Van der Waals epitaxy and remote epitaxy of LiNbO₃ thin films by pulsed laser deposition

Nonlinear optical phenomena such as the Kerr effect, frequency mixing processes, and cross-phase modulation result from nonlinear interaction of the electromagnetic field with noncentrosymmetric crystals. Major applications of nonlinear optical phenomena include ultrashort pulsed lasers, sensors, laser amplifiers, and digital optical information processing, among many others.¹ A recently proposed application lies in the field of quantum computing, which includes spontaneous parametric downconversion that is used to produce entangled photons.² 

LiNbO₃ (R3c, a = b = 5.212 Å, c = 14.356 Å, α = β = 90°, and γ = 120°) is one of the most widely used materials for nonlinear optics.³ Apart from having a wide transmission window (0.35–5 μm), high refractive indices (n_o= 2.286, n_e= 2.203 at 632.8 nm), and high second-order susceptibilities (d₃₁= −4.88 pm/V, d₃₃= 34.0 pm/V), it is also ferroelectric in nature, making the reconfigurability of its nonlinear property by an electric field possible.^4,5 For applications in photonic integrated circuits, LiNbO₃ in thin film form as opposed to the bulk form is necessary. Thin films can be applied to reduce device dimensions and the operating voltage, making it easier to modulate electrical and optical properties.

Over the past few decades, several methods have been used to grow LiNbO₃ thin films, including pulsed laser deposition (PLD),⁶ RF sputtering,⁷ metal-organic chemical vapor deposition,⁸ as well as solgel process.⁹ Highly demanded epitaxial growth of LiNbO₃ has been reported on many substrates, such as ZnO/Si,¹⁰ MgO,^11,12 sapphire,^6,13 and LiTaO₃.^14,15 In the pursuit of high-quality and transferable thin films, van der Waals epitaxy^16,17 (in which film-substrate interaction is believed to be van der Waals or quasi van der Waals based) and remote epitaxy^18,19 (in which film-substrate interaction is remotely controlled by a buffer layer of 2D materials such as graphene) have been regarded as promising solutions. In the past few years, we have demonstrated the van der Waals and remote epitaxy of several halide perovskites, chalcogenides, and complex oxides.^18,20–27 In some of these works, we have found that remote and van der Waals epitaxial films could be mechanically exfoliated and transferred onto arbitrary substrates.¹⁸ 

Here, we report the growth of epitaxial LiNbO₃ thin films based on van der Waals epitaxy as well as remote epitaxy. For van der Waals epitaxy, muscovite mica (C2/c, a = 5.199 Å, b = 9.027 Å, c = 20.106 Å, α = γ = 90°, and β = 95.78°) is chosen as the substrate. Prior to growth, the heterostructure of LiNbO₃/mica is proposed in Figs. 1(a)–1(c). The pseudohexagonal lattice structure of mica with a ≈ 5.18 Å matches well with that of LiNbO₃ (less than 0.5% mismatch), as shown in Fig. 1(c). For remote epitaxy, graphene covered sapphire (sapphire: $R3¯c$⁠, a = b = 4.760 Å, c = 12.993 Å, α = β = 90°, and γ = 120°) is chosen as the substrate. Figures 1(d)–1(f) presents our proposed heterostructure of LiNbO₃/graphene/sapphire. As shown in Fig. 1(f), the lattice mismatch of lattice parameter a between sapphire and LiNbO₃ is about 8.7%, which is large but still feasible for the execution of epitaxy.

All thin film growth was executed by a home-built PLD system (KrF excimer laser, λ = 248 nm) using a 2 in.-diameter commercial LiNbO₃ target. The laser power density during deposition was about 7.5 J/cm². The target-to-substrate distance was set as 5 cm. LiNbO₃ thin films were grown on freshly cleaved muscovite mica at 550 °C with the oxygen pressure ranging from 300 to 600 mTorr. LiNbO₃ was deposited on graphene-buffered sapphire (0001) at an optimized condition of 550 °C and 500 mTorr oxygen pressure. All the films were postannealed at 550 °C in 760 Torr oxygen pressure.

Monolayer graphene, synthesized on Cu foils by chemical vapor deposition, was purchased from Graphene Laboratories Inc. (Calverton, New York, USA). For graphene transfer, poly(methyl methacrylate) (PMMA) was first spin-coated onto the graphene/Cu foil as a support. The rear side of the Cu foil was then treated in an O₂ plasma etcher to remove unwanted graphene. The Cu foils were etched in an ammonium persulfate aqueous solution (60 g/l). The PMMA/graphene assembly was rinsed in water several times and picked up by the sapphire (0001) substrate. The PMMA/graphene/sapphire (0001) stack was dried in air and followed by the dissolution of PMMA in acetone. After these procedures, the graphene/sapphire substrate is ready for PLD growth.

The structure of the grown LiNbO₃ films was characterized by x-ray diffraction (XRD) with a Cu K_α radiation (Panalytical X’pert PRO MPD system). X-ray pole figures were obtained using point focus optics with a polycapillary x-ray lens to study the in-plane orientation. High-resolution transmission electron microscopy (HRTEM) images of the LiNbO₃ on mica were acquired by an FEI F20 TEM at 200 kV. A Helios G4 UX focused ion beam system was used to cut a cross section of the thin film. Raman spectra (WITec Alpha 300R Confocal Raman imaging system) were collected to confirm the composition of LiNbO₃ on the graphene-buffered sapphire (0001) and check the integrity of the graphene buffer layer after film growth.

Figure 2(a) shows the XRD pattern of LiNbO₃ thin films grown on mica. The peak at 39.0^o denotes the LiNbO₃ (0006) plane. The peak at 36.0° is recognized as muscovite mica (008)²⁸ and peaks at 29.1°, 32.4°, and 40.8° are all from the cleaved mica substrate (see the supplementary material³⁵ for detailed indices). A series of oxygen partial pressures were applied during growth to minimize the Li-deficient phase, LiNb₃O₈. Only for films grown in 500 mTorr oxygen pressure, no peaks other than the LiNbO₃ (0006) peak are observed. There is an additional peak at 38.1° for films grown in relatively low oxygen pressure (300 and 400 mTorr), which corresponds to the LiNb₃O₈ (⁠$6¯02$⁠) peak.²⁹ This observation can be explained by the suppression of oxidation of Li atoms in the oxygen-rich environment.¹³ When the oxygen pressure is increased to 600 mTorr, the LiNbO₃ peak disappears since collisions between the ablated species and oxygen atoms increase and crystallization may be hindered by the reduced kinetic energy of the ablated atoms.^30,31

Figure 2(b) shows the x-ray pole figure used to determine the in-plane orientation of the LiNbO₃ thin film. Two sets of (⁠$011¯2$⁠) planes are seen in the pole figure, which indicates a 60°-rotated twin structure. The two reflections with an abnormal intensity were further studied along their χ direction. The χ scan result at 2θ = 23.7° is shown in Fig. 2(c). It can be seen that the (023) plane of mica and the (⁠$011¯2$⁠) reflection of LiNbO₃ overlap, which reveals the in-plane epitaxial relation as mica [010] || LiNbO₃ [$011¯0$] and mica [010] || LiNbO₃ [$101¯0$]. The cross-sectional HRTEM images of the LiNbO₃ thin film with a zone axis of [$011¯0$] in Figs. 2(e) and 2(f) also confirm the 60°-rotated twin structure. From the inset fast Fourier transform (FFT) pattern, the interplanar spacing of (⁠$101¯2¯$⁠) planes is 0.37 nm, which corresponds well with that of bulk LiNbO₃. The twin structure of the atomic model is shown in Fig. 2(d).

Figure 3(a) shows the XRD pattern of LiNbO₃ thin films grown at 500 °C and 500 mTorr oxygen pressure on a graphene-buffered sapphire (0001) substrate. A clear LiNbO₃ (0006) peak can be seen. The peak at 41.7° is attributed to the sapphire (0006) peak. The peak at 37.5° is also from the substrate. The Raman spectrum in Fig. 3(b) shows the 9E(TO) + 4A₁(LO) modes of LiNbO₃,^32,33 except for the E(7TO) at ∼430 cm⁻¹ and E(9TO) at ∼660 cm⁻¹ modes, which have intensities too small to be identified. The inset of Fig. 3(b) shows the Raman spectrum ranging from 1100 to 2800 cm⁻¹ of the same measurement. The G (∼1580 cm⁻¹) and 2D (∼2680 cm⁻¹) peaks of graphene can be clearly identified in the spectrum indicating the existence of graphene after growth. The existence of a small D (∼1350 cm⁻¹) peak indicates that the graphene buffer layer has been partially damaged by the oxygen atmosphere and the high-energy plasma at high temperatures.³⁴ 

The in-plane symmetry of the grown film was determined from the x-ray pole figure in Fig. 3(c). The two sets of (⁠$011¯2$⁠) planes indicate a twin structure in LiNbO₃ as also seen in the LiNbO₃ film grown on mica. Thus, there are two sets of in-plane symmetry relations, sapphire [$101¯0$] || LiNbO₃ [$101¯0$] and sapphire [$101¯0$] || LiNbO₃ [$112¯0$] in the 60°-rotated twin structure, as shown in Fig. 3(d).

In summary, we demonstrated the growth of epitaxial LiNbO₃ film via van der Waals and remote epitaxy by pulsed laser deposition. Structural analysis has revealed the epitaxy relations and identified the presence of twin structures of LiNbO₃ films in both LiNbO₃/mica and LiNbO₃/graphene/sapphire systems. In remote epitaxy, Raman spectroscopy has confirmed the existence of graphene after deposition suggesting the active role of graphene during the deposition process. The demonstration of the feasibility of van der Waals and remote epitaxy of LiNbO₃ suggests a promising processing route toward the development of free-standing LiNbO₃ films in the near future.

This work was supported by the U.S. Air Force Office of Scientific Research under Grant No. FA9550-18-1-0116 (R.J. and J.S.) and the NYSTAR Focus Center at Rensselaer Polytechnic Institute (RPI) with Contract No. C150117 (X.S., T.-M.L, M.W., and J.S.). The authors declare no competing interests. This work is also supported by U.S. National Science Foundation under Award No. of 1712752 (D.G.). This paper is also supported by the U.S. National Science Foundation [Platform for the Accelerated Realization, Analysis, and Discovery of Interface Materials (PARADIM)] under Cooperative Agreement No. DMR-1539918 and made use of the Cornell Center for Materials Research (CCMR) Shared Facilities, which are supported through the NSF MRSEC Program (No. DMR-1719875).

The data that support the findings of this study are available from the corresponding author upon reasonable request.