The effect of annealing on the properties of diamond-like carbon protective antireflection coatings

ARTICLE IN PRESS

Renewable Energy 33 (2008) 226–231 www.elsevier.com/locate/renene

The effect of annealing on the properties of diamond-like carbon protective antireflection coatings Won Seok Choia, Byungyou Honga,b, a

Center for Advanced Plasma Surface Technology, Sungkyunkwan University, Suwon 440 746, South Korea School of Information and Communication Engineering, Sungkyunkwan University, Suwon 440 746, South Korea

b

Available online 27 June 2007

Abstract In addition to its similarity to genuine diamond film, diamond-like carbon (DLC) film has many advantages, including its wide band gap and variable refractive index. Therefore, as one of the diverse applications, DLC film can be utilized as a protective coating for IR windows and an anti-reflective coating for solar cells. For this study, DLC films were prepared by the radio frequency-plasma enhanced chemical vapor deposition (RF-PECVD) method on silicon substrates using methane (CH4) and hydrogen (H2) gas. We examined the effects of the post-annealing temperature and the annealing ambient on structural, electrical and optical properties of DLC films. The films were annealed at temperatures ranging from 300 to 900 1C in steps of 200 1C using rapid thermal annealing equipment in nitrogen ambients. The thickness of the film was observed by scanning electron microscopy (SEM) and surface profile analysis. The variation of structure according to the annealing treatment was examined using Raman spectroscopy, X-ray photoelectron spectroscopy (XPS) and high-resolution transmission electron microscopy (HRTEM). The reflectance of DLC thin film was investigated by UV–vis spectrometry and its electrical properties were investigated using a four point probe and I–V meter. The carrier lifetime of the film was also checked. r 2007 Elsevier Ltd. All rights reserved. Keywords: Plasma enhanced chemical vapor deposition; Diamond-like carbon film; Antireflection coating; Annealing treatment; Electrical properties

1. Introduction Due to its excellent properties, such as its high hardness, high electrical resistivity, low IR absorption, transparency to visible light and chemical inertness, diamond-like carbon (DLC) film has found several important applications in optics and electricity [1,2], solid state electronics and other fields. In addition to its similarity to genuine diamond film, DLC film has many advantages [3], including its wide band gap and variable refractive index. Therefore, DLC film can be used as a protective coating for IR windows and an antireflective coating for solar cells [4,5]. For the synthesis of DLC thin films, the radio frequencyplasma enhanced chemical vapor deposition (RF-PECVD) method has been widely used because it uses standard

Corresponding author. Center for Advanced Plasma Surface Technology, Sungkyunkwan University, Suwon 440 746, South Korea. E-mail address: [email protected] (B. Hong).

0960-1481/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2007.05.022

plasma processing technology, which is simple, relatively inexpensive and operates at low temperatures. Moreover, RF-PECVD allows the film to be uniformly coated on substrates of different shapes and sizes [6] and DLC properties to be easily controlled by adjusting the hydrogen contents in the film [7]. However, in certain applications, DLC films have some disadvantages, such as their high residual stress and thermal degradation at high temperature [8,9]. Also, in the case where DLC film is used for antireflection coatings for Si solar cells, it has to endure a firing procedure at a temperature of about 700 1C. In this work, the RF-PECVD method was used to prepare DLC thin films with methane (CH4) and hydrogen (H2) gas. The synthesized DLC films were annealed at temperatures ranging from 300 to 900 1C in steps of 200 1C using rapid thermal annealing (RTA) equipment in nitrogen ambient. We measured and compared the properties of the post-annealed DLC films using structural, electrical and optical methods as a function of the annealing temperature.

ARTICLE IN PRESS W.S. Choi, B. Hong / Renewable Energy 33 (2008) 226–231

2. Experiments DLC films were deposited on p-type (1 0 0) silicon substrates using the 13.56 MHz RF-PECVD method. The substrates were cleaned using the usual RCA method followed by in situ H2 plasma pre-treatment for 10 min at a gas pressure of 133 Pa (1 torr) and an RF power of 150 W without additional heating, in order to remove any contaminants on the surface and to activate the surface. Methane and nitrogen gases were then introduced into the reaction chamber to grow the films at room temperature. The substrate was heated only by the plasma during the growth of the film and a constant deposition time of 5 min 30 s was used for each sample. Fig. 1 shows cross-sectional scanning electron microscopy (SEM) image of DLC film, which was used in this experiment. The thicknesses of all the samples were kept constant at 120 nm. After the deposition of DLC film, it was post-annealed with RTA equipment in nitrogen ambient. DLC films were annealed at various temperatures ranging from 300 to 900 1C in steps of 200 1C. The annealing temperature was maintained for 1 min and the rising time and falling time were both 30 s. The experimental parameters used in this study are shown in Table 1. The thickness and surface morphology of DLC thin films were measured using a field emission scanning electron microscope (FESEM: Jeol, JSM-6700F) and an atomic force microscope (AFM: Seiko, SPA-400), respectively. Raman spectrometry (Jasco, MRS-3000) and X-ray photoelectron spectroscopy (XPS: VG Microtech, ESCA2000) were used to characterize the structure of DLC films as a function of the annealing temperature. High-resolution transmission electron microscopy (HRTEM: JEOL, JEM3010) was used to compare the structural variation of DLC films and the interface between film and substrate as a function of the annealing temperature. The optical properties of DLC thin films were investigated by UV–vis spectrometry (Scinco, S-3100). The electrical properties

Fig. 1. Cross-sectional SEM image of as-deposited DLC film.

227

Table 1 Deposition conditions Substrates

p-type Si (1 0 0)

Pre-treatment gas

H2: 80 sccm

Deposition gas

CH4: 20 sccm H2: 80 sccm

Working pressure RF power Deposition time Thickness of the film Substrate temperature

1 torr 150 W 5 min 30 s 120 nm No external heating

were investigated by means of a four point probe (AIT, CMT-ST1000), an I–V meter (HP, 4140B) and the measurement of the carrier life time (LEO, LTA-700) as a function of the annealing temperature. 3. Results and discussion The Raman spectra (excitation wavelength of 532.01 nm and laser power of 3.5 mW) for DLC thin films as a function of the annealing temperature are shown in Fig. 2(a). Raman data are widely used for the characterization of the structure of DLC films, because of their ability to distinguish sp3 and sp2 bonding types [10]. The Raman spectra in the present study were deconvoluted into two Gaussian peaks and summarized in Fig. 2(b). The ID/IG ratio and G-peak wavenumber increased with annealing temperature. Such an increase in the ID/IG ratio means that there is an increase and enlargement of the sp2 clusters in DLC films [11]. The G-peak position in the Raman spectra is related to the sp2/sp3 ratio [12] and the film stress [13]. The G-peak position for DLC films in this study shifted to a high wavenumber with annealing temperature, which corresponded to an increase in the sp2 bonding fraction [7]. Therefore, the Raman spectra showed that the thermal annealing caused the DLC structure to be transformed into the graphite structure [14]. And also, they indicate that increasing the annealing temperature led to a reduction in the hydrogen contents in DLC films and to a transition from sp3 to sp2 carbon hybridization, i.e., partial tetrahedral bonds have been broken and transformed to trigonal bonds. And this resulted in an increase in the number of sp2 clusters in the film, i.e., the graphitization of the film. In order to determine the detailed bonding structure, XPS analysis was employed. Fig. 3(a) shows the XPS C1s spectra of DLC films as a function of the annealing temperature. The C1s peak positions of diamond and graphite were 291.4 and 284.4 eV, respectively [15]. The C1s peak position shifted from a high binding energy (285.05 eV) to a low binding energy (284.4 eV) as the annealing temperature increased. The C1s spectra were deconvoluted into three components centered at 284.4, 285. 8 and 287.5 eV, which were assigned to the sp2 C–C, sp3

ARTICLE IN PRESS W.S. Choi, B. Hong / Renewable Energy 33 (2008) 226–231

300°C 500°C 700°C 900°C

700 600 500

ID/IG ratio G-peak position

3.5

ID/IG ratio

800 Intensity (Cts)

1560

4.0

900

400 300 200

1550

3.0

1540

2.5

1530

2.0

1520

1.5

1510

100 0

300

900 1000 1100 1200 1300 1400 1500 1600 1700 1800

Wavenumber (cm-1)

500

700

G-peak position (cm-1)

228

900

Annealing temperature (°C)

sp2 C-C

80000

Intensity (Cts)

300 °C 500 °C 700 °C 900 °C

sp3 C-C

70000 60000 50000 40000 30000 20000

C-O, C=O

10000

sp2 content in the C1s (%)

90000

90

90

80

80

70

70

60

60

50

50

40

40

30

30

20

20

sp3 content in the C1s (%)

Fig. 2. (a) Raman spectra of DLC films as a function of annealing temperature, (b) ID/IG ratio and G-peak positions extracted from Raman data.

0 282

283

284

285

286

287

288

289

290

Binding energy (eV)

300

500

700

900

Annealing temperature (°C)

Fig. 3. XPS analysis as a function of annealing temperature, (a) C1s spectra, (b) sp2 and sp3 content in %.

C–C and C–O (CQO) bonds, respectively [16]. These results are summarized in Fig. 3(b). The sp3 content decreased from 75.2% to 24.1% and the sp2 content increased from 24.8% to 75.9% as the annealing temperature increased from 300 to 900 1C. This result is consistent with the Raman analysis. The post annealing treatment resulted in an increase in the graphitic fraction in the film and the clustering of the sp2 bonded carbon [17]. This indicates the phase transition from diamond-like to graphite-like structure as a function of annealing temperature. This is verified from Fig. 4. High-resolution crosssectional TEM image of as-deposited DLC film is shown in Fig. 4(a) and (b) shows the cross-sectional TEM image of DLC film annealed at 900 1C. There are shown many crystallized clusters with the diameter of about 5 nm in DLC film annealed at 900 1C and the crystallized region is enlarged in Fig. 5(c). From this figure and the Raman data, the crystallized regions are interpreted as the graphite clusters and these clusters were formed during the high post-annealing process. In order to observe the antireflection properties, UV–vis spectroscopic analysis was employed. Fig. 5(a) shows the reflectance within the wavelength range of 400–1000 nm as a function of the annealing temperature and Fig. 5(b) shows the average reflectance within the same wavelength

range. The average reflectance of DLC film increased from 8.75% to 18.27% as the annealing temperature increased from 300 to 900 1C and an abrupt increase was observed at the temperature above 700 1C. The phase transition of DLC film from diamond-like carbon to graphite-like carbon influences its optical properties. The increase in the graphitic fraction and the clustering of the sp2 bonded carbon in the film has the effect of reducing the transmittance and increasing the reflectance. Fig. 6 shows the electrical properties of DLC film as a function of the annealing temperature and Fig. 6(a) shows the variation of the sheet resistance as the annealing temperature increased from 500 to 900 1C in steps of 100 1C, and we could not obtain the sheet resistance below the annealing temperature at 500 1C because of their high sheet resistance. The sheet resistance abruptly decreased between 500 and 600 1C and became very small (o10 kO/cm2) at the temperature of 700 1C and above. The sp2 carbon bonds were gradually overlapped as the annealing temperature increased and would be connected to each other in the long run, while the electrical resistance suddenly decreased. Fig. 6(b) shows the leakage current characteristics of DLC films as a function of the annealing temperature. The thickness of all of the samples in this experiment is 120 nm. The leakage current densities at 1 MV/cm were 4.6  10 6, 8.7  10 6, 7.2  10 5,

ARTICLE IN PRESS W.S. Choi, B. Hong / Renewable Energy 33 (2008) 226–231

229

50 45 40 35 30 25 20 15 10 5 0

20 900 °C 500 °C As-depo.

700 °C 300 °C

Average reflectance (%)

Reflectance (%)

Fig. 4. Cross-sectional high-resolution TEM images, (a) as-deposited DLC film, (b) DLC film annealed at 900 1C, (c) enlargement of the tetragonal region in (b).

18 16 14 12 10

As-depo.

8 400

500

600 700 800 Wavelength (nm)

900

1000

300 500 700 Annealing temperature (°C)

900

Fig. 5. Reflectance variation of the annealed DLC film, (a) as a function of wavelength, (b) average reflectance within the wavelength range of 400–1000 nm.

5.1  10 4 and 0.3 A/cm2 for the as-deposited film and DLC films annealed at 300, 500, 700 and 900 1C, respectively. The leakage current increased with the annealing temperature and the film lost its electrical resistivity at 900 1C. Figs. 6(a) and (b) show the differences between the sheet resistances and leakage currents at 700 and 900 1C. It is clear that the sheet resistance was almost same at both the temperatures, whereas the leakage current

properties were quite different. This result implies that the sp2 carbon bonds gradually overlapped each other starting from the surface by the annealing treatment, and it makes difference because the sheet resistance and the leakage current were dominantly affected by surface resistivity and inside resistivity of the film, respectively. Fig. 7 shows the carrier lifetime of DLC film as a function of the annealing temperature. The carrier lifetime

ARTICLE IN PRESS W.S. Choi, B. Hong / Renewable Energy 33 (2008) 226–231

6000

100

5000

10-1

Current density (A/cm2)

Sheet resistance (kΩ /cm2)

230

50 40 30 20 10

900 °C 500 °C

10-2

700 °C 300 °C

As-depo.

10-3 10-4 10-5 10-6 10-7

0 500

600 700 800 900 Annealing temperature (°C)

10-8 0.0

0.2

0.4 0.6 0.8 Electric field (MV/cm)

1.0

Fig. 6. Electrical properties of DLC film as a function of annealing temperature, (a) sheet resistance, (b) leakage current characteristics.

TEM, which showed that the graphitic clusters were formed after annealing at 900 1C. The average reflectance of the annealed DLC film within the wavelength range of 400–1000 nm increased from 8.75% to 18.27% as the annealing temperature was increased from 300 to 900 1C. The sheet resistance decreased with increasing annealing temperature, while the leakage current increased. The carrier lifetime gradually decreased at higher annealing temperatures, and this might be due to the extraction of the hydrogen inside the film.

9 8

Carrier lifetime (μs)

7 6 5 4 3 2

Acknowledgments

1 0 500

600

700

800

900

Annealing temperature (°C) Fig. 7. Carrier lifetime of DLC film as a function of annealing temperature.

This work was supported by Grant no. R-11-2000-0860000-0 from the Center of Excellency Program of the Korea Science and Engineering Foundation and Ministry of Science & Technology (2006). References

of DLC film decreased with increasing annealing temperature. In general, surface passivation with hydrogen treatment increases the carrier lifetime of the film by reducing recombination of electron–hole pairs near the surface of the film [18,19]. Therefore, the formation of vacancies in the surfaces and bulk of the DLC film caused by the hydrogen extrication during the annealing treatment is supposed to cause a decrease in the carrier lifetime [20].

[1] [2] [3] [4] [5] [6] [7]

4. Conclusions

[8]

We investigated the effect of post-annealing on the properties of DLC antireflection films prepared using PECVD. The structural properties of the annealed DLC films were characterized by Raman spectroscopy and XPS. The ID/IG ratios and G-peak wavenumbers increased with increasing post annealing temperature and the sp3 content decreased from 75.2% to 24.1% while the sp2 content increased from 24% to 75.9% as the annealing temperature increased from 300 to 900 1C. These were confirmed by

[9] [10] [11] [12] [13] [14]

Grill A. Thin Solid Films 1999;355:189. Milne WI. Semicond Sci Technol 2003;18:S81. Robertson J. Surf Coat Technol 1992;50:185. Kutsay OM, Gontar AG, Novikov NV, Dub SN, Tkach VN, Gorshtein BA, et al. Diamond Relat Mater 2001;10:1846. Choi WS, Hong B, Jeon Y, Kim K, Yi J. J Korean Phys Soc 2004;45:S864. Martinu L, Poitras D. J Vac Sci Technol A 2000:2619. Cheng C-L, Chia C-T, Chiu C-C, Wu C-C, Lin I-N. Diamond Relat Mater 2001;10:970. Friedmann TA, McCarty KF, Barbour JC, Siegal MP, Dibble DC. Appl Phys Lett 1996;68:1643. Anders S, Diaz J, Ager III JW, Lo RY, Bogy DB. Appl Phys Lett 1997;71:3367. Ferrari AC. Diamond Relat Mater 2002;11:1053. Tang Z, Zhang ZJ, Narumi K, Xu Y, Naramoto H, Nagai S, et al. J Appl Phys 2001;89:1959. Anders S, Anders A, Brown IG, Wei B, Komvopoulos K, Ager III JW, et al. Surf Coat Technol 1994;68–69:388. Ager III JW, Anders S, Anders A, Brown IG. Appl Phys Lett 1995;66:3444. Beeman D, Silverman J, Lynds R, Anderson MR. Phys Rev B 1984;30:870.

ARTICLE IN PRESS W.S. Choi, B. Hong / Renewable Energy 33 (2008) 226–231 [15] Leung TY, Man WF, Lim PK, Chan WC, Gaspari F, Zukotynski S. J Non-Cryst Solids 1999;254:156. [16] Huang L-Y, Lu J, Xu K-W. Mater Sci Eng A 2004;373:45. [17] Chan HL, Ekanayake U, Kumar A, Alam MR, You Q, Inturi RB, et al. Appl Surf Sci 1997;109–110:58.

231

[18] Kakiuchi H, Kobayashi T, Terai T. Nucl Instrum and Methods B 2000;166–167:415. [19] Ma¨ckel H, Lu¨demann R. J Appl Phys 2002;92:2602. [20] Leguijt C, Lo¨lgen P, Eikelboom JA, Weeber AW, Schuurmans FM, Sinke WC, et al. Sol Energ Mater Sol Cells 1996;40:297.