In 1962, Armstrong et al. first proposed the concept of QPM (Quasi-phase-match), which uses the inverted lattice vector provided by superlattice to compensate phase mismatch in optical parametric process. The polarization direction of ferroelectrics influences the nonlinear polarization rate χ². QPM can be realized by preparing ferroelectric domain structures with opposite periodic polarization directions in ferroelectric bodies, including lithium niobate, lithium tantalate, and KTP crystals. LN crystal is the most widely used material in this field.

In 1969, Camlibel proposed that the ferroelectric domain of LN and other ferroelectric crystals could be reversed by using a high voltage electric field above 30 kV/mm. However, such a high electric field could easily puncture the crystal. At that time, it was difficult to prepare fine electrode structures and accurately control the domain polarization reversal process. Since then, attempts have been made to construct the multi-domain structure by alternating lamination of LN crystals in different polarization directions, but the number of chips that can be realized is limited. In 1980, Feng et al. obtained crystals with periodic polarization domain structure by the method of eccentric growth by biasing the crystal rotation center and the thermal field axisy-symmetric center, and realized the frequency doubling output of 1.06 μm laser, which verified the QPM theory. But this method has great difficulty in the fine control of periodic structure. In 1993, Yamada et al. successfully solved the periodic domain polarization inversion process by combining the semiconductor lithography process with the applied electric field method. Applied electric field polarization method has gradually become the mainstream preparation technology of periodic poled LN crystal. At present, the periodic poled LN crystal has been commercialized and its thickness can be more than 5 mm.

The initial application of periodic poled LN crystal is mainly considered for laser frequency conversion. As early as 1989, Ming et al. proposed the concept of dielectric superlattices based on the superlattices constructed from ferroelectric domains of LN crystals. The inverted lattice of the superlattice will participate in the excitation and propagation of light and sound waves. In 1990, Feng and Zhu et al. proposed the theory of multiple quasi matching. In 1995, Zhu et al. prepared quasi-periodic dielectric superlattices by room temperature polarization technique. In 1997, experimental verification was carried out, and effective coupling of two optical parametric processes - frequency doubling and frequency summing was realized in a quasi-periodic superlattice, thus achieving efficient laser triple frequency doubling for the first time. In 2001, Liu et al. designed a scheme to realize three-color laser based on quasi-phase matching. In 2004, Zhu et al realized the optical superlattice design of multi-wavelength laser output and its application in all-solid-state lasers. In 2014, Jin et al. designed an optical superlattice integrated photonic chip based on reconfigurable LN waveguide optical path (as shown in figure), achieving efficient generation of entangled photons and high-speed electro-optic modulation on the chip for the first time. In 2018, Wei et al and Xu et al prepared 3D periodic domain structures based on LN crystals, and realized efficient nonlinear beam shaping using 3D periodic domain structures in 2019.

Integrated active photonic chip on LN (left) and its schematic diagram(right)

The development of dielectric superlattice theory has promoted the application of LN crystal and other ferroelectric crystals to a new height, and given them important application prospects in all-solid-state lasers, optical frequency comb, laser pulse compression, beam shaping and entangled light sources in quantum communication.