DSpace at VNU: Low defect interface study of intrinsic layer for c-Si surface passivation in a-Si:H c-Si heterojunction solar cells

Low defect interface study of intrinsic layer for c-Si surface passivation in a-Si:H/c-Si heterojunction solar cells Sangho Kima, Vinh Ai Daob,c, Chonghoon Shina, Jaehyun Chob, Youngseok

Low defect interface study of intrinsic layer for c-Si surface passivation in a-Si:H/c-Si heterojunction solar cells

Sangho Kima, Vinh Ai Daob,c, Chonghoon Shina, Jaehyun Chob, Youngseok Leea, Nagarajan Balajia, Shihyun Ahnb, Youngkuk Kimb, Junsin Yia,b,⁎

a

Department of Energy Science, Sungkyunkwan University, Suwon, 440-746, Republic of Korea

b

School of Information and Communication Engineering, Sungkyunkwan University, Suwon, 440-746, Republic of Korea

c

Faculty of Materials Science, College of Science, Vietnam National University, Hochiminh, Viet Nam

a b s t r a c t

a r t i c l e i n f o

Available online 27 March 2012

Keywords:

Passivation

Spectroscopy ellipsometry

Heterojunction solar cells

Fraction depletion

VHF-PECVD

High quality hydrogenated intrinsic amorphous silicon [a-Si:H(i)] layer with adequate hydrogen content and lesser void fraction is the key to obtaining good surface passivated crystalline silicon (c-Si), with high open-circuit voltage (Voc), which will ultimately make the heterojunction with intrinsic thin layer (HIT) solar cell highly efficient In this study, we performed good surface passivation of a HIT solar cell by depositing a-Si: H(i) layers at different working pressures from 26.7 to 107 Pa by using very high frequency of 60 MHz plasma-enhanced chemical vapor deposition Based on spectroscopic ellipsometry and gas depletion analysis,

we discuss the influence of the working pressure on the deposition mechanism, interface passivation and ul-timately cell efficiency Highest minority lifetime of about 4 ms was achieved at the highest working pressure

of 107 Pa The decrease in working pressure results in less denser and/or incorporation of epitaxy layer inside the a-Si:H(i)films, and leads to decrease in c-Si surface passivation The performance of heterojunction solar cell device was improved with the increase of working pressure and the best photo voltage parameters of the device were found to be Vocof 647 mV, short-circuit current density of 32.28 mA/cm2and efficiency of 15.57%

at working pressure of 107 Pa

© 2012 Elsevier B.V All rights reserved

1 Introduction

Heterojunction with intrinsic thin-layer (HIT) solar cells, developed

by Sanyo Ltd in 1994, offers low-cost fabrication for high-efficiency

solar cells compared to crystalline silicon (c-Si) solar cell with diffused

p–n junctions[1] The world record efficiency of 23% fabricated base

on HIT structures, in which stacks of hydrogenated intrinsic amorphous

silicon [a-Si:H(i)] and doped a-Si:H layers help to form both the emitter

and the back surfacefield, is also held by Sanyo group[2] The a-Si:H(i)

layers, which have high amorphicity, adequate hydrogen content and

low void fraction, enables to suppress surface recombination at the

a-Si:H/c-Si heterointerface, to ensures high cell performance[3]

Deposi-tion condiDeposi-tion, however, strongly affects the structural and physical

properties of the thin a-Si:H(i) layer Completely amorphous silicon

can be obtained at low temperatures (~100 °C), but, it is believed that

an ultra-thin layer on c-Si surface is epitaxial Again, hydrogen dilution

is reported to be a key deposition parameter that controlsfilm quality

and phase[1] During deposition, working pressure could also affect

the properties of materials deposited by plasma-enhance vapor

deposi-tion via silane depledeposi-tion[4] For good surface passivation, as

well-known, a low order silane radical such as SiH3is preferred, which can

be obtained by depositing of a-Si:H(i)films at a high working pressure

as well as silane depletion In spite of the effectiveness of the deposition parameter, fewer studies have focused on the optimization of the deposition condition, especially working pressure, for the excellent properties of a-Si:H(i), Thus, the present paper focuses on such an optimization

Moreover, to obtain high performance devices, it is importance to monitor the properties of the a-Si:H(i) layers This monitoring is usually difficult for heterojunction solar cells, since the thickness of the a-Si:H(i) layer is of the order of nano-scale[3,5] Owing to the measurement of phase, spectroscopic ellipsometry (SE) is sensitive

to subnanometer changes and the precision of measurements is suf fi-cient to allow determination of dielectric function information at the shorter wavelengths from few nanolayers[6] The measured data of the dielectric function of a-Si:H(i) wasfirst simulated by using the tetrahedron model in combination with the Bruggeman effective me-dium approximation The hydrogen content, crystalline fraction and void fraction of thefilm were determined This method also suggests

a way to obtain good quality passivation a-Si:H(i) layer with informa-tion of interface property such asfilm density and included defects in thefilm

In this paper, the fabrication of heterojunction solar cell with a dif-ferent working pressure during the deposition of a-Si:H(i) layer is

⁎ zCorresponding author at: Department of Energy Science, Sungkyunkwan University,

Suwon, 440-746, Republic of Korea Tel.: +82 31 290 7139; fax: +82 31 290 7159.

Contents lists available atSciVerse ScienceDirect

Thin Solid Films

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / t s f

reported Spectroscopic ellipsometry studies and fractional depletion

analysis have been utilized as a tool to optimize the a-Si:H(i) layer

property

2 Experimental details

The solar cell devices fabricated for this experiment consist of a

with a specular surface on the front side and an unpolished surface

on the rear side that is treated by a sequence of, acetone/methanol/

deionized water, RCA Immediately before a-Si:H deposition, native

oxide was removed from the wafer by dipping the wafer into 1%

hydrofluoric acid for 1 min Then, the a-Si:H(i) layer was deposited

at various working pressures on n-type c-Si by 60 MHz very high

fre-quency plasma-enhanced chemical vapor deposition (VHF-PECVD) in

order to investigate the effect of c-Si wafer passivation and on solar

cell performance The details of the deposition condition are provided

inTable 1 The VHF-PECVD system had a load-lock system and three

different chambers for the respective depositions of p, n and intrinsic

i layers The deposition temperature of VHF-PECVD was 200 °C Radio

frequency magnetron sputtering was used to deposit the Indium Tin

Oxide (ITO) thinfilms at substrate temperature of 200 °C The ITO

films thickness was fixed at 80±5 nm The deposition of

aluminu-m(Al)/silver(Ag)finger as the emitter contacts has been performed

using the evaporation system Al was evaporated on backside prior

to area defining with mesa etching Our final HIT cell structure is

con-structed as following: (Al/Ag)/ ITO/a-Si:H(p)-7 nm/a-Si:H(i)-6.5 nm/

c-Si (n)/a-Si:H(i)-6.5 nm/a-Si:H(n)-10 nm/Al

All the depositedfilms were measured by using SE (VASE®, J A

Woollam, 240 nmbλb1700 nm) at room temperature The dielectric

function, thefilm thickness and the band gap energy were obtained

by using Tauc–Lorentz model In order to verify the SE analyses,

cross-section of the transmission electron microscopy (TEM)

micro-graphs was acquired by a JEM-ARM200F-JEOL system operated at an

acceleration voltage of 200 kV The samples were prepared by

me-chanical polishing followed by 4 keV Ar ion milling using a Gatan

PIPS ion miller The silane fraction depletion (F) is defined as

F=(Po−P)/Po, wherePoandP are the partial pressures of the silane

with the plasma turn off and on respectively To measureF during a

film deposition, the plasma was first turned on with the required

very high frequency input power, and the working pressure After

the plasma stabilized the partial pressure of the silane (P) was

measured The plasma was then turned off while the computer

continued to monitor the Po [7,8] In order to determine the

heterointerface quality, the effective lifetime (τeff) was measured by

the quasi-steady-state Photoconductance (QSSPC) method, using a

commercial Photo conductance set-up from Sinton Consulting

(WCT-120) The quantum efficiency (QE) of the heterojunction solar

cells was measured by using a xenon lamp, a monochromator, and

opticalfilters, which filtered out the high orders with a light probe

beam impinging normal on the samples Solar cell performance was

characterized by current–voltage measurements under illuminated

AM1.5, 100 mW/cm2conditions

3 Result and discussion

Fig 1shows the effect of the a-Si:H(i) layer deposited at different working pressures on the real and imaginary parts of the pseudo-dielectric function, in which the pseudo-dielectric function of thefilms is obtained multiplying the equation of the Lorentz oscillator by the equation of the Tauc join density of states[9]:

εimTLð Þ ¼E A:Eo:C: E−Eg

E2−Eg

where E0is the peak transition energy, Egis the band gap energy and

C is a broadening parameter which can be related to the degree of dis-order in the material A is another parameter which is proportional to the height ofbεim,max> that is related to thefilm density At the low-est working pressure of 26.7 Pa, peaks started to appear near 3.4 eV and 4.2 eV, indicative of partial epitaxy growth With further increase

in working pressure, a smooth, single-peaked dielectric function of amorphous silicon is observed at around 3.7 eV and the amplitude

of A gradually increases, indicating the increase of density of the films[9] On the other hand, the minimum amplitude ofbεreal,min>, which is related to the void fraction in thefilm, decreases with the in-crease of working pressure, particularly around 4.75 eV[5] Therefore, high quality a-Si:H(i) can be obtained for c-Si surface passivation layer by increasing the working pressure because the observed in-crease and dein-crease ofbεim,max> andbεreal,min>, respectively, led to low percentage void fraction inside the material with adequate hy-drogen content[9]

In order to evaluate the validity of our SE analysis, the TEM image

of the a-Si:H/c-Si heterointerface was performed for the samples with the lowest and the highest working pressure As shown inFig 2, the film quality from the TEM image shows excellent agreement with the SE result FromFig 2a, it can be seen that the epitaxial growth occurs at lowest working pressure of 26.7 Pa, encompassing the in-trinsic layer (~5 nm) and extending well into the p-layer For the higher working pressure of 107 Pa, abrupt amorphous-silicon growth

is observed as clearly shown inFig 2b

Surface recombination velocity (Seff) was estimated from theτeff We assume that the device structure is symmetrical (Seff= Sfront= Sback) The Seffwas estimated by the equation

Seff¼w2 τ1

effþτ1

b

!

ð3Þ

whereτbis the recombination lifetime in the bulk and W is the wafer thickness The uncertainty of Seffdepends on the value used forτb The Shockley–Read–Hall recombination was not considered

to calculate the upper limit of the surface recombination velocity So

τbhas an unlimited value Theτeffand Seffas a function of working pressure are depicted inFig 3 The minimum value of Seffwhich

Table 1

Deposition condition of hydrogenated intrinsic amorphous silicon layer in this study.

(Pa)

Gas ratio

Temp.

(°C)

Electrode distance (mm)

Power (W)

Deposition time (s)

Thickness (nm)

Deposition rate (nm/min)

corresponds to the highest value of effective lifetime, was determined

from QSSPC measurements, and is in agreement with the result of

Olibet et al.[10].Fig 3shows that the lowest effective lifetime is

exhibited by the sample deposited at the lowest working pressure

of 26.7 Pa As reported by Wang et al with incorporation of the

epi-taxy on c-Si surface, it extends through the a-Si:H(i) layer, so the

de-fect interface will come in contact with the emitter layer, leading to

lesser passivation than that of the pure a-Si:H(i) layer[11] Theτeff

linearly increased with increasing working pressure, because of the

lower void fraction inside thefilm with adequate hydrogen content,

at higher working pressures[5]

The plasma diagnostics and growth mechanism were analyzed

quantitatively to provide further insight into how the a-Si:H(i)

prop-erties, and subsequently, the surface passivation change with

work-ing pressure Fig 4 shows the silane fractional depletion and

deposition rate as a function of the working pressure We found

that the fractional depletion is increased with the increase of the

working pressure while the deposition rate is decreased This

oppo-site trend can be attributed to the hydrogen effect with changing

working pressure During the growth offilms, atomic hydrogen can

move in or out of the surface The hydrogen can react with the silicon

network in the subsurface region after the silicon has attached to the

surface The hydrogen, then, terminates dangling bonds and removes

weak bonds while excess hydrogen is evolved from thefilm[12] With increasing working pressure resulting in increasing silane de-pletion, the atomic hydrogen, thus, dominates thefilm This phenom-enon leads to the decomposition of H2resulting in hydrogen coverage

of the growing film surface, and then enhances the precursors absorbed at the growing surface to have enough time tofind their en-ergetically suitable sites that decreased the structural disorder in the films as well as the deposition rate[13,14] Therefore, with increasing working pressure, the silane depletion is more accomplished, leading

to denserfilms with moderating hydrogen content, which is a merit property for c-Si surface passivation[5]

The performance of the solar cell devices is indicated inFig 5 This figure shows an apparent increase in both open-circuit voltage (Voc) and short-circuits current density (Jsc) with the increase in working pressures The increase of Vocand Jsc, consequently, leads to an in-crease in solar cell efficiency from 11.95 to 15.57% with increasing working pressure from 26.7 Pa to 107 Pa As mentioned earlier, the lowest Vocof the sample deposited at the lowest working pressure could be attributed to the incorporation of epitaxy into the a-Si:H(i) layer, resulting in lesser surface passivation of the a-Si:H/c-S hetero-interface With further increase in working pressure, however, the better a-Si:H(i)film properties with adequate hydrogen content and less void fraction lead to longer effective carrier lifetime of the

0

10

20

30

-10 0 10 20 30

εrea

εim

Photon Energy (eV)

26.7 Pa 53.3 Pa 80.0 Pa 107.0 Pa

deduced from the Tauc–Lorentz analysis of the experiment data.

0.1 1 10

Working pressure (Pa)

τeff

10 0

10 1

10 2

S eff

passivated wafers, which in turn leads to higher open-circuit voltage

of the HIT solar cell

To elucidate the reason for the increase in Jsc with increasing

working pressure, the QE of the four different cells was characterized

and shown inFig 6 The solar cells fabricated with lower working

pressure indicate better spectral response at shorter wavelengths of

about 400 nm, especially at 26.7 Pa In this case, the a-Si:H(i) layer

may be thinner than the other due to the growth of the epitaxial on

the c-Si surface This thinness leads to lower light absorption,

result-ing in better in QE at shorter wavelengths Solar cell devices, however,

fabricated at higher working pressures indicate better spectral

re-sponse above 500 nm wavelength.Fig 7shows the energy band gap

of the a-Si:H layer as a function of working pressure The a-Si:H(i)

layer deposited at higher pressures show the higher energy band

gap, which leads to more light absorption by the wafer and the

crea-tion of more electron hole pairs Hence, the higher spectral response

at wavelengths around 500 to 600 nm may be expected On the

back surface, in addition, the higher energy band gap of the a-Si:

H(i) layer deposited at higher working pressure leads to the larger

va-lence band offset, and this provides a back surface“mirror” for holes,

however does not much hinder electron transport due to the

small-ness of the offset in the conduction band edges[15] Thus, both higher

energy band gap and merit properties for c-Si surface passivation of

the a-Si:H(i) layer deposited at higher working pressure conduct to

a better back-surface passivation, thus, in turn improved QE in the

longer wavelength; which is in accord with Yamanaka et al.[16]

4 Conclusions

In summary, the effect of working pressure on a-Si:H/c-Si inter-face passivation during deposition of the a-Si:H(i) layer on a-Si: H(p)/c-Si(n) interface passivation and hence on device performance was investigated by spectroscopic ellipsometry as well as by fraction depletion analysis From SE, we found the formation of mixing phase, epitaxy and hydrogenated amorphous silicon, on the c-Si surface at the low working pressure of 26.7 Pa This incorporation of the epitax-ial layer in a-Si:H(i) led to the suppression of the surface passivation

of the c-Si surface, and hence to the reduction of Voc, and also cell per-formance For working pressures above 26.7 Pa, the silane depletion was increased, which led to denserfilms with adequate hydrogen content, which resulted in the longer lifetime of the passivation of the c-Si surface Highest minority lifetime of about 4 ms was achieved

at the highest working pressure of 107 Pa At the optimum condition,

we obtained a conversion efficiency of 15.57% Thus, the high quality

of the a-Si:H(i) layer with low defects and low void fraction in the film is very important for the a-Si:H/c-Si heterointerface, one of the major factors to improve the junction properties and enable the high Vocand also enhance cell efficiency

Acknowledgments This work was supported by the New & Renewable Energy of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Knowledge Economy (No 20113010010100)

5

6

7

8

9

10

10 15 20 25 30

Working pressure (Pa)

Fig 4 Fractional depletion and deposition rate as a function of working pressure.

27

30

33

585

650

12

14

16

Working pressure (Pa)

J sc

2 )

V oc

effi-η, measured on heterojunction solar cells as function of working pressure.

20 30 40 50 60 70 80

Wavelength (nm)

Working pressure 26.7 Pa 53.3 Pa 80.0 Pa 107.0 Pa

Fig 6 The QE curves of HIT solar cells as function of working pressure.

1.80 1.85 1.90 1.95

E g

Working pressure (Pa)

[1] V.A Dao, J Heo, H Choi, Y Kim, S Park, S Jung, N Lakshminarayan, J Yi, Solar

En-ergy 84 (2010) 777.

[2] T Mishima, M Taguchi, H Sakata, E Maruyama, Sol Energy Mater Sol Cells 95

(2011) 18.

[3] H Fujiwara, M Kondo, Appl Phys Lett 90 (2007) 013503.

[4] A Descoeudres, L Barraud, R Barlome, C Ballif, Appl Phys Lett 97 (2010)

183505.

[5] V.A Dao, Y.S Lee, S.H Kim, J Yi, J Electrochem Soc 158 (2011) H312.

[6] C Pickering, Surf Interface Anal 31 (2001) 927.

[7] J.R Doyle, Y Xu, R Reedy, A.H Mahan, Thin Solid Films 516 (2008) 526.

[8] B Strahm, A.A Howling, L Sansonnens, C Hollenstein, Plasma Source Sci

Tech-nol 16 (2007) 80.

[9] A Fontcuberta i Morral, P Roca i Cabarrocas, Phys Rev B 69 (2004) 125307.

[10] S Olibet, E.V Sauvain, C Ballif, L Fesqust (Eds.), Proceedings of

2007, p 141.

[11] T.H Wang, E Iwaniczko, M.R Page, D.H Levi, Y Yan, H.M Branz, Q Wang, 31st IEEE Photovoltaics Specialists Conf and Exhibition, Florida, USA, 2005, p 1 [12] R.A Street, Hydrogenated Amorphous Silicon, Cambridge University Press, New York, 1991.

[13] Q.S Lei, Int J Mod Phys B 20 (2006) 2035.

[14] W.M.M Kessels, A.H.M Smets, D.C Marra, M.C.M van de Sanden, Thin Solid Films

383 (2011) 154.

[15] T.H Wang, M.R Page, E Iwaniczko, D.H Levi, Y Yan, H.M Branz, Q Wang, V Yelundur, A Rohatgi, 14th Workshop on Crystalline Silicon Solar Cells and Mod-ules, Colorado, USA, 2004, p 1.

[16] M Yamanaka, I Sakata, R Shimokawa, H Takato, Conf Record of the 2006 IEEE 4th World Conf on Photovoltaic Energy Conversion, Hawaii, USA, 2006, p 1421.