c-Si wafer. Previous studies haverevealed that the DWS c-Si wafer has a quite complex surface phasecomposition consisting of stable crystalline silicon, amorphous silicon(a-Si) and some other metastable phases, due to the different modes ofstress release incurred during wafer slicing (Yu et al., 2012). In detail,shallow grooves and flat areas on the raw DWS c-Si wafer are coveredwith a layer of a-Si, while deep grooves and cracks on the surface areconcentrated with defects but which remain crystalline silicon in gen-eral (Chen et al., 2015, 2014). For the Cu-assisted chemical etchingprocess, it has been proven that the Cu-NPs will preferentially depositon regions where the defects are concentrated because the electrons aremuch easier to be captured in these areas (Zhong et al., 2011). On theother hand, the a-Si layer on the surface of the Si wafers is often used asa mask for etching processes with a much lower etching rate (Meinelet al., 2012). As a result, this difference in deposition rates of Cu-NPsbetween different areas leads to more Cu-NPs being deposited in deep

Fig. 1. Simulation results of V-groove textured silicon. (a) Simulated reflectance of the V-groove textured silicon over the AOI range of 0–80°. (b) Schematic diagramsof the light transmitting paths on V-groove textured silicon when light is incident in direction parallel to the V-grooves. (c) Schematic diagrams of the lighttransmitting paths on V-groove textured silicon when light is incident in direction perpendicular to the V-grooves. Different colors in the legend represent differentrelative intensity of the light. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. SEM images of V-groove structures prepared through the Cu-assisted chemical etching, (a) before the removal of Cu-NPs, (b) after the removal of Cu-NPs.

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

134

grooves and cracks compared with shallow grooves and flat areas onthe surface of the Si wafers during the texturization process.

However, the formation of the V-groove structure is probably notonly the result of the selective deposition of Cu-NPs but also the su-perior anisotropic etching ability of the Cu-NPs, which plays an im-portant role. In the Cu-assisted etching process, the silicon under theline-shaped gathered Cu-NPs will be gradually etched away and thecrystal planes with lower etching rates are exposed as etching time isprolonged (Kawasegi et al., 2005; Wang et al., 2017; Wu et al., 2018).For the DWS c-Si wafers with (1 0 0) orientation, the surface is coveredwith shallow smooth grooves that are parallel to the ⟨1 0 0⟩ crystalorientation. Therefore, the line-shaped gathered Cu-NPs are depositedalong the ⟨1 0 0⟩ crystal orientation as well, leading to the formation ofthe textured V-grooves with the same extending directions. Under thiscircumstance, the sidewalls of the textured V-grooves are composed oflattice planes, and the included angles are 90°.

3.3. Omnidirectionality of the V-groove textured silicon

The optical properties of the V-groove textured structure that wasprepared through Cu-assisted chemical etching were studied. The re-flectance spectrum of the V-grooves textured sample in the wavelengthrange 300–1000 nm was measured, as seen in Fig. 4. Moreover, thereflectance spectra for a raw DWS c-Si wafer and a pyramid textured c-Si wafer are provided as references in this figure. The average re-flectance R̄ in the wavelength range of 300–1000 nm is calculated asfollows,

∫

∫=R

R λ S λ dλ

S λ dλ¯

( ) ( )

( )nm

nm

nmnm

3001000

3001000

(2)

where λ and R(λ) represent the wavelength of the incident light and theexperimental reflectance corresponding to λ, respectively, while S(λ)

Fig. 3. Formation process of the V-groove textured silicon. (a) SEM image of raw DWS c-Si wafer. SEM images of textured Si wafers which have been etched for (b)3 s, (c) 30 s, (d) 60 s, (e) 180 s and (f) 300 s.

Fig. 4. Reflectance spectra of the V-groove textured surface, raw DWS c-Sisurface and random pyramid surface in the wavelength range of 300–1000 nm.

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

135

represents the solar photon flux spectral distribution under the standardAM1.5 conditions. The test results show that the average reflectance R̄calculated using the expression (2) of the V-groove textured sample is12.33%. This result is slightly higher when compared with the con-ventionally used random pyramid sample for which the R̄ is 10.75% inthe same wavelength range. Nevertheless, the V-groove texturedstructure can still be recognized as an effective anti-reflection structure,since the light absorption is dramatically increased; theR̄ of a raw DWSc-Si wafer without any texturing treatment reaches up to 34.23% in thewavelength range of 300–1000 nm.

However, the measured reflectance of the V-groove textured Si atthe wavelength of 650 nm is slightly lower than the simulated resultunder normal incidence shown in Fig. 1a. The difference may probablyresulted from the distinctions between the ideal analytical model weused in the simulations and the V-groove textured sample we obtained.In simulations, all the grooves had the same size and perfect V-shapedcross-sections, and they were all arranged parallel to each other.However, the V-groove textured c-Si sample we fabricated were notexactly the same as the model we used in the simulations. As can beseen in Fig. 2, the sizes of the textured V-shaped grooves, including thewidths and depths, are slightly different. The extending directions ofsome grooves were not so straight. In addition, the nanostructures onthe surface of V-groove textured Si, which were formed during themetal-assisted chemical etching process, also had influence on thelower measured reflectance.

Reflectance spectra as a function of AOI and wavelength for the V-groove textured structure were acquired in this work. Similar to thesimulation results, the experimental results also demonstrate that the V-groove structure has a remarkable omnidirectional anti-reflectionproperty over a wide range of AOI. Fig. 5a shows the first case when theincident light changes in the parallel direction. In the short wavelengthregion, the reflectance of the V-grooves sample stays at approximately30% over the AOI range of 0–80°, while the reflectance in the longwavelength region is maintained at approximately 10%; this shows aprominent broadband and omnidirectional light absorption ability.Fig. 5b shows the other case when the AOI changes in the directionperpendicular to the grooves correspondingly. We can determine in thefigure that the reflectance increases gradually as the AOI is varied from0° to 80°, even approaching 50% in the short wavelength range. Thismeans that the anti-reflection ability of the V-grooves textured structurewill deteriorate when the incident light is obliquely in the perpendi-cular direction, which is different from case of parallel incident light.

The same experiments and analyses were also performed on arandom pyramid sample for comparison. It is worth noting that there isonly one figure for the random pyramid sample to characterize therelationship between the reflectance, wavelength and AOI, as the dif-ferent incident directions are equivalent for this uniform randomstructure. Fig. 5c shows the reflectance of the random pyramid sampleas a function of the wavelength and AOI. The changing tendency for thepyramid structure resembles the variation trend for the V-groovestructure in the perpendicular direction, as the reflectance will increasewith increasing AOI both in the short and long wavelength regions. Thischanging tendency indicates that the light absorption for the randompyramid textured c-Si wafer will be decreased dramatically when lightis obliquely incident, thus resulting in a reduction in conversion effi-ciencies for the solar cells.

The relationship between the average reflectanceR̄ and AOI in thewavelength range of 300–1000 nm for the V-groove and pyramidsamples is provided in Fig. 5d. As seen in Fig. 5d, the experimental anti-reflection performance of the V-groove textured sample is in ac-cordance with the simulation results declared in the beginning of thepaper, showing an extraordinary broadband omnidirectionality in thecase when light is incident along the extending direction of the V-grooves. Another two light incident cases, in which the deflection anglebetween the component of incident light on xoy plane and x-axis wasset to 30° and 60° respectively, have been studied and the results are

provided in Fig S2.However, quantitative analyses of the omnidirectionality should be

introduced for precise evaluation. The impact of the AOI has not beentaken into consideration in the calculation of R̄ using expression (2), soit is necessary to use a new parameter to fully describe the reflectionperformance at various AOIs. Therefore, we put forward Rave which isdefined as

∫

∫= °

°

°

°R

R θ θdθ

θdθ

¯ ( ) cos

cosave

080

080

(3)

where θ and R θ¯ ( ) stand for the AOI and the corresponding averagereflectance respectively. The introduction of cosθ is due to the photonflux being reduced to cosθ times compared to the normal incidence. Asa result, the omnidirectionality of the samples can be adequatelycharacterized through Rave. The calculated value of Rave for the V-grooves sample is 9.25% when light is incident along the parallel di-rection, and Rave is 13.86% when light is incident perpendicular to thegrooves. As a reference, the pyramid sample shows a Rave value of11.04% in the experiments. In summary, the V-groove textured struc-ture can significantly suppress light reflection over a wide range of AOIand wavelength, with the exhibited omnidirectional anti-reflectionability much superior to that observed for the random pyramid struc-ture. Hence, the broadband omnidirectional V-groove textured struc-ture has more advantages over the conventional random pyramidstructure since higher conversion efficiencies and power output can beachieved for V-groove textured c-Si solar cells when light is obliquelyincident.

4. Conclusions

Ray tracing simulation was utilized in this study to evaluate theomnidirectional anti-reflection ability of the V-groove textured Si. Thesimulation results indicate that the Si V-groove structure has an om-nidirectionality over a wide AOI range when the direction of incidentlight changes in the plane parallel to the V-grooves and perpendicularto the surface of the Si wafer at the same time. To prove the simulationresults in experiments, we have fabricated a V-groove textured struc-ture on c-Si wafers by employing the Cu-assisted chemical etchingmethod. Tightly arranged parallel V-shaped grooves with wide open-ings were obtained after texturization. In addition, the formation me-chanism for the V-groove structure and the strong correlation betweenthe V-groove structural characteristics and slicing parameters of theraw DWS c-Si wafers were studied.

The omnidirectional anti-reflectance performance of the V-groovetextured Si was studied. Experimental results show that the averagereflectance of the V-groove textured samples is 12.33% in the wave-length range of 300–1000 nm, which is comparable to the reflectance ofconventional random pyramidal c-Si wafers. Moreover, the experi-mental results show a similar changing tendency with the simulationresults as the V-groove textured structure maintains high light absorp-tion over a wide AOI range of 0–80° when the incident light changes inthe parallel direction. The Rave, which was introduced to describe thereflection performance at various AOIs, is 9.25% for the V-groovessample when light is incident along the parallel direction, showingmore advantages on omnidirectionality when compared with therandom pyramid c-Si wafers. As a consequence, the V-groove texturedstructure fabricated through the Cu-assisted chemical etching exhibitsan excellent broadband omnidirectional anti-reflection property, whichindicates great potentials for the fabrication of omnidirectional solarcells.

Acknowledgements

This work was supported by the National Natural ScienceFoundation of China (grant nos. 11675280, 61874139, and 11674405)

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

136

and Jiangsu Science and Technology Department (TechnologicalAchievements Transformation Project, grant nos. BA2017137).

Appendix A. Supplementary material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.solener.2019.09.048.

References

Backlund, Y., Rosengren, L., 1992. New shapes in (100) Si using KOH and EDP etches. J.Micromech. Microeng. 2, 75–79. https://doi.org/10.1088/0960-1317/2/2/002.

Borojevic, N., Lennon, A., Wenham, S., 2014. Light trapping structures for silicon solarcells via inkjet printing: light trapping structures for silicon solar cells. Phys. Stat.Solidi A 211, 1617–1622. https://doi.org/10.1002/pssa.201330654.

Campbell, P., Green, M.A., 1987. Light trapping properties of pyramidally textured sur-faces. J. Appl. Phys. 62, 243–249. https://doi.org/10.1063/1.339189.

Chen, K., Liu, Y., Wang, X., Zhang, L., Su, X., 2015. Novel texturing process for diamond-wire-sawn single-crystalline silicon solar cell. Sol. Energy Mater. Sol. Cells 133,148–155. https://doi.org/10.1016/j.solmat.2014.11.016.

Chen, W., Liu, X., Li, M., Yin, C., Zhou, L., 2014. On the nature and removal of saw markson diamond wire sawn multicrystalline silicon wafers. Mater. Sci. Semicond. Process.27, 220–227. https://doi.org/10.1016/j.mssp.2014.06.049.

Chhajed, S., Schubert, M.F., Kim, J.K., Schubert, E.F., 2008. Nanostructured multilayergraded-index antireflection coating for Si solar cells with broadband and omnidir-ectional characteristics. Appl. Phys. Lett. 93, 251108. https://doi.org/10.1063/1.3050463.

Dobrzański, L.A., Drygała, A., 2008. Surface texturing of multicrystalline silicon solarcells. J. Achiev. Mater. Manuf. Eng. 31, 6.

Haiyuan, X., Sihua, Z., Yufeng, Z., Wenzhong, S., 2018. Controllable nanoscale invertedpyramids for highly efficient quasi-omnidirectional crystalline silicon solar cells.Nanotechnology 29, 015403. https://doi.org/10.1088/1361-6528/aa9a96.

Huang, Y.-F., Chattopadhyay, S., Jen, Y.-J., Peng, C.-Y., Liu, T.-A., Hsu, Y.-K., Pan, C.-L.,Lo, H.-C., Hsu, C.-H., Chang, Y.-H., Lee, C.-S., Chen, K.-H., Chen, L.-C., 2007.Improved broadband and quasi-omnidirectional anti-reflection properties with bio-mimetic silicon nanostructures. Nat. Nanotechnol. 2, 770–774. https://doi.org/10.1038/nnano.2007.389.

Kawasegi, N., Morita, N., Yamada, S., Takano, N., Oyama, T., Ashida, K., 2005. Etch stopof silicon surface induced by tribo-nanolithography. Nanotechnology 16, 1411–1414.https://doi.org/10.1088/0957-4484/16/8/073.

Kuang, P., Eyderman, S., Hsieh, M.-L., Post, A., John, S., Lin, S.-Y., 2016. Achieving anaccurate surface profile of a photonic crystal for near-unity solar absorption in asuper thin-film architecture. ACS Nano 10, 6116–6124. https://doi.org/10.1021/acsnano.6b01875.

Lin, H., Xiu, F., Fang, M., Yip, S., Cheung, H.-Y., Wang, F., Han, N., Chan, K.S., Wong, C.-Y., Ho, J.C., 2014a. Rational design of inverted nanopencil arrays for cost-effective,broadband, and omnidirectional light harvesting. ACS Nano 8, 3752–3760. https://doi.org/10.1021/nn500418x.

Lin, Q., Leung, S.-F., Lu, L., Chen, X., Chen, Z., Tang, H., Su, W., Li, D., Fan, Z., 2014b.Inverted nanocone-based thin film photovoltaics with omnidirectionally enhancedperformance. ACS Nano 8, 6484–6490. https://doi.org/10.1021/nn5023878.

Meinel, B., Koschwitz, T., Acker, J., 2012. Textural development of SiC and diamond wiresawed sc-silicon wafer. Energy Proc. 27, 330–336. https://doi.org/10.1016/j.egypro.2012.07.072.

Nussbaumer, H., Willeke, G., Bucher, E., 1994. Optical behavior of textured silicon. J.Appl. Phys. 75, 2202–2209. https://doi.org/10.1063/1.356282.

Savin, H., Repo, P., von Gastrow, G., Ortega, P., Calle, E., Garín, M., Alcubilla, R., 2015.Black silicon solar cells with interdigitated back-contacts achieve 22.1% efficiency.Nat. Nanotechnol. 10, 624–628. https://doi.org/10.1038/nnano.2015.89.

Scheibe, C., Obermeier, E., 1995. Compensating corner undercutting in anisotropicetching of (100) silicon for chip separation. J. Micromech. Microeng. 5, 109–111.

Fig. 5. Reflectance spectra as a function of AOIs and wavelength. (a) Reflectance spectra as a function of AOIs and wavelength of V-groove textured c-Si surface whenlight is incident in direction parallel to the V-grooves. (b) Reflectance spectra as a function of AOIs and wavelength of V-groove textured c-Si surface when light isincident in direction perpendicular to the V-grooves. (c) Reflectance spectra as a function of AOIs and wavelength on random pyramid c-Si surface. AOI is in the rangeof 0–80° while the wavelength is ranging from 300 nm to 1000 nm. (d) The average reflectance as a function of AOI on the V-grooves sample and random pyramidsample.

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

137

https://doi.org/10.1088/0960-1317/5/2/013.Smith, A.W., Rohatgi, A., 1993. Ray tracing analysis of the inverted pyramid texturing

geometry for high efficiency silicon solar cells. Sol. Energy Mater. Sol. Cells 29,37–49. https://doi.org/10.1016/0927-0248(93)90090-P.

Spinelli, P., Verschuuren, M.A., Polman, A., 2012. Broadband omnidirectional antire-flection coating based on subwavelength surface Mie resonators. Nat. Commun. 3.https://doi.org/10.1038/ncomms1691.

Untila, G., Palov, A., Kost, T., Chebotareva, A., Stepanov, A., Zaks, M., Sitnikov, A.,Saprykin, D., Grishaev, A., 2013. Crystalline silicon solar cells with laser ablatedpenetrating V-grooves: modeling and experiment. Phys. Status Solidi A 210,760–766. https://doi.org/10.1002/pssa.201200707.

van Dam, D., van Hoof, N.J.J., Cui, Y., van Veldhoven, P.J., Bakkers, E.P.A.M., GómezRivas, J., Haverkort, J.E.M., 2016. High-Efficiency nanowire solar cells with omni-directionally enhanced absorption due to self-aligned indium–tin–oxide mie scat-terers. ACS Nano 10, 11414–11419. https://doi.org/10.1021/acsnano.6b06874.

Vangbo, M., Bäcklund, Y., 1996. Terracing of (100) Si with one mask and one etching stepusing misaligned V-grooves. J. Micromech. Microeng. 6, 39–41. https://doi.org/10.1088/0960-1317/6/1/006.

Wang, Y., Liu, Y., Yang, L., Chen, W., Du, X., Kuznetsov, A., 2017. Micro-structured in-verted pyramid texturization of Si inspired by self-assembled Cu nanoparticles.Nanoscale 9, 907–914. https://doi.org/10.1039/C6NR08126F.

Wang, Y., Yang, L., Liu, Y., Mei, Z., Chen, W., Li, J., Liang, H., Kuznetsov, A., Xiaolong, D.,2015. Maskless inverted pyramid texturization of silicon. Sci. Rep. 5, 10843. https://doi.org/10.1038/srep10843.

Wei, W.-R., Tsai, M.-L., Ho, S.-T., Tai, S.-H., Ho, C.-R., Tsai, S.-H., Liu, C.-W., Chung, R.-J.,He, J.-H., 2013. Above-11%-efficiency organic-inorganic hybrid solar cells withomnidirectional harvesting characteristics by employing hierarchical photon-trap-ping structures. Nano Lett. 13, 3658–3663. https://doi.org/10.1021/nl401540h.

Wilbers, J.G.E., Berenschot, J.W., Tiggelaar, R.M., Dogan, T., Sugimura, K., van der Wiel,W.G., Gardeniers, J.G.E., Tas, N.R., 2018. 3D-fabrication of tunable and high-densityarrays of crystalline silicon nanostructures. J. Micromech. Microeng. 28, 044003.https://doi.org/10.1088/1361-6439/aaab2d.

Willeke, G., Nussbaumer, H., Bender, H., Bucher, E., 1992. A simple and effective lighttrapping technique for polycrystalline silicon solar cells. Sol. Energy Mater. Sol. Cells

26, 345–356. https://doi.org/10.1016/0927-0248(92)90054-S.Winderbaum, S., Reinhold, O., Yun, F., 1997. Reactive ion etching (RIE) as a method for

texturing polycrystalline silicon solar cells. Sol. Energy Mater. Sol. Cells 46, 239–248.https://doi.org/10.1016/S0927-0248(97)00011-1.

Wu, J., Liu, Y., Chen, Q., Chen, W., Yang, L., Wang, Y., He, M., Du, X., 2018. The or-ientation and optical properties of inverted-pyramid-like structures on multi-crys-talline silicon textured by Cu-assisted chemical etching. Sol. Energy 171, 675–680.https://doi.org/10.1016/j.solener.2018.07.011.

Yagi, T., Uraoka, Y., Fuyuki, T., 2006. Ray-trace simulation of light trapping in siliconsolar cell with texture structures. Sol. Energy Mater. Sol. Cells 90, 2647–2656.https://doi.org/10.1016/j.solmat.2006.02.031.

Yang, L., Liu, Y., Wang, Y., Chen, W., Chen, Q., Wu, J., Kuznetsov, A., Du, X., 2017a.18.87%-efficient inverted pyramid structured silicon solar cell by one-step Cu-as-sisted texturization technique. Sol. Energy Mater. Sol. Cells 166, 121–126. https://doi.org/10.1016/j.solmat.2017.03.017.

Yang, S., Zhang, L., Ge, P., 2017b. Experimental study of the pyramidal texturization onthe surface of photovoltaic silicon with cemented carbide micro-milling tool. Int. J.Adv. Manuf. Technol. 92, 3187–3196. https://doi.org/10.1007/s00170-017-0388-9.

Yu, X., Wang, P., Li, X., Yang, D., 2012. Thin Czochralski silicon solar cells based ondiamond wire sawing technology. Sol. Energy Mater. Sol. Cells 98, 337–342. https://doi.org/10.1016/j.solmat.2011.11.028.

Zhang, J., Zhao, L., Rosenkranz, A., Song, C., Yan, Y., Sun, T., 2019. Nanosecond pulsedlaser ablation of silicon—finite element simulation and experimental validation. J.Micromech. Microeng. 29, 075009. https://doi.org/10.1088/1361-6439/ab208b.

Zhao, J., Wang, A., Green, M.A., 1999. 24·5% Efficiency silicon PERT cells on MCZsubstrates and 24·7% efficiency PERL cells on FZ substrates. Prog. Photovolt. Res.Appl. 7, 471–474. https://doi.org/10.1002/(SICI)1099-159X(199911/12)7:6<471::AID-PIP298>3.0.CO;2-7.

Zhong, S., Wang, W., Tan, M., Zhuang, Y., Shen, W., 2017. Realization of quasi-omni-directional solar cells with superior electrical performance by all-solution-processedSi nanopyramids. Adv. Sci. 4, 1700200. https://doi.org/10.1002/advs.201700200.

Zhong, X., Qu, Y., Lin, Y.-C., Liao, L., Duan, X., 2011. Unveiling the formation pathway ofsingle crystalline porous silicon nanowires. ACS Appl. Mater. Interfaces 3, 261–270.https://doi.org/10.1021/am1009056.

Y. Zhao, et al. Solar Energy 193 (2019) 132–138

138