Micro-structure of high dose He-implanted x-cut KTP is investigated. Rutherford backscattering spectroscopy/channeling (RBS/Channeling) and transmission electron microscopy (TEM) are used to examine the structural and lattice damage properties in KTP after 200keV He+-implantation and following thermal annealing. Lattice crack, lattice disorder and He-bubble are observed in different implantation regions. The results show that strain induced by implantation is released through non-elastic lattice deformation in KTP. The implications of these observations for KTP thin film fabrication by smart-cut method are discussed.
© 2015 Optical Society of America
As optical components continue to replace electronics for optical signal processing applications, there is a growing impetus to integrate more photonic devices onto a single chip. Photonic devices offer several advantages compared to electronics, such as large bandwidth operation, wavelength division multiplexing, and also absence of electro-magnetic interference . A waveguide with large refractive index contrast is essential for the realization of highly integrated optics devices based on photonic wires [2,3] and photonic band gap structures . This waveguide can be produced with single-crystalline thin films which could be structured and embedded in low-index dielectric materials. In endeavor of film fabrication, many techniques have been studied in the past, such as chemical vapor deposition , RF sputtering , molecular beam epitaxy , sol-gel , and pulsed laser deposition . However, all these techniques have difficulty in producing high crystalline quality materials. In addition, the use of substrate with required properties is limited, in particular for epitaxial growth due to lattice matching constraints. More recently, a method called “Smart Cut” is becoming a promising technique for fabrication of single-crystalline thin films, which uses high-dose implantations of H+ and/or He+ ions (Dose of 1016-1017/cm2) for cleaving films from a bulk material with bonding technique and thermal treatment . This method was originally discovered and applied for silicon-on-insulator (SOI) wafers production , and has been successfully applied to ferroelectric materials, such as lithium niobate [12,13]. But no reports regarding to investigation on thin film fabrication of KTP with “smart-cut” method have been found yet.
Potassium titanyl phosphate (KTiOPO4 or KTP) has a variety of attractive optical applications spanning from nonlinear optics to electro-optics due to its excellent properties [14,15]. Fabrication of KTP thin film or hetero-structure is highly demanded because of the potential applications in nonlinear optics and optoelectronic integrated devices [16,17]. KTP crystal has a complex orthorhombic structure, with helical chains of TiO6 octahedra that are linked at two corners and are separated by PO4 tetrahedra, which results in a net c-directed polarization and is the major contribution to the large nonlinear optic and electrooptic coefficients. Its complex structure makes it very difficult to grow KTP film with single crystalline quality using conventional film growth methods. Smart-cut technique has been used to slicing LiNbO3 and other optical crystals, but not KTP. Although a few results about the waveguide properties or radiation effects in KTP induced by ion implantation can be found in several papers [18–21], the possibility of delamination or exfoliation in KTP by high dose ion implantation has not been investigated yet. In the present experiment, we implanted 200 keV He+ into x-cut KTP samples and imposed annealing on the samples to examine lattice structural modifications, sample surface variation, and then to investigate the possibility of KTP thin film fabrication through ion implantation technique or “smart-cut” method.
2. Materials and methods
200keV He+ ions with fluences of 6 × 1016, 8 × 1016 and 1 × 1017 ions/cm2 were implanted into x-cut KTP samples at room temperature, and the beam current density was kept around 5~6 μA/cm2 during implantation. Table 1 shows the implantation parameters. Before implantation, all the samples with size of 10mm × 7mm × 1mm were optically polished and cleaned. During the implantation the ion beam was electrically scanned to ensure a uniform implantation over the samples, samples were tilted by 7° off the beam direction in order to minimize the channeling effect. The implantation process was performed at LC-4 Ion Implanter in Institute of Semiconductors of Chinese Academy of Sciences.
Damage analysis was carried out by using RBS/channeling measurement, which was performed by 2.1MeV He2+ ions at a scattering angle of 165° in 1.7 MV tandem accelerator of Shandong University. Stopping and Range of Ion Matter (SRIM 2008) was used to simulate the vacancy distribution in ion-implanted KTP. Optical microscopy (OM) was used to observe morphology modifications at sample surface. Post-implant annealing was employed to all the implanted samples to see the crystal structural modification and lattice damage evolution. Annealing parameters are given in Table 2.The lattice structural modifications were detected by transmission electron microscopy (TEM) method in both cross-section and plan-view geometry, which was performed using Tecnai G2 F20 S-Twin at 200kV with a field emission gun. TEM samples were prepared with conventional polishing and ion-beam milling techniques.
3. Results and discussion
3.1. Lattice damage properties in as-implanted and annealed samples
The x-cut KTP samples were implanted with He+ at ion fluences of 6 × 1016, 8 × 1016 and 1 × 1017 ions/cm2. After implantations, samples were investigated using RBS/Channeling for lattice damage analysis and OM for surface modification observation. No evident change can be observed on the samples' surface through OM method. Figure 1(a) shows the RBS/channeling spectra of implanted samples, as well as the aligned and random spectra from a virgin KTP sample for comparison. It is evident that the backscattering spectra of as-implanted samples are higher than aligned spectra from virgin KTP, which demonstrates the presence of damage from induced defects and crystalline disorder. It can be found that the 6 × 1016 sample and the 8 × 1016 sample have almost the same spectra, while the 1 × 1017 sample shows a little higher backscattering yield at near surface region. At the end of ion range, the backscattered signals of these samples all reach the level of purely random scattering, indicating that a nearly amorphous layer is formed at this depth. Damage profiles are extracted from RBS by using a multiple-scattering dechanneling model, which is based on Feldman’s procedure and is applied for all target elements in the crystal [22,23]. The results are presented in Fig. 1(b). A nearly 100% damage ratio from a depth of 0.7~1.0μm was found in all three samples. Damage ratio corresponds to the extent of disorder in lattice structure. The layer with 100% damage ratio—a saturated lattice damage representing an almost amorphous structure, is formed by 200keV He+ implantation even at fluence of 6 × 1016 ions/cm2. However, implantations using the same parameters are far from causing amorphization in LiNbO3, LiTaO3 and other optical crystals. Instead, a blistering or exfoliation is induced in theses crystals in as-implanted or annealed state. While surface bulge, blistering or exfoliation is not observed in our He-implanted KTP samples, neither in as-implanted nor after annealing treatments.
The number of displacements per atom (dpa) vs depth (z) can be calculated using equation of dpa(z) = Ndispl(z) × N1/N0, with the ion fluence N1, the number of displacements per incident ion and unit length Ndispl (taken from SRIM 2008) and the atomic density of KTP N0. The distribution profile of dpa is also presented in Fig. 1(b) in solid line. Comparing the calculated damage profile with the experimental results, some distinct differences can be seen in Fig. 1(b). In the near surface region lattice disorder increase more rapidly than the estimation. A complete disorder region starts from 700nm below surface, extending about 300nm in depth. It is consistent with the calculated position of the damage region with a big damage ratio but much broader than the FWHM of the simulated damage profile, which can be attributed to the errors in both damage extraction and SRIM simulation. The calculated dpa peak is 1.72 for the 6 × 1016 sample, 2.29 for the 8 × 1016 sample and 2.87 for the 1 × 1017 sample. If comparing the damage profile from RBS with the simulation result, the dpa number for causing amorphization in 6 × 1016 He-implanted KTP is 0.75 as shown in Fig. 1(b), which is larger than the data of 0.5 in Ref .
To study the lattice structural modification and helium behavior dependence on thermal process, samples were annealed in a heating process from 200°C to 600°C. Real time monitoring on sample surface through OM was carried out during annealing process, but no significant change on surface morphology was found. Annealing at rather low temperature (200°C) was applied to these samples first, RBS spectra keep almost unchanged compared to as-implanted samples. To verify the results of RBS, 6 × 1016 He-implanted sample was analyzed using TEM technique, the results are discussed in the following part.
After annealing at 350°C for 60mins, damage profile was measured by RBS once again. RBS results of 8 × 1016 sample are shown in Fig. 2.Obviously different from the annealing at 200°C, annealing at 350°C promotes a lattice structure recovery in the whole irradiated damage region, even in the damage layer. At the same time, OM monitoring revealed that no surface modification was found during annealing process. It is different from the cases in many He-implanted materials, such Si, SiC, LiNbO3, BaTiO3 etc. In He-implanted semiconductors and oxide crystals, He ions aggregation is induced at about 200°C. Inner pressure in He-bubble increases as temperature rises, thus surface upward bulge or rupture is caused. However, such phenomenon didn’t occur in He-implanted KTP. No bubble or blistering was observed during annealing treatment.
3.2. Lattice structural modification in annealed samples
We used TEM technique to detect the lattice structural evolution with annealing process. Figure 3 presents the TEM images in cross-section geometry of KTP sample with ion fluence of 6 × 1016 ions/cm2 after annealing at 200°C for 1hr. It can be found that the thickness of the total region affected by implantation is about ~1000nm. A damage layer (bright region) was produced at the depth of 630~900nm from surface, which is consistent with the extracted damage profile from RBS results. In the near surface region (~200nm from surface) the crystal structure is almost unperturbed. A slight dark contrast compared to the substrate is related to the remaining defects. Beneath this layer a very dark contrast region (~430nm) extending to the bright damage layer can be seen. In comparison with the damage profiles of as-implanted samples in Fig. 1(b) which is consistent with 200°C-annealed results, the TEM results show good consistency between the color contrast in TEM and the lattice disorder in RBS. In the surface region, lattice disorder is less than 20% according to RBS. Then it increases gradually up to almost 100% at the depth ~700nm. From this depth to ~900nm, a bright damage layer can be clearly seen in TEM, which corresponds to the damage range in RBS with damage ratio as much as 100%. The detailed lattice structure was observed with high-resolution TEM, which is shown in Fig. 4(a)-(d). Figure 4(a)-4(d) correspond to the detected regions marked by (a), (b), (c) and (d) in Fig. 3, respectively.
A relative intact lattice structure can be seen in region (a), which corresponds to the slightly damaged region at surface measured by RBS. Even near the boundary of damage layer, in region (b) and (d), good ordered structure is also visible. A complete disorder lattice structure (damage layer) is produced in region (c), consistent with the result of RBS.
To investigate the dynamics of implanted He ions, all the samples were further annealed up to 600°C.
After annealed in furnace at 600°C for 1h, significant lattice modification can be seen in TEM results, which are shown in Fig. 5(a)-5(d). From Fig. 5(a) and 5(b) it can be seen that the implanted region are clearly divided into three parts: (1) a relative undisturbed lattice structure from surface to ~360nm; (2) a region with high density of disorder or dislocation located from 360nm to 700nm beneath surface. The lattice disorder with lateral size of 100nm~300nm are mostly connected in a chain and parallel to the implanted surface; (3) the region from 700 to 900nm, corresponding to the damage layer in Fig. 3, some cracks can be clearly seen off and on along the damage layer. The more clear TEM image is presented in Fig. 5(b).
Figure 5(a) shows that He-cavities or cracks distribute intermittently along the damage layer. The width of crack is about 100nm, much smaller than that of damage layer in Fig. 3. Locations of cracks correspond with the peak of He ions density. Apparently they are caused by the He ions aggregation. Considering the RBS results in Fig. 2, which shows that a damaged lattice recovers to some extent after annealing at 350°C, it can be deduced that small He-cavities may be aggregated and formed initially in annealing temperature of at least 350°C. With the increase of temperature some cavities grow and merge with each other, and cracks result at these places. In the vicinity of cracks, three regions labeled A, B and C are shown in detail in Fig. 5(c)-(d). Outside the crack boundary in A and B, basic lattice structure and lattice deformation can be clearly seen in Fig. 5(c). In those remaining connected regions, partially restored lattice structure also can be clearly seen in Fig. 5(d), which is consistent with our RBS results in Fig. 2. However, different from most other He-implanted crystals, the strain produced by He-cavities does not result in elastic expansion of the KTP crystal's surface. Instead of causing surface bulge and blistering, strain or stress is released in KTP crystal through a non-elastic deformation in surrounding lattice, which can be seen in Fig. 6.
High density and random distribution of lattice disorder can be found direct near the crack edge in region A and B, as is shown in Fig. 6(a) and 6(b). However, outside this region, lattice disorders show obvious directional properties, as shown in Fig. 6 (c). Linear lattice disorders become dominant. Fold shaped lattice deformations distribute from here to the surface, gradually becoming sparse and disappear finally. This can explain why there is no bulge or blister on surface. It seems like strain is propagating in a form of a wave toward the sample's surface and is released along with a non-elastic lattice deformation. So lattice distortions are mostly aligned parallel to the sample's surface (100). In region C, similar results also can be seen. Since the applied strain or stress in region C is normal to the strain in region A and B, instead of parallel to the surface, lattice deformation propagates laterally along damage layer, as is shown in Fig. 6(d).
Different from He-implanted LiNbO3 and other optical crystals, He+ implanted x-cut KTP has no sign of bulging or blistering on surface, although much higher He+ fluence was applied in KTP than in LiNbO3. Based on the results and analysis above, we attribute this result to three factors: (1) He ions has a higher solubility in KTP crystal than in other optical crystals. Loose lattice structure of KTP makes He ion aggregation more difficult. (2) Lattice structure tends to deform non-elastically when strain or stress exists. These deformations appear mostly at the locations where cracks take place. (3) He ions diffuse more easily in KTP than in other crystals, especially at a high annealing temperature. This can be confirmed by our TEM observation. Except for the big He-gas cavities at the buried damage layer, some small bubbles of He ions also can be seen near sample surface by plane-view TEM, which is shown in Fig. 7.The size of bubbles is around 10~20nm. They are scattered all over the implanted region.
Smart cutting by ion implantation has been demonstrated in several optical crystals [24,25]. Film splitting occurs during the annealing treatment with the help of elastic or plastic deformation of the crystal’s surface. Such a situation is very difficult to achieve in He-implanted x-cut KTP, where non-elastic deformation plays the main role. Although there are He-cavities and cracks formed in KTP at damage layer after a series of annealing treatments, it is not clear if these cavities can grow big enough to merge together when appropriate experimental parameters are applied. This is the key point whether film can be split from the crystal in large areas. However, several experimental factors are considered to be favorable to the He-cavity formation and aggregation, for example reducing beam current to suppress He diffusion during implantation, increasing total He fluence or with a help of external force, such as by bonding it on a stiff substrate. Currently, we are trying to test the influence of ion beam flux on the lattice structural modifications and layer-splitting by implantations under different beam densities, meanwhile the effect of co-implantation from He+/H+ is also investigated. The research is in progress.
This paper has revealed how He+-implantation influences the lattice structure in KTP with the following thermal treatments, which may enable ion-induced thin-film exfoliation from a KTP crystal. 200KeV He+-implantations with ion fluences of 6 × 1016, 8 × 1016 and 1 × 1017 ions/cm2 can cause substantial structural changes at the end of ion track. After a series of annealing, with a partial recovery of the damaged lattice some bubbles and cracks formed by He+ aggregation appear at the region close to the end of ion track. TEM observation show that strain or stress induced by implantation is released gradually by non-elastic deformation of KTP lattice during the annealing process. No bulge or blistering was caused on the sample's surface. The buried cracking phenomenon provides the possibility of splitting KTP thin film from the bulk of the crystal. The experimental parameters will be adjusted and optimized. Some additional techniques such as bonding or wet etching will be tried to realize thin film exfoliation in our later work.
This work was supported by the National Nature Science Foundation of China (Grant No.51272135). The authors thank Shanghai Doesun Energy Technology Co. Ltd and Department of Electronic Materials Engineering in Australian National University(ANU) for their efforts on TEM results, and China Scholarship Council for their financial support of studying in ANU.
References and links
1. G. Poberaj, M. Koechlin, F. Sulser, A. Guarino, J. Hajfler, and P. Günter, “Ion-sliced lithium niobate thin films for active photonic devices,” Opt. Mater. 31(7), 1054–1058 (2009). [CrossRef]
2. R. U. Ahmad, F. Pizzuto, G. S. Camarda, R. L. Espinola, H. Rao, and R. M. Osgood, “Ultracompact corner-mirrors and T-branches in silicon-on-insulator,” IEEE Photon. Technol. Lett. 14(1), 65–67 (2002). [CrossRef]
4. B. S. Song, S. Noda, T. Asano, and Y. Akahane, “Ultra-high-Q photonic double-heterostructure nanocavity,” Nat. Mater. 4(3), 207–210 (2005). [CrossRef]
5. Y. Sakashita and H. J. Segawa, “Preparation and characterization of LiNbO3 thin films produced by chemicalvapor deposition,” J. Appl. Phys. 77(11), 5995–5999 (1995). [CrossRef]
6. X. Lansiaux, E. Dogheche, D. Remiens, M. Guilloux-viry, A. Perrin, and P. Ruterana, “LiNbO3 thick films grown on sapphire by using a multistep sputtering process,” J. Appl. Phys. 90(10), 5274–5277 (2001). [CrossRef]
7. F. Gitmans, Z. Sitar, and P. Günter, “Growth of tantalum oxide and lithium tantalate thin films by molecular beam epitaxy,” Vacuum 46(8-10), 939–942 (1995). [CrossRef]
8. J. G. Yoon and K. Kim, “Growth of highly textured LiNbO3 thin film on Si with MgO buffer layer through the sol-gel process,” Appl. Phys. Lett. 68(18), 2523–2525 (1996). [CrossRef]
9. Y. Nakata, S. Gunji, T. Okada, and M. Maeda, “Fabrication of LiNbO3 thin films by pulsed laser deposition and investigation of nonlinear properties,” Appl. Phys., A Mater. Sci. Process. 79(4-6), 1279–1282 (2004). [CrossRef]
10. Q. Y. Tong and U. Gösele, “Semiconductor wafer bonding: Science and Technology,” John Wiley & Sons, New York, (1999).
11. M. Bruel, “Application of hydrogen ion beams to silicon on insulator material technology,” Nucl. Instrum. Methods Phys. Res. B 108(3), 313–319 (1996). [CrossRef]
12. I. Szafraniak, I. Radu, R. Scholz, M. Alexe, and U. Gösele, “Single-crystalline ferroelectric thin films by ion implantation and direct wafer bonding,” Integr. Ferroelectr. 55(1), 983–990 (2003). [CrossRef]
13. P. Rabiei and P. Günter, “Optical and electro-optical properties of submicrometer lithium niobate slab waveguides prepared by crystal ion slicing and wafer bonding,” Appl. Phys. Lett. 85(20), 4603–4605 (2004). [CrossRef]
14. Y. S. Liu, D. Dentz, and R. Belt, “High-average-power intracavity second-harmonic generation using KTiOPO4 in an acousto-optically Q-switched Nd:YAG laser oscillator at 5 kHz,” Opt. Lett. 9(3), 76–78 (1984). [CrossRef] [PubMed]
15. J. D. Zumsteg, Bierlein, and T. E. Gier, “KxRb1-xTiOPO4: A new nonlinear optical material,” J. Appl. Phys. 47(11), 4980–4985 (1976).
16. X. Gu, M. Makarov, Y. J. Ding, J. B. Khurgin, and W. P. Risk, “Backward second-harmonic and third-harmonic generation in a periodically poled potassium titanyl phosphate waveguide,” Opt. Lett. 24(3), 127–129 (1999). [CrossRef] [PubMed]
17. L. Ma, O. Slattery, T. Chang, and X. Tang, “Non-degenerated sequential time-bin entanglement generation using periodically poled KTP waveguide,” Opt. Express 17(18), 15799–15807 (2009). [CrossRef] [PubMed]
18. P. Bindner, A. Boudrioua, J. C. Loulergue, and P. Moretti, “Formation of planar optical waveguides in potassium titanyl phosphate by double implantation of protons,” Appl. Phys. Lett. 79(16), 2558–2560 (2001). [CrossRef]
19. F. Chen, Y. Tan, L. Wang, D.-C. Hou, and Q.-M. Lu, “Optical channel waveguides with trapezoidal-shape cross sections in KTiOPO4 crystal fabricated by ion implantation,” Appl. Surf. Sci. 254(6), 1822–1824 (2008). [CrossRef]
20. R. Zhang, F. Lu, J. Lian, X. Liu, H. Liu, and Q. Lu, “Buried planar waveguide in KTiOPO4 formed by Cs+ ion exchange and Si+ ion implantation,” Mod. Phys. Lett. B 22(10), 755–762 (2008). [CrossRef]
21. F. Schrempel, Ch. Beeker, J. Fick, and W. Wesch, “Waveguide barriers with adjustable refractive index produced in KTP by irradiation with He and Li ions,” Nucl. Instrum. Methods Phys. Res. B 257(1-2), 484–487 (2007). [CrossRef]
22. K. Sato, Y. Fujino, S. Yamaguchi, H. Naramoto, and K. Ozawa, “Ion channeling studies of C+ irradiated TiC single crystals,” Nucl. Instrum. Methods Phys. Res. B 47(4), 421–426 (1990). [CrossRef]
23. L. C. Feldman and J. M. Rodgers, “Depth profiles of the lattice disorder resulting from ion bombardment of silicon single crystals,” J. Appl. Phys. 41(9), 3776–3782 (1970). [CrossRef]
24. T. Izuhara, I. L. Gheorma, R. M. Osgood Jr, A. N. Roy, H. Bakhru, Y. M. Tesfu, and M. E. Reeves, “Single-crystal barium titanate thin films by ion slicing,” Appl. Phys. Lett. 82(4), 616–618 (2003).
25. A. Tauzin, J. Dechamp, F. Madeira, F. Mazen, M. Zussy, C. Deguet, L. Clavelier, J. S. Moulet, C. Richtarch, T. Akatsu, M. Yoshimi, and A. Rigny, “3-inch single-crystal LiTaO3 films onto metallic electrode using smart cut technology,” Elec. Lett. 44(13), 822 (2008). [CrossRef]