Influence of hydrogen on the properties of titanium doped hydrogenated amorphous silicon prepared by sputtering

Accepted Manuscript Influence of hydrogen on the properties of titanium doped hydrogenated amorphous silicon prepared by sputtering Tianwei Zhou, Yuhua Zuo, Kai Qiu, Jun Zheng, Qiming Wang PII:

S0042-207X(15)30050-6

DOI:

10.1016/j.vacuum.2015.08.024

Reference:

VAC 6783

To appear in:

Vacuum

Received Date: 26 July 2014 Revised Date:

26 August 2015

Accepted Date: 27 August 2015

Please cite this article as: Zhou T, Zuo Y, Qiu K, Zheng J, Wang Q, Influence of hydrogen on the properties of titanium doped hydrogenated amorphous silicon prepared by sputtering, Vaccum (2015), doi: 10.1016/j.vacuum.2015.08.024. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Influence of hydrogen on the properties of titanium doped hydrogenated amorphous silicon prepared by sputtering

RI PT

Tianwei Zhou, YuhuaZuo*, Kai Qiu,Jun Zheng, Qiming Wang State Key Laboratory on Integrated Optoelectronics, Institute of Semiconductors, ChineseAcademy of Sciences, Beijing 100083, China E-Mail: [email protected] Abstract

Titanium doped hydrogenated amorphous silicon (a-Si:Ti(H)) with different gas ratio

SC

(rH=H2/Ar) were prepared by rf co-sputtering in mixture of hydrogen and argon, with

M AN U

fixed Ti content. The optical, structural and optoelectronic properties of the prepared a-Si:Ti(H) were investigated systematically by Spectroscopic Ellipsometry (SE), photo/dark conductivity and Fourier Transform Infrared (FTIR) measurements. With hydrogen introducing into a-Si: Ti network, the optical bandgap of a-Si:Ti(H), dark

TE D

current and photosensitivity (Ratio of photo conductivity to dark conductivity) improved. However, although detectable photosensitivity occurs at rH of 0.3, it has relatively larger microstructure factor R and less compact structure compared with rH

EP

of 0.2 and 0.4. It is found that unintended oxygen content plays an important role in

AC C

improving photosensitivity, and the relationship between microstructure, optical and photosensitivity of a-Si:Ti (H) was discussed in detail.

Keywords

Titanium doped hydrogenated amorphous silicon, photosensitivity, microstructure factor, unintended oxygen

ACCEPTED MANUSCRIPT 1. Introduction

Intermediate band solar cell (IBSC) has attracted much attention as the third generation solar cell due to its high efficiency above the S-Q limitation [1]. IBSC based

RI PT

on the use of a material that possesses an electronic energy band of allowed states within the conventional bandgap. The advantage of IBSC is proposed to increase the

realizing the IBSC is impurity doping

[2, 3]

SC

current of solar cells while at the same time preserve the output voltage. One way of . Ti element seems to be a promising

M AN U

impurity and several groups have studied heavily doped Ti in crystalline silicon

[4-5]

.

[6]

.

The issue of nonradiative recombination has also been resolved by Ti doping

However, the efficiency is limited by the intrinsic bandgap of host materials which immediately impacts the open circuit voltage of solar cells. The amorphous silicon

TE D

with bandgap of 1.7eV is an appropriate host material for IBSC

[7]

. It is necessary to

study the properties of Ti doped amorphous silicon. Although PECVD is a universal

EP

method to prepare a-Si:H, it is difficult to introduce Ti source in PECVD system. On the contrary, the co-sputtering method appears as an interesting technique to study the

AC C

dopant properties of solid nontoxic dopants, such as aluminum, indium, iron and some transition metals

[8]

. However, there is a problem that the preparation of a-Si by

sputtering has poor photoelectrical properties due to a high density of localized states formed by dangling bonds. One way of improving the quality of sputtered a-Si is depositing the films in Ar/H2 mixtures, as hydrogen can compensate the defects and dangling bond. Much work has been done to prepare low defect density films of a-Si:H by rf magnetron sputtering[9,10]. However, few reports about Ti doped a-Si:H 2

ACCEPTED MANUSCRIPT are reported. Previously, we have studied Ti doped amorphous silicon without hydrogen by sputtering and have proved that low content of Ti can partly compensate the dangling bond in amorphous silicon

[11]

, therefore, we fix the concentration of Ti

RI PT

at low content of 2% and plan to improve the quality of Ti doped amorphous silicon by adding hydrogen.

In this work, we propose to study the effects of incorporated H atoms on the titanium

SC

doped amorphous silicon prepared by reactive magnetron sputtering. Raman

M AN U

scattering spectra, Infrared absorption, ESR, optical bandgap, and photosensitivity are used to characterize the properties of films deposited by sputtering. The effect of gas ratio on the microstructure of the films is systematically investigated.

TE D

2. Material and Methods

Titanium-doped hydrogenated amorphous silicon (a-Si:Ti(H)) films were deposited on

EP

silicon and quartz substrates in the hydrogen-Argon gas mixture by rfco-sputtering magnetron sputtering. The target materials were 5N-pure single crystal silicon and

AC C

metal titanium with diameter of 5cm. Base pressure of the system before sputtering was kept below 5×10-7Torr. Sputtering pressure, rf power of Si and Ti targets, and substrate temperature were 3mTorr, 100W, 4W and 300℃ for each deposition, respectively. The parameter rH, which is defined as the ratio of the flow rate of hydrogen to that of argon,varied from 0 to 0.5. The thickness and absorption property of films were studied byJ.A.Woollam M-2000 DI Spectroscopic Ellipsometry (SE) in the spectrum range of 200-1700 nm. The 3

ACCEPTED MANUSCRIPT optical bandgap was obtained from the optical absorption coefficient from the Tauc method [12]. Fourier Transform Infra-red (FTIR) spectroscopy was performed for samples on c-Si

RI PT

substrates in the range from 400 to 4000cm-1 to obtain information about hydrogen content and analyze the bonding configurations. All the transmission measurements were made relative to an uncoated reference Si substrate. In order to analyze the

SC

structural properties, Micro-Raman scattering spectra were recorded from the films on

M AN U

quartz glass upon the excitation of an argon laser with 488 nm wavelength. And the power is kept to below 5mW with focusing diameter of about 10 µm(power density is below 25µW/µm2) to avoid the unintended crystallization during measurements. The measurement of photosensitivity was carried underAM1.5G illumination

TE D

provided by a solar simulator(Sol3A, Newport Oriel, USA).Dark conductivity (σD)and photoconductivity (σph) measurements were made by recording the current for a fixed voltage (50 V) applied across the gap (0.1 cm) of a planar geometry with

EP

Al electrodes configuration. The photosensitivity is defined as the ratio of

AC C

photoconductivityto dark conductivity.

3. Results

3.1 Deposition rate and Ti content

Figure 1 shows the deposition rate of a-Si: Ti(H) as a function of hydrogen flow rate ratio rH. It is predicted that in magnetron sputtered a-Si: H films, the deposition rate

4

ACCEPTED MANUSCRIPT decreases generally by the addition of hydrogen to argon [13], owing to the decrease in the net sputtering yield. However, in our system, the deposition rate of a-Si: Ti(H) films decreases slightly from 0.254 to 0.243Å/swith increasing H flow rate and even

RI PT

increases a little at rH of 0.3 to 0.25 Å/s. It may be contributed to two reasons. First, there is measurement error for thickness obtained from SE data, especially for thin films with slow deposition rate of about 0.25Å/s. Second, some parts of the hydrogen

SC

gas introduced into the Ar-H2 gas mixture may be ionized and contribute to the

M AN U

sputtering process in the present system. Similar result can be found in the paper [14]. As Ti is introduced by co-sputtering of Ti and Si targets, Ti content is independent of rH, with the constant value of about 2%, indicated by RBS(Rutherford Back

TE D

Scattering)test (data not shown here).

3.2 Microstructure Analysis

EP

3.2.1 FTIR

AC C

The total hydrogen content and silicon-hydrogen bonding configurations can be determined by FTIR. Fig. 2 shows IR absorption spectra of five samples with different rH. Strong absorbance peaks can be observed at ~640cm-1, and between 2000 and 2100 cm-1. The peak at 640 cm-1 represent the Si-H bending and Si-H2 wagging mode, while the stretching broad absorption band between 2000 and 2100 cm-1, can be deconvoluted into two satellite peaks: one at 2000 cm-1, associated with mono-hydrides (Si-H) bonds, and the other at 2090cm-1, associated withpoly-hydrides

5

ACCEPTED MANUSCRIPT (Si-H2, Si-H3,(Si-H2)n…) bonds. The strongest absorption peak of 640 cm-1 mode has been used to determine the total bonded hydrogen content, CH, which is proportional to the integrated absorption area

RI PT

of 640 cm-1 (I): CH=A×I/N[15], where A640=1.6×1019cm-2,and N=5.0×1022 cm-3, the atomic number density of pure silicon. The hydrogen contents are 9.58%, 10.2%, 11%, 11.5% and 11.9% for rH varying from 0.1 to 0.5, respectively, as shown in Fig. 3. It is

SC

evident that with increasing hydrogen ratio, the hydrogen content increases. Although

M AN U

other work has reported [16] that with higher hydrogen flow rates, the hydrogen content would decrease due to the extraction of hydrogen from the deposition surface by excessive H atoms, it is not observed here due to the limitation of rH in this work. As hydrogen can be incorporated into the amorphous Si network in different ways

TE D

(SiH, SiH2), the microstructure factor R[17] is defined as the fraction of hydrogen with di-hydride bonding among all hydrogen incorporated in the films:R= I2100/ (I2000 + I2100), where I2000 and I2100 represent the integrated band intensities at 2000cm-1 and

EP

2100cm-1 respectively. It is noted that R fluctuate with rH and have a minimum at rH of

AC C

0.2 in Fig. 3. Since SiH2 and (Si-H2)n bonding configurations is often correlated to microvoids in disordered regions of amorphous silicon networks[18-20], the initial decrease of R may be associated with the reduction of undesirable microvoids in the films, while the increase in R at higher rH indicates more similar microvoids occurs with more hydrogen atoms bonded to Si atoms in di-hydride form. Different from the literature[21], there is a broad absorption band between 700-1100 cm-1, and its intensity and shape change obviously with increasing rH. Especially for 6

ACCEPTED MANUSCRIPT rH of 0.3 and 0.4, they have evolved into two sharp and weak peaks near 900 and 1100 cm-1. However, compared with the absorption band between 900-1100 cm-1 of rH=0.3, a broader band of rH=0.4 is observed. That broad absorption band is a complex one,

RI PT

which may be related to poly-hydrides (Si-H2, Si-H3,(Si-H2)n…) bending vibration mode (700-900 cm-1), asymmetric stretching vibration mode of O in Si-O-Si (900-1200cm-1)with different bond angles[22] and possible Ti-O-Si vibration mode

SC

(930 cm-1) [23].The peak near 1100cm-1 represents the Silica peak with the bond angle

M AN U

θ=180°. As indicated in reference [22], if two oxygen atoms be bonded to a Si atom, it contributes towards 1020 cm-1 absorption peak, for three atoms the contribution is towards 1060 cm-1 and for four it is close to 1120 cm-1. Shown in the inset of Fig. 2, it can be seen that the peak of 1100cm-1 gradually disappears, while the broad peak of

TE D

1020 cm-1 gradually becomes obvious with higher rH of 0.5. It indicates that with higher hydrogen flow, the number of oxygen atoms being bonded to a Si atom changes from four to two and the reconstruction of a-Si:Ti network occurs at higher rH.

EP

The function of higher flow hydrogen on the microstructure of a-Si:Ti network can be

AC C

also proven in the following context in Fig. 4. Since the vibration of H or O is modified depending on atomic environment around Si atoms, the initial broad resonant absorption band indicates the nature of environment disorder with Ti doping in amorphous Si network. With rH increasing, more weak bonds (such as poly-hydrides or weak bonded O) are broken and the amorphous Si network is reconstructed with stronger Si-H and more stable Si-O-Si bonds. 7

ACCEPTED MANUSCRIPT The unexpected oxygen related band near 1000cm-1 is mainly due to oxygen contamination, which may be introduced during the sputtering process and/or storage in the air for some time. Besides the base pressure of sputtering system, the content of

that C

[O]

[O])

is dependent on the deposition conditions. It is reported

[24]

RI PT

unintended oxygen(C

increases with increasing growth temperature and increasing H2 to SiH4

ratio in PECVD-grown a-Si: H films. There was no oxygen peak in the a-Si: H films

SC

grown in 95 °C in the literature [16]by magnetron sputtering, while the a-Si: Ti(H)

temperature on C

[O]

M AN U

films in our work grown in 300 °C has apparent oxygen peak. The effect of growth is similar as PECVD for sputtering. However, it is clear in our

work that C [O] (considering the integrated area of 1000 cm-1 peak) initially decreases at rH of 0.3 and then increases again with rH, which indicates a different growth

TE D

mechanism for hydrogen-argon sputtering process.

3.2.2 Raman scattering spectra

EP

The Raman scattering spectra of all samples are similar (as shown in the inset of Fig.

AC C

4), characterized by a broad amorphous peak at around 480cm-1, without any sharp peak or small shoulder at around 520cm-1(characteristic of nanocrystalline Si). It proves that even with increasing hydrogen flow rate, no amorphous-to- anocrystalline transition occurs, and the a-Si: Ti(H) films remain amorphous. The broad peak of 480 cm-1can be decomposed into four peaks, corresponding to TA (Transverse Acoustical, 150cm-1), LA (Longitudinal Acoustical, 310cm-1), LO (Longitudinal Optical, 400cm-1) and TO (Transverse Optical, 480cm-1) mode of

8

ACCEPTED MANUSCRIPT amorphous Si, respectively.And theratio of integrated peak intensity of TA mode to that of TO mode (ITA/ITO) can be taken as the degree of short-range order [25]. The LA-like mode and the LO-like mode are related to the presence of defects in the films,

RI PT

and the more defects the larger values of ILA+LO/ITO ratio[26]. As shown in Fig.4, the variations for both ITA/ITO and ILA+LO/ITO are similar, and for titanium-doped hydrogenated amorphous silicon films, the short-range order is best

SC

and the defects are least when rH is 0.2. It can be seen that when rH is higher to 0.3 and

M AN U

0.4, both of the ratio of ITA/ITO and ILA+LO/ITO increases. It indicates that excess hydrogen atoms incorporated in the films are harmful to the local order of a-Si: Ti network, with more Si=H2 or (Si=H2)n chain forms, which is proven in Fig.3. However, when rH is 0.5, both of the ratios decreases again, indicating that

TE D

short-range order and defects have improved when hydrogen atoms amount are larger than a certain level. It has been reported in the literature

[27]

that high flow of

hydrogen atoms can remove the weak bonds and reconstruct the amorphous Si

EP

network. The same effect of high flow of hydrogen on the local atom arrangements of

AC C

amorphous Si: Ti network can be verified by our work, from both Fig. 2 and Fig. 4. The disappear of 1100 cm-1 absorption peak and gradually obvious broad 1020 cm-1 peak at rH of 0.5 in Fig.2 are related with the re-arrangement of oxygen and silicon bonds, while the change of ITA/ITO and ILA+LO/ITO at rH of 0.5 in Fig. 4 are related with reconstruction in amorphous Si: Ti network with more weak bonds removed. In summary, the function of hydrogen to the microstructure of amorphous Si: Ti(H) is quite complex, and is dependent on the content of hydrogen and reactions with the 9

ACCEPTED MANUSCRIPT amorphous silicon network.

3.3 Optical absorption

RI PT

The optical bandgap can be obtained from the optical absorption spectra of SE measurements. According to Tauc formula, the optical bandgap is defined as the intercept of the linear part of the plot of (αhν) 1/2 versus hν, where α is the absorption

SC

coefficient and hν is the photon energy. In addition, the absorption coefficient at the

M AN U

photon energy below the optical bandgap depends exponentially on the photon energy: α(hν)～exp (hν/EU), where EU is called Urbach energy.

The plot indicates a bandgap of 1.54eV for films containing no hydrogen, while the bandgap increases to around 1.8eV when hydrogen is incorporated into a-Si film. The

TE D

change of Urbach energy is opposite to that of the bandgap, and it has the trend of decreasing from 893 to 549 meV with rH increasing.

It is reported in the literature [28]

EP

that Urbach energy is related to the width of band tail and the degree of disorder. The less Urbach energy, the narrower band tail width and less degree of disorder. It can be

AC C

seen that the introducing of hydrogen in a-Si: Ti can change the band tail states, and reconstruct a new network of amorphous Si doping with Ti.

3.4 Photosensitivity

Photosensitivity is an important parameter for many optoelectronic materials to be studied

[29-30]

. As shown in Fig.6, with hydrogen introduced in a-Si: Ti network, both

10

ACCEPTED MANUSCRIPT the dark and light current density drops from 14nA/cm2 to 3nA/cm2 at rH=0.1;with rH increasing to 0.2, the dark and light current density continue to decrease to 0.6nA/cm2and 0.88nA/cm2. However, when rH increases to 0.3, the dark current

RI PT

density almost remains unchanged, while the photocurrent density increases obviously to about 5nA/cm2. Thus, a photoconductivity gain of 8 can be obtained for the a-Si: Ti(H) film grown at rH of 0.3.WithrH increasing, the dark and light current densities

SC

continue to decrease.As a contrast, for the undoped a-Si: H film grown by the same

of 104 reported in the literature

[16]

M AN U

sputtering system, the photosensitivity is about 100, which is much less than the value . The reason will be discussed later. It is clear that

there is a bad influence on the photosensitivity of a-Si:H with Ti adding,maybe more hydrogen atoms needed to modify the defects and micro-voids introduced by Ti atoms

TE D

in theamorphous Si networks. Attempt has been made to correlate the microstructure, optical gap and photosensitivity of a-Si: Ti (H) films in the following section.

EP

4 Discussions

AC C

To clarify the effect of hydrogen flow rate ratio on the properties of a-Si: Ti(H) films, the information about microstructure, optical and electrical properties of the samples as a function of rH has been listed in Table 1. It is clear that hydrogen content and optical gaptend to increase with rH increasing, while unintended oxygen content, Urbach energy and σDare inclined to decrease. It indicates that the introducing of hydrogen in the amorphous Si network with high Ti doping can improve the degree order of band tail, including Eg and EU. The dark conductivity can also be reduced 11

ACCEPTED MANUSCRIPT with fewer defects hopping places for carriers. However, microstructure factors R and the degree of short-range order ITA/ITOin a-Si: Ti (H) have the minimum value at rH of 0.2. It is reported in other literatures that

RI PT

proper hydrogen content of around 10% and less microstructure factor R with less di-hydrogen (Si-H2) and complexes is beneficial to photosensitivity. However, it seems not the truth for our case, since the growth condition with rH of 0.3 for best

SC

photosensitivity has a relatively larger R compared with rH of 0.2. Another important

M AN U

factor which has been neglected before should be pay attention to. That is the unintended oxygen content. It is reported in the literature

[31]

that the sample with

more oxygen content will have less photosensitivity. With decreasing oxygen content, light-induced conductivity as well as photosensitivityincreases apparently.Moreover, [16]

without

TE D

the n-doped samples also grown by magnetron sputtering in the literature

oxygen peak in FTIR has the photosensitivity of about 1000, while our best samplehere has the photosensitivity of only around 10. It indicates that besides

EP

hydrogen content and microstructure factor R, unintended oxygen content is of

AC C

importance to the photosensitivity of doped amorphous Si films grown by magnetron sputtering. It should be noted that the calculated C[O] may be overestimated with nominally too high C[O], since the peak of FTIR between 900-1100 cm-1 is so broad and it contains more information besides oxygen bond, but the tendency of C[O] can be trusted. The way to reduce the oxygen content may be reducing the base oxygen content of the sputtering system by some pre-sputtering measurements or grown at lower temperature. The details of the origins of the unintended oxygen are under 12

ACCEPTED MANUSCRIPT study now.

5. Conclusion

RI PT

Titanium doped hydrogenated amorphous silicon films in mixture of Ar and H2 with different gas ratio have been prepared and investigated.A strong correlation between the structure, optical and optoelectronic properties of the films was found. The

SC

introducing of H in a-Si: Ti(H) materials have the benefits in broadening optical

M AN U

bandgap, decreasing band tail states; while the unintended oxygen content plays an important role in improving photosensitivity. The optimum growth condition investigated is of hydrogen gas ratio of 0.3, with photosensitivity of 8 and the least oxygen content. This indicates that reactive magnetron sputtering a-Si: Ti(H) films

EP

Acknowledgements

TE D

may have the potential application in optoelectronics, such as low cost solar cells.

AC C

This work was supported by the National Natural Science Foundation of China (No. 51072194, 61021003, 61036001 and 61376057).

References

[1] W.Shockley and H.J.Queisser. J. Appl. Phys. 32 (1961) 510-519. [2] A.S.Brown and M.A.Green.J. Appl. Phys. 92 (2002) 1329. [3]

G.Beaucarne,

A.S.Brown,

M.

J. 13

Keevers,

R.

Corkish

and

M.

A.

ACCEPTED MANUSCRIPT Green.Prog.Photovoltaics.10 (2002) 345. [4] J. Olea, M. Toledano-Luque, D. Pastor, E. San-Andrés, I. Mártil, and G. González-Díaz.J. Appl. Phys. 107 (2010) 103524.

RI PT

[5] Ma Z H, Cao Q, Zuo Y H, Zheng J, Xue C L, Cheng B W, Wang Q M.Chin. Phys. B. 20 (2011) 106104.

[6] J.Olea, M.Toledano-Luque, D.Pastor, G. González-Díaz and I. Mártil.J. Appl.

SC

Phys. 104 (2008) 016105.

M AN U

[7] Cao Q, Ma Zhi-Hua, Xue Chun-Lai, Zuo Yu-Hua, Wang Qi-Ming. Chin. Phys. B.20 (2011) 097103.

[8] T Shimizu, M Kumeda, I Watanabe and Y Nouwi. Philosophical Magazine B. 44 (1981) 159-174.

TE D

[9] N. Sawides.J. Appl. Phys. 56 (1984) 2788.

[10] M.M. de Lima Jr, F.C. Marques.Journal of Non-Crystalline Solids. 299-302 (2002) 605-609.

EP

[11] Tianwei Zhou, YuhuaZuo*, Leliang Li, Kai Qiu, Jun Zheng, Qiming Wang.

AC C

Vaccum.104 (2014) 65-69.

[12] J. Tauc, R. Grigorovici, A.Vancu, Phys. Stat. Sol. 15 (1996) 627. [13] N. Savvides. J. Appl. Phys. 55 (1984) 4232-4238. [14] N. Saito. J. Appl. Phys., 61(1987) 2962. [15] C. J. Fang, K. J. Gruntz, L. Ley, M. Cardona, F. J. Demond, G. Miiller, and S. Kalbitzer.J. Non-Cryst. Solids. 35-36 (1980) 255-260. [16] D. Girginoudi, C. Tsiarapas, N. Georgoulas. Appl. Surf. Sci. 257 14

ACCEPTED MANUSCRIPT (2011)3898-3903. [17] J. Mouro, A. Gualdino, V. Chu, and J. P. Conde.J. Appl. Phys.114 (2013) 184905. [18] J. Daey Ouwens and R. E. I. Schropp. Phys. Rev. B. 54 (1996) 17759-17762.

RI PT

[19] Biswas P, and Timilsina R. J. Phys.: Condens.Matter.23 (2011) 65801. [20] Smets A.H.M., Kessels W.M.M and Van de Sanden M.C.M. Appl. Phys. Lett.82 (2003) 1547.

SC

[21] M. H. Srodsky, Manuel Cardona and J. J. Cuomo.Phys. Rev. B. 16 (1977)

M AN U

3556-3571.

[22] S. M. Iftiquar. High Temp. Mat.Proc. 6(2002) 35.

[23] D. Brassard, D. K. Sarkar, M. A. El Khakani, and L. Ouellet.Journal of Vacuum Science & Technology A. 22 (2004) 851.

TE D

[24] Xiao-dan Zhang*, Ying Zhao, Feng Zhu, et al.Applied Surface Science 245 (2005) 1-5.

[25] Gerbi JE, Voyles PM, Treacy MMJ, Gibson JM, Abelson JR. ApplPhysLett82

EP

(2003)3665-3667.

AC C

[26] Zotov N, Marinov M, Mousseau N, et al. J. Phys.: Condens Matter. 11(1999) 9647-9658.

[27] R.A.Street, Biegelsen, in: Hydrogenated Amorphous Silicon, E-Publishing, Inc. the PressSyndicate of the University of Cambridge, 1991, pp.58-60. [28] G. D. Cody, T. Tiedje, B. Abeles, B. Brooks, and Y. Goldstein. Phys. Rev. Lett. 47 (1981) 1480-1483. [29] Baojun Yan, Lei Zhao, Wenjing Wang, etal.Vacuum. 89 (2013) 43-46. 15

ACCEPTED MANUSCRIPT [30] Nobuo Saito.

Appl. Phys. Lett. 46 (1985) 61.

[31] JiEun Lee, JooHyung Park, JinsuYoo, Kyung Hoon Yoon, Donghwan Kim ,

AC C

EP

TE D

M AN U

SC

RI PT

Jun-Sik Cho.Solid State Sciences 20(2013) 70-74.

16

ACCEPTED MANUSCRIPT Table 1 Hydrogen content CH, unintended oxygen content C[O], microstructure factor R, the degree of short-range order ITA/ITO, Eg, EU, σD,σph,σph /σD of the a-Si: Ti(H) samples

AC C

EP

TE D

M AN U

SC

RI PT

grown at different hydrogen to argon flow rate ratio rH

ACCEPTED MANUSCRIPT Table 1

0 0.1 0.2 0.3 0.4 0.5

CH (%)

C[O] (%)

R=I2100/ (I2000+I2100)

ITA /ITO

Eg (eV)

EU (meV)

0 9.58 10.2 11 11.5 11.9

-8.4* 5.2* 1.6* 2.59* 5.61*

-0.57 0.1 0.38 0.202 0.205

0.78 0.68 0.57 0.63 0.787 0.74

1.54 1.71 1.74 1.8 1.77 1.79

893 592 565 546 559 549

σD

σph -1

(Ω⋅cm ) 5.34×10-4 1.31×10-5 3.75×10-6 3.75×10-6 1.15×10-6 1.48×10-6

(Ω⋅cm ) 5.34×10-4 1.44×10-5 5.63×10-6 3.24×10-5 1.27×10-6 1.61×10-6

EP

TE D

M AN U

SC

* The calculation of C [O] is used the formula shown in the literature [22].

AC C

-1

σph /σD 1 1.09 1.5 8.64 1.1 1.08

RI PT

rH

ACCEPTED MANUSCRIPT Fig. 1. The deposition rate of a-Si: Ti(H) films as a function of rH. Fig. 2.

RI PT

IR absorption spectra of the films with: (a) rH=0.1 (b) rH=0.2 (c) rH=0.3 (d) rH=0.4 (e) rH=0.5. Fig. 3.

SC

Content of [H] as well as microstructure factor R as a function of rH.

M AN U

Fig. 4. ITA/ITO and ILA+LO/ITO as a function of rH. Fig. 5.

Optical bandgap and Urbach energy of samples as a function of rH.

TE D

Fig. 6.

AC C

EP

Dark and light current density as well as photosensitivity as a function of rH.

ACCEPTED MANUSCRIPT Fig.1.

0.30 0.29

RI PT

0.27 0.26 0.25 0.24 0.23

SC

Growth rate (Å/s)

0.28

0.22

0.20

0.0

0.1

M AN U

0.21 0.2

0.3

0.4

AC C

EP

TE D

Hydrogen flow rate ratio rH

0.5

ACCEPTED MANUSCRIPT Fig. 2. 1020

e

1120

Absorption Coefficient(cm-1)

d

800

900

1000

1100

RI PT

c

e

1200

d

400

600

M AN U

SC

c

800 1000 1800

2000

AC C

EP

TE D

-1 Wavenumber(cm )

b a

2200

ACCEPTED MANUSCRIPT Fig. 3.

12.0

0.6 CH

0.5

R

RI PT

0.4

11.0

0.3

10.5

SC

0.2

10.0

0.2 0.3 0.4 Hydrogen flow rate ratio rH

M AN U

0.1

AC C

EP

TE D

9.5

Microstructure R

CH(%)

11.5

0.5

0.1 0.0

ACCEPTED MANUSCRIPT Fig. 4. 1.0

1.00

400

a-Si:H(Ti) rH =0.3

300

0.95

250 200

TA

150

0.8

LO LA

0.90

100 100

200

300 400 500 600 Raman Shift(cm -1 )

700

RI PT

ITA/ITO

0.9

0.85

0.7

0.80

SC

0.6

0.1 0.2 0.3 0.4 Hydrogen flow rate ratio rH

AC C

EP

TE D

M AN U

0.0

ILA+LO/ITO

Intensity(a.u.)

350

TO

0.5

0.75

ACCEPTED MANUSCRIPT Fig. 5. 1.80

Eg

900

EU

1.75

1000

)

1.55

a-Si:Ti(H) rH=0.3 a-Si:Ti(H) rH=0.4

600

a-Si:Ti(H) rH=0.5

700

400 200

600

0

0.0

0.1

RI PT

1/2

1.60

800

a-Si:Ti(H) rH=0.2

1.5

0.2

2.0

0.3

2.5

0.4

AC C

EP

TE D

M AN U

rH

hν(eV)

3.0

SC

-1/2 1/2

(α hν ) (ev cm

1.65

1.50

a-Si:Ti(H) a-Si:Ti(H) rH=0.1

800

EU(meV)

Eg(eV)

1.70

0.5

ACCEPTED MANUSCRIPT Fig. 6.

a-Si:H(Ti)

14 12

RI PT

6

10 8

4

0.1

0.2

0.3

EP

TE D

rH

AC C

0.4

M AN U

0.0

SC

6

2

0

Current density(nA/cm2)

8

Photosensitivity

16

photosensitivity dark light

0.5

4 2 0

ACCEPTED MANUSCRIPT 1.Hydrogen content, Eg and photosensitivity of a-Si: Ti(H) films increase with rH. 2.The unintended oxygen content, Urbach energy and σD decrease with gas ratio rH.

RI PT

3.Photosensitivity of the films increases with rH and is maximized to 8 at rH=0.3. 4.Microstructure factor R of the films is minimum at rH of 0.2

AC C

EP

TE D

M AN U

SC

5.The unintended oxygen content plays a vital role in improving photosensitivity.