• Energy Research

AbstractCrystallographic phase engineering plays an important part in the precise control of the physical and electronic properties of materials. In two-dimensional transition metal dichalcogenides (2D TMDs), phase engineering using chemical lithiation with the organometallization agent n-butyllithium (n-BuLi), to convert the semiconducting 2H (trigonal) to the metallic 1T (octahedral) phase, has been widely explored for applications in areas such as transistors, catalysis and batteries1–15. Although this chemical phase engineering can be performed at ambient temperatures and pressures, the underlying mechanisms are poorly understood, and the use of n-BuLi raises notable safety concerns. Here we optically visualize the archetypical phase transition from the 2H to the 1T phase in mono- and bilayer 2D TMDs and discover that this reaction can be accelerated by up to six orders of magnitude using low-power illumination at 455 nm. We identify that the above-gap illumination improves the rate-limiting charge-transfer kinetics through a photoredox process. We use this method to achieve rapid and high-quality phase engineering of TMDs and demonstrate that this methodology can be harnessed to inscribe arbitrary phase patterns with diffraction-limited edge resolution into few-layer TMDs. Finally, we replace pyrophoric n-BuLi with safer polycyclic aromatic organolithiation agents and show that their performance exceeds that of n-BuLi as a phase transition agent. Our work opens opportunities for exploring the in situ characterization of electrochemical processes and paves the way for sustainably scaling up materials and devices by photoredox phase engineering.

Abstract Li-ion batteries (LIBs) have powered society for decades since their first commercialization in 1991. However, the current Li-ion chemistry deploying traditional graphite anode is approaching its energy density limit and struggling to meet the growing demand. The use of pure metallic Li with almost ten folds of anodic specific capacity is therefore critical to realize a higher energy density Li metal battery (LMB). A pure Li metal anode faces great challenges before its readiness for commercial applications. In addition to safety issues, which are a subject researched and reviewed widely, drastic Li loss (including the loss from active utilization or storage) is often overlooked, resulting in reduced capacity and eventually limited battery longevity. The Li loss in conventional liquid electrolyte settings, refers to the proportion of Li not taking part in an electrochemically active role for generating energy, and can be mainly categorized as inactive metallic Li, solid–electrolyte interphase (SEI) dissolution, and Li corrosion. To date, the underlying mechanisms involving these Li loss pathways and their dependence on each other are subject of ongoing investigations. This paper summarizes the major forms of Li loss processes when using a Li metal anode in an LMB, and existing strategies to mitigate these losses. Graphical abstract

Defect engineering in transition metal dichalcogenide (TMD) monolayers enables applications in single-photon emission, sensing, and photocatalysis. These functionalities critically depend on defect type, density, spatial distribution, relative energy, and the dynamics of exciton trapping at the defect sites. The latter are mediated by coupling to optical phonons through mechanisms not yet fully understood. Traditionally, exciton or carrier trapping at defects in inorganic crystals has been described by incoherent multiphonon emission within the Born–Oppenheimer approximationan approach that underpins the widely used Shockley–Read–Hall framework for nonradiative recombination. Here, we use impulsive vibrational spectroscopy to investigate exciton trapping in defect-modified monolayers of WS(2) grown through metal–organic chemical vapor deposition. We find that the phonon coherences of the Raman-active A’ and E’ modes persist throughout the ultrafast (∼100 fs) exciton trapping process, indicating a continuous evolution of the excitonic wave function. This observation is consistent with a conical intersection-mediated trapping process, in which a potential energy surface crossing between the free and trapped excitonic states acts as a funnel to drive this nonadiabatic transition. Such a molecular-like, vibronically coherent mechanism lies beyond the Born–Oppenheimer approximation, in stark contrast to classical, incoherent trapping models in solids. Moreover, the faster dephasing of the E’ mode in the trapped exciton state compared to the free exciton suggests it acts as a vibrational coordinate that promotes the trapping process. These findings provide mechanistic insights into exciton–phonon interactions at defects in TMD monolayers and inform strategies for engineering quantum and energy functionalities.