Cu(In,Ga)(S,Se) 2

Subscriber access provided by - Access paid by the | UCSB Libraries

Surfaces, Interfaces, and Applications 0.9

0.1

0.8

0.2

2

20%-Efficient Zn Mg O:Al/Zn Mg O/Cu(In,Ga)(S,Se) Solar Cell Prepared by All Dry Process through Combination of Heat-Light Soaking and Light Soaking Process Jakapan Chantana, Takuya Kato, Hiroki Sugimoto, and Takashi Minemoto ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b01247 • Publication Date (Web): 13 Mar 2018 Downloaded from http://pubs.acs.org on March 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

20%-Efficient Zn0.9Mg0.1O:Al/Zn0.8Mg0.2O/Cu(In,Ga)(S,Se)2 Solar Cell Prepared by All Dry Process through Combination of Heat-Light Soaking and Light Soaking Process

Jakapan Chantana*1, Takuya Kato2, Hiroki Sugimoto2, Takashi Minemoto*1 1

Department of Electrical and Electronic Engineering, Ritsumeikan University, 1-1-1

Nojihigashi, Kusatsu, Shiga 525-8577, Japan

2

Atsugi Research Center, Solar Frontier K. K., Atsugi, Kanagawa 243-0206, Japan

*Corresponding authors:

Jakapan Chantana, E-mail: [email protected]

Takashi Minemoto, E-mail: [email protected]

Co-authors:

Takuya Kato, E-mail: [email protected]

Hiroki Sugimoto, E-mail: [email protected]

* To whom all correspondence should be addressed. Tel/Fax: +81-77-561-4836;

E-mails: [email protected] and [email protected]

1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 45

ABSTRACT Development of Cd-free Cu(In,Ga)(S,Se)2 (CIGSSe)-based thin-film solar cells fabricated by all dry process is intriguing to minimize optical loss at a wavelength shorter than 520 nm owing to absorption of CdS buffer layer and to be easily integrated into an in-line process for cost reduction. Cd-free CIGSSe solar cells are therefore prepared

by

all

dry

process

with

a

structure

of

Zn0.9Mg0.1O:Al/Zn0.8Mg0.2O/CIGSSe/Mo/Glass. It is demonstrated that Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al are appropriate as buffer and transparent conductive oxide layers with large optical band-gap energy values of 3.75 and 3.80 eV, respectively. The conversion efficiency (η) of the Cd-free CIGSSe solar cell without K-treatment is consequently increased to 18.1%. To further increase the η, Cd-free CIGSSe solar cell with K-treatment is next fabricated and followed by post treatment called HLS+LS process, including heat-light soaking (HLS) at 110 oC followed by light soaking (LS) under AM 1.5G illumination. It is disclosed that the HLS+LS process gives rise to not only the enhancement of carrier density but also the decrease in carrier recombination rate at the buffer/absorber interface. Ultimately, the η of the Cd-free CIGSSe solar cell with K-treatment prepared by all dry process is enhanced to the level of 20.0%.

2


Page 3 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


KEYWORLD: Cd-free thin-film solar cell, Cu(In,Ga)(S,Se)2, dry process, heat-light soaking, light soaking, carrier recombination rate

3



1. INTRODUCTION Polycrystalline chalcopyrite compound Cu(In,Ga)Se2 (CIGSe) and Cu(In,Ga)(S,Se)2 (CIGSSe) are interesting semiconductor materials as the absorber of thin-film photovoltaic devices with high conversion efficiencies (η) above 20% with small area (0.5 cm2).1-3 The conventional structure of CIGSe and CIGSSe-based thin-film solar cells with high η (> 20%) is ZnO:Al/ZnO/CdS/(CIGSe or CIGSSe)/Mo/Glass.1,2 Nevertheless, the CdS buffer layer, which is generally prepared by chemical bath deposition (CBD), absorbers light with a wavelength shorter than 520 nm owing to its band-gap energy (Eg) of about 2.40 eV, thereby limiting short-circuit current density (JSC).4 In addition, liquid waste from the CBD process has a detrimental effect on the environment.5 The research on Cd-free CIGSe and CIGSSe-based thin-film solar cells has been intensively conducted to minimize the optical loss at a wavelength shorter than 520 nm and reduce the toxicity of the materials involved.6-11 Moreover, the CBD, which is wet process, for the CdS buffer layer is not readily integrated into an in-line process. Consequently, the alternative deposition methods by dry process have to be developed. Among Cd-free CIGSe solar cell, the 18.1%-efficient Cd-free CIGSe solar cell with atomic-layer-deposited Zn1-xMgxO buffer layer (dry process) was achieved, presently the highest value for this material combination.12,13 Moreover, with In2S3 buffer

4


Page 4 of 45

Page 5 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


prepared by the thermal evaporation, the highest η for the Cd-free CIGS solar cell fabricated by all dry process was further enhanced to 18.2%.14,15 It is known that the sputtering method can be a part of in-line industrial production process.11 The Zn1-xMgxO as buffer (or window) layer and Zn1-xMgxO:Al as transparent conductive oxide (TCO) layer have been already applied into CIGSSe solar cell with η above 20%,4,16-18 where the Zn1-xMgxO and Zn1-xMgxO:Al layers with controllable Mg content can be easily deposited by radio frequency (RF) magnetron co-sputtering methods.4,16,19 It has been additionally reported that post-treatment processes after the fabrication of the solar cell such as light soaking (LS) and heat-light soaking (HLS) can lead to the enhancement of the η since the carrier concentration is increased and the carrier recombination rates in the device are decreased.20-24 To develop the solar cell, the understanding of carrier recombination in the solar cell subjected to LS and/or HLS process

is

therefore

important.

It

has

been

recently

reported

that

temperature-illumination-dependent open-circuit voltage (VOC) method can separately quantify the recombination rates at the buffer/absorber interface, in space-charge region (SCR) and in quasi-neutral region (QNR).25,26 In this work, Zn1-xMgxO:Al/Zn1-xMgxO/CIGSSe solar cells, which are the Cd-free solar cells, were fabricated by all dry process. The Zn1-xMgxO:Al and Zn1-xMgxO layers

5



Page 6 of 45

were deposited by the RF magnetron co-sputtering. The post-treatment process named HLS+LS process composed of HLS followed by LS under AM1.5 G illumination was then performed on the as-fabricated CIGSSe solar cell with K-treatment. The effect of the HLS+LS process on recombination rates at the buffer/absorber interface, in SCR, and in QNR of the CIGSSe solar cell was examined, thereby providing a meaningful interpretation to understand the state-of-the-art and future improvement of the photovoltaic performances.

2. EXPERIMENTAL SECTION According to our previous work, Zn0.9Mg0.1O:Al/Zn0.8Mg0.2O/Cd0.75Zn0.25S/CIGSSe solar

cell

with

η

above

20%

was

obtained.16

In

this

work,

Zn0.9Mg0.1O:Al/Zn0.8Mg0.2O/CIGSSe solar cells with a designated area (da) of approximately 0.5 cm2 were therefore fabricated by all dry process as shown in Figure 1. In Figure 1, CIGSSe layers with and without K-treatment on Mo-coated glasses were fabricated using the two-step process at Solar Frontier K. K., where the sputter-deposited Cu-In-Ga precursors were first prepared and followed by selenizanition and sulfurization processes.18,27 For the CIGSSe sample with K-treatment, a thin (approximately 3-15 nm) KF was deposited on the CIGSSe surface at room

6


Page 7 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


temperature. 18 The sample was then followed by annealing at around 250-350 oC in S-containing ambient.18 Zn0.8Mg0.2O buffer layers with varying the thickness from 0 (without (w/o) the buffer layer) to 155 nm were then deposited on CIGSSe absorbers by RF magnetron co-sputtering from ZnO (99.99%) and MgO (99.99%) targets. The samples during the co-sputtering for Zn0.8Mg0.2O films were water-cooled to maintain a substrate temperature at room temperature. The target diameter, working pressure, and Ar flow rate were 7.62 cm, 0.1 Pa, and 3.9 sccm, respectively. To obtain Mg content of 0.2 in the Zn0.8Mg0.2O buffers, the powers applied to ZnO and MgO targets were 82.6 and 85 W, respectively.4 After preparing the Zn0.8Mg0.2O buffer layers, Zn0.9Mg0.1O:Al as TCO layer with a thickness of about 700 nm was next deposited by the RF magnetron co-sputtering from ZnO:Al (Al2O3: 2 wt%-doped) (99.99%) and MgO (99.99%) targets at room temperature. The target diameter, working pressure, and Ar flow rate were 7.62 cm, 0.25 Pa, and 10 sccm, respectively. The powers applied to ZnO:Al and MgO targets for the Zn0.9Mg0.1O:Al layers were 110 and 83 W, respectively.16,19 In this work, it is noted that the aim of using lower power applied to ZnO (82.6 W) target for the Zn0.8Mg0.2O buffer layer than that applied to ZnO:Al target (110 W) for the Zn0.9Mg0.1O:Al layer (TCO) is to minimize the sputtering damages on CIGSSe surface.

7



The samples during Zn0.8Mg0.2O buffer/ Zn0.9Mg0.1O:Al (TCO) sputtering were placed at approximately 4 cm from the centers of the targets (target diameter of 7.62 cm) to further decrease the sputtering damages caused by severely exposing the Ar plasma. In addition, the Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al films with a thickness of about 1 µm were prepared on soda-lime glass (SLG) substrates to examine their Mg contents, which were observed by energy dispersive spectroscopy operated at 4 kV. The Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al films on SLG substrates were additionally investigated by ultraviolet/visible near-infrared spectrophotometer (UV-3600, Shimadzu) for their transmittance and reflectance spectra. The optical Eg values were therefore derived from plot of (αhν)2 as a function of photon energy (hν),19,28 where α is the absorption coefficient of their films. Moreover, the sheet resistance of the films on SLG substrates were investigated by four point probe method (NPS resistivity processor Model Σ-5+). After the Zn0.9Mg0.1O:Al (TCO) deposition, 100-nm-thick Ni and 4 µm-thick-Al grids of the solar cells were formed by electron-beam evaporation. The In contacts on Mo layers were finally prepared by soldering method in Figure 1. The cell performance parameters, which are JSC, VOC, fill factor (FF) and η, in the CIGSSe solar cells without anti-reflective layer were evaluated from photo current density-voltage (J-V) characteristics under AM 1.5G illumination (100 mW/cm2). External quantum

8


Page 8 of 45

Page 9 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


efficiency (EQE) was measured using the measurement system (CEP-25RR, Bunkoukeiki).

The

CIGSSe

solar

cell

with

a

conventional

structure

of

ZnO:Al/ZnO/CdS/CIGSSe/Mo/Glass was also fabricated as a reference (Ref.) (see Figure S1, Supporting Information), where the detailed fabrication process was discussed in Ref. [29]. To further enhance photovoltaic performances, the HLS+LS process (post-treatment) was proposed and conducted on the Cd-free CIGSSe solar cell with K-treatment prepared by all dry process with a structure shown in Figure 1. For the HLS+LS process (see Figure S2, Supporting Information), after J-V measurement of the as-fabricated Cd-free CIGSSe solar cell, the HLS at 110 oC under AM 1.5G illumination was first performed for 4 hours (h), and the solar cell was cooled down to 25 oC for 5 min before J-V measurement ((i) HLS at 110 oC for 4 h). The LS under AM 1.5G illumination was then conducted for 2 h and J-V measurement ((ii) LS for 2 h) was performed. Temperature dependent J-V characteristic of the CIGSSe solar cell before and after the HLS+LS process was additionally conducted under AM 1.5G illumination using a cryostat cooled with liquid-N2 and heated by temperature controller (Model 9700, Scientific Instruments). It was reported that the graph of VOC as a function of temperature (T) presents a straight line and the extrapolation of this line to T at 0 K

9



leads to qVOC at T = 0 K, which is defined as activation energy of recombination (EA).30 Furthermore, the illumination dependence of VOC of the solar cells (Suns-VOC) was characterized by WCT−120 (Sinton Instruments) before and after the HLS+LS process. The sheet resistance of Zn0.9Mg0.1O:Al (TCO) in the structure of the solar cell in Figure 1 was investigated by four point probe method (NPS resistivity processor Model Σ-5+). To estimate carrier concentration (NA) in the CIGSSe absorber and built-in potential (φbi) of the solar cell, capacitance−voltage (C−V) measurement was performed using LCR meter (Hewlett Packard 4284A) with a frequency and amplitude of 10 kHz and 10 mV, respectively.

3. RESULTS AND DISCUSSION 3.1. Optical properties and impacts of Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al on photovoltaic performances. Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al films on SLG substrates were first investigated. It is noted that ZnO is normally used as the buffer (window) layer of the CIGSe solar cell with high η.1,2 Figure 2 therefore shows (αhν)2 plots of ZnO (as reference), Zn0.8Mg0.2O, and Zn0.9Mg0.1O:Al films on SLG substrates as a function of hν to estimate their optical Eg values, which were determined by linearly extrapolating the (αhν)2 to hν (x-axis), where the hν-intercept gives the optical Eg. It is

10


Page 10 of 45

Page 11 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


demonstrated that the derived optical Eg values are 3.30 eV for pure ZnO, and 3.75 eV for Zn0.8Mg0.2O, which well agree with the literature,28 where the Mg content was investigated by induced coupled plasma photoluminescence spectroscopy. The larger optical Eg of Zn0.8Mg0.2O as the buffer layer could lead to the minimization of the optical loss at a short wavelength caused by conventional ZnO buffer (window) layer in the solar cell. In addition, the sheet resistance is significantly increased from 650 Ω/sq for ZnO to 3.7x106 Ω/sq for Zn0.8Mg0.2O. The higher sheet resistance of the Zn0.8Mg0.2O as the buffer layer could effectively suppress the direct contact between TCO and absorber (shunt path) as reported by Ishizuka et al..31 The results have shown that the Zn0.8Mg0.2O is suitable as the buffer layer of the solar cell. In addition, the large optical Eg of 3.80 eV for Zn0.9Mg0.1O:Al on SLG substrate is shown in Figure 2, which is well corresponding to the literature,16 and its low sheet resistance of 18.7 Ω/sq is observed. The Zn0.9Mg0.1O:Al layer is consequently appropriate for TCO layer of the solar cell. The Cd-free CIGSSe solar cells without (w/o) K-treatment were therefore fabricated by all dry process with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (0-155 nm)/CIGSSe/Mo/Glass in Figure 1, where the thickness of Zn0.8Mg0.2O buffer layer was varied. Figure 3 depicts photovoltaic performance parameters as a function of thickness of Zn0.8Mg0.2O buffer layer of the Cd-free CIGSSe solar cells with a structure

11



in Figure 1. The dashed lines denote the cell parameters of the conventional CIGSSe solar cell with a structure of ZnO:Al/ZnO/CdS/CIGSSe/Mo/Glass for reference (Ref.) (see Figures S1 and S3 in Supporting Information). Table 1 shows the corresponding maximum cell parameters and diode parameters in Figure 3. According to Figure 3 and Table 1, all photovoltaic parameters are low with the η of 5.5% when the thickness of Zn0.8Mg0.2O buffer layer is 0 nm (w/o buffer layer). On the other hand, when the thickness of Zn0.8Mg0.2O buffer layer of the solar cell is increased to 124 nm, JSC, VOC, and FF are enhanced to 38.2 mA/cm2, 0.650 V, and 73.2%, thus increasing the η to 18.1%. Figure 4 illustrates corresponding EQE spectra of the Cd-free CIGSSe solar cells with Zn0.8Mg0.2O buffer thicknesses of 0 (w/o buffer layer) and 124 nm. It is shown that the EQE of the solar cell with Zn0.8Mg0.2O buffer thickness of 124 nm at short wavelength is limited at about 330 nm (or 3.75 eV), well corresponding to the optical Eg of 3.75 eV for Zn0.8Mg0.2O buffer layer, thus enhancing JSC up to 38.2 mA/cm2. This JSC of 38.2 mA/cm2 is higher than the JSC of 36.1 mA/cm2 in the reference solar cell (ZnO:Al/ZnO/CdS/CIGSSe) owing to no optical loss by CdS layer and larger optical Eg of the Zn0.8Mg0.2O buffer layer. The optical loss caused by CdS layer in the reference solar cell is seen in Figure S4a in supporting information, where the first derivative of EQE for the reference solar cell is depicted in Figure S4b to

12


Page 12 of 45

Page 13 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


demonstrate the Eg values of ZnO (3.30 eV), CdS (2.40 eV) and CIGSSe (1.07 eV) in the structure of the reference solar cell, respectively. Moreover, the EQE with the Zn0.8Mg0.2O buffer thickness of 124 nm in Figure 4, especially at a wavelength range of approximately 400-800 nm, is much higher than that with the Zn0.8Mg0.2O buffer thickness of 0 nm (w/o buffer layer). It is consequently implied in Figure 4 that the sputtering damages near CIGSSe surface are severely formed with the Zn0.8Mg0.2O buffer thickness of 0 nm (w/o buffer layer) and the damages are reduced with increasing the Zn0.8Mg0.2O buffer thickness to 124 nm. It is noted that the lower powers during the Zn0.8Mg0.2O sputtering, where the powers applied to ZnO and MgO targets were 82.6 and 85 W, are intentionally applied as compared with those during the Zn0.9Mg0.1O:Al sputtering to use the Zn0.8Mg0.2O to partly reduce sputtering damage near CIGSSe surface. In Table 1, saturation current density (J0), and ideality factor (n) are additionally reduced and shunt resistance (RSH) is increased with enhancing the Zn0.8Mg0.2O buffer thickness to 124 nm. The results suggest that the carrier recombination is decreased with the Zn0.8Mg0.2O buffer thickness up to 124 nm, thereby increasing VOC and FF up to 0.650 V and 73.2%. This is because the 124-nm-thick Zn0.8Mg0.2O buffer can partly reduce sputtering damages near CIGSSe surface and/or completely cover the CIGSSe surface, thus giving rise to the suppression of the direct

13



contact between TCO and CIGSSe absorber (shunt path). The highest η for 18.1% in the Cd-free solar cell (w/o K-treatment) with a structure in Figure 1 by all dry process is consequently achieved with the optimized Zn0.8Mg0.2O buffer thickness of 124 nm, which is as high as η of 18.2% in the reference solar cell. On the other hand, all cell parameters are decreased with too thick high-resistive Zn0.8Mg0.2O buffer layer above 155 nm in Table 1.

3.2. Zn0.9Mg0.1O:Al/Zn0.8Mg0.2O/Cu(In,Ga)(S,Se)2 solar cell with K-treatment and analysis of its carrier recombination rates. It has been reported that CIGSSe with K-treatment gives rise to the increase in the η of the CIGSSe solar cell.18 According to section 3.1, the thickness of Zn0.8Mg0.2O buffer layer is optimized at 124 nm. Therefore, the CIGSSe solar cell with K-treatment is fabricated by all dry process with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass in Figure 1. Since Zn0.8Mg0.2O buffer has the large optical Eg of 3.75 eV almost the same as that of Zn0.9Mg0.1O:Al (TCO) layer, there is no optical loss at short wavelength (approximately 520 nm) with its thickness of 124 nm. As a result, the η of the as-fabricated Cd-free CIGSSe solar cell (with K-treatment) is enhanced to 18.7% seen in Table 2. It was reported that post-treatment processes such as LS and HLS gives rise to the increase in

14


Page 14 of 45

Page 15 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the η of the as-fabricated solar cell.20-24 The LS process under AM 1.5G illumination at 25 oC for 2 h was therefore conducted on the as-fabricated Cd-free CIGSSe solar cell (with K-treatment). However, it is found that its η is slightly increased from 18.7 to approximately 19.0%, mainly attributed to the small decrease in series resistance (RS), thereby slightly enhancing FF. Alternatively, the HLS process under AM 1.5G illumination at 110 oC for 4 h, named (i) HLS for 4 h (see Figure S2, Supporting Information), was thus performed on the as-fabricated Cd-free CIGSSe solar cell (with K-treatment). It is disclosed in Figure 5a and Table 2 that (i) HLS for 4 h leads to the increase in VOC from 0.680 to 0.693 V but the severe decrease in FF from 70.2 to 67.0%, thus slightly decreasing η from 18.7 to 18.3%. The enhancement of VOC after (i) HLS for 4 h is feasibly caused by the increase in NA32 and/or the decrease in carrier recombination, implied by the reductions of J0 and n as well as the increase in RSH in Table 2. In addition, (i) HLS for 4 h results in the J-V roll-over (distortion of J-V curve) in Figure 5a and the increase in sheet resistance of Zn0.9Mg0.1O:Al from 65 to 76 Ω/sq in the solar cell structure, thus decreasing FF to 67.0%. The J-V roll-over in the CIGSSe solar cell was clearly studied in Ref. [33], where the CIGSSe absorber was deposited by the same process as the CIGSSe absorber in this work. The study has disclosed that the CIGSSe absorber with higher S content

15



near Mo layer gives rise to the lower valence band minimum of CIGSSe near Mo layer.33 Therefore, valence band offset (Schottky barrier or back diode) at CIGSSe/Mo back junction becomes larger, thus forming the J-V roll-over. 33 The J-V roll-over with constant S content moreover becomes severe when the parallel resistance of the back diode (Rpbc) is increased, which is considered as the deterioration of CIGSSe/Mo back junction.33 Consequently, the J-V roll-over after (i) HLS for 4 h in Figure 5a is attributed to the increase in Rpbc, thereby deteriorating CIGSSe/Mo back junction. It has been reported that the resistivity of ZnMgO and J-V roll-over can be reduced after LS.33,34 Therefore, after (i) HLS for 4 h, (ii) LS under AM 1.5G illumination at 25 o

C for 2 h, called (ii) LS for 2 h, was successively conduced, where the process,

including (i) HLS for 4 h followed by (ii) LS for 2 h (see Figure S2, Supporting Information), is named HLS+LS process in this work. It is revealed that the sheet resistance of Zn0.9Mg0.1O:Al is decreased from 76 to 62 Ω/sq in the solar cell structure after (ii) LS for 2 h.34 Moreover, the J-V roll-over after (ii) LS for 2 h in Figure 5a is successfully disappeared because of the decrease in Rpbc, thereby recovering CIGSSe/Mo back junction.33 The FF is subsequently recovered and further increased to 73.1% after (ii) LS for 2 h as depicted in Figure 5a and Table 2, thus increasing the η up to 20.0%. According to the result, the HLS+LS process, including (i) HLS for 4 h

16


Page 16 of 45

Page 17 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


followed by (ii) LS for 2 h, is an effective way to increase photovoltaic performances of the Cd-free CIGSSe solar cell (K-treatment) with a structure in Figure 1. Figure 6 depicts (a) Mott-Schottky plot (C−2 vs. Bias (V)) and (b) VOC as a function of T (VOC-T) for the Cd-free CIGSSe solar cell (K-treatment) with a structure in Figure 1 before (as-fabricated) and after the HLS+LS process. Near-surface Eg of the CIGSSe in this work is constant at approximately 1.26 eV. In Figure 6a, NA is increased from 1.54x1017 to 2.98x1017 cm-3 after subjected to the HLS+LS process. The amount of K used in our K-treatment (around 3-15 nm in KF thickness) is smaller than that utilized for KF-post deposition treatment (normally 15-35 nm).2,18 Therefore, it is considered that the increase in the NA after the HLS+LS process to 2.98x1017 cm-3 is feasibly attributable to divacancy complex (VSe-VCu) in CIGSSe absorber converting from a shallow donor to a shallow acceptor (the metastable hole) by electron capture (photon absorption).35,36 Therefore, the EQE after the HLS+LS process in a range from about 900 to 1100 nm in Figure 5b is slightly decreased. The corresponding first derivative of EQE in Figure 5b is additionally shown in Figure S5 in Supporting Information to observe the Eg values of Zn0.8Mg0.2O buffer (3.75 eV) and CIGSSe absorber (1.07 eV) in the structure of Cd-free CIGSSe solar cell with K treatment. It is seen in Figure 5b that EQE at short wavelength (about 330 nm) and long wavelength (approximately 1158

17



nm) is limited by the Eg values of Zn0.8Mg0.2O (3.75 eV) and CIGSSe (1.07 eV), respectively, thereby resulting in the increase in JSC up to 39 mA/cm2 in Table 2. Moreover, the increase in the NA from 1.54x1017 to 2.98x1017 cm-3 after the HLS+LS process in Figure 6a should result in the difference in VOC of ∆VOC=8.2 mV, where the calculation method is shown in a literature.37 However, the observed increase in ∆VOC of up to 15 mV (VOC increase from 0.680 to 0.695 V) in Table 2 after the HLS+LS process is higher than expected only as a result from the enhanced NA. The increase in VOC after the HLS+LS process is thus considered to be attributed to not only the increase in NA but also the decrease in carrier recombination. Therefore, it is intriguing to investigate the carrier recombination at various regions in the Cd-free solar cell before and after the HLS+LS process. It has been reported that the recombination rates at the buffer/absorber interface (Ri), in SCR (Rd) and in QNR (Rb) can be quantitatively calculated after knowing the voltage-independent recombination rates at the buffer/absorber interface (Ri0), in SCR (Rd0), and in QNR (Rb0).25,26 To estimate the recombination rates, the C-V and Voc-T measurements in Figure 6 and Suns-Voc measurement in Figure 7 were performed before (as-fabricated) and after the HLS+LS process. The NA and φbi values were extracted from C-V measurement in Figure 6a and EA values were obtained from Voc-T measurement in Figure 6b.

18


Page 18 of 45

Page 19 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Moreover, the Rd0 and Ri0+Rb0 was calculated by fitting the Suns-Voc in Figure 7, and the Ri0/Rb0 is calculated based on NA, φbi, and EA, where the estimation method is explained in literatures,25,26 except that the near-surface Eg (1.26 eV) of CIGSSe absorber is used. All values are therefore summarized in Table 3. According to the essential parameters in Table 3, the Ri0, Rd0, Rb0, Ri, Rd, and Rb before (as-fabricated) after the HLS+LS process in Table 4 were ultimately derived with the previous reported methodology.25,26 It is disclosed that the HLS+LS process gives rise to not only the enhanced NA in Table 3 but also the decrease in the recombination rates at the buffer/absorber interface (Ri0 and Ri) in Table 4. The increased NA is caused by divacancy complex (VSe-VCu) converting from a shallow donor to a shallow acceptor,35,36 whereas the reductions of Ri0 and Ri are possibly attributable to the migrated alkaline metals during the HLS+LS process, thereby passivating the recombination centers at the buffer/absorber interface as also explained by Nishinaga et al..24 Ultimately, the η of the Cd-free CIGSSe solar cell (K treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass by all dry process is enhanced to 20.0% after the HLS+LS process as illustrated in Figure 8, where its JSC of 39.3 mA/cm2 is equal to its integrated photocurrent density in Figure 5b. The optimizations on both device structure and HLS+LS process in this work gives rise to

19



the η (20.0%) of the level of record efficiency for Cd-free CIGSSe solar cell prepared by all dry process.12,14,15 The further development to more increase the VOC and FF for η above 20.0% can be conducted as follow. The conduction band offsets among CIGSSe absorber, Zn1-xMgxO buffer layer, and Zn1-xMgxO:Al (TCO) layer should be examined and the deposition condition of Zn1-xMgxO buffer layer should be optimized such as the further decrease in the deposition power to avoid sputtering damage. These could result in the decrease in the carrier recombination at the buffer/absorber interface, thereby increasing VOC and FF.

4. CONCLUSIONS Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al are deposited by the co-sputtering method. The Zn0.8Mg0.2O has the optical Eg of 3.75 eV and high sheet resistance of 3.7x106 Ω/sq, whereas Zn0.9Mg0.1O:Al has optical Eg of 3.80 eV and low sheet resistance of 18.7 Ω/sq. Therefore, Zn0.8Mg0.2O and Zn0.9Mg0.1O:Al are suitable for buffer and TCO layers, respectively. The lower powers during the Zn0.8Mg0.2O sputtering are intentionally applied as compared with those during the Zn0.9Mg0.1O:Al sputtering to use the Zn0.8Mg0.2O to partly reduce sputtering damage near CIGSSe surface. As a result, the

20


Page 20 of 45

Page 21 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


18.1%-efficient CIGSSe solar cell (without K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass by all dry process is obtained. To further enhance the η, the Cd-free CIGSSe solar cell with K-treatment is fabricated by all dry process with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass and followed by HLS+LS process. Ultimately, the η of the Cd-free CIGSSe solar cell (K-treatment) is enhanced to 20.0%. According to the investigation of carrier recombination rates, the HLS+LS process gives rise to not only the enhanced NA but also the most decrease in the recombination rates at the buffer/absorber interface (Ri0 and Ri). The HLS+LS process is consequently proved to be an effective method to further enhance the η of the Cd-free CIGSSe solar cell.

21



Page 22 of 45

Tables Table 1. Corresponding maximum values of photovoltaic parameters of the Cd-free CIGSSe solar cells (w/o K-treatment) with different thickness of Zn0.8Mg0.2O buffer layer in Figure 3. The diode parameters of the solar cells such as series resistance (RS), shunt resistance (RSH), saturation current density (J0), and ideality factor (n) were obtained by the curve fitting of their J-V characteristics. The w/o denotes “without”. Zn0.8Mg0.2O JSC VOC FF η J0 RSH RS n thickness (nm) (mA/cm2) (V) (%) (%) (mA/cm2) (Ω.cm2) (Ω.cm2) 0 (w/o buffer layer)

20.6

0.441 61.2 5.5 3.4x10-6 1.97

180

0.72

34

37.2

0.634 71.1 16.8 1.9x10-8 1.70

290

0.42

60

36.2

0.647 72.4 16.7 1.8x10-8 1.72

370

0.52

124

38.2

0.650 73.2 18.1 1.1x10-8 1.67

380

0.32

155

34.9

0.631 71.2 15.6 2.5x10-8 1.71

300

0.81

36.1

0.673 75.6 18.2 3.4x10-9 1.60

1100

0.35

Reference solar cell ZnO/CdS/CIGSSe

22


Page 23 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Corresponding cell parameters (before (as-fabricated) and after the HLS+LS process including (i) HLS for 4 h followed by (ii) LS for 2 h) for Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass. JSC VOC FF η J0 RSH RS Post-treatment n (mA/cm2) (V) (%) (%) (mA/cm2) (Ω.cm2) (Ω.cm2) As-fabricated

39.4

0.680 70.2 18.7 3.01x10-9 1.65

201

0.35

HLS+LS (i) HLS for 4h

39.3

0.693 67.0 18.3 1.71x10-9 1.59

210

1.28

process

39.3

0.695 73.1 20.0 1.69x10-9 1.58

250

0.37

(ii) LS for 2 h

Table 3. Essential parameters for extracting Ri0, Rd0, and Rb0 for the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass before (as-fabricated) and after the HLS+LS process, which were obtained from VOC−T, Suns−VOC, and C−V measurements. Near-surface Eg of CIGSSe absorber is 1.26 eV. R d0 Ri0+Rb0 (cm-2s-1) (cm-2s-1)

NA

φbi Near-surface Eg EA

(cm-3)

Ri0/Rb0

(V)

(eV)

(eV)

As deposited 8.19x1011 3.38x106 1.54x1017 0.87

1.26

1.02

9.16

8.01x1011 7.84x105 2.98x1017 1.00

1.26

1.14

3.60

HLS + LS

Table 4. Summary of device performances of the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass and the extracted recombination parameters before (as-fabricated) after the HLS+LS process. η

Ri0

Near-surface Eg VOC

(%)

(eV)

As deposited 18.7

1.26

Rd0

(V) (cm-2s-1) (cm-2s-1)

Rb 0

Ri

Rd

Rb

(cm-2s-1)

(cm-2s-1)

(cm-2s-1)

(cm-2s-1)

0.680 3.05x106 8.19x1011 3.33x105 6.96x1017 3.91x1017 7.60x1016

23



HLS + LS

20.0

1.26

Page 24 of 45

0.695 6.13x105 8.01x1011 1.71x105 2.49x1017 5.11x1017 6.93x1016

ASSOCIATED CONTENT AUTHOR INFORMATION Corresponding Authors: *E-mails: [email protected] and [email protected]

Notes The authors declare no computing financial interest.

ACKNOWLEDGEMENT This work is partly supported by NEDO (the New Energy and Industrial Technology Development Organization) in Japan.

SUPPORTING INFORMATION Schematic structure, photo-J-V characteristic, EQE, and first derivative of EQE for reference CIGSSe solar cell (without K-treatment), HLS+LS process, first derivative of EQE for Cd-free CIGSSe solar cell (with K-treatment)

24


Page 25 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


REFERENCES (1) Jackson, P.; Wuerz, R.; Hariskos, D.; Hariskos, D.; Lotter, E.; Witte, W.; Powalla, M. Effects of heavy alkali elements in Cu(In,Ga)Se2 solar cells with efficiencies up to 22.6%. Phys. Status Solidi RRL 2016, 10, 583-586. (2) Chirila, A.; Reinhard, P.; Pianezzi, F.; Bloesh, P.; Uhl, AR.; Fella, C.; Kranz, L.; Keller, D.; Gretener, C.; Hagendorfer, H.; Jaeger, D.; Erni, R.; Nishiwaki, S.; Buecheler, S.; Tiwari, A. N. Potassium-induced surface modification of Cu(In,Ga)Se2 thin films for high-efficiency solar cells. Nat. Mater. 2013, 12, 1107-1111. (3) Kamada, R.; Yagioka, T.; Adachi, S.; Handa, A.; Tai, K. F.; Kato, T.; Sugimoto, H. New world record Cu(In,Ga)(Se,S)2 thin film solar cell efficiency beyond 22%. in: Conference Record of the 43rd IEEE Photovoltaic Specialists Conference 2016, pp. 1287-1291. (4) Chantana,

J.;

Kato,

T.;

Sugimoto,

H.;

Minemoto,

T.

Thin-film

Cu(In,Ga)(S,Se)2-based solar cell with (Cd,Zn)S buffer layer and Zn1-xMgxO window layer. Prog. Photovol. Res. Appl. 2017, 25, 431-440.

25



Page 26 of 45

(5) Mizumoto, Y.; Chantana, J.; Hironiwa, D.; Yamamoto, A.; Yabuki, K.; Nakaue, A.;

Minemoto,

T.

Evaluation

of

junction

quality

of

buffer-free

Zn(O,S):Al/Cu(In,Ga)Se2 thin-film solar cells. Appl. Phys. Express 2014, 7, 125503. (6) Kobayashi, T.; Zacharie, J. L. K.; Kato, T.; Sugimoto, H.; Nakada, T. A comparative

study

of

Cd-

and

Zn-compound

buffer

layers

on

Cu(In1-x,Gax)(Sy,Se1-y) thin film solar cells. Prog. Photovol. Res. Appl. 2016, 24, 389-396. (7) Nakada, T.; Mizutani, M. 18% efficiency Cd-free Cu(In,Ga)Se2 thin-film solar cells fabricated using chemical bath deposition (CBD)-ZnS buffer layer. Jpn. J. Appl. Phys. 2002, 41, L165-L167. (8) Ennaoui, A.; Bar, M.; Klaer, J.; Kropp, T.; Saez-Araoz, R.; Lux-Steiner, M. Ch. Highly-efficient Cd-free CuInS2 thin-film solar cells and mini-modules with Zn(S,O) buffer layers prepared by an alternative chemical bath process. Prog. Photovol. Res. Appl. 2006, 14, 499-511. (9) Nakashima, K.; Kumazawa, T.; Kobayashi, T.; Nakada, T. Wide-gap Cu(In,Ga)Se2 solar cells with Zn(O,S) buffer layers prepared by atomic layer deposition. Jpn. J. Appl. Phys. 2012, 51, 10NC15.

26


Page 27 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(10) Platzer-Bjorkman, C.; Torndahl, T.; Abou-Ras, D.; Malmstrom, J.; Kessler, J.; Stolt, L. Zn(O,S) buffer layers by atomic layer deposition in Cu(In,Ga)Se2 based thin film solar cells: band alignment and sulfer gradient. J. Appl. Phys. 2006, 100, 044506. (11) Powalla, M.; Dimmler, B. Scaling up issues of CIGS solar cells. Thin Solid films 2000, 361-362, 540-546. (12) Hultqvist, A.; Platzer-Bjorkman, C.; Torndahl, T.; Ruth, M.; Edoff, M. Optimization of i-ZnO window layers for Cu(In,Ga)Se2solar cells with ALD buffer, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference 2007, pp. 2381-2384. (13) Negami, T.; Aoyagi, T.; Satoh, T.; Shimakawa, S.; Hayashi, S.; Hashimoto, Y. Cd free CIGS solar cells fabricated by dry process. in: Conference Record of the 29th IEEE Phovoltaic Specialists Conference 2002, pp. 656-659. (14) Spiering, S.; Nowitzki, A.; Kessler, F.; Igalson, M.; Maksoud, H. A. Optimization of buffer-window layer system for CIGS thin film devices with indium sulphide buffer by in-line evaporation. Sol. Energy Mater Sol. Cells 2016, 144, 544-550.

27



(15) Bauer, A.; Sharbati, S.; Powalla, M. Systematic survey of suitable buffer and high resistive window layer materials in CuIn1-xGaxSe2 solar cells by numerical simulations. Sol. Energy Mater Sol. Cells 2017, 165, 119-127. (16) Chantana, J.; Kato, T.; Sugimoto, H.; Minemoto, T. Aluminum-doped Zn1-xMgxO as transparent conductive oxide of Cu(In,Ga)(S,Se)2-based solar cell for minimizing surface carrier recombination. Prog. Photovol. Res. Appl. 2017, 25, 996-1004. (17) Tai, K. F.; Kamada, R.; Yagioka, T.; Kato, T.; Sugimoto, H. From 20.9 to 22.3% Cu(In,Ga)(S,Se)2 solar cell: Reduced recombination rate at the heterojunction and the depletion region due to K-treatment. Jpn. J. Appl. Phys. 2017, 56, 08MC03. (18) Kato, T.; Handa, A.; Yagioka, T.; Matsuura, T.; Yamamoto, K.; Higashi, S.; Wu, J. L.; Tai, K. F.; Hiroi, H.; Yoshiyama, T.; Sakai, T.; Sugimoto, H. Enhanced efficiency of Cd-free Cu(In,Ga)(S,Se)2 minimodule via (Zn,Mg)O secondary buffer layer and alkali metal post-treatment. IEEE J. Photovolt. 2017, 7, 1773-1780.

28


Page 28 of 45

Page 29 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(19) Chantana, J.; Ishino, Y.; Kawabata, K.; Minemoto, T. Examination of electrical and optical properties of Zn1-xMgxO:Al fabricated by radio frequency magnetron co-sputtering. Thin Solid Films 2018, 646, 105-111. (20) Kushiya, K.; Tachiyuki, M.; Kase, T.; Sugiyama, I.; Nagoya, Y.; Okumura, D.; Sato, M.; Yamase, O.; Takeshita, H. Fabrication of graded band-gap Cu(In,Ga)Se2 thin-film mini-modules with a Zn(O,S,OH)x buffer layer. Sol. Energy Mater Sol. Cells 1997, 49, 277-283. (21) Nakada, T.; Furumi, K.; Kunioka, A. High-efficiency cadmium-free Cu(In,Ga)Se2 thin- film solar cells with chemically deposited ZnS buffer layers. IEEE Trans. Electron Devices 1999, 46, 2093-2097. (22) Kobayashi, T.; Kumazawa, T.; Jehl, Z.; Nakada, T. Post-treatment effects on ZnS(O,OH)/Cu(In,Ga)Se2 solar cells deposited using thioacetamide-ammonia based solution. Sol. Energy Mater Sol. Cells 2014, 123, 197-202. (23) Kobayashi, T.; Kao, Z. J. L.; Nakada, T. Temperature dependent current-voltage and admittance spectroscopy on heat-light soaking effects of Cu(In,Ga)Se2 solar cells with ALD-Zn(O,S) and CBD-ZnS(O,OH) buffer layers. Sol. Energy Mater Sol. Cells 2015, 143, 159-167.

29



(24) Nishinaga, J.; Koida, T.; Ishizuka, S.; Kamikawa, Y.; Takahashi, H.; Iioka, M.; Higuchi, H.; Ueno, Y.; Shibata, H.; Niki, S. Effect of long-term heat-light soaking on Cu(In,Ga)Se2 solar cells with KF postdeposition treatment. Appl. Phys. Express 2017, 10, 092301. (25) Grover, S.; Li, J. V.; Young, D. L.; Stradings, P.; Branz, H. M. Reformulation of solar cell physics to facilitate experimental separation of recombination pathways. Appl. Phys. Lett 2013, 103, 093502. (26) Li, J. V.; Grover, S.; Contreras, M. A.; Ramanathan, K.; Kuciauskas, D.; Noufi, R. A recombination analysis of Cu(In,Ga)Se2 solar cells with low and high Ga compositions. Sol. Energy Mater Sol. Cells 2014, 124, 143-149. (27) Nakada, T.; Ohbo, H.; Watanabe, T.; Nakazawa, H.; Matsui, M.; Kunioka, A. Improved Cu(In,Ga)(S,Se)2 thin film solar cells by surface sulfurization. Sol. Energy Mater Sol. Cells 1997, 49, 285-290. (28) Minemoto, T.; Negami, T.; Nishiwaki, S.; Takakura, H.; Hamakawa, Y.; Preparation of Zn1-xMgxO films by radio frequency magnetron sputtering. Thin Solid Films 2000, 372, 173-176. (29) Chantana, J.; Kato, T.; Sugimoto, H.; Minemoto, T. Time-resolved photoluminescence of Cu(In,Ga)(Se,S)2 thin films and temperature dependent

30


Page 30 of 45

Page 31 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


current density-voltage characteristics of their solar cells on surface treatment effect. Curr. Appl. Phys. 2017, 17, 461-466. (30) Turcu, M.; Pakma, O.; Rau, U. Interdependence of absorber composition and recombination mechanism in Cu(In,Ga)(Se,S)2 heterojunction solar cells. Appl. Phys. Lett 2002, 80, 2598-2600. (31) Ishizuka, S.; Sakurai, K.; Yamada, A.; Matsubara, K.; Fons, P.; Iwata, K.; Nakamura, S.; Kimura, Y.; Baba, T.; Nakanishi, N.; Kojima, T.; Niki, S. Fabrication of wide-gap Cu(In1-xGax)Se2 thin film solar cells: a study on the correlation of cell performance with highly resistive i-ZnO layer thickness. Sol. Energy Mater Sol. Cells 2005, 87, 541-548. (32) Khatri, I.; Shudo, K.; Matsuura, J.; Sugiyama, M.; Nakada, T. Impact of heat-light soaking on potassium fluoride treated CIGS solar cells with CdS buffer

layer.

Prog.

Photovol.

Res.

Appl.

2017,

1-8.

http://doi.org/10.1002/pip.2962 (33) Kato, T.; Kitani, K.; Tai, K. F.; Kamada, R.; Hiroi, H.; Sugimoto, H. Characterization of the back contact of CIGS solar cell as the origin of “rollover” effect. in: Proceedings of the 32nd European Photovoltaic Solar Energy Conference and Exhibition 2016, pp. 1085-1088.

31



(34) Kim, S.; Lee, C. S.; Kim, S.; Chalapathy, R. B. V.; Al-Ammar, E. A.; Ahn, B. T. Understanding the light soaking effect of ZnMgO buffer in CIGS solar cell. Phys. Chem. Chem. Phys. 2015, 17, 19222-19229. (35) Rau, U.; Schmitt, M.; Parisi, J.; Riedl, W.; Karg, F. Persistent photoconductivity in Cu(In,Ga)Se2 heterojunctions and thin films prepared by sequential deposition. Appl. Phys. Lett 1998, 73, 223-225. (36) Lany, S.; Zunger, A. Light- and bias-induced metastabilities in Cu(In,Ga)Se2 based solar cells caused by the (VSe-VCu) vacancy complex. J. Appl. Phys. 2006, 100, 113725. (37) Wuerz, R.; Eicke, A.; Kessler, F.; Paetel, S.; Efimenko, S.; Schlegel, C. CIGS thin-film solar cells and modules on enameled steel substrates. Sol. Energy Mater Sol. Cells 2012, 100, 132-137.

32


Page 32 of 45

Page 33 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure captions

Figure 1. Schematic layer structure of Cd-free CIGSSe solar cell fabricated by all dry process.

Figure 2. (αhν)2 plots of non-doped ZnO, Zn0.8Mg0.2O, and Zn0.9Mg0.1O:Al films on SLG substrates as a function of photon energy (hν) to estimate the optical Eg. The sheet resistances of non-doped ZnO layer and Zn0.8Mg0.2O layer are 650 and 3.7x106 Ω/sq, while that of Zn0.9Mg0.1O:Al layer is 18.7 Ω/sq.

33



Figure 3. Photovoltaic performance parameters as a function of thickness of Zn0.8Mg0.2O buffer layer of the Cd-free CIGSSe solar cells without K-treatment, where their structure is shown in Figure 1. The cell parameters of standard CIGSSe solar cell (without K-treatment) with a structure of ZnO:Al/ZnO/CdS/CIGSSe/Mo/Glass for reference (Ref.) are indicated by dashed lines.

Figure 4. EQE spectra of the Cd-free CIGSSe solar cells (without K-treatment) with Zn0.8Mg0.2O buffer thicknesses of 0 (without buffer layer) and 124 nm, where their cell structure is shown in Figure 1.

Figure 5. (a) Photo-J-V characteristics (before (as-fabricated), and after HLS+LS process including (i) HLS for 4 h followed by (ii) LS for 2 h) as well as (b) EQE spectra and integrated photocurrent densities (before (as-fabricated) and after the HLS+LS process) for the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass in Figure 1.

34


Page 34 of 45

Page 35 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. (a) Mott-Schottky plot (C−2 vs. Bias (V)) and (b) VOC as a function of temperature (T) for the Cd-free CIGSSe solar cell (with K treatment) with a structure of Zn0.9Mg0.1O:Al

(700

nm)/Zn0.8Mg0.2O

(124

nm)/CIGSSe/Mo/Glass

before

(as-fabricated) and after the HLS+LS process. Near-surface Eg of CIGSSe is 1.26 eV.

Figure 7. Dependences of VOC on light-intensity G (Suns-VOC) for the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass before (as-fabricated) and after the HLS+LS process. G denotes the illumination intensity (or optical generation).

Figure 8. Photo-J-V characteristic of the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass after the HLS+LS process.

35



Figures

36


Page 36 of 45

Page 37 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Schematic layer structure of Cd-free CIGSSe solar cell fabricated by all dry process.

37



Figure 2. (αhν)2 plots of non-doped ZnO, Zn0.8Mg0.2O, and Zn0.9Mg0.1O:Al films on SLG substrates as a function of photon energy (hν) to estimate the optical Eg. The sheet resistances of non-doped ZnO layer and Zn0.8Mg0.2O layer are 650 and 3.7x106 Ω/sq, while that of Zn0.9Mg0.1O:Al layer is 18.7 Ω/sq.

38


Page 38 of 45

Page 39 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Photovoltaic performance parameters as a function of thickness of Zn0.8Mg0.2O buffer layer of the Cd-free CIGSSe solar cells without K-treatment, where their structure is shown in Figure 1. The cell parameters of standard CIGSSe solar cell (without K-treatment) with a structure of ZnO:Al/ZnO/CdS/CIGSSe/Mo/Glass for reference (Ref.) are indicated by dashed lines.

39



Figure 4. EQE spectra of the Cd-free CIGSSe solar cells (without K-treatment) with Zn0.8Mg0.2O thicknesses of 0 (without buffer layer) and 124 nm, where their cell structure is shown in Figure 1.

40


Page 40 of 45

Page 41 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. (a) Photo-J-V characteristics (before (as-fabricated), and after HLS+LS process including (i) HLS for 4 h followed by (ii) LS for 2 h) as well as (b) EQE spectra and integrated photocurrent densities (before (as-fabricated) and after the HLS+LS process) for the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass in Figure 1.

41



Page 42 of 45

Figure 6. (a) Mott-Schottky plot (C−2 vs. Bias (V)) and (b) VOC as a function of temperature (T) for the Cd-free CIGSSe solar cell (with K treatment) with a structure of Zn0.9Mg0.1O:Al

(700

nm)/Zn0.8Mg0.2O

(124

nm)/CIGSSe/Mo/Glass

before

(as-fabricated) and after the HLS+LS process. Near-surface Eg of CIGSSe is 1.26 eV. 42


Page 43 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7. Dependences of VOC on light-intensity G (Suns-VOC) for the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass before (as-fabricated) and after the HLS+LS process. G denotes the illumination intensity (or optical generation).

43



Figure 8. Photo-J-V characteristic of the Cd-free CIGSSe solar cell (with K-treatment) with a structure of Zn0.9Mg0.1O:Al (700 nm)/Zn0.8Mg0.2O (124 nm)/CIGSSe/Mo/Glass after the HLS+LS process.

44


Page 44 of 45

Page 45 of 45 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents (TOC) graphic

45


Cu(In,Ga)(S,Se) 2

Recommend Documents