So far, visible light-induced domain reversal (LIDOR) in LiNbO3 has only been observed to occur under assistance of an external electric field. Here, we develop a two-step technique to directly achieve green-LIDOR for the first time. In the first step, an external electric field as well as the green laser of 532 nm wavelength is applied on the LiNbO3 crystal. In the second step, direct writing of domain structures in LiNbO3 is realized by another 532 nm laser without assistance of external electric field. Green-LIDOR results from a light-induced space-charge field generated by the prior application on an external electric field in the first step. Due to the unique two-step method, our further experiments show that light-induced space-charge field in LIDOR enhances with increasing the applied external electric field. Therefore, we propose a carrier-drift model to this phenomenon.
© 2014 Optical Society of America
Domain engineering is one of the most important issues for ferroelectric LiNbO3 (LN) crystals. Electric field poling (EFP), i.e., by applying an external electric field above the coercive electric field, is the most popular method to fabricate various domain patterns. E.g., domain structures such as the one of periodically poled LiNbO3 (PPLN) are fabricated by this way and have found various applications like nonlinear frequency conversion devices, optical parametric oscillators (OPO) or electro-optic Bragg modulators [1–3]. Recently, light-induced domain reversal (LIDOR) was widely investigated because it can directly fabricate domain patterns in LN compared to EFP. Etch frustration is observed in congruent LN in the region irradiated by a femtosecond Ultraviolet (UV) laser (248 nm) . But it is not sure whether domain reversal occurs. Domain patterns are fabricated in LN using an Ar+ laser (488 nm) with assistance of an external electric field . Light-induced space-charge field is the dominant explanation in the visible-LIDOR. A systematic study of the wavelength dependence of light-induced domain nucleation reveals that the applied external electric field required for nucleation reduces with decreasing wavelength of the input light , which greatly supports the light-induced space-charge field theory. Domain reversal directly induced by a focused continuous wave UV laser (244 nm) is achieved in LN without assistance of an applied external electric field . It is attributed to the temperature change caused by the strong absorption in LN at 244 nm. Thus the pyroelectric effect is the dominant reason for domain reversal induced by UV light. Although the visible has low absorption in LN, pyroelectric effect due to absorption in the green was proven to play an important role in green-LIDOR . Therefore, the essential mechanism of LIDOR is still unclear and needs further investigation.
Reduction of the coercive field induced by illumination is widely studied. It was reported that the coercive electric field was reduced by as much as 30% at 488 nm laser in 2004.5 Then A decrease by an order of magnitude was achieved by a focused laser beam at the wavelengths of 514, 488, and 475 nm in Mg-doped LiNbO3 in 2005 . In 2010, an applied external electric field of 30 V/mm at 532 nm was reported to achieve LIDOR in near-stoichiometric Mg-doped LiNbO3, which is the lowest value of applied electric field reported so far . In this paper, we develop a two-step technique to first achieve green-LIDOR without assistance of an external electric field in the second step. Direct writing of domain patterns in LN by the green are realized like the direct-write domain patterns by deep UV light. Our further experimental results show that the light-induced space-charge field has a close relationship with the applied external electric field in the first step. Therefore, we give a carrier-drift model to explain our results.
2. Experimental setup and results
2.1 Two-step method to achieve green-LIDOR
The samples used in this study are 0.5 mm thick z-cuts of 1.0 mol% Mg-doped stoichiometric LiNbO3 provided by the R&D Center for Photo-Electro Materials of Nankai University. The experimental setup mainly involves a pair of insulating holders filled with water that allow applying a uniform electric field as well as illumination. Two frequency-doubled diode-pumped Nd: VYO4 lasers of 532 nm (laser 1 for the first step, and laser 2 for the second step) were used separately to perform the experiments. There are two methods to observe the process of domain reversal. One is an etching technique using HF acid and the other is a real-time imaging system. The real-time imaging system consists of a CCD connected with a computer to present real-time microscopic images of the domain reversal process and a polarized white light source as a reference light . There will be a bright spot emerging on the CCD screen in the reversed region due to the refractive index difference between the domain and the original region.
In the first step: an external electric field of 1.8kV/mm (which is less than coercive electric field Ec = 2kV/mm) was applied to the sample. Then a 532 nm beam (laser 1) was focused by a focal 70 mm lens to produce a spot of 200 μm diameter at an intensity of 0.44kW/cm2 on the sample as shown in Fig. 1(a).Then we first switched off irradiation and subsequently removed the external electric field. This sequence is crucial. Otherwise, nothing would happen in the next step. In the second step: a focused 532 nm beam (laser 2) of 5 μm diameter and 3.1 kW/cm2 intensity was applied on the -z surface of the crystal by a 20 × objective as shown in Fig. 1(b). Domain reversal occurred at the focus without assistance of an external electric field. Figures 1(c)-1(f) are the etching results of reversed domains after applying different external electric fields of 1, 2, 3, and 4kV/mm respectively. The writing sequence in each figure is from left to right and from bottom to top. It is clear that the size of the domain at high external electric field is larger than the one at the low external electric field.
Another experiment was designed that we turned over the LN samples in the two-step LIDOR and kept other parameters same to the above experiment. In this experiment, a big domain emerged in the first step as Fig. 2(a) because the applied electric field assisted the domain reversal. Then laser 2 was focused on the big domain as Fig. 2(b). Figure 2(c) is a microscopic image after the sample was etched in HF in 10 minutes. Figure 2(d) is a real-time image taken by the CCD in the process of domain reversal. Compared with each other, both images correspond very well. In fact, this experiment is in same mechanism to the first experiment. Domain reversal happened in the –z face in both two experiments.
2.2 Role of laser 1 intensity in the two-step LIDOR
To clarify how the intensity of laser 1 influences domain reversal in the two-step LIDOR, we changed the intensity of the first laser and kept the second laser constant. The results are showed in Figs. 3(a)-3(d). The intensity of the second laser is 28kW/cm2 and the applied electric field is 1.5kV/mm. The quantities of the first laser intensity from Figs. 3(a)-3(d) are 256, 128, 64, and 32kW/cm2, respectively. The domain sizes just slightly vary with different intensity quantities of the first laser.
2.3 Role of the external electric field in the two-step LIDOR
In the traditional LIDOR, the external electric field is used to assist to finish domain reversal. However, in the two-step LIDOR the external electric field prevents the domain reversal in the first step. We measured the minimum value of the external electric field to induce domain reversal in two-step LIDOR at the certain intensity of laser 1. Domain reversal is monitored by the real-time observing system. The intensity (laser 2) in the second step was kept constant at 2.6kW/cm2. The minimum values of the external electric field vary with different intensity of laser 1 as Fig. 4(a).Besides, the external electric field as high as 14kV/mm is applied in the first step. The results are as Fig. 4(b) (Media 1) and Fig. 4(c) (Media 1) which are excerpted from a movie. We found that domain reversal happened in the focus of laser 1 once laser 1 and the external electric field were switched off in the first step.
2.4 Investigation of domain etching and domain depth in the two-step method
The depth of ferroelectric domains is measured by x face polishing and etching in HF acid for 48h. Figure 5(a) is –z face domain pattern. Figure 5(b) is x face domain pattern corresponding to Fig. 5(a). The size of domains from left to right is 41, 31, 17 μm, respectively. The depth of domains from left to right is 214, 109, 36 μm, respectively. We can see that the first domain is almost half of crystal thickness and not just surface domain. The black pits in Fig. 5(b) should be related to crystal imperfections. Another interesting phenomenon is that the cycles of etching influences the size of some domains. Figure 5(c) is the domain patterns after 2 cycles of etching. Each etching time is about 30 minutes. We can see that there are two marks around the biggest domains. Figure 5(d) is the same domain pattern after 5 cycles of etching. There are five marks of the biggest domains. Each marks means that the domain becomes a little bigger than the former. After five cycles of etching, two growing domains merge together and become a big domain. We supposed that etching broke the balance of space-charge field and surface charge field. Therefore, domain had to grow bigger to reach a new balance.
3.1 Space-charge field in the two-step LIDOR
According to our knowledge, there is no report that domain reversal in LiNbO3 can be induced by visible light alone, i.e., without assistance of an external electric field. However, in our experiment, green-LIDOR indeed occurred in the second step without assistance of an external electric field. Both the external electric field and irradiation of laser 1 are crucial to two-step LIDOR in the first step. Either the external electric field or irradiation of laser 1 cannot achieve the two-step LIDOR. Carriers which are excited by laser 1 drift to different directions under the external electric field. The space-charge field depends on the separation of carriers excited by photons. As is known, green-light cannot cause domain reversal directly in LN. There should be an electric field which assists the domain reversal in the second step. Therefore we assume that the space-charge field produced in the reversed region in the first step assists domain reversal in the second step.
The establishment of the space-charge field in the first step depends on both the external electric field and the laser 1. From Figs. 1(c)-1(f), the first domain becomes bigger and bigger with increasing the external electric field. It indicates that the space-charge field enhances with the increase of the external electric field. From Fig. 4 (a), we can see that the space-charge field can be high enough to achieve two-step LIDOR by increasing the external electric field in the weak intensity of laser1. We increased the external electric field to really high value and LIDOR happened after the first step immediately without the second step as Figs. 4 (b) and 4 (c). Therefore, there is no threshold of the external electric field to increasing the space-charge field. This result is very important to the establishment mechanism of the space-charge field. The space-charge field increases with the increase of the input light intensity before the light intensity is up to the saturated value . From Fig. 4 (a), we know that there is a threshold of laser 1 intensity. When the intensity of laser 1 is larger than the threshold, the space-charge field changes little with different intensity of laser 1. That is why the domain sizes change little with different intensity of laser 1 in Figs. 3 (a)-3 (d).
In two-step LIDOR, domain sizes vary with the writing sequence in the same sample as Figs. 1(c)-1(f). It indicates that space-charge field becomes weaker and weaker with the appearance of more and more domains. It is reported that the dark conductivity of LN can weaken the light-induced space-charge field with time . However, we believe that the space-charge field is weakened by the second laser irradiation because the same situation happens in different regions (about 100 μm away from each other) of the same sample. Domain sizes change does not depend on time but the writing sequence. Moreover, we found that domain reversal still occurred in the second step up to 24 hours after the first step. Therefore, the space-charge field here is really stable.
3.2 Model to explain the two-step LIDOR
According to the light-induced charge transport model, there are three contributions to light-induced charge transport processes: drift in an external field, the bulk photovoltaic effect, and diffusion . The bulk photovoltaic effect is considered to be responsible for the generation of the space-charge field . The bulk photovoltaic current is given by, where K is Glass coefficient, α is the absorption coefficient and I is intensity of light. Carriers excited by light produce a bulk photovoltaic current. Since its direction is always opposite to the direction of the spontaneous polarization, it would be hard to explain the relation between the space-charge field and the external electric field. Diffusion of carriers from the illuminated region to the dark region is another contribution to the space-charge field . The diffusion current is given by, where Q is carriers conductivity, D is diffusion tensor, Ne,h is the density of free charge carriers. The diffusion current is influenced by inhomogeneous illumination and has nothing to do with the external electric field. As is known, the drift current is given by, i.e. proportional to the external electric field. In our experiments we proved that the space-charge field depends on the external electric field. The direction of the space-charge field is reverse to the direction of applied external electric field. Space-charge field can exist for a long time because of the low conductivity of LN and the capture of electrons by defects. Thus we considered that the drift of carriers is the main source of the space-charge field.
We proposed a schematic model to explain the two-step LIDOR as Figs. 6(a)-6(d). In the original LN crystal, the spontaneous polarization induces a depolarization electric field Edep . A screen charge field Escr is to balance Edep. In the first step, carriers (electrons and holes) are excited by laser 1 and drift to different directions by the external electric field as Fig. 6(a). The space-charge field Eph is formed by the drift of carriers. After the laser1 and the external electric field are switched off, a new screen charge field E’scr which is larger than Escr is to balance the Eph in the irradiation region as Fig. 6(b). In the second step, carriers are relocated in the region irradiated by laser 2 as Fig. 6(c). Domain reversal is induced by both the laser 2 and Eph as Fig. 6(d). Around the domain, Eph is weakened by laser 2. Thus domains become smaller and smaller with the writing sequence.
Our explanation is not contrary to the traditional LIDOR. It is hard to discover the influence of the drift of carriers in the traditional LIDOR because the external electric field assists domain reversal. In our experiments, the relation between space-charge field and external electric field is easy to obtain. Only carrier-drift model can explain this phenomenon. Therefore, we think that the establishment of space-charge field in LIDOR results from the competition of photovoltaic effect, carrier diffusion, and the drift of carriers. In traditional LIDOR, carrier diffusion plays a major role in the establishment of space-charge field. Thus the direction of space-charge field is coincident with the direction of external electric field. In our experiments, due to two-step process, space-charge field is only contributed to by the drift of carriers. Thus the direction of space-charge field is opposite to the direction of the external electric field.
In conclusion, we develop a two-step technique to achieve direct writing green-LIDOR in Mg-doped LiNbO3 without assistance of an external electric field. This technique can also be applied to large-scale domain engineering which is crucial for practical applications. Also, this method was tested to work in 5 mol% Mg-doped congruent LiNbO3 too. Due to our two-step method, the light-induced space-charge field is proven to have close relationship with the external electric field applied beforehand. The direction of the space-charge field is also proven to be opposite to the direction of the external electric field. Therefore, we propose a carrier-drift model to explain this phenomenon. Further research is needed to improve this method to precisely control the size of the domains.
This work was financially supported by the National Basic Research Program of China (No. 2013CB328706), International S&T Cooperation Program of China (S2013GR0402), the National Natural Science Foundation of China (91222111), Taishan Scholar Construction Project Special Fund, and the Fundamental Research Funds for the Central Universities (65030091 and 65010961).
References and links
1. G. P. Banfi, P. K. Datta, V. Degiorgio, and D. Fortusini, “Wavelength shifting and amplification of optical pulses through cascaded second-order processes in periodically poled lithium niobate,” Appl. Phys. Lett. 73(2), 136–138 (1998). [CrossRef]
2. V. Bermúdez, J. Capmany, J. G. Solé, and E. Diéguez, “Growth and second harmonic generation characterization of Er-doped bulk periodically poled LiNbO3,” Appl. Phys. Lett. 73(5), 593–595 (1998). [CrossRef]
3. J. A. Abernethy, C. B. E. Gawith, R. W. Eason, and P. G. R. Smith, “Demonstration and optical characteristics of electro-optic Bragg modulators in periodically poled lithium niobate in the near-infrared,” Appl. Phys. Lett. 81(14), 2514–2516 (2002). [CrossRef]
4. S. Mailis, P. T. Brown, C. L. Sones, I. Zergioti, and R. W. Eason, “Etch frustration in congruent lithium niobate single crystals induced by femtosecond ultraviolet laser irradiation,” Appl. Phys., A Mater. Sci. Process. 74(2), 135–137 (2002). [CrossRef]
5. V. Dierolfa and C. Sandmann, “Direct-write method for domain inversion patterns in LiNbO3,” Appl. Phys. Lett. 84(20), 3987–3989 (2004). [CrossRef]
6. Y. Zhi, D. Liu, W. Qu, Z. Luan, and L. Liu, “Wavelength dependence of light-induced domain nucleation in MgO-doped congruent LiNbO3 crystal,” Appl. Phys. Lett. 90(4), 042904 (2007). [CrossRef]
7. A. C. Muir, C. L. Sones, S. Mailis, R. W. Eason, T. Jungk, A. Hoffman, and E. Soergel, “Direct-writing of inverted domains in lithium niobate using a continuous wave ultra violet laser,” Opt. Express 16(4), 2336–2350 (2008). [CrossRef] [PubMed]
8. S. Zheng, Y. Kong, H. Liu, S. Chen, L. Zhang, S. Liu, and J. Xu, “Pyroelectric effect in green light-assisted domain reversal of Mg-doped LiNbO3 crystals,” Opt. Express 20(27), 29131–29136 (2012). [CrossRef] [PubMed]
9. C. L. Sonesa, M. C. Wengler, C. E. Valdivia, S. Mailis, R. W. Eason, and K. Buse, “Light-induced order-of-magnitude decrease in the electric field for domain nucleation in MgO-doped lithium niobate crystals,” Appl. Phys. Lett. 86(21), 212901 (2005). [CrossRef]
10. H. Zeng, Y. Kong, H. Liu, S. Chen, Z. Huang, X. Ge, and J. Xu, “Light-induced superlow electric field for domain reversal in near-stoichiometric magnesium-doped lithium niobate,” J. Appl. Phys. 107(6), 063514 (2010). [CrossRef]
11. W. Wang, Y. Kong, H. Liu, Q. Hu, S. Liu, S. Chen, and J. Xu, “Light-induced domain reversal in doped lithium niobate crystals,” J. Appl. Phys. 105(4), 043105 (2009). [CrossRef]
13. K. Buse, “Light-induced charge transport processes in photorefractive crystals I: Models and experimental methods,” Appl. Phys. B 64(3), 273–291 (1997). [CrossRef]
14. P. Hou, Y. Zhi, and L. Liu, “Laser-induced preferential domain nucleation in hafnium-doped congruent lithium niobate crystal,” Appl. Phys., A Mater. Sci. Process. 99(1), 105–109 (2010). [CrossRef]
15. H. Zeng, Y. Kong, T. Tian, S. Chen, L. Zhang, T. Sun, R. Rupp, and J. Xu, “Transcription of domain patterns in near-stoichiometric magnesium-doped lithium niobate,” Appl. Phys. Lett. 97(20), 201901 (2010). [CrossRef]