Fig. 1. Schematic diagrams of various linearization strategies: (a) mixed-polarization Mach–Zehnder modulators (MZMs); (b) dual-polarization power-combined MZMs; (c) X-assisted MZMs (where X represents a new structure, such as a ring or a racetrack); (d) parallel or cascaded MZMs. PSR: polarization splitter rotator; PRC: polarization rotator combiner.

Fig. 2. (a) Schematic of the proposed linear Mach–Zehnder modulator (LMZM) based on thin-film lithium niobate (TFLN) photonics. The on-chip optical processing consists of five main steps. (b) Top view of the proposed LMZM, showing several building blocks such as edge couplers, an optical power splitter, 2×2 multimode interference (MMI) couplers, and mode converters. Three phase shifters made of Ni-Cr are labeled as DC1, DC2, and DC3.

Fig. 3. (a)–(d) Simulated output RF power (signal and IMD3) as a function of the input RF power for both the LMZM and the reference CMZM at optical powers (OPs) of 10, 5, 0, and −5  dBm into the PD. (e) Comparison of the normalized transmission curves of the ideal LMZM and CMZM with the same value of Vπ. (f) Normalized transmission curves as the parameter R varies from 0 to 1 in increments of 1/18.

Fig. 4. (a) Microscope image of the fabricated device. (b) Measured small-signal EO responses S21 of the TE0-MZM and TE1-MZM. (c) Measured transmission curve of the LMZM based on EO effect, showing a Vπ of 6.37 V. (d) Measured transmission curve of the LMZM compared with the theoretical model. (e), (f) Measured transmission curve versus voltage for the TE0-MZM and TE1-MZM, based on the thermo-optic (TO) effect, using phase shifters DC2 and DC3 in Fig. 2(b). (g) Measured optical power ratio R versus voltage and power for the adjustable optical power splitter, using the phase shifter DC1 in Fig. 2(b).

Fig. 5. (a) Schematic of the setup for SFDR measurements. EDFA: erbium-doped fiber amplifier; PC: polarization controller; DUT: device under test; EPC: electrical power combiner; ESA: electrical spectrum analyzer. (b) Image of the DUT. (c) Image of the experimental setup for SFDR measurements. (d)–(f) Measured output RF powers of the signal and IMD3 as functions of input RF power at different OPs, using two frequencies at 1 GHz and 1.01 GHz, with OPs of about 5.5 dBm, 0.1 dBm, and −4.7  dBm, respectively.

Fig. 6. (a)–(c) Measured output RF powers of the signal and IMD3 as a function of input RF power for the LMZM, TE0-MZM, and TE1-MZM under close power conditions (∼0  dBm). (d)–(f) Simulation and experimental fitting results of SFDR values at different OPs. (g) Simulation and experimental results of SFDR improvement as a function of OP.

Fig. 7. (a)–(e) Measured output RF powers of the signal and IMD3 as a function of input RF power for the LMZM and CMZM at various frequency pairs F and F+0.01  GHz (F=1, 5, 10, 15, 20 GHz), under similar power conditions of 0 dBm. (f) Experimental results of SFDR improvement as a function of frequency.

Fig. 8. (a) Simulated SFDR as a function of the optical ratio R with different OPs. (b) SFDR as a function of the voltage applied to the phase shifter DC1. (c), (d) Measured and simulated SFDRs as a function of the voltage applied to the phase shifter DC1. (e), (f) SFDRs improvement with respect to heating power fluctuations of the phase shifters DC2 and DC3, corresponding to the TE0-MZM and TE1-MZM at their bias points.

Table 1. Comparison of Various Linearization Strategies in Optical Domain