Lithium–sulfur (Li–S) battery, with an ultra-high theoretical energy density of 2600 Wh/kg, is inevitably deemed as one of the most promising alternatives for the state-of-the-art lithium-ion (Li-ion) batteries [1]. As a typical conversion-type of batteries, Li–S batteries comply multi-electron transfer and multiphase transformation between elemental S and lithium sulfide (Li₂S) during charging/discharging, accompanied by a series of dissolved LiPSs with a general formula of Li₂S_n (n = 4, 6 or 8) [2]. The dissolved Li₂S_n are traditionally considered to exhibit high ionization degree to dissociate into polysulfide anions (S_n²⁻), behaving as a typical weak electrolyte. The S_n²⁻ species can accelerate the interfacial charge transfer via the equilibrium-driven solubilization of the inert S solids and the redistribution of active S, which enables fast solid–liquid–solid S conversion and electrochemical redox at the cathode/electrolyte interface [3]. However, the dissolved LiPS species also simultaneously shuttle between electrodes during the aforementioned processes, i.e., the so-called "shuttle effect", resulting in active S loss, continuous anode corrosion, and the resultant fast cell failure [4]. Therefore, it remains essential to acquire deeper understandings of the intrinsic properties, behaviors (including dissolution, diffusion, and shuttling), and solvation structures of LiPSs.

In the past years, numerous efforts have been devoted to investigate the LiPSs in Li–S batteries. Various spectral characterizations such as nuclear magnetic resonance spectroscopy (NMR), liquid chromatography, ultraviolet-visible spectroscopy (UV–vis), and X-ray absorption near edge structure spectroscopy have been applied to investigate the structure of LiPSs [5–7]. However, the insights on individual LiPS species obtained through solely spectral characterization still remain rather limited, due to the inherent complexity of LiPSs. In addition, theoretical calculations such as density functional theory (DFT) and molecular dynamics simulations are found effective in revealing the interactions between LiPSs and Li salts or solvent molecules, which help to further deepen the understanding of the dissociation/association behaviors of LiPSs [8]. However, experimental research works are still lacking to convincingly validate the theoretical results and elucidate the actual existing form of LiPSs.

Recently, a pioneering study by Zhang and co-workers was published in Chem, providing reliable experimental and theoretical evidence on the existing form of LiPSs in the electrolyte (Fig. 1) [9]. The authors first performed conductivity analysis on Li₂S₆ LiPSs, which revealed a triple-ion ionization behavior in the concentration region above 100 mmol/L; three neutral Li₂S₆ molecules reform to one Li₃S₆⁺ cation and one Li₃S₁₂⁻ anion [10]. A combination of electron spray ionization mass spectrometry, ⁷Li NMR spectroscopy, and theoretical calculations further confirm the existence of cationic LiPSs with the presence of Li salts. Neutral LiPSs are found to preferentially bond with the extra Li⁺ from the Li salts. DFT calculations verified that the extra Li⁺ and LiPSs interact via ionic bonds. The sulfur-chain length tunes such interactions as the extra Li⁺ inserts in the long-chain Li₂S_n (n = 6 or 8), serving as a linker, while for the short-chain Li₂S₄, the additional Li⁺ is coordinated to terminal S atoms without chain breaking, forming a more compact structure than the long-chain ones. The formation of the compact structure is attributed to the small steric hinderance of original Li₂S₄.

The effects of cationic LiPSs on electrochemical behaviors were investigated at both S cathode and Li anode sides. On one hand, UV–vis spectroscopy indicated that the cationic LiPSs are inherently more difficult to adsorb on cathodic carbon substrate due to the electrostatic repulsion. Such difficulty in adsorption resulted in inferior cathode kinetics. On the other hand, improved cationicity of LiPSs by adding extra Li salt gave rise to two-fold higher shuttle current density (a measure to quantitatively monitor the reactivity between LiPSs and Li), indicating a more aggressive reactivity of the cationic LiPSs with Li metal and consequently severer irreversible Li consumption. All above results evidence that the cationic LiPSs remain unfavorable for the cathode redox kinetics and rather reactive towards the Li metal anode, due to their electrostatic interactions with the EDLs at both interfaces. Accordingly, a strategy of decreasing the Li salt concentration was proposed to efficiently reduce the proportion of LiPS cations, and the Li–S battery showed low polarization with superior discharge capacity of 1124 mAh/g and excellent energy density of 353 Wh/kg.

In summary, this work experimentally and theoretically confirmed the actual existing form of LiPS species in the electrolyte of Li–S batteries, yielding "new flowers blooming on old trees". It overturns the conventional view that the dissolved LiPS anions are negatively charged and provides a whole new understanding that can inspire fresh ideas for Li–S battery research, such as the redesign of Li⁺-selective membranes to retard LiPS diffusion, the regulation of cathodic EDL to anchor LiPSs, the regulation of anionic EDL to suppress parasitic reactions, and the mechanism of Li-ion transport with cationic LiPSs. The development and construction of high-performance Li–S batteries can thereby be greatly accelerated.