The manuscript by Tsai et al. “Multilevel Storage and Linear Optoelectronic Response in Mixed-Dimensional Photomemory” reports a mixed-dimensional photomemory device based on a CsPbBr3/Al2O3/MoS2 architecture. The authors demonstrate multilevel storage capability and a nearly linear optoelectronic response, supported by detailed structural, optical, and electrical analyses. The work is clearly presented, the results are significant, and the topic is highly relevant to the readership of Nanoscale Horizons. The device architecture provides an innovative route toward energy-efficient photomemory for advanced AI applications. The integration of perovskite quantum dots with a MoS2 channel through an optimized Al2O3 dielectric represents a novel and impactful concept. The device exhibits excellent performance metrics, including low optical energy per bit and 4-bit multilevel storage. The manuscript is well written, with comprehensive characterization supporting the conclusions.
Overall, I recommend addressing some minor points before acceptance. My remarks are as follows:
The introduction could better emphasize the technological relevance of mixed-dimensional architectures in bridging photonic and electronic domains for AI hardware applications.
A clearer explanation of how the Al2O3 dielectric thickness modulates the charge transfer dynamics would strengthen the conceptual framework introduced at the beginning.
The introduction could also benefit from a more detailed discussion of the role of interfacial charge trapping and hysteresis in MoS2-based photomemories. Since MoS2 is used as the channel material, a brief comparison with previous observations in similar MoS2 memory and phototransistor devices would help contextualize the results; for example doi: 10.1002/adfm.201800657 explores asymmetric Schottky contacts in bilayer MoS2 FETs. Furthermore, they should expand on the physical mechanisms responsible for the nearly linear photocurrent response over two orders of magnitude, clarifying whether surface trap states or photogating effects contribute to this behavior, and refer to similar findings reported in MoS2 optoelectronic memory devices (doi:10.1016/j.jpcs.2023.111406).
The authors may consider adding more experimental details on the optical power calibration and the reproducibility of the multilevel storage states, to strengthen the reliability of the reported results.
Figure 1 could be improved by including scale bars and clearer labeling of the layers in the schematic cross-section, facilitating the reader’s understanding of the device architecture.
In the optical characterization figures, adding a comparative PL spectrum of bare MoS2 would highlight the effect of the CsPbBr3/Al2O3 interface on exciton dynamics.