We have designed and grown triply doped LiNbO3:Zr,Cu,Ce crystal and investigated its characteristics of nonvolatile holographic storage. It’s observed that the photorefractive sensitivity of LiNbO3:Zr,Cu,Ce has improved to 0.099 cm/J, which is about one order of magnitude larger than that of congruent LiNbO3:Cu,Ce. And LiNbO3:Zr,Cu,Ce also has high suppression to light-induced scattering. Our results indicated that triply doped LiNbO3:Zr,Cu,Ce is an excellent candidate for nonvolatile holographic data storage.
©2010 Optical Society of America
Owing to huge data capacities, short access times, and high data transfer rates, photorefractive crystals have been extensively investigated for applications in holographic data storage [1–4]. However, there remains a crucial problem of volatility: the stored information will be erased during readout. Several techniques such as thermal fixing , electrical fixing  and two-step recording  have been developed to overcome this problem. Unfortunately, it is not practical for heating or application of large external fields during holographic storage, and the stored information cannot be rapidly refreshed. Two-step recording, as an all-optical approach for nonvolatile storage, causes the substantial information losses due to the Bragg condition of diffraction when recording with one light and reading with another light of a longer wavelength . The outlined disadvantage makes the nonvolatile holographic storage impractical.
Recently, using two coherent recording beams in the presence of a sensitizing beam with shorter wavelength, good nonvolatile holographic storage was achieved in doubly doped LiNbO3:Fe,Mn . Later, Liu et al.  reported LiNbO3:Cu,Ce was another crystal that can be used for nonvolatile holographic storage. Cu and Ce ions occur in the valence states Cu+/2+  and Ce3+/4+ , respectively, which are regarded as shallow and deep centers . They achieved nonvolatile holographic storage with two illumination schemes, i.e., an ultraviolet beam for sensitizing and a red interfering pattern for recording or a blue beam for sensitizing and a red interfering pattern for recording. This all-optical method makes practical application of nonvolatile storage more possible. However, the sensitivity obtained by this process still remains very low. To realize real-time nonvolatile read-write memory, improving the sensitivity seems to be extremely important.
One method to improve the sensitivity of two-color nonvolatile holographic recording is to increase the ratio of [Li/Nb], in other words, growing near-stoichiometric crystals. Li et al.  have obtained LiNbO3:Cu,Ce crystals with various compositions through vapor transport equilibration (VTE) treatment and achieved nonvolatile holographic storage. The sensitivity of near-stoichiometric LiNbO3:Cu,Ce (sLiNbO3:Cu,Ce) is one order of magnitude larger than that of congruent LiNbO3:Cu,Ce crystals. However, one can only obtain thin plates by VTE technique, and it is very difficult to grow large size sLiNbO3:Cu,Ce crystals with good optical quality, so it is not practical to use near-stoichiometric crystals in nonvolatile holographic storage.
On the other hand, doping optical damage resistant ions with concentration above threshold can greatly shorten the response time . Recently, it has been reported that the response time of LiNbO3 codoped with Zr and Fe in norrnal volatile photorefractive storage is only 2 s while the saturation diffraction efficiency still remains at a high level, hence the sensitivity is as large as 12 cm/J, which is much better than other codoped crystals such as LiNbO3:Mg,Fe . Similar situation happened when doping Zr into LiNbO3:Fe,Mn, the sensitivity of nonvolatile holographic recording is also improved effectively to 1.31 cm/J .
In addition, as propagating in LiNbO3 crystals, a laser beam often leads to the appearance of different patterns of scattered light known as light-induced scattering (LIS) , which arises from amplification of the scattered light due to the nonlinear coupling with the pump beam on noisy photorefractive gratings. The light-induced scattering strongly impedes the progress of holographic volume storage in LiNbO3 crystals. Thus it becomes very important for nonvolatile holographic storage to find a material which has weak light-induced scattering as well as large sensitivity. Although the sensitivity of LiNbO3:Zr:Fe,Mn is improved, the light-induced scattering is still very large. It has been reported that LiNbO3:Cu,Ce has weak light-induced scattering as compared with LiNbO3:Fe,Mn . If the high sensitivity of LiNbO3:Zr:Fe,Mn and weak LIS of LiNbO3:Cu,Ce can be combined, it will be very useful for nonvolatile holographic recording. In order to achieve this propose, we grew the LiNbO3 triply doped with Zr, Cu and Ce (LiNbO3:Zr,Cu,Ce). By doping Zr into LiNbO3:Cu,Ce, the sensitivity is greatly improved as compared with LiNbO3:Cu,Ce, meanwhile, the LIS is obviously suppressed as compared with LiNbO3:Zr:Fe,Mn.
2. Nonvolatile holographic storage
The conventional Czochralski method was used to grow LiNbO3:Zr,Cu,Ce crystals along the c axis. The ratio of [Li]/[Nb] was selected as 48.38/51.62. 0.011 wt% CuO, 0.085 wt% Ce2O3 and 2 mol% ZrO2 were doped into the melt. For characterization of two-color holographic recording, the crystal were cut to 3 mm plates along the y faces and optically polished after annealing treatment and artificial polarization.
The traditional setup for two-color holographic recording was utilized (as shown in Ref. 13). A 400 mercury lamp and a semiconductor laser were used as the sensitizing and recording lights, respectively. After being preexposed to the UV light (wavelength 365 nm, intensity 40 mW/cm2) for at least 1 h, LiNbO3:Zr,Cu,Ce specimens were used to record holograms using two coherent green beams (wavelength 532nm, intensity per beam 200, 300, and 400 mW/cm2, respectively) with the UV sensitizing light on (recording process), until saturation was reached. Then specimens were illuminated by only one green beam to fix holograms (readout process). The diffraction efficiency, as the ratio between the diffracted and incident light intensities, was monitored via blocking one green beam from time to time and diffracting the second beam from the written grating simultaneously. The typical time dependence of diffraction efficiency is shown in Fig. 1 . The saturation and nonvolatile diffraction efficiency (denoted as ηsat and ηnon, respectively) is 69.1% and 5.9% respectively. Extrapolation of this readout experiment shows that the diffraction efficiency drops to approximately 5.7% for continuous readout over one week. As shown in Fig. 1, the nonvolatile holographic storage is achieved in LiNbO3:Zr,Cu,Ce.
The measure of recording speed, sensitivity (S), is normally calculated as Eq. (1) does not include the effect of partial erasure during readout in two-color holographic recording. It can be modified to obtain a better measure of sensitivity in two-color holographic recording :
The measured S and S’ of LiNbO3:Zr,Cu,Ce oxidized 13 h in air at 700 °C is 0.312 cm/J and 0.099 cm/J, respectively. The sensitivity S of LiNbO3:Zr,Cu,Ce is exceptionally improved one order of magnitude compared with doubly doped congruent LiNbO3:Cu,Ce (the same dopant concentrations of Cu and Ce as our sample, respectively) , and the value of S’ is comparable to that of near-stoichiometric LiNbO3:Cu,Ce . Furthermore, during the last decade some doped LiNbO3 materials were developed for the nonvolatile holographic recording, in the view of comparison, the sensitivity of different doped lithium niobate crystals are outlined in Table 1 .
It was reported the oxidation/reduction state of crystal and the ratio of sensitizing and recording intensities (Isen/Irec) have profound impacts on the nonvolatile holographic recording properties [23–25]. Thus LiNbO3:Zr,Cu,Ce specimens were oxidized by annealing at a temperature of 700 °C in air for 13 h and 24 h, respectively. Holographic recording was measured with different Isen/Irec. The experimental results were listed in Table 2 , we can see when Isen/Irec equals to 40/400 the 13 h oxidized sample has higher photorefractive sensitivity.
For the nonvolatile holographic recording in LiNbO3:Cu,Ce crystal, Cu+/2+ and Ce3+/4+ ions act as shallow and deep photorefractive centers respectively. At the beginning only homogeneous UV light is present, electrons are excited from Ce3+ ions to Cu2+ ions. During recording, some electrons are excited to conduction band from Cu+ ions by the interference light, moving a short distance, and then trapped by Cu2+ ions. After cycles of excitation, movement, and trapping, the space charge field is built-up in the Cu+/2+ ions. And the electrons trapped by Ce3+/4+ ions also built up a space charge field opposite to that built-up in Cu+/2+ ions. The holograms are recorded in both Cu+/2+ and Ce3+/4+ ions. In the period of readout, the green light removes the electrons from Cu+ ions until all of them are trapped in Ce4+ ions, and then readout becomes non-volatile. As we know lithium niobate is a non-stoichiometric crystal with plenty of intrinsic defects such as Li-site vacancies (VLi -) and charge compensated anti-site Nb5+ ions () . In nominally pure congruent LiNbO3 the photorefractive centers are mainly small polarons () and bipolarons () . Therefore in doubly doped congruent LiNbO3:Cu,Ce crystals electrons may be also trapped by intrinsic photorefractive centers during moving in conduction band, which greatly increases the average time of each excitation, movement, and trapping cycle of Cu+/2+ ions, resulting in a very low sensitivity in two-color holographic recording. When Zr ions are doped into LiNbO3:Cu,Ce crystals with concentrations above threshold, these Zr ions will substitute antisite Nb ions completely, therefore small polarons and bipolarons are mostly disappeared. Consequently the moving distances as well as lifetimes of electrons excited from dopants become longer. Most electrons trapped by Cu+/2+ ions directly. The average time of each cycle will be decreased greatly as compared with that in LiNbO3:Cu,Ce. So the response speed and the sensitivity of two-color holographic storage in LiNbO3:Zr,Cu,Ce is improved.
From Table 2 we can see that the nonvolatile diffraction efficiency of LiNbO3:Zr,Cu,Ce is a little lower than that of LiNbO3:Cu,Ce . That might be related to the site occupation change of Ce ions when Zr ions were codoped into. Because part of Ce ions have the same tetravalence as Zr ions, just like what happens in LiNbO3:Mg,Fe , when the concentration of Zr exceeds its doping threshold, a part of Ce4+ ions will be pushed to Nb sites from Li sites, these Ce4+ ions cannot trap electrons and lost their ability as photorefractive acceptors. As a result, the hologram stored in Ce3+/4+ centers is decreased and the nonvolatile diffraction efficiency is a little lower. And the same reason may also contribute to the lower sensitivity of LiNbO3:Zr,Cu,Ce as compared with LiNbO3:Zr,Fe,Mn.
3. Light-induced scattering
The experimental arrangement measuring light-induced scattering is schematically shown in Fig. 2 . The pump light (e-polarized, wavelength 532nm, beam diameter 1.5mm) is incident normally onto the sample and perpendicularly to the crystalline c-axis. In order to investigate the dependence of LIS on the incident light intensity, we use a neutral density filter to vary the incident light intensity and a screen to block the scattered light. Moreover, a reference light, detected by D2 (shown in Fig. 2), is taken out in front of the sample to monitor the laser power fluctuation. The sample used here is the same as the above nonvolatile holographic recording. For comparison, LiNbO3:Zr,Fe,Mn (0.075 wt% Fe2O3, 0.01 wt% MnO and 2 mol% ZrO2) was also investigated, which was observed having a short response time for nonvolatile holographic recording in our recent work .
During experiment the transmitted pump light was monitored by D1 connected to a computer. The measured change of the transmitted light is shown in Fig. 3 , from which we can get the scattered light intensity Is=It0-It1, where It0 is the transmitted light intensity just at the beginning and It1 the intensity at saturation. A unified scattered light intensity value R is defined as the strength of LIS, which is proportional to Is/Ii, where Ii is the incident light intensity. Here we define R as R=Is/It0 in order to eliminate the absorption through crystal. The light intensity dependence of LIS is shown in Fig. 4 . We can see the LIS of LiNbO3:Zr,Cu,Ce is only 0.17 at the light intensity of 1400 mW/cm2, which is much smaller than that of LiNbO3:Zr,Fe,Mn. Hence LiNbO3:Zr,Cu,Ce exhibits an enhanced ability to suppress LIS, which is very promising for holographic recording application.
In summary, we have designed and grown triply doped LiNbO3:Zr,Cu,Ce crystal. As compared with congruent LiNbO3:Cu,Ce crystal, the photorefractive sensitivity has been greatly improved in LiNbO3:Zr,Cu,Ce. And the light-induced scattering in LiNbO3:Zr,Cu,Ce is suppressed effectively, which is much lower than that in LiNbO3:Zr,Fe,Mn. According to our experimental results, triply doped LiNbO3:Zr,Cu,Ce has the low light induced scattering and high photorefractive sensitivity, therefore it is an outstanding candidate for nonvolatile holographic storage.
This work was supported by Chinese National Key Basic Research Special Fund (No.2006CB921703), National Basic Research Program of China (No. 2007CB307002), and the National Advanced Materials Committee of China (No. 2007AA03Z459).
References and links
1. L. Hesselink, S. S. Orlov, A. Liu, A. Akella, D. Lande, and R. R. Neurgaonkar, “Photorefractive materials for nonvolatile volume holographic data storage,” Science 282(5391), 1089–1094 (1998). [CrossRef] [PubMed]
2. Y. S. Bai and R. Kachru, “Nonvolatile Holographic Storage with Two-Step Recording in Lithium Niobate using cw Lasers,” Phys. Rev. Lett. 78(15), 2944–2947 (1997). [CrossRef]
4. L. Dhar, K. Curtis, and T. Fäcke, “Holographic data storage: coming of age,” Nat. Photonics 2(7), 403–405 (2008). [CrossRef]
5. J. J. Amodei and D. L. Staebler, “Holographic pattern fixing in electro-optic crystals,” Appl. Phys. Lett. 18(12), 540–542 (1971). [CrossRef]
6. F. Micheron and G. Bismuth, “Electrical control of fixation and erasure of holographic patterns in ferroelectric materials,” Appl. Phys. Lett. 20(2), 79–81 (1972). [CrossRef]
7. D. von der Linde, A. M. Glass, and K. F. Rodgers, “Multiphoton photorefractive processes for optical storage in LiNbO3,” Appl. Phys. Lett. 25(3), 155–157 (1974). [CrossRef]
8. H. C. Külich, “A new approach to read volume holograms at different wavelengths,” Opt. Commun. 64(5), 407–411 (1987). [CrossRef]
9. D. Psaltis, K. Buse, and A. Adibi, “Non-volatile holographic storage in doubly doped lithium niobate crystals,” Nature 393(6686), 665–668 (1998). [CrossRef]
10. Y. Liu, L. Liu, C. Zhou, and L. Xu, “Nonvolatile photorefractive holograms in LiNbO3:Cu:Ce crystals,” Opt. Lett. 25(12), 908–910 (2000). [CrossRef]
11. T. Vitova, J. Hormes, K. Peithmann, and T. Woike;, “X-ray absorption spectroscopy study of valence and site occupation of copper in LiNbO3:Cu,” Phys. Rev. B 77(14), 144103 (2008). [CrossRef]
12. A. Darwish, M. D. Aggarwal, J. Mortis, J. Choi, J. C. Wang, P. Venkateswarlu, A. Willimas, P. P. Banerjee, D. McMillen, and T. D. Hudson, “Investigations of the charge transfer and the photosensitivity in single and double doped LiNbO3 single crystals: an optical-electron paramagnetic resonance study: I,” Proc. SPIE 3137, 63–74 (1997). [CrossRef]
13. X. Li, Y. Kong, Y. Wang, L. Wang, F. Liu, H. Liu, Y. An, S. Chen, and J. Xu, “Nonvolatile holographic storage of near-stoichiometric LiNbO3:Cu:Ce with green light,” Appl. Opt. 46(31), 7620–7624 (2007). [CrossRef] [PubMed]
14. G. Zhang, J. Xu, S. Liu, Q. Sun, G. Zhang, Q. Fang, and C. Ma, “Study of resistance against photorefractive light-induced scattering in LiNbO3:Fe,Mg crystals,” Proc. SPIE 2529, 14–17 (1995). [CrossRef]
15. Y. Kong, S. Wu, S. Liu, S. Chen, and J. Xu, “Fast photorefractive response and high sensitivity of Zr and Fe codoped LiNbO3 crystals,” Appl. Phys. Lett. 92(25), 251107 (2008). [CrossRef]
16. Y. Kong, F. Liu, T. Tian, S. Liu, S. Chen, R. Rupp, and J. Xu, “Fast responsive non-volatile holographic storage in LiNbO3 triply doped Zr, Fe, and Mn,” Opt. Lett. 34(24), 3896–3898 (2009). [CrossRef] [PubMed]
17. M. Goulkov, S. Odoulov, T. Woike, J. Imbrock, M. Imlau, E. Krätzig, C. Bäumer, and H. Hesse, “Holographic light scattering in photorefractive crystals with local response,” Phys. Rev. B 65(19), 195111 (2002). [CrossRef]
18. P. Günter, and J.-P. Huignard, Photorefractive materials and their applications 1: Basic Effects, (Springer, Berlin, 2005).
19. A. Adibi, K. Buse, and D. Psaltis, “Sensitivity improvement in two-center holographic recording,” Opt. Lett. 25(8), 539–541 (2000). [CrossRef]
20. Q. Dong, L. Liu, D. Liu, C. Dai, and L. Ren, “Effect of dopant composition ratio on nonvolatile holographic recording in LiNbO3:Cu:Ce crystals,” Appl. Opt. 43(26), 5016–5022 (2004). [CrossRef] [PubMed]
22. W. Zheng, Q. Gui, and Y. Xu, “Defect structure and optical fixing holographic storage of Mg:Mn:Fe:LiNb3 crystals,” Cryst. Res. Technol. 43(5), 526–530 (2008). [CrossRef]
23. A. Adibi, K. Buse, and D. Psaltis, “Effect of annealing in two-center holographic recording,” Appl. Phys. Lett. 74(25), 3767–3769 (1999). [CrossRef]
24. Y. Liu, L. Liu, D. Liu, L. Xu, and C. Zhou, “Intensity dependence of two-center nonvolatile holographic recording in LiNbO3:Cu:Ce crystals,” Opt. Commun. 190(1-6), 339–343 (2001). [CrossRef]
26. O. F. Schirmer, O. Thiemann, and M. Wohlecke, “Defects in LiNbO3-I: experimental aspects,” J. Phys. Chem. Solids 52(1), 185–200 (1991). [CrossRef]