Toward Omnidirectional Light Absorption by Plasmonic Effect for High-Efficiency Flexible Nonvacuum Cu(In,Ga)Se-2 Thin Film Solar Cells

(1)

August 05, 2014

C 2014 American Chemical Society

Toward Omnidirectional Light

Absorption by Plasmonic E

ﬀect for

High-E

ﬃciency Flexible Nonvacuum

Cu(In,Ga)Se

₂

Thin Film Solar Cells

Shih-Chen Chen,‡_{Yi-Ju Chen,}§_{Wei Ting Chen,})_{Yu-Ting Yen,}§_{Tsung Sheng Kao,}†_{Tsung-Yeh Chuang,}^

Yu-Kuang Liao,‡Kaung-Hsiung Wu,‡Atsushi Yabushita,‡Tung-Po Hsieh,^Martin D. B. Charlton,#

Din Ping Tsai,),XHao-Chung Kuo,†,_{* and Yu-Lun Chueh}§,_*

†_{Department of Photonics and Institute of Electro-Optical Engineering and}‡_{Department of Electrophysics, National Chiao-Tung University, Hsinchu 30010, Taiwan,}

§

Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan, )Department of Physics, National Taiwan University,

Taipei 10617, Taiwan,^Compound Semiconductor Solar Cell Department, Next Generation Solar Cell Division, Green Energy and Environment Research Laboratories,

Industrial Technology Research Institute, Hsinchu, Taiwan,#School of Electronics and Computer Science, University of Southampton, Highﬁeld,

Southampton SO17 1BJ, United Kingdom, andXResearch Center for Applied Sciences, Academia Sinica, Taipei, Taiwan

T

he energy crisis that has arisen from

the decreasing reserve of combustible fuel is one of great impacts to the world that we are currently facing. Greenhouse

eﬀects resulting from the side products of

combustible fuel signiﬁcantly accelerate

haz-ardous climate change. Therefore, ﬁnding

environmentally friendly and sustainable en-ergy to meet enen-ergy demands and reduce climate change is crucial. Energy harvesting by photovoltaic (PV) methods has been garded as a favorable candidate among re-newable energies because the amount of solar radiation reaching to the earth's surface

in one hour can suﬃciently provide one year's

worth of humanity's energy demand. For materials as PV, silicon-based materials, in-cluding single and poly crystalline structures, which are the leading materials, have been extensively studied. However, drawbacks

such as indirect bandgap and poor light

absorption could result in unsatisﬁed

conver-sion eﬃciency for Si-based solar cells.1

Alternatively, Cu(In,Ga)Se2(CIGS) is the most

promising material owing to its excellent light-trapping ability, broadband light absorp-tion, and environmentally friendly manufac-turing processes. The chalcopyrite compound materials have abundant advantages,

includ-ing high environmental tolerance,2

long-term stability,3,4and remarkable absorption

characteristics.5 _{The highest e}_{ﬃciency of}

20.9% on CuIn1 xGaxSe2(CIGS) thin-ﬁlm

so-lar cells has been achieved, which is the

highest recorded eﬃciency among all

thin-ﬁlm photovoltaics (TFPVs).6

Various

deposi-tion methods for CIGS thinﬁlms have been

developed, including vacuum processes

(coevaporation,7 sputtering,8 and pulsed

laser deposition9) and nonvacuum processes

* Address correspondence to [email protected], [email protected]. Received for review June 18, 2014 and accepted August 5, 2014. Published online

10.1021/nn503320m

ABSTRACT We have successfully demonstrated a great advantage of

plas-monic Au nanoparticles for eﬃcient enhancement of Cu(In,Ga)Se2(CIGS)ﬂexible

photovoltaic devices. The incorporation of Au NPs can eliminate obstacles in the

way of developing ink-printing CIGSﬂexible thin ﬁlm photovoltaics (TFPV), such as

poor absorption at wavelengths in the high intensity region of solar spectrum, and that occurs signiﬁcantly at large incident angle of solar irradiation. The enhancement of external quantum eﬃciency and photocurrent have been

systematically analyzed via the calculated electromagnetic ﬁeld distribution.

Finally, the major beneﬁts of the localized surface plasmon resonances (LSPR) in visible wavelength have been investigated by ultrabroadband pump probe spectroscopy, providing a solid evidence on the strong absorption and reduction of surface recombination that increases electron hole

generation and improves the carrier transportation in the vicinity ofpn-juction.

KEYWORDS: solar Cell . Cu(In,Ga)Se2. omnidirectional light absorption . plasmonic eﬀect . nanoparticles . ﬂexible

(2)

(ink-printing,10 and electrochemical deposition11). Among these fabrication processes, the ink-printing method is an attractive approach for development of ﬂexible CIGS TFPV devices because of mass produc-tion and low cost nonvacuum device fabricaproduc-tion. To further promote the CIGS TFPV light absorption ability throughout the full solar spectrum range, extensive

eﬀorts have been devoted, including bandgap

man-agement through diﬀerent doping concentrations12

and various postprocessing procedures, to placing or

creating nanostructures on the CIGS absorber layer,13

while some unwanted defects may be generated,

hindering further practical applications.14

The plasmonic eﬀect, for which the surface charges

of mental nanostructures are triggered by electromag-netic waves, can be collectively oscillated as set of

resonances known as surface plasmons.15 The

plas-mons appearing in a localized area at the interface between two materials are known as localized surface plasmon resonances (LSPR), which are created in

a limited volume of metal nanostructures.16 Metal

conduction electrons in high conduction metals, particularly gold (Au), silver (Ag), and copper (Cu), can coherently trigger the LSPR as a collective. By adjustment of particle diameter, shape, and surround-ing materials, a large fraction of oscillator strength of electrons into the proper spectral ranges can be

controllably tuned.17 22_{The optical cross sections for}

absorption and scattering can be enhanced due to the massive engaged metallic conduction electrons, such that the projected geometrical area of a plasmonic particle can even span over an substantial spectral

range.20,23 Thus, manipulation of the plasmon

reso-nances into the required spectral range is a principal

issue for photovoltaic applications.24 26In this regard,

we demonstrate the plasmonic eﬀect based on gold

nanoparticles, which are embedded between the pn-junction interface of CIGS TFPV, to increase the electron-hole generation and thus enhance the photo-current. With the properly designed layer structure, for which the interface of CdS/CIGS has been found to be the suitable choice for distribution of Au NPs, the

conversion eﬃciency of the plasmonic CIGS ﬂexible

thin-ﬁlm solar cells with embedded Au nanoparticles

can be promoted, especially at large incidence angles. Analyses of the enhancement of external quantum eﬃciency and photocurrent by the plasmonic Au nano-particles (NPs) were investigated in detail. To further investigate the enhancement mechanisms, the

ele-ctromagnetic ﬁeld distribution of plasmon modes

at diﬀerent spectral ranges has been calculated via

ﬁnite element (FEM) simulation. Finally, the ultrafast

carrier dynamics of plasmonic CIGS thin ﬁlm have

been delineated by the ultrabroadband femtosecond pump probe spectroscopy.

RESULTS AND DISCUSSION

Figure 1a shows the absorption spectrum (green color) and the corresponding photo image of a CIGS pn-junction fabricated by ink-printing process on a stainless steel substrate. Obviously, in the range of 500 600 nm from the absorbance spectra of the CIGS pn-junction, poor light absorption can be observed, which may result from the undesirable coverage and the poor quality of the CdS layer on the rough CIGS

surface.27_{It is a critical problem because this range}

belongs to wavelength regions of high intensity in the solar spectrum. As a result, great enhancement of

light-to-electricity eﬃciency can be expected if this

drawback can be tackled. For such a purpose, we

introduced the Au NPs with a diameter of∼10 nm,

whose plasmonic resonance is located at a wavelength

Figure 1. (a) Absorption spectra and photo images of CdS/CISGpn-junction of ink-printing CIGS TFPV and gold nanoparticles solution. (b) Schematic layer structure of plasmonic CIGS TFPV. (c) TEM image of the plasmonic CIGS TFPV. (d) High-resolution TEM image of a selected area at the interface between CdS and CIGS.

(3)

of∼530 nm as shown in the red curve in Figure 1a. The inset also shows the corresponding photo image of the Au NPs solution. For the fabrication of the CIGS solar

cell, Mo back contact deposited onﬂexible stainless

steel foil was prepared as the substrate. A precursor layer was prepared via the nonvacuum nanoparticle method and coated on the substrate by an ink-printing process. After the postselenization and the KCN washing process, the p-type CIGS absorber layer was accomplished. Then, plasmonic Au nanoparticles with a diameter of 10 nm were sprayed on the CIGS layer via

air brush with low nitrogen gas ﬂow. Through the

chemical bath deposition, a cadmium sulﬁde (CdS)

thinﬁlm was deposited on the Au nanoparticles layer

where plasmonic nanoparticles were located at the

pn-junction interface of CIGS/CdS. The buﬀer layer and

transparent conductive oxides, i-ZnO and Al:ZnO(AZO), were capped on the pn-junction. Finally, the silver grid was coated by a printing process. The ultimate device structure is schematically shown in Figure 1b. To

conﬁrm the existence of the Au NPs between CIGS

pn-junction, a TEM image was conducted as shown in Figure 1c. Figure 1d shows the corresponding high-resolution TEM image taken from rectangular area in Figure 1c. Obviously, the Au NPs (∼10 nm) embedded between CdS and CIGS can be observed and are marked by the white dash circles in Figure 1d.

The eﬀect of Au NPs on light absorption on the

CdS/CIGS pn-junction was examined before the device was accomplished. The absorption enhancement

factor deﬁned by ratio of absorption intensity with

and without Au NPs (Awith Au NPs/Awithout Au NPs) in

the CdS/CIGS pn-junction are shown in Figure 2a at

diﬀerent incident light wavelengths. The absorption

enhancement factor >1 can be obviously observed, indicating the enhanced light absorption in the CdS/ CIGS pn-junction, which is attributed to the

super-position of various plasmonic modes of diﬀerent arrays

of Au NPs by comparing the absorption spectrum with Au NPs (Figure 1a) while the red-shifted behavior due to diﬀerent surrounding materials can be found

as well.24 _{Note that the peak enhancement factor}

arisen from plasmonic resonance of Au NPs is broad

due to the random distribution of Au NPs.24 _{On the}

other hand, a small decrease at shorter wavelengths (<450 nm) and longer wavelengths (>800 nm) are most

likely attributed to the energy dissipation28,29and the

backward scattering30,31by Au NPs, respectively. The

latter one can be compensated and is beneﬁcial after

deposition of i-ZnO and AZO layers on the junction owing to the increase of absorption through the

Fabry Perot interferences.32,33To shed light on

en-hanced absorption by the plasmonic eﬀect for possible

application on ﬂexible TFPV, the measurements of

optical properties on the fabricated plasmonic CIGS thin-ﬁlm samples were conducted by using

angle-resolved reﬂectance spectrometer. The corresponding

reﬂectance spectra of the conventional and plasmonic

CIGS thin-ﬁlm solar cells obtained at diﬀerent angle of

incidence (AOI) are shown in parts b and c of Figure 2, respectively. Obviously, a strong elimination of light

reﬂectance in the visible range of 500 700 nm can be

found in the CIGS photovoltaic device with Au NPs

(plasmonic CIGS thin-ﬁlm solar cell) compared with the

CIGS photovoltaic device without Au NPs (conventional CIGS thin-ﬁlm solar cell). The signiﬁcant absorption region following the high optical intensity region in

solar spectrum reﬂects the great feasibility of the

photocurrent enhancement. Thus, with the strong plas-monic resonance of Au nanoparticles, one may improve the poor light absorption, which results from the undesirable coverage and poor quality of the CdS layer on the rough CIGS nanoparticle surface. On the other

hand, the conventional CIGS exhibits higher reﬂectance

at lager incident angles that give a restriction on the

application of theﬂexible TFPV. Such high reﬂectance at

large incident angles can be signiﬁcantly suppressed

after the incorporation of Au NPs, which enables inte-gration of CIGS TFPV to infrastructures with various

Figure 2. (a) Absorption enhancement factor of the plas-monic CIGS TFPV. Angle-resolved re_{ﬂectance spectra of} the (b) conventional and (c) plasmonic CIGS TFPVs. A strong elimination of light reﬂectance can be found in the spectra of the plasmonic CIGS TFPV in the visible range of 500 700 nm.

(4)

geometrical conditions. In addition, we also demon-strated another layer structure design whose Au NPs were placed in the interface of the CdS/i-ZnO layer rather than in the interface of the CdS/CIGS pn-junction (Figure S1a, Supporting Information). The TEM and the corresponding high-resolution TEM images as shown in

Figure S1b (Supporting Information) distinctly conﬁrm

the existence of Au NPs marked by white circles, at

which the diameters of Au NPs with∼10 nm was found.

The angle-resolved reﬂectance of Au NPs in the

inter-face of the CdS/i-ZnO layer as shown in Figure S2 (Supporting Information) reveals that the reﬂectance can be decreased in the wavelengths between 500 and 600 nm due to surface plasma resonance of Au NPs but

cannot be eﬀectively decreased in omnidirectional

directions as the best layer structure design.

In addition to the optical properties, the electric

properties of the plasmonic CIGS thin ﬁlm solar cell

were also quantitatively characterized. The results with a comparison to the conventional CIGS photovoltaic device are shown in Figure 3a. The J V characteristics

show that the light-to-electricity eﬃciency can be

enhanced from 8.31% to 10.36% with the incorpora-tion of Au NPs to harvest more incident light energy at the plasmonic resonance regions. The corresponding Voc, Jsc, FF, and efficiency are listed in Table 1. The inset shows a photo image of a fabricated plasmonic CIGS flexible TFPV. It is worth noting that the strong near-field signal of the plasmonic modes not only enhanced the photocurrent by generating more electron hole pairs but also improved the open-circuit voltage

from 0.44 to 0.45 V with an enhanced ﬁlling factor

from 54.77 to 61.25%. The improved Voc and FF may be

related to signiﬁcant reduction of surface

recombina-tion at the interface of the CdS/CIGS pn-juncrecombina-tion and

will be discussed later.34,35 A comparison between

the external quantum eﬃciency (EQE) spectra of

con-ventional and plasmonic ﬂexible CIGS photovoltaic

devices are shown in Figure 3b. The EQE spectrum of the plasmonic CIGS device shows signiﬁcant enhance-ment throughout wavelengths ranging from 500 to 1100 nm. In order to investigate the mechanism of the

eﬃciency enhancement in the plasmonic CIGS

photo-voltaic device, enhancement factor of the EQE by

ratio of EQEwith Au NPs/EQEwithout Au NPsand enhanced

photocurrentΔJph= Jph(with Au NPs) Jph(without

Au NPs) were plotted by red and blue lines as the func-tion of the light wavelengths as shown in Figure 3c.

Obviously, the enhanced EQE of∼5 10% in a

broad-band range of 500 1100 nm can be conﬁrmed, which

is most likely attributed to the localized surface plas-mon resonance (LSPR) of Au NPs embedded in the

interface of the CdS/CIGS pn-junction.32 In addition,

the LSPR can trigger more electron hole pairs, and

thus, distinctly enhanced ΔJph at wavelengths of

500 800 nm can be achieved for the plasmonic CIGS PV cell (blue curve in Figure 3c). However, as for the region of longer wavelengths (>800 nm), the enhanced EQE factor may result from the expected rescattering of incident light by Au nanoparticles. However, there are

no signiﬁcant enhancement in photocurrent due to

the weaker intensity of solar spectrum in this region. Abrupt peaks in this region correspond to the spec-trum of the light source. The poor performance at wavelength <450 nm as marked at the red region in Figure 3c is most likely attributed to energy dissipation and backward scattering of light by Au NPs, which is consistent with the absorption spectra in Figure 2a.

Figure 3. (a) Comparison of theJ V characteristics between conventional (black) and plasmonic (red) ﬂexible CIGS TFPV. (b) Comparison of the EQE curve between conventional (black) and plasmonic (red)ﬂexible CIGS TFPV. (c) Enhancement factor of the EQE(red) and photocurrentΔJph(blue) of plasmonic CIGS TFPV.

TABLE 1.Comparison of the J V Characteristics between Conventional and Plasmonic Flexible CIGS Photovoltaic Devices

ﬂexible CIGS TFPV Voc(V) Jsc(mA/cm 2

) FF (%) eﬃciency (%)

without Au NPs 0.44 34.84 54.77 8.31

with Au NPs 0.45 37.81 61.25 10.36

(5)

All of the PV characterizations discussed above were performed with normal incident light. However, the conversion of light into electric energy occurring at a wide range of incident light should be considered for practical PV applications. To investigate the

enhance-ment of conversion eﬃciency at diﬀerent angles of

light irradiation, the angle-resolved J V characteristics measurements were conducted as shown in Figure 4a.

The average of acquired eﬃciency enhancement

factors deﬁned by the ratio of eﬃciency

calcula-tion with and without Au NPs in the interface of the

CdS/CIGS pn-junction (ηwith Au NPs/ηwithout Au NPs) are

shown in Figure 4b at diﬀerent irradiation angles from

normal to 48 with respect to the surface of the device.

Clearly, the conversion eﬃciency can be more

signiﬁ-cantly enhanced at a large operating angle (>40)

from the plasmonic CIGS photovoltaic devices due to the excellent omnidirectional absorption. The results demonstrate that the plasmonic CIGS TFPV are quite

suitable for the practical applications ofﬂexible TFPV.36

In addition, the angle-resolved eﬃciency

enhance-ment from J V measureenhance-ments for another layer struc-ture design (see Figure S1a,b, Supporting Information) was also measured (Figure S3, Supporting Information). Obviously, the plasmonic CIGS TFPV with Au NPs in the interface of the CdS/i-ZnO layers cannot maintain the similar angle-resoled J V characteristic, with which

the enhanced factors of conversion eﬃciency were

suppressed with the large angle of incident angles. The results are also agreement with the angle-resoled

reﬂectance mapping at diﬀerent incident angles that

the enhanced absorption by Au NPs is not omnidirec-tional (Figure S2, Supporting Information). As a result, it can be concluded that the best location for placing Au NPs is in the interface of the CdS/CIGS pn-juction of the plasmonic CIGS TFPV.

Furthermore, a ﬁnite element method (FEM) was

adopted to calculate the electromagneticﬁeld

distri-bution in order to elucidate the plasmonic modes, which enhance the performance of plasmonic CIGS

TFPV.37,38Note that we found that the LSPR of Au NPs

was excited with diﬀerent ﬁeld distributions in regions

II and III indicated as the green and blue color regions from Figure 3c in parts a and b, respectively, of Figure 5. At region II (Figure 3c), as this LSPR mode was excited,

there were very strong near-ﬁeld regions in the vicinity

of the gold particle, especially near the interface of the

Figure 4. (a) Schematic of angle-resolvedJ V characteris-tics measurements. (b) Enhancement factor of conversion eﬃciency at diﬀerent incident angles of light irradiation.

Figure 5. Distributions of electricﬁelds for plasmon resonance at (a) region II and (b) region III calculated by ﬁnite element (FEM) method.

(6)

CdS/CISG pn-junction (Figure 5a), and the strong near-ﬁeld region of the LSPR mode can penetrate into

the CIGS absorber layer. This strong near-ﬁeld region

can be signiﬁcantly enhanced by gap resonance, while

there is a gap between Au NPs and the interface of

pn-juction.39We also demonstrated the electromagnetic

ﬁeld distribution in Figure S4 (Supporting Information). Thus, a larger amount of incident light was

concen-trated and absorbed in this ﬁeld, generating more

photoexcited carriers (electron hole pairs). There are more aﬀected zones in the CIGS absorber layer by the tilted near-ﬁeld region, which arise from the incident light at large angle. That explains the omnidirectional absorption behavior, resulting in a great

enhance-ment in conversion eﬃciency at higher incident angle

(Figure 4b). In addition, the strong ﬁeld nearby the

interface of the CdS/CIGS pn-junction can trigger an additional external force to enhance the carrier extrac-tion, by which surface recombination of carriers can be

signiﬁcantly reduced, resulting in the improvement

in electrical transport properties to enhance Voc and

FF (Table 1), respectively.32,34However, in region III,

the conventional ﬁeld distribution of LSPR occurred

(Figure 5b). The near-ﬁeld proﬁles of the dipolar mode

are distributed on the both sides of the gold nano-particle, which mainly result in an increase of the scattering cross-section, triggering more rescattering

activities. Such a phenomenon is beneﬁcial to the

utilization of incident light for a solar cell. Therefore, the enhanced photocurrent can be also acquired in the long wavelength region (blue region in

Figure 3c).25

To further verify our speculation, we investigate the ultrafast carrier dynamics of plasmonic CIGS thin ﬁlm by ultrabroadband pump probe spectroscopy. Pump probe spectroscopy has been commonly used for carrier recombination studies, particularly

Shockley Read Hall (SRH) recombination.40 42Here,

the samples were pumped and probed by an ultra-broadband femtosecond pulsed laser. The transient absorptivity changes (ΔA/A) corresponded to the re-laxation process of photoexcited carriers.

Quantita-tively, the higher intensity of (ΔA/A) indicates a larger

amount of excited carriers, while the longer relaxation

time implies the less SRH recombination rate.40 42_The

corresponding ultrabroadband pump probe spectra

for the CIGS thinﬁlms without and with Au NPs are

shown in parts a and b, respectively, of Figure 6. After the incorporation of Au NPs, we found more excited carrier at the wavelengths where LSPR occurs in region

II, which is attributed to strong near-ﬁeld region near

the Au NPs. We have quantitatively conﬁrmed the

increase of excited carrier by using moment analysis

as shown in Figure S5 (Supporting Information).43

In addition, longer carrier relaxation time can also be observed in this region, which is ascribed to the

strong ﬁeld in the vicinity of the interface of the

CdS/CISG pn-juction. The strong ﬁeld improved the

poor transportation of carriers at the interface by the

eﬀective reduction of surface recombination, which

is consistent with the observations of the signi

ﬁ-cant improvement in electrical transportation in J V measurements when the Au NPs was incorporated. The demonstration of the LSPR in visible wavelength

inﬂexible CIGS provides solid evidence of the strong

absorption and reduction of surface recombina-tion that increases electron hole generarecombina-tion and im-proves the carrier transportation in the vicinity of the pn-juction.

CONCLUSION

In this work, we have successfully demonstrated the great advantage of plasmonic Au nanoparticles

for eﬃciency enhancement of CIGS ﬂexible

photovol-taic devices. The incorporation of Au NPs can eliminate obstacles in the way of developing ink-printing CIGS ﬂexible TFPV, such as the poor absorption at wave-lengths in the high intensity region of the solar spec-trum, which occurs at a large incident angle of solar irradiation. The enhancement of external quantum

eﬃciency and photocurrent have been systematically

analyzed via the calculated distributions of electric ﬁeld. Finally, the major beneﬁts the LSPR at shorter wavelength have been investigated by ultrabroad-band pump probe spectroscopy, which gives solid evidence of the strong absorption and reduction of surface recombination that increases electron hole generation and improves the carrier transportation

Figure 6. Ultrabroadband pump probe spectra mapping images for the (a) conventional and (b) plasmonic CIGS TFPV at diﬀerent wavelengths.

(7)

in the vicinity of pn-juction. These results suggest a promising method for rapidly improving the

performance of CIGS ﬂexible photovoltaic devices

at low cost.

EXPERIMENTAL METHODS

Preparation CIGS Absorber Layer by Ink-Printing Process. The CIGS thin films were prepared by a nonvacuum nanoparticle synthe-sis method. The nanoingredients of copper oxide, indium oxide, and gallium oxide were dispersed in water-based solution by utilizing the agent. The solid content of the stable ink was about 20 g/100 mL. A scalpel was employed to coat the precursor film on the Mo/Cr/stainless steel substrate. Then, the precursor was re-duced under H2atmosphere at 450C for 30 min. At the end, the

precursor was annealed in H2Se atmosphere at 400C for 30 min.

Characterizations. A UV vis NIR spectrophotometer (Hitachi U4100) with standard mirror optics and an integrating sphere was used to measure the absorptance of the CIGS pn-juction and Au NPs in the 350 800 nm range at normal incidence. The angle-resolved reflectance spectroscopic system used a custom-built 15 cm-radius integrating sphere with a motor-controlled rotational sample stage in the center, and a broad-band 300 W xenon lamp was employed to carry out the angle-resolved mapping image. Then, the collected reflectance photons were resolved by a spectrometer (SPM-002-ET, Photon-control Inc.) to acquire the reflective spectrum with respect to different AOIs. The system was calibrated by the reflective spectrum of a NIST-standard, intrinsic Si at normal incidence.

In order to analyze the characterization of devices, J V measurements were conducted, closely following the proce-dure described in international standard CEI IEC 60904-1. Both the CIGS solar cells and the reference cell were measured under a simulated Air Mass 1.5, Global (AM1.5G) illumination with a power of 1000 W/m2_{. The temperature was actively maintained}

at 25( 1 C during the measurements. Power conversion efficiency (PCE) measurements were performed for further characterization. The PCE measurement system consisted of a power supply (Newport 69920), a 1000 W Class A solar simulator (Newport 91192A) with a xenon lamp (Newport 6271A), an AM1.5Gfilter (Newport 81088A), a probe stage, and a source meter with a four-wire mode (Keithley 2400), respectively. The spectrum of the solar simulator was measured by a calibrated spectroradiometer (Soma S-2440) in the wavelength range of 300 to 1350 nm. The EQE was acquired by a 300 W xenon lamp (Newport 66984) light source with a monochromator (Newprot 74112). The EQE measurement was performed by using a lock-in amplifier (Standard Research System, SR830), an optical chopper unit (SR540) operated at 260 Hz chopping frequency, and a 1Ω resistor in shunt connection to convert the photo-current into voltage. TEM (JEOL, JEM-3000F) equipped with EELS was utilized to observe the nanoscale layer structure and composition distribution.

Ultrabroadband Pump Probe Measurements. For ultrabroadband pump probe measurements, a noncollinear optical parametric amplifier (NOPA) was built to produce visible laser pulses, whose spectral width is adequately broad to sustain sub-10 fs visible pulses for the measurement of ultrafast time-resolved spectro-scopy. We used a regenerative chirped pulse amplifier (Legend-USP-HE; Coherent) seeded with a Ti:sapphire laser oscillator (Micra 10; Coherent: wavelength = 800 nm, pulsed width = 40 fs, repetition rate = 5 kHz) for pumping and seeding the NOPA. The beam from the regenerative amplifier was divided into two beams by using a beam splitter. One beam was utilized to generate a second harmonic of 400 nm to pump the NOPA. Another beam was focused on a sapphire plate to generate white light by inducing self-phase modulation and regarded as the seed beam of the NOPA. Finally, the NOPA generated a broad visible spectrum span over 520 to 700 nm with a nearly constant phase. A beam sampler separated the visible laser pulse into pump and probe beams. A polychromator (SP2300i; Princeton Instruments) was conducted for dispersing the probe pulse into a 128-branch fiber bundle, whose other end was split into 128 fiber branches and linked to avalanche photodiodes (APDs).

Three-Dimensional FEM Electromagnetic Simulation. Three-dimen-sional FEM electromagnetic simulation (COMSOL Multiphysics) was use to simulate the electromagnetic field distribution of Au NPs with 10 nm diameter embedded into CIGS device. The size of gold nanoparticles are much smaller than the incident light wavelength. We choose periodic boundary condition with 200 nm periodicity along both x- and y-directions that match the average distance of Au NPs in experimental sample. Normal incident light with linear polarization transverses to the plane of incidence (i.e., TM polarization) radiated from the top of the gold nanoparticle. TM (transverse magnetic) polarized light means the electromagnetic wave in which the magnetic field vector is everywhere perpendicular to the plane of incidence (the plane of incidence is the plane spanned by the surface normal and the propagation vector of light).

Conflict of Interest: The authors declare no competing ﬁnancial interest.

Supporting Information Available: Another layer structure design for plasmonic CIGS solar cells and the corresponding TEM image; angle-resolved reﬂectance of Au NPs in the inter-face of the CdS/i-ZnO layer; angle-resoled J V characteristics; simulated electromagneticﬁeld distribution of gap resonance; moment analysis for ultrabroadband pump probe measure-ments. This material is available free of charge via the Internet http://pubs.acs.org/.

Acknowledgment. The research was supported by the Ministry of Science and Technology through Grant Nos. 102-2112-M-009-006-MY3, 101-2218-E-007-009-MY3, 102-2633-M-007-002, and 103-2745-M-002-004-ASP and the National Tsing Hua University through Grant No. 102N2022E1. Y.L.C. greatly appreciates the use of the facility at CNMM, National Tsing Hua University, through Grant No. 102N2744E1. Funding for this work was also provided by the Department of Industrial Technology, Ministry of Economic Aﬀairs, Taiwan.

REFERENCES AND NOTES

1. Yang, J.; Luo, F.; Kao, T. S.; Li, X.; Ho, G. W.; Teng, J.; Luo, X.; Hong, M. Design and Fabrication of Broadband Ultralow Re-ﬂectivity Black Si Surfaces by Laser Micro/Nanoprocessing. Light Sci. Appl. 2014, 3, 185.

2. Kawakita, S.; Imaizumi, M.; Sumita, T.; Kushiya, K.; Ohshima, T.; Yamaguchi, M.; Matsuda, S.; Yoda, S.; Kamiya, T. Super Radiation Tolerance of CIGS Solar Cells Demonstrated in Space by MDS-1 Satellite. Proc. 3rd WCPEC 2003, 693–696. 3. del Cueto, J. A.; Rummel, S.; Kroposki, B.; Osterwald, C.; Anderberg, A. Stability of CIS/CIGS Modules at the Outdoor Test Facility Over Two Decades. Proc. 33th IEEE Photovoltaic Spec. Conf. 2008, 1–6.

4. Kim, D. I.; Park, S. W.; Park, S. R.; Baek, J. Y.; Yun, T. Y.; Han, H. J.; Jeon, C. W. Long-Term Eﬃciency Gain of Cu(In,Ga)Se2

Solar Cell. 39th IEEE Photovoltaic Spec. Conf. 2013, 1980– 1982.

5. Rockett, A.; Birkmire, R. W. CuInSe2for Photovoltaic

Appli-cations. J. Appl. Phys. 1991, 70, 81–97.

6. Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D. Solar Cell Eﬃciency Tables (Version 44). Prog. Photovolt: Res. Appl. 2014, 22, 701–710.

7. Jackson, P.; Hariskos, D.; Lotter, E.; Paetel, S.; Wuerz, R.; Menner, R.; Wischmann, W.; Powalla, M. New World Record Eﬃciency for Cu(In,Ga)Se2Thin-Film Solar Cells Beyond

20%. Prog. Photovoltics: Res. Appl. 2011, 19, 894–897. 8. Chang, J. C.; Chuang, C. C.; Guo, J. W.; Hsu, S. C.; Hsu, H. R.;

Wu, C. S.; Hsieh, T. P. An Investigation of CuInGaSe2Thin

Film Solar Cells by Using CuInGa Precursor. Nanosci. Nanotechnol. Lett. 2011, 3, 200–203.

(8)

9. Chen, S. C.; Hsieh, D. H.; Jiang, H.; Liao, Y. K.; Lai, F.-I.; Chen, C. H.; Luo, C. W.; Juang, J. Y.; Chueh, Y. L.; Wu, K. H.; et al. Growth and Characterization of Cu(In,Ga)Se2Thin Films by

Nanosecond and Femtosecond Pulsed Laser Deposition. Nanoscale Res. Lett. 2014, 9, 280.

10. Kapur, V. K.; Bansal, A.; Le, P.; Asensio, O. I. Non-Vacuum Processing of CuIn1‑xGaxSe2 Solar Cells on Rigid and

Flexible Substrates Using Nanoparticle Precursor Inks. Thin Solid Films 2003, 431 432, 53–57.

11. Calixto, M. E.; Dobson, K. D.; McCandless, B. E.; Birkmire, R. W. Controlling Growth Chemistry and Morphology of Single-Bath Electrodeposited Cu(In,Ga)Se2Thin Films for

Photovoltaic Application. J. Electrochem. Soc. 2006, 153, G521–G528.

12. Wei, S. H.; Zhang, S. B.; Zunger, A. Eﬀects of Ga Addition to CuInSe2on Its Electronic, Structural, and Defect Properties.

Appl. Phys. Lett. 1998, 72, 3199–3201.

13. Liu, C. H.; Chen, C. H.; Chen, S. Y.; Yen, Y. T.; Kuo, W. C.; Liao, Y. K.; Juang, J. Y.; Lai, C. H.; Chen, L. J.; Chueh, Y. L. Large Scale Single-Crystal Cu(In,Ga)Se2Nanotip Arrays for High

Eﬃciency Solar Cell. Nano Lett. 2011, 11, 4443–4448. 14. Igalson, M.; Zabierowski, P.; Przado, D.; Urbaniaka, A.;

Edoﬀ, M.; Shafarmanc, W. N. Understanding Defect-related Issues Limiting Eﬃciency of CIGS Solar Cells. Sol. Energy Mater. Sol. Cells 2009, 93, 1290–1295.

15. Ritchie, R. H. Plasma Losses by Fast Electrons in Thin Films. Phys. Rev. 1957, 106, 874–881.

16. Jain, P. K.; El-Sayed, M. A. Plasmonic Coupling in Noble Metal Nanostructures. Chem. Phys. Lett. 2010, 487, 153– 164.

17. Auguie, B.; Barnes, W. L. Collective Resonances in Gold Nanoparticle Arrays. Phys. Rev. Lett. 2008, 101, 143902. 18. Jensen, T. R.; Malinsky, M. D.; Haynes, C. L.; Van Duyne, R. P.

Nanosphere Lithography: Tunable Localized Surface Plas-mon Resonance Spectra of Silver Nanoparticles. J. Phys. Chem. B 2000, 104, 10549–10556.

19. Chan, G. H.; Zhao, J.; Hicks, E. M.; Schatz, G. C.; Van Duyne, R. P. Plasmonic Properties of Copper Nanoparticles Fabri-cated by Nanosphere Lithography. Nano Lett. 2007, 7, 1947–1952.

20. Hägglund, C.; Apell, S. P. Plasmonic Near-Field Absorbers for Ultrathin Solar Cells. J. Phys. Chem. Lett. 2012, 3, 1275– 1285.

21. Chen, W. T.; Wu, P. C.; Chen, C. J.; Weng, C. J.; Lee, H. C.; Yen, T. J.; Kuan, C. H.; Mansuripur, M.; Tsai, D. P. Manipulation of Multidimensional Plasmonic Spectra for Information Storage. Appl. Phys. Lett. 2011, 98, 171106.

22. Chen, W. T.; Yang, K. Y.; Wang, C. M.; Huang, Y. W.; Sun, G.; Chiang, I. D.; Liao, C. Y.; Hsu, W. L.; Lin, H. T.; Sun, S.; et al. High-Eﬃciency Broadband Meta-Hologram with Polarization-Controlled Dual Images. Nano Lett. 2014, 14, 225–230. 23. Jain, P. K.; Lee, K. S.; El-Sayed, I. H.; El-Sayed, M. A. Calculated

Absorption and Scattering Properties of Gold Nanoparti-cles of Diﬀerent Size, Shape, and Composition: Applications in Biological Imaging and Biomedicine. J. Phys. Chem. B 2006, 110, 7238–7248.

24. Pillai, S.; Catchpole, K. R.; Trupke, T.; Green, M. A. Surface Plasmon Enhanced Silicon Solar Cells. J. Appl. Phys. 2007, 101, 093105.

25. Paris, A.; Vaccari, A.; Calà Lesina, A.; Serra, E.; Calliari, L. Plasmonic Scattering by Metal Nanoparticles for Solar Cells. Plasmonics 2012, 7, 525–534.

26. Nakayama, K.; Tanabe, K.; Atwater, H. A. Plasmonic Nano-particle Enhanced Light Absorption in GaAs Solar Cells. Appl. Phys. Lett. 2008, 93, 121904.

27. Orgassa, K.; Rau, U.; Nguyen, Q.; Schock, H. W.; Werner, J. H. Role of the CdS Buﬀer Layer as an Active Optical Element in Cu(In,Ga)Se2Thin-Film Solar Cells. Prog. Photovol. Res.

Appl. 2002, 10, 457–463.

28. Govorov, A. O.; Zhang, W.; Skeini, T.; Richardson, H.; Lee, J.; Kotov, N. A. Gold Nanoparticle Ensembles as Heaters and Actuators: Melting and Collective Plasmon Resonances. Nanoscale Res. Lett. 2006, 1, 84–90.

29. Richardson, H. H.; Hickman, Z. N.; Govorov, A. O.; Thomas, A. C.; Zhang, W.; Zhang, M. E. Thermooptical Properties of

Gold Nanoparticles Embedded in Ice: Characterization of Heat Generation and Melting. Nano Lett. 2006, 6, 783–788. 30. Bohren, C. F.; Huﬀman, D. R. Absorption and Scattering of

Light by Small Particles; Wiley: New York, 2008.

31. Mertz, J. Radiative Absorption, Fluorescence, and Scatter-ing of a Classical Dipole Near a Lossless Interface: A Uniﬁed Description. J. Opt. Soc. Am. B 2000, 17, 1906–1913. 32. Atwater, H. A.; Polman, A. Plasmonics for Improved

Photo-voltaic Devices. Nat. Mater. 2010, 9, 205–213.

33. Zhu, J.; Hsu, C. M.; Yu, Z.; Fan, S.; Cui, Y. Nanodome Solar Cells with Eﬃcient Light Management and Self-Cleaning. Nano Lett. 2010, 10, 1979–1984.

34. Ferry, V. E.; Munday, J. N.; Atwater, H. A. Design Considera-tions for Plasmonic Photovoltaics. Adv. Mater. 2010, 22, 4794–4808.

35. Sze, S. M.; Ng, K. K. Physics of Semiconductor Devices; John Wiley & Sons: New York, 2007.

36. Reinhard, P.; Chirila, A.; Blosch, P.; Pianezzi, F.; Nishiwaki, S.; Buecheler, S.; Tiwari, A. N. Review of Progress Toward 20% Eﬃciency Flexible CIGS Solar Cells and Manufacturing Issues of Solar Modules. IEEE J. Photovolt. 2013, 3, 572–580. 37. Chen, M. W.; Chau, Y. F.; Tsai, D. P. Three-Dimensional Analysis of Scattering Field Interactions and Surface Plas-mon Resonance in Coupled Silver Nanospheres. PlasPlas-monics 2008, 3, 157–164.

38. Chau, Y. F.; Chen, M. W.; Tsai, D. P. Three-dimensional Analysis of Surface Plasmon Resonance Modes on a Gold Nanorod. Appl. Opt. 2009, 48, 617–622.

39. Su, K.-H.; Wei, Q.-H.; Zhang, X.; Mock, J. J.; Smith, D. R.; Schultz, S. Interparticle Coupling Eﬀects on Plasmon Re-sonances of Nanogold Particles. Nano Lett. 2003, 3, 1087– 1090.

40. Chen, S. C.; Liao, Y. K.; Chen, H. J.; Chen, C. H.; Lai, C. H.; Chueh, Y. L.; Kuo, H. C.; Wu, K. H.; Juang, J. Y.; Cheng, S. J.; et al. Ultrafast Carrier Dynamics in Cu(In,Ga)Se2Thin Films Probed

by Femtosecond Pump-Probe Spectroscopy. Opt. Express 2012, 20, 12675–12681.

41. Othonos, A. Probing Ultrafast Carrier and Phonon Dy-namics in Semiconductors. J. Appl. Phys. 1998, 83, 1789– 1830.

42. Wilke, R. I.; Cho, S.; Lu, H.; Schaﬀ, W. J. Ultrafast Recombina-tion in Si-doped InN. Appl. Phys. Lett. 2006, 88, 112111. 43. Liu, J.; Yabushita, A.; Taniguchi, S.; Chosrowjan, H.; Imamoto,

Y.; Sueda, K.; Miyanaga, N.; Kobayashi, T. Ultrafast Time-Resolved Pump Probe Spectroscopy of PYP by a Sub-8 fs Pulse Laser at 400 nm. J. Phys. Chem. B 2013, 117, 4818– 4826.