Study of Al-Ti/Si bi-layer as the recording media for write-once HD-DVD optical disks

Contents lists available atScienceDirect

Materials Chemistry and Physics

Study of Al–Ti/Si bi-layer as the recording media for

write-once HD-DVD optical disks

Hung-Chuan Mai

_{, Tsung-Eong Hsieh}

_{, Shiang-Yao Jeng}

_{, Chong-Ming Chen}

_{, Jen-Long Wang}

b_{Prodisc Technology Inc., No. 35, Wu-Kung 6th Road, Wugu Shiang, Taipei 242, Taiwan, ROC}

a r t i c l e i n f o

Received 9 May 2007

Received in revised form 15 July 2008 Accepted 9 August 2008

Keywords:

Write-once optical disk Bi-layer recording structure Signal properties Microstructure

a b s t r a c t

Optimum structure for HD-DVD optical disks containing Al–Ti/Si bi-layer recording system was identified by reflectivity simulation and dynamic test of disk samples. For the disk sample with optimized structure, the maximum partial response signal-to-noise ratio (PRSNR) of 19.1 dB, minimum simulated bit error rate (sbER) of 1.7× 10−7_{and modulation >0.6 were achieved at the writing power (Pw) = 11.2 mW. Transmission} electron microscopy (TEM) revealed that the polycrystalline granular clusters constitute the recording marks. Subsequent analyses evidenced that element mixing/alloy reactions occur in between Si and Al–Ti layers and the formation of Al3.21Si0.47crystalline phase is responsible for the signal recording in the disk samples.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Optical disks are the most important external data storage devices for multimedia or personal data files because of their large recording capacity, interchangeability and media portabil-ity[1–3]. One of the main trends for the development of optical disks is to increase the recording capacity as it can be seen from the evolution of compact disk (CD) [4]into digital versa-tile disks (DVDs)[5–6]. At present, high-density DVD (HD-DVD) and Blu-ray (BD-R) disk are recognized as the optical recording media for next-generation and many studies relating to recording materials have been reported[7–17]. In addition to phase-change alloys and organic dye materials, single-layer alloys such as AlSi [7–9], BiGe [10,11] and bismuth iron oxide (BiFeO) [12,13], bi-layered structures such as Cu/Si [14], Cu/a-Si[15], ZnO/Ge [16], and Ge/Au[17]have been proposed for the write-once type opti-cal recording. Their recording mechanisms are generally resulted from the thermally activated element mixing and/or alloy reac-tions when the recording stack is subjected to the irradiation of laser beam.

In this work, disk structure optimization, signal properties, microstructure change and recording mechanism of write-once HD-DVD disk containing Al–Ti/Si bi-layer as the recording media were investigated. The Al–Ti alloy and Si were selected since they are relatively inexpensive materials and can be easily acquired.

∗ Corresponding author.

E-mail address:[email protected](T.-E. Hsieh).

In addition, Al and its alloys form thin, dense oxide layers to inhibit further oxidation in their interiors. Such a self-protecting feature implies a better reliability of disks containing Al alloys. Fur-thermore, from the production point of view, this allows an easy maintenance of vacuum deposition chambers handling Al alloys in comparison with those for the production of other disks types containing recording media such as Cu/Si. Satisfactory signal prop-erties, PRSNR = 19.1 dB, sbER = 1.7_{× 10}−7and modulation >0.6 at the optimized writing power (Pw) of 11.2 mW, were achieved in the

disk samples containing Al–Ti/Si bi-layer with optimized structure. TEM characterizations revealed that polycrystalline agglomerates constitute the recording marks regime. The alloy phase type and element distribution were analyzed and the mechanisms of signal recording was discussed.

2. Experimental

2.1. Optimization of disk structure

Optimization of disk structure was carried out by, first, substituting the opti-cal constants including refraction index (n) and absorption coefficient (k) for each layer of optical disk into Spectrum Simulation, a self-developed software, to identify the preliminary disk structures that possess reflectivities in the range of 16–32% as required by HD-DVD Specifications[18].Table 1lists the optical constants adopted for the simulation. According to the preliminary simulation result, thickness ranges for each of layers were then designated and various disk samples were constructed by DOE (Design of Experiment) method as depicted inTable 2. Accordingly, the disk samples were fabricated via sputtering process and their reflectivities and PRSNR were measured by the dynamic test. The measured results are shown inTable 2and they were subsequently substituted into Minitab Release 14 Statistical Software to identify the primary affect factor on the disk signal properties. After determining the layer type that affects most on the signal properties, the disk layer structure was 0254-0584/$ – see front matter © 2008 Elsevier B.V. All rights reserved.

Table 1

Optical constants for disk structure simulation

Layer type Optical constants (for = 405 nm)

n k ZnO–SiO2 2.14 0.01 Si 4.18 0.72 Al–Ti 0.41 3.80 Ag 0.23 2.40 PC 1.62 0.00

fine-tuned accordingly, and in association with the reflectivity and PRSNR measure-ments, in order to identify the optimum disk structure.

2.2. Disk sample preparation

Optical disk samples were prepared according to the HD-DVD Specifications

[18]by using a self-designed six-target sputtering system at a background pressure better than 10−6_{torr. The five-layered disk structure was constituted by pre-grooved}

PC substrate (0.6 mm), lower ZnS–SiO2dielectric layer, Si/Al–Ti bi-recording layer,

upper ZnS–SiO2dielectric layer and Ag reflection layer with cross-sectional structure

depicted inFig. 1. After the deposition of Ag layer, another 0.6-mm PC substrate was attached on to complete the preparation of optical disk samples.

2.3. Dynamic test

A dynamic tester (ODU-1000, Pulstec Industrial) equipped with a 405-nm laser diode and a numerical aperture (NA) of 0.65 was adopted to characterize the sig-nal properties of optical disk samples. The characterization was performed at clock frequency = 64.8 MHz, linear speed = 6.61 m s−1_{and the track pitch = 0.4␮m.} 2.4. Microstructure

The microstructures and phase identification of the recording marks were exam-ined by a transmission electron microscope (TEM, JEOL FX-II 2010) equipped with an energy dispersive spectrometer (EDS, Link ISIS 300). During the examination, selected area electron diffraction (SAED) was also performed in order to identify the crystallinity of phases. The plan-view TEM (PTEM) samples were prepared in accordance with the method reported by Chen et al.[19]. The disk sample was first cut into small pieces by using scissors. After dissolving the PC substrate in CH2Cl2

solution, the specimen was mounted on a copper (Cu) mesh and transferred to the TEM for microstructure characterization. The cross-sectional TEM (XTEM) samples were also prepared by using an ultramicrotomy (Leica Reichert Ultracut S) so as to identify the layer thickness of optical samples.

3. Results and discussion

3.1. Optimization of disk structure

Table 2lists the disk samples constructed by DOE method and the signal properties measured by the dynamic test. It is noted that these are the preliminary values measured prior to the fine tune of

Fig. 1. Cross-sectional structure of optical disk samples.

disk layer structure and they simply serve as the guideline for the identification of optimum disk structure described as follows.

An analysis of signal properties shown inTable 2indicates that thickness increment of lower/upper ZnS–SiO2dielectric layer and

Al–Ti recording layer by 1 nm causes 0.5% decrease of reflectivity and thickness increment of Si recording layer by 1 nm causes 0.4% decrease of reflectivity.Table 2also shows that the PRSNR increases with the increase of reflectivity that the PRSNR becomes higher than 2.5 dB when reflectivity exceeds 4%.Fig. 2(a)–(e) depict reflec-tivities and PRSNR’s of the disk samples as a function of thickness. Among them, the steepest slopes occur in the plot for Al–Ti layer (seeFig. 2(c)). Since it possesses higher thermal conductivity and lower melting point than that of Si, the Al–Ti layer is thus the key element to ignite the alloy reaction with Si to form the recording marks when the laser beam irradiates on the bi-layer structure.

Table 2

The disk samples constructed by DOE method and the measured reflectivities and PRSNR by the dynamic test

Sample designation Thickness (nm) Reflectivity (%) PRSNR (dB) Sensitivity (mW)

Lower ZnS–SiO2 Si Al–Ti Upper ZnS–SiO2 Ag

1 30 12 9 10 100 3.99 <1 11 2 35 12 9 10 80 3.59 <1 11 3 30 16 9 10 80 4.96 2.5 11 4 35 16 9 10 100 1.84 <1 11 5 30 12 14 10 80 1.37 <1 11 6 35 12 14 10 100 2.13 <1 11 7 30 16 14 10 100 1.63 <1 11 8 35 16 14 10 80 < 1 <1 11 9 30 12 9 15 80 6.62 18.5 11 10 35 12 9 15 100 4.3 2.5 11 11 30 16 9 15 100 5.20 3 11 12 35 16 9 15 80 4 2.5 11 13 30 12 14 15 100 1.49 <1 11 14 35 12 14 15 80 1.35 <1 11 15 30 16 14 15 80 2.67 <1 11 16 35 16 14 15 100 1.51 <1 11

Fig. 2. The reflection of disk sample as a function of thickness of (a) lower ZnO–SiO2layer, (b) Si recording layer, (c) Al–Ti recording layer, (d) upper ZnO–SiO2layer and (e)

Ag reflection layer calculated by Minitab software.

Hence, the thickness of Al–Ti layer is the primary affect factor for the structure of HD-DVD disk samples as illustrated byFig. 2(c).

The iteration of sample preparation by fine tuning the Al–Ti layer thickness of optical disk sample to achieve the highest reflectivity (i.e., sample 9 in Table 2), and with the aid of reflectivity mea-surement and XTEM observation, eventually led to the optimum optical disk structure as: PC substrate/ZnS–SiO2(35 nm)/Si(9 nm)/

Al–Ti(9 nm)/ZnS–SiO2(15 nm)/Ag(90 nm)/PC substrate. Fig. 3

shows the XTEM image of the optimized optical disk structure. Subsequent signal characterizations and microstructure obser-vations were carried out based the disk samples with such an optimized structure.

3.2. Signal properties

Fig. 4presents the variation of PRSNR and sbER with the writing power (Pw) of disk samples obtained by dynamic test. According

to the HD-DVD Specifications[18], the PRSNR must be higher than 15 dB and the sbER must be lower than 10−5. As seen inFig. 4, the optical disk prepared in this work achieved the maximum PRSNR of 19.1 dB and the minimum sbER of 1.7× 10−7 _{at the optimized}

Pw of 11.2 mW. Further, the modulation of optimized disk

sam-ple was found to be greater than 0.6. As stated previously, various

Fig. 4. Variation of PRSNR and sbER of disk samples versus the writing power (Pw).

blue-laser era[7–17]. For single-layer AlSi alloy applied to HD-DVD, PRSNR≈ 18.5 dB, sbER ≈ 2 × 10−5_{, jitter value <7% and the}

modula-tion >0.4 were reported[7–9]. The HD-DVD disk containing Ge/Au bi-layer achieved PRSNR = 16 dB and the sbER = 9.9× 10−7 _{at 1×}

recording speed[17]. The results presented above show that the Al–Ti/Si bi-layer disk samples possess a better disk signal properties in comparison with those reported by previous studies. As revealed by the TEM analysis presented in next section, the Al–Ti/Si bi-layer produced by sequential sputtering deposition is amorphous and such a uniform background is beneficial to the signal contrast when marks are written in the disk samples. Compared with the disks containing single-layer AlSi alloy, it is speculated that the as-deposited recording layer, if it were prepared by using a hyper-eutectic AlSi alloy target[9], is likely to be a mixture of Al–12.6 wt.% Si eutectic phase and solid solution Si according to the Al–Si binary phase diagram[20]. The two alternative phases might result in some structural inhomogeneity. This raises the background noise and hence the inferior signal properties of optical disks in compar-ison with those achieved in this study[7–9].

Eye pattern corresponding to the signals ranging from 2 T to 11 T read directly from oscilloscope is shown inFig. 5. Though the pat-tern looks somewhat blurry, it illustrates that the random signals could be satisfactorily written in the disk samples containing the Al–Ti/Si bi-layer structure. We note that this work is a preliminary study on the feasibility of Al–Ti/Si bi-layer as the write-once record-ing medium and it was carried out by adoptrecord-ing the PC substrate and writing strategy available at hand without further optimizations. This leads to some seemingly low signal values, for instance, the “highest” reflectivity listed inTable 2is only 6.62% which might result from the deep land-and-groove geometry of PC substrate adopted for disk sample preparation. It is believed that the better signal properties and sharp eye patterns could be accomplished if

Fig. 5. Eye patterns for random signals read directly from oscilloscope.

Fig. 6. PTEM image of a recording mark in optical disk sample. The SAED patterns

attached at upper and lower right-hand corners were taken from the areas outside and right on the mark, respectively.

the writing strategy and PC substrate geometry were subsequently adjusted.

3.3. Microstructure

Fig. 6shows the PTEM image of signal marks in a recorded disk sample. The inserted SAED pattern at the upper right-hand corner is taken form the area outside of the mark while the pat-tern at the lower right-hand corner corresponds to the signal mark regime. TEM characterization shown inFig. 6clearly reveals that the polycrystalline agglomerate with grain sizes about 30–40 nm embedded in amorphous matrix constitutes the signal mark. We calculated the d-spacing corresponding to the each of the diffrac-tion rings in SAED pattern of polycrystalline agglomerate, and with the aid of Joint Committee of Powder Diffraction Standard (JCPDS file card No. 41-1222), to identify the alloy phase in the signal marks as Al3.21Si0.47. Apparently, the alloy reaction of Al with Si induced

by laser irradiation in the bi-layer structure generated such a poly-crystalline phase and was responsible for the signal recording in our disk samples.

Fig. 7shows the variation of composition in the signal mark. The composition analysis was carried out along a line across the mid-dle of mark regime by using the EDS attached to TEM. At the edges of mark (i.e., positions 1 and 5), distinct signal intensities of Si and Al were observed. This should result from the bi-layer recording structure that the upper layer of TEM sample (in this case Si should be the upper layer) produces more X-ray signal when detecting electrons reach the sample. The EDS signals of Al and Si gradually merge together in the middle of mark, indicating the occurrence of element mixing of Al and Si. This also confirms the alloy reac-tions and formation of crystalline phase described above. As shown inFig. 7, the EDS analysis also revealed a rather smooth signal variation of Ti element across the signal mark. In this work, the Al–1.5 wt.%Ti sputtering target was adopted for disk sample prepa-ration. The comparatively small Ti content implies an insignificant concentration change of Ti around the signal marks. Thus, Ti ele-ment plays a negligible role on the signal properties of disk samples and phase constitution of signal marks.

Fig. 7. EDS analysis of composition variation of the recording mark shown inFig. 6.

4. Conclusions

The disk structure optimization and recording mechanism of write-once HD-DVD disk containing Al–Ti/Si bi-layer recording sys-tem were investigated in this work. Simulation study indicated that the thickness of Al–Ti recording layer is the primary factor affecting the optical reflectivities of disk samples. For the disk sample with optimized structure deduced by simulation study and preliminary signal property measurement, PRSNR = 19.1 dB, sbER = 1.7× 10−7

and modulation >0.6 were achieved at Pw= 11.2 mW as revealed

by dynamic test. The TEM characterization revealed the forma-tion of polycrystalline recording mark in the disk samples. Phase identification and EDS composition analysis evidenced that ele-ment mixing/alloy reactions occurs in between Si and Al–Ti layers

and formation of Al3.21Si0.47crystalline phase is responsible for the

signal recording of disk samples.

Acknowledgement

This work was supported by the National Science Council of the Republic of China under contract NSC93-2216-E-009-008.

References

[1] M. Takenaga, N. Yamada, K. Nishiuchi, N. Akahira, T. Ohta, S. Nakamura, T. Yamashita, J. Appl. Phys. 54 (1983) 5376.

[2] Y. Unno, K. Goto, Proc. SPIE 382 (1983) 32.

[3] M. Terao, S. Horigome, K. Shigematsu, Y. Miyauchi, M. Nakazawa, J. Appl. Phys. 62 (1987) 1029.

[4] B. Tieke, G.R. Langereis, E.R. Meinders, J.G.F. Kablau, R. Woudenberg, R.A.J. van Kollenburg, Jpn. J. Appl. Phys. 41 (2002) 1735.

[5] K.S. Min, Y.J. Huh, H.K. Shim, Jpn. J. Appl. Phys. 38 (1999) 1675.

[6] K. Nishiuchi, H. Kitaura, N. Yamada, N. Akahira, Jpn. J. Appl. Phys. 37 (4B) (1998) 2163.

[7] B.M. Chen, R.L. Yeh, Proc. SPIE 5380 (2004) 141.

[8] B.M. Chen, H.F. Chen, R.L. Yeh, J.M. Chung, Jpn. J. Appl. Phys. 43 (7B) (2004) 5018.

[9] B.M. Chen, J.H. Chen, L.C. Chung, Patent No. 220523, Taiwan, ROC, August (2004). [10] Y. Hosoda, T. Izumi, A. Mitsumori, F. Yokogawa, S. Jinno, H. Kudo, Jpn. J. Appl.

Phys. 42 (2003) 1040.

[11] Y. Hosoda, A. Mitsumori, M. Sato, M. Yamaguchi, Jpn. J. Appl. Phys. 43 (2004) 4997.

[12] N. Sasa, Y. Hayashi, T. Fujii, K. Otaka, A. Watada, H. Kamezaki, H. Komoda, Jpn. J. Appl. Phys. 43 (2004) 4972.

[13] N. Sasa, Y. Hayashi, T. Fujii, A. Watada, H. Komoda, Jpn. J. Appl. Phys. 44 (2005) 3643.

[14] A.E.T. Kuiper, R.J.M. Vullers, D. Pasquariello, E.P. Naburgh, Appl. Phys. Lett. 86 (2005) 221921.

[15] Y.C. Her, C.L. Wu, Jpn. J. Appl. Phys. 43 (2006) 1013.

[16] C.P. Liu, Y.T. Hung, J.S. Tsai, G.J. Huang, T.R. Jeng, Jpn. J. Appl. Phys. 46 (2007) 7365.

[17] T.H. Wu, P.C. Kuo, Y.H. Fang, J.P. Chen, P.F. Yen, T.R. Jeng, C.Y. Wu, D.R. Huang, Appl. Phys. Lett. 90 (2007) 151111.

[18] DVD Specifications for High-Density Recordable Disc (HD DVD-R) for Blue Laser Optical System, Version 0.9, The DVD Forum, July, 2004.

[19] H.W. Chen, T.E. Hsieh, J.R. Liu, H.P.D. Shieh, Jpn. J. Appl. Phys. 38 (1999) 1691. [20] Thaddeus B. Massalski (editor-in-chief), Binary Alloy Phase Diagrams, 2nd ed.,