Indium gallium zinc oxide (IGZO) and similar wide bandgap metal oxides are among the most widely used channel materials for drive transistors in displays due to their excellent electronic mobility and their ultra-high transparency1. However, industry-standard processing involves expensive vacuum deposition and elevated activation temperatures to produce semiconducting thin films. Liquid metal printing (LMP) is an emerging technique for oxide semiconductor fabrication poised to overcome these drawbacks via scalable vacuum-free transfer of the native oxide layers formed by spontaneous surface oxidation of molten metals2–4. Heterostructures of these 1-4 nm 2D oxide layers provide unprecedented opportunities for engineering electrostatic control of multilayers in thin film transistors, leading to improved mobility, Ion/Ioff ratios, and faster switching capabilities. Likewise, the backchannel is of high importance to these devices, as selection of an appropriate capping layer can enhance performance via remote doping while also mitigating bias stress effects. Herein, we compare the results of heterostructure InOx/GaOx with pure InOx TFTs, which demonstrates the mobility enhancement provided by GaOx modulation doping.Bottom gate thin film transistors (TFTs) (Figures 1a and 1b) were fabricated on Si substrates with 100 nm of thermally grown SiO2. 4 nm thick InOx and GaOx were deposited at 240 ˚C and 180 ˚C, respectively, using a linear printing speed of 8 cm/s. InOx and GaOx were printed in less than 10 s, with no post annealing necessary. Figure 1c illustrates the proposed mechanism for electron donation from GaOx at the heterointerface with InOx. The conduction band offset between these materials results in band bending at the interface and an increased carrier concentration in the InOx layer. The substoichiometric, defective GaOx is expected to further enhance this effect. Figure 1d demonstrates the transfer characteristics of heterostructure InOx/GaOx in comparison with pure InOx. The improved mobility for the heterostructure (7.8 cm2/Vs) vs pure InOx (3.8 cm2/Vs) channels can be attributed to modulation doping provided by GaOx and can be analyzed by extracting the electronic density of states (eDOS). These results illustrate a unique capability of LMP, which is to engineer the electronic structure of highly conductive 2D oxides while maintaining electrostatic control.This work also investigates the material properties of these 2D oxide heterostructures by UV-vis, XRD and XPS characterization. UV-Vis analysis revealed that the GaOx capping layer induces band gap widening and enhanced transparency, which can be explained by the Burstein-Moss effect from modulation doping. Unique to this LMP process is also the low temperature crystallization of the InOx films. XRD showed that even with low deposition temperatures (200 – 240 ˚C), these InOx films are highly crystalline with grain sizes substantially larger than the film thickness. Finally, XPS analysis of the O1s peak was utilized to understand the stoichiometry and interactions between the InOx and GaOx layers.This work demonstrates an effective pathway to enhance electronic transport in semiconducting metal oxides through liquid metal printed 2D heterostructures. The ultrathin films produced by LMP are well suited for thin film devices requiring nm-scale electrostatic control for effective gating. Combining this 2D nature of LMP InOx with a 2D GaOx backchannel capping layer is shown to yield high-performance printed transistors. This approach demonstrates a rapid, open-air compatible and low temperature manufacturing method, elucidating the broad impact of this technology in display fabrication, low-cost and flexible electronics. H. Hosono, Nat Electron, 1, 428–428 (2018).K. A. Messalea et al., ACS Nano, 15, 16067–16075 (2021).R. S. Datta et al., Nat Electron, 3, 51–58 (2020).A. Jannat et al., ACS Nano, 15, 4045–4053 (2021).A. Goff et al., Dalton Transactions, 50, 7513–7526 (2021).C.-H. Choi, Y.-W. Su, L.-Y. Lin, C.-C. Cheng, and C. Chang, RSC Advances, 5, 93779–93785 (2015). Figure 1