Role of an electrolyte and substrate on the stability of porous silicon

Đây là một bài báo khoa học về dây nano silic trong lĩnh vực nghiên cứu công nghệ nano dành cho những người nghiên cứu sâu về vật lý và khoa học vật liệu.Tài liệu có thể dùng tham khảo cho sinh viên các nghành vật lý và công nghệ có đam mê về khoa học

Trang 1

Physica E 28 (2005) 264–272

Role of an electrolyte and substrate on the stability

of porous silicon Shailesh N Sharma , R.K Sharma, S.T Lakshmikumar Materials Division, National Physical Laboratory, Dr K S Krishnan Marg, New Delhi-110012, India

Received 14 March 2005; accepted 21 March 2005

Available online 6 June 2005

Abstract

Porous silicon (PS) layers were prepared by anodization on polished and textured substrates of (1 0 0) Si for a fixed anodization time at different current densities in different HF-based electrolytes Highly stable, mechanically strong, hydrogen-passivated surface and thick porous silicon films have been obtained using HF:ethanol-based electrolyte on textured silicon substrates Porous silicon formed using HF:ethanol as an electrolyte exhibits superior properties compared to porous silicon formed using HF:H2O2-based electrolyte at the same current density, time of anodization and type of substrate Porous silicon films formed on textured substrates exhibits higher porosity and photoluminescence efficiency, negligible PL decay, better mechanical strength, adherence to the substrate, non-fractured surface morphology and lower stress compared to porous silicon formed on polished silicon substrates at the same current density for both ethanol and H2O2-based electrolytes, respectively Use of textured silicon substrate and ethanol-based electrolyte is a key parameter for the formation of tailored-made porous silicon films for device applications

PACS: 61.43.Gt; 81.05.Rm; 82.45.Gj

Keywords: Porous silicon layers; HF-electrolytes: Si substrates

1 Introduction

Porous silicon (PS) exhibits visible

photolumi-nescence and electrolumiphotolumi-nescence which has

gen-erated considerable interest [1] The potential of

porous silicon for various technological applica-tions such as chemical sensors [2], optoelectronic devices [3], displays[4]and photodetectors[5]has been extensively investigated Recent emphasis has been on the utilization of the large surface area of the porous layers for chemical and biological applications[6] It is possible to control the degree

of porosity of the porous layers formed by electro-chemical etching in HF-containing electrolytes (ethanol, hydrogen peroxide, etc.) However, the

www.elsevier.com/locate/physe

doi:10.1016/j.physe.2005.03.020

Corresponding author Tel.: 91 11 25742609 14x2409;

fax: 91 11 25726938, 25726952.

E-mail address: shailesh@mail.nplindia.ernet.in

(S.N Sharma).

Trang 2

nanoscale structure of PS leads to an enormous

increase in surface area and the presence of large

number of unpaired bonds at the surface which

alter the surface recombination rates and

conse-quently the PL efﬁciency, surface reactivity and

stability[7] Several approaches have been tried for

preparing uniformly bonded stable surfaces The

formation of a high-quality oxide surface layer is

now accepted as a good solution to the formation

of a stable surface and improved luminescent

silicon particles in an optically transparent

med-ium is another way of isolating the surface from

the ambient and providing a stable luminescence

[9] Recently, the use of alkyl-terminated

mono-layers as a mean of stabilizing the PS surface has

received attention where Si–H bonds at the surface

during PS formation are replaced by a hydrophilic

alkyl termination[10]

The electrolyte composition is one of the most

important fabrication parameter for well-deﬁned

porous layers The pore dimensions and porosity

change with different ratios of electrolytes

Var-ious electrolytes have been used for the fabrication

of porous silicon viz, HF, ethanol, H2O2 and

HNO3 [1,11,12] HF is mainly used for the

dissolution of silicon, ethanol is basically used to

reduce the surface tension of the electrolytic

mixture since surface wetting is important for

good pore uniformity Recently thrust has been

given on H2O2-based electrolytes preferably as an

oxidizing agent [12] The photochemical etching

method with H2O2 solution does not generate a

toxic material unlike in the case of HNO3 [11]

Moreover, the addition of H2O2 to the etching

mixture raises the pH of the solution and produces

ideal Si surfaces terminated with Si–H bonds thus

resulting in a homogeneous PS surface with low

defect density[12]

Recently, we have demonstrated by means of

high-resolution XRD studies that texturization of

silicon surface is an effective method for the

formation of stable and thick porous silicon ﬁlms

[13] In this paper, using PL decay as a probe, we

are evaluating the degradation of stability of PS on

electrolyte (HF–C2H5OH and HF–H2O2) and

current density formed on textured and polished

Si substrates, respectively The emphasis is mainly

on the development of PS with high and stable PL, control of pore size distribution and therefore a better control on the formation process

2 Experimental

Boron-doped p-type Si wafers of (1 0 0) orienta-tion, 8–10 ohm cm resistivity and 400 mm thickness were used for preparing PS The wafers were polished in 40% NaOH for 2 min These wafers were textured using 2% NaOH at 85 1C for 30 min For forming the back contact, Ag–Al paste was screen printed on the wafer and dried at 250 1C The wafer was then heated to 750 1C for 2 min in

an IR furnace PS was formed by the standard anodization process using Si as the anode and Pt as the counter electrode in an acid resistant container The anodization was carried out at 20–50 mA cm2 for 30 min, in two different electrolytes The ﬁrst is

which is almost universally used [1]and would be abbreviated as electrolyte A The second is a mixture of HF and H2O2 (1:1 by volume) which was extensively used by Nafeh et al.[12]and would

be abbreviated as electrolyte B After the anodiza-tion, the ﬁlms were washed in deionized water and ethanol and dried in nitrogen The samples were subjected to continuous agitation in an ultrasonic cleaner to evaluate the speed with which the sample

is destroyed The weight of the sample is con-tinuously monitored The PL was measured using a home assembled system consisting of a two-stage monochromator, a photomultiplier tube (PMT) with a lock-in ampliﬁer for PL detection, and an

(corresponding to 0.125 W cm2) for excitation in all the measurements Decay of PL intensity has been used as a measure of the stability of the

studies, the sample was continuously exposed to the laser radiation and PL measurements were carried out at regular intervals

3 Results and discussion

Good porous silicon ﬁlms exhibiting high photoluminescence intensity could be formed on

Trang 3

both textured and polished substrates at various

current densities corresponding to both

electro-lytes A and B, respectively The porosity (45–80%)

and thickness (12–96 mm) of PS ﬁlms were

Fig 1 shows porosity values as a function of Id

for PS ﬁlms formed on textured and polished

substrates corresponding to both electrolytes A

and B, respectively As shown inFig 1, porosity of

PS ﬁlms increases with increase in current density

As evident fromFig 1, PS ﬁlms corresponding to

electrolyte B exhibits higher porosity as compared

to the corresponding ﬁlms of electrolyte A for both

textured and polished substrates

Fig 2(A) shows the weight loss of PS ﬁlms

prepared using electrolyte A at different Id, as a

function of time of ultrasonic treatment There is a

substantial weight loss of PS samples on polished

substrates when subjected to an ultrasonic

treat-ment for an hour by which time the entire porous

layer has been removed and the loss of weight

saturates However, for textured PS ﬁlms, the

weight loss is marginal The rate of weight loss

increases with increase in Id and this effect is felt

more on PS ﬁlms prepared on polished substrates

Results of weight loss for PS ﬁlms prepared

using electrolyte B are shown inFig 2(B) In this

case some loss is observed for textured samples also However, the rate of weight loss increases with Id and is much higher for the untextured samples (Fig 2(B))

Typical PL curves for PS ﬁlms formed at

50 mA cm2) on textured and polished substrates corresponding to electrolytes A and B are shown

in Figs 3(A) and (B) As evident from Figs 3(A) and (B), the absolute PL intensity is higher for the porous silicon formed on textured substrates and for PS ﬁlms corresponding to electrolyte B owing

40

50

60

70

80

(d) (c) (b) (a)

Fig 1 Porosity of PS as a function of current density (I d ); (a)

textured substrate, electrolyte B; (b) polished substrate,

electrolyte B; (c) textured substrate, electrolyte A; (d) polished

substrate, electrolyte A.

0.3450 0.3455 0.3460 0.3465 0.3470 0.3475 0.3480 0.3485

(c)

(e)

(f)

(b)

(d)

(a)

Time of Ultrasonic treatment (mins.)

0.3450 0.3455 0.3460 0.3465 0.3470 0.3475 0.3480 0.3485 0.3490

(b)

(d) (e)

(f)

(a) (c)

Time of ultrasonic treatment (mins.)

(A)

(B)

Fig 2 Weight loss of porous silicon samples prepared at different current densities (I d ) for (A) electrolyte A and (B) electrolyte B; (a) textured substrate, I d ¼ 20 mA cm 2 ; (b) polished substrate, I d ¼ 20 mA cm 2 ; (c) textured substrate,

I d ¼ 35 mA cm 2 ; (d) polished substrate, I d ¼ 35 mA cm 2 ; (e) textured substrate, I d ¼ 50 mA cm 2 and (f) polished substrate,

I d ¼ 50 mA cm 2

Trang 4

to its higher porosity Fig 3(A) shows that with

increase in Id from 20 to 50 mA cm2for

electro-lyte A, the PL peak position shifts towards low-l

side for PS ﬁlms formed on both textured and

polished substrates Similarly, for PS samples

corresponding to electrolyte B, the blue-shift of

the PL peak position is more prominent with

the PL peak being at 650 nm as compared to

610 nm for PS ﬁlms prepared on textured

sub-strates corresponding to electrolyte A at higher

Id50 mA cm2 (Fig 3(B)) This trend is quite prominent for PS ﬁlms formed on textured substrates as compared to the corresponding ﬁlms formed on polished substrates for both electrolytes

A and B, respectively (Figs 3(A) and (B)) These results are in accordance with quantum conﬁne-ment effects[1] It is known that the peak position

of the PL intensity is blue shifted when HF-H2O2

is used as the electrolyte [15] A marginal shift in

PL peak position towards low l side is also observed upon texturization (Figs 3(A) and (B)) Visual observation shows that the porous silicon ﬁlms corresponding to electrolyte A formed on textured surfaces appear more uniform and strong

as compared to the corresponding films prepared using electrolyte B The PS films at higher current densities (IdX35 mA cm2) on polished substrates shows a break off in PL curves as these films are powdery in nature and hence unstable corre-sponding to both electrolytes A and B PS films prepared using B are more powdery in nature and shows peeling-off tendency particularly for films prepared on polished substrates This is even more obvious for films formed at higher Id

(X50 mA cm2)

Decay of PL intensity is a good indication of the stability of porous silicon particularly of the surface bond configurations [3,16] In Fig 4(A), decay of the PL intensity at the peak wave-length due to exposure to the laser radiation for porous silicon films formed at different Id ¼ ð20250 mA cm2Þon textured and polished silicon substrates for electrolyte A are compared Simi-larly, the corresponding PL-decay curves for electrolyte B are shown inFig 4(B) The PL peak position was recorded for different times corre-sponding to a fixed wavelength As shown inFigs 4(A) and (B), significant decay of the PL intensity

is observed for PS ﬁlms formed on polished substrate and the rate of decay increases with increase in Id This is observed for both A and B-based electrolytes with the rate of PL decay being higher for electrolyte B as compared to electrolyte

A at all current densities However, for PS ﬁlms formed on textured silicon, no PL decay was observed when ethanol was used as an electrolyte and a very marginal decay was noted when H2O2 -based electrolyte is used (Figs 4(A) and (B)) To

0

1

2

3

4

5

6

7

(a)

(f)

(e)

(d) (c) (b)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

(d)

(c) (e)

(f)

(a)

(b)

Wavelength (nm)

(A)

(B)

Fig 3 PL spectra of porous silicon samples prepared at

different current densities (I d ) for (A) electrolyte A and (B)

electrolyte B; (a) textured substrate, Id¼ 20 mA cm 2 ; (b)

polished substrate, Id¼ 20 mA cm 2 ; (c) textured substrate,

Id¼ 35 mA cm 2 ; (d) Polished substrate, Id¼ 35 mA cm 2 (e)

textured substrate, Id¼ 50 mA cm 2 and (f) polished substrate,

Id¼ 50 mA cm 2

Trang 5

ensure the reproducibility of this PL decay,

measurements were done repeatedly and for

several hours and the PL decay trend was found

to be the same This is a direct evidence for the

formation of stable surface and correlates with the

superior mechanical stability of porous silicon

formed on textured substrates

SEM was used to identify the surface

morphol-ogy of the porous silicon formed on textured and

polished Si-substrates at different current den-sities for electrolytes A and B, respectively Silicon nanowires are not visible at these magniﬁca-tions Figs 5 (A) and (B) show the surface of porous silicon formed on polished silicon at

Id10 mA cm2 corresponding to electrolytes A and B, respectively A plain featureless surface morphology is observed at Id10 mA cm2 for electrolyte A while a cracked surface morphology

is obtained for electrolyte B for the same current density Similar observations on the fragility of thick and highly porous ﬁlms had been noted earlier [8,17] For electrolyte A-based samples at lower Id, lack of cracking indicates lower stress while the corresponding electrolyte B-based sam-ple exhibits higher stress At Id ¼35 mA cm2, distinct cracking and disintegration is observed for

PS ﬁlms formed on polished substrates for both electrolytes A and B with the cracking being more pronounced for the latter than for the former (Figs 5(C) and (D)) The higher current density results in increased porosity and the inability of the silicon nanowires to withstand the stress leads

to cracking

The surface morphology of PS ﬁlms formed on textured substrates is signiﬁcantly different as compared to polished substrates Figs 6(A) and

35 mA cm2 corresponding to electrolytes A and

B, respectively Here, the smooth surface morphol-ogy consists of randomly sized and spaced pyramids homogeneously distributed on the sur-face The pyramids appear to be more sharply separated but no macroscopic cracking is observed even for electrolyte B-based sample unlike in the case of PS ﬁlm formed polished silicon substrate for the same current density (Figs 6(A) and (B)) This surface morphology does not essentially differ from the textured silicon substrate (not shown) and is not affected by current density On polished silicon substrates, PS layers showed a tendency to have a mechanically weak structure at higher current densities (Id50 mA cm2) owing to its higher porosity resulting in many cracks or peeling off the ﬁlm from the substrate This effect

is more prominent for electrolyte B-based samples than for electrolyte A-based samples However,

0.0

0.8

1.6

2.4

3.2

4.0

(d)

(c) (b) (a)

Time (mins)

0

1

2

3

4

5

6

(f)

(b) (e)

(d)

(c) (a)

Time (mins.)

(A)

(B)

Fig 4 PL decay of porous silicon samples prepared at different

current densities (I d ) as a function of time of laser exposure for

(A) electrolyte A and (B) electrolyte B; (a) textured substrate,

Id¼ 20 mA cm 2 ; (b) polished substrate, Id¼ 20 mA cm 2 ; (c)

textured substrate, Id¼ 35 mA cm 2 ; (d) polished substrate,

Id¼ 35 mA cm 2 ; (e) textured substrate, Id¼ 50 mA cm 2 ;

and (f) polished substrate, Id¼ 50 mA cm 2

Trang 6

this is not so in the case of textured substrates The

cracks observed for PS ﬁlms formed on polished

substrates for both electrolytes A and B indicates

higher stress and as a consequence, higher PL

decay is observed Whereas PS samples formed on

textured substrates are marked by smooth surface

morphology, lower stress and consequently, neg-ligible PL decay

In order to identify the chemical composition of our samples, we have investigated the Fourier transform infrared (FTIR) absorption spectra From our FTIR data (Fig 7) obtained for freshly

Fig 5 Scanning electron micrographs of porous silicon prepared on polished substrates at different current densities (I d ); (A)

I d ¼ 10 mA cm 2 , electrolyte A; (B) I d ¼ 10 mA cm 2 , electrolyte B; (C) I d ¼ 35 mA cm 2 , electrolyte A; (D) I d ¼ 35 mA cm 2 , electrolyte B.

Fig 6 Scanning electron micrographs of porous silicon prepared on textured substrates at Id¼ 35 mA cm 2 ; (A) electrolyte A; (B) electrolyte B.

Trang 7

prepared samples, it is clear that there are a

number of distinct peaks with different intensities

Figs 7(a) and (b)shows FTIR absorption spectra

for PS samples prepared using electrolyte A at

Id ¼20 mA cm2 on textured and polished

sub-strates, respectively PS ﬁlms prepared on textured

substrates exhibit mainly Si–H related modes at

910 cm1due to Si–H2scissors or Si–H3symmetric

660 cm1 due to Si–H2and Si–H wagging [19,20]

while for Si–O related modes are marked by a

interstitial Si–O–Si asymmetric stretching mode

[18] However, PS ﬁlms prepared on polished

substrates exhibits mainly Si–O-related peaks with

a doublet showing peaks at 2256 cm1which is

attributed to Si–H stretching modes when the

silicon is backbonded to oxygen atoms[21]and at

2117 cm1due to Si–H stretching mode, broad

1010 cm1due to Si–O–Si stretching mode and

non-stretching Si–H modes[20]and no signal of Si–H

observed It is worthwhile to note that there is no

signature of any O atoms backbonded to Si–H

related mode at 2250 cm1for PS ﬁlms prepared

on textured substrates (Fig 7(a)) Another inter-esting difference noted in the FTIR spectra of PS ﬁlms using electrolyte A prepared on textured and polished substrates is the shift of Si–O related

increase in the oxidation state (x) of the SiOx

species [22] For H2O2-based (B) samples formed

(Fig 7(c)) shows characteristic peaks of both Si–H and Si–O-related modes with a doublet comprising of peak at 2256 cm1(O backbonded

to Si in SiH stretching mode) and at 2117 cm1 (SiH stretching mode), a distinct broad peak at

1215 cm1(Si–O–Si) stretching mode, a broad peak doublet comprising of peaks at 940 and

bending modes and a shoulder at 650 cm1due

to Si–H wagging modes, respectively However, for the corresponding PS sample formed on polished substrate, the FTIR spectrum (Fig 7(d)) exhibits mainly Si–O-related modes at 2250 cm1 (O backbonded to SiH mode), a broad peak

(Si–O–Si stretching mode) with weak contribu-tions at 880 and 805 cm1(Si–H-related bending and wagging modes) Here inFig 7(d), the notable feature is the absence of Si–H stretching at

Thus, silicon–hydrogen-related modes are stronger for PS samples prepared on textured substrates while silicon–oxygen-related modes are stronger for the corresponding ﬁlms prepared on polished substrates for the same current density and electrolyte The effect of oxidation is felt more for H2O2-based PS ﬁlms particularly formed on polished substrates as compared to ethanol-based

PS ﬁlms From the above results, it can be conjectured that there is a change in the surface passivation from hydrogen to oxygen-like species

as we go from textured to polished substrate for

(Id20 mA cm2) for both the electrolytes A and

B, respectively In case of H2O2-based PS ﬁlms (B),

a signiﬁcant blue shift in PL spectra as compared

to the corresponding ethanol based ﬁlms could be due to the enhanced oxidation of surface of nanocrystalline Si resulting in an increase of SiOx

thickness surrounding the Si-core Oxidation of

2500 2250 2000 1750 1500 1250 1000 750 500

1.0

1.5

2.0

2.5

3.0

3.5

(d) (c)

(b)

(a)

Wavenumber (cm-1)

Fig 7 FTIR absorption spectra of porous silicon prepared at

current density Id¼ 20 mA cm 2 ; (a) textured substrate,

electrolyte A; (b) polished substrate, electrolyte A; (c) textured

substrate, electrolyte B; (d) polished substrate, electrolyte B.

Trang 8

nanocrystalline Si causes shrinkage of the Si-core

due to the breaking of Si–Si bonds resulting in a

blue-shift in PL spectra[11] However, apart from

interpretation in terms of quantum conﬁnement in

silicon clusters that decrease in size upon

oxida-tion, the PL blue shift can also be related to Si–O

species or due to defects and the silica networks on

which OH groups are absorbed as suggested by

others [23] These results are in accordance with

our PL and SEM studies where a signiﬁcant PL

decay and cracked surface morphology was

observed for PS ﬁlms formed on polished

sub-strates which underlines the importance of

tex-tured substrates and ethanol-based PS ﬁlms which

exhibits stable PL, smooth surface morphology

and H-passivated surfaces

Previous measurements showed that using H2O2

in a HF-based electrolytic mixture results in the

termination of Si surfaces mainly with

silicon-monohydrides leading to the formation of stable

contrary to other studies, we have found that

ethanol-based PS ﬁlms formed on textured

sub-strates are relatively more mechanically strong,

stable, stress-free and highly passivated with

hydrogen than the corresponding H2O2-based PS

ﬁlms as elucidated by our weight loss

measure-ments, PL, SEM and FTIR studies It seems that

the improved luminescent properties of our PS

ﬁlms is more an artifact of the substrate (textured

one) rather than that of the electrolyte alone On

the textured surface, the nucleation of nanopores

is preferentially initiated at the boundaries

be-tween the pyramids This would be assisted by the

faceted surfaces compared to the /1 0 0S surface

exposed at the boundaries This may lead to

partial merging of nanopores and the formation of

a high porosity region which can deform and

release the stress at dimensions small enough to

prevent macroscopic crack formation and fragility

Thus high porosity of PS ﬁlms formed on textured

substrates can be explained However, in case of

PS ﬁlms formed on polished substrates, the etching

is not preferential but random thus resulting in

lower porosity of PS layers However, the proper

choice of both the substrate (textured) and the

electrolyte (ethanol-based) in conjunction can have

a profound effect in improving the luminescent properties and stability of porous silicon ﬁlms

4 Conclusions

The visual observation of mechanically strong, stable surface bond configuration, smooth surface morphology and hydrogen-passivated PS surfaces essentially conforms the viability of textured substrates and ethanol-based electrolyte as a requisite condition for the formation of highly luminescent, thick and stable porous silicon films Porous silicon using ethanol-based electrolyte is superior to porous silicon formed using H2O2 -based electrolyte at the same current density on both textured and polished substrates, respec-tively A proper choice of a substrate and an electrolyte are essential for the formation of highly porous silicon films with lower fragility, superior stability and long-term usability

Acknowledgements

We thank Director NPL for permission to publish this work supported by CSIR network project on custom tailored special materials RKS thanks CSIR for providing a research associate-ship We acknowledge the help of Dr Ramkishore and Shri K.N Sood for SEM work and of Dr V.K Kaul (CEL) for sample preparation

References [1] L.T Canham, Appl Phys Lett 57 (1992) 1046 [2] V.S.Y Lin, K Motesharie, K.P.S Dancil, M.J Sailor, M.R Ghadiri, Science 278 (1997) 840.

[3] B Hamilton, Semicond Sci Technol 10 (1995) 1187 [4] V.V Doan, M.J Sailor, Science 256 (1992) 1791 [5] M.J Sailor, J.L Heinrich, J.M Lauerhaas, in: P.V Kamat, D Meisel (Eds.), Semiconductor Nanocrystals, Elsevier, New York, 1996, p 103.

[6] M.P Stewart, J.M Buriak, Adv Mater 12 (2000) 859 [7] S.T Lakshmikumar, P.K Singh, J Appl Phys 92 (2002) 3413.

[8] A.G Cullis, L.T Canham, P.D.J Calcott, J Appl Phys.

82 (1997) 909.

[9] J.L Heinrich, C.L Curtis, G.M Credo, K.L Cavanagh, M.J Sailor, Science 255 (1992) 66.

[10] J.M Buriak, M.J Allen, J Am Chem Soc 120 (1998) 1339.

Trang 9

[11] N Yamamoto, H Takai, Jpn J Appl Phys 38 (1999)

5706.

[12] Z Yamani, W.H Thompson, L AbuHassan, M.H.

Nayfeh, Appl Phys Lett 70 (1997) 3404.

[13] G Bhagavannarayana, S.N Sharma, R.K Sharma, S.T.

Lakshmikumar, communicated.

[14] O Bisi, S Ossicini, L Pavesi, Surf Sci Rep 38 (2000) 1.

[15] Z Yamani, S Ashhab, A Nayfeh, W Thompson, M.

Nayfeh, J Appl Phys 83 (1998) 3929.

[16] P.K Singh, S.T Lakshmikumar, Semicond Sci Technol.

17 (2002) 1123.

[17] S.N Sharma, R Banerjee, S Chattopadhyay, A.K Barua,

Proceedings of the 11th International Workshop on the

Physics of Semiconductor Devices (IWPSD) 2001, Allied Publishers Limited, p 1444.

[18] W.H Thompson, Z Yamani, L AbuHassan, O Gurdal,

M Nayfeh, Appl Phys Lett 73 (1998) 841.

[19] G Belomoin, J Therien, M Nayfeh, Appl Phys Lett 77 (2000) 779.

[20] D.R Kwon, S Ghosh, C Lee, Mater Sci Eng B (2003) 1 [21] V.M Dubin, F Ozanam, J.-N Chazalviel, Thin Solid Films 255 (1995) 87.

[22] S.N Sharma, R Banerjee, A.K Barua, Curr Appl Phys.

3 (2003) 269.

[23] H Tamura, M Ruckschloss, T Wirschem, S Veprek, Appl Phys Lett 65 (1994) 1537.

Định dạng
Số trang	9
Dung lượng	449,93 KB