Introduction

An interface refers to the contact boundary plane, separating two phases that are chemically and/or structurally distinct, each of which may be solid, liquid or gaseous. The ubiquitous interfaces play a vital role in determining the properties and processing of almost all materials, which have promoted great interest in areas including catalysis1, electrochemistry2, nucleation and growth3, batteries4, optoelectronics5, and biological reactions6. For example, a thin liquid-like layer can serve as an intermediate for the mass transport from metal nanoparticles (NPs) to the liquids, differing from the traditional transfer behavior7. Similarly, the hot electron transfer has been found to be faster at the solid−liquid interfaces than that at the solid−gas ones in the Pt/n-Si system8. The oxidation of isopropyl alcohol to acetone catalyzed by platinum has been reported to be enhanced at the solid/liquid interfaces upon comparison with that at the solid/gas ones9, while the enhancement in the oxidation of alloys (e.g., nickel–chromium) occurs with the existence of water-vapor interfaces compared to that of pure oxygen10. Similar increment of etching rates has been observed in the solid–liquid–gas interfaces formed at the gold nanorods in liquids with the assistance from the adsorbed oxygen11.

Despite the aforementioned progress achieved in the leading materials with the solid form, the behaviors of liquid nanodroplets at different interfaces have been relatively less reported and understood. Unraveling the fundamental physicochemical processes of interfacial liquid droplets offer unique and promising applications in the fields including nanoreactors12, microfluidics13, nanomotors14, nanotransporters15, 3D additive manufacturing16, nanowelding17, nanowetting18, spraying19, and flexible/wearable devices20. It has been shown that the motion velocities of water microdroplets are increased dramatically on the superhydrophobic interfaces21, and the chemical reactions can be also facilitated on the surface of liquid droplets22 as well as the solid-liquid interfaces23,24. Moreover, the introduction of additional gradient external fields including electric force, temperature, light and surface free energy would further modulate the dynamical processes of droplets. For instance, the water droplets demonstrate the exceptional uphill motion through tuning the gradient surface energy caused by the interfacial morphologies25. Nevertheless, the full picture of nanoscale droplets dynamics from initial formation to subsequent evolution as well as the simultaneous structural/elemental identification remains challenging and unclear, leading to the isolated understanding and even uncertainty in the results. Therefore, direct visualization of the whole evolution of nanodroplets with in-situ capability is indispensable, which enables a thorough understanding of fundamental interfacial processes at complex multiphase boundaries (e.g., solid–liquid–gas and solid–liquid–liquid), and further manipulation of the interfacial behaviors at nanoscale.

Herein, using the massive cell fabrication method26, we encapsulated the HgS nanocrystals (NCs) in gas and liquid environments, and statistically investigated the full evolution dynamics of metal nanodroplets (Hg) at nanoscale solid-gas and solid-liquid interfaces by in-situ transmission electron microscopy (TEM). The choice of HgS as the studied system was based on the following considerations: (i) Facile formation of metal Hg nanodroplets from HgS upon the electron-beam (e-beam) excitation so that the full dynamics from the initial formation (birth) to the later evolutions were allowed; (ii) HgS NCs were water soluble, leading to the easy preparation of liquid cells; and (iii) Nanoscale saw-tooth morphology of HgS NCs enabled the investigations of motion behavior of the Hg nanodroplets. In the gas cells, the nucleation, growth and coalescence of the voids were observed, while simultaneously the Hg nanodroplets occurred, then moved rapidly on the ratchet surface, and finally coalesced into the large ones through the bridges. Interestingly, in the case of liquid cells, the ink-like liquid Hg jetted from the solid-liquid (HgS-water) interface. Determined by the competition between the reductive electrons and the oxidative species derived from the radiolysis of liquid, consecutive ink-jetting behavior happened with the interval time between two neighboring ones in the range from several to several tens of seconds. Further ex-situ control experiments verified the above behaviors. The underlying mechanisms for the above phenomena were thoroughly discussed.

Results and discussion

HgS characteristics for gas/liquid cell studies

Chiral α-HgS NCs were prepared by the seed-mediated epitaxial growth technique (see “Methods”). Figure 1a shows the helical arrangement of Hg and S atoms along the crystallographic c axis in HgS with the P3121 space group, which represents an atomic-scale primary chiral unit. Figure 1b shows the typical structure of the as-synthesized NCs. The measured interplanar spacings of 3.163 and 2.863 Å in high-resolution TEM correspond to the (003) and (102) planes of α-HgS that is indicated from the diffraction pattern (Fig. 1b, inset). The chirality of the NCs would be tuned through the enantioselective synthesis strategy, where an example is demonstrated in Fig. 1c. Using the massive cell fabrication method developed recently, we have encapsulated the HgS NCs in gas and liquid environments for later dynamics studies. The structural evolution of the HgS nanostructures in gas and liquid is visualized by in-situ TEM (Fig. 1d). An e-beam is utilized to excite the NCs, while simultaneously the structural information involved during the dynamical processes is recorded by a digital camera in real-time. The schematic diagram of the gas/liquid cell is illustrated in Fig. 1e. The difference in fabrication was due to the addition of the glycerol (retard the volatilization of water) into the core liquid precursor during electrospinning of the liquid cells. Figure 1f presents a typical TEM image of the as-prepared cell with a HgS NC inside. With the e-beam excitation, a certain number of dark clusters (marked in the area A) were steadily formed on the NC (Fig. 1g). The dark clusters were composed of Hg nanodroplets, which were verified by the energy dispersive X-ray spectroscopy (EDS). When the HgS NC located in vacuum (without the isolation by the cell), however, only the morphological framework remained after the long-term e-beam irradiation because of the evaporation of Hg droplets (Fig. 1h). The EDS analysis suggested that the remaining white feature (marked in the area B) was almost from S, verifying the evaporation of Hg.

Fig. 1: Structural characterization of HgS for in-situ gas/liquid cell TEM.
Fig. 1: Structural characterization of HgS for in-situ gas/liquid cell TEM.
Full size image

a Schematic illustration of the atomic structure of α-HgS. b Typical TEM image of the as-synthesized HgS and high-resolution image with the d-value of 3.163 and 2.863 Å corresponding to the (003) and (102) planes of α-HgS. The inset shows the corresponding diffraction pattern. c Circular dichroism (CD) spectra of the chiral and achiral HgS. d Schematic of the time-resolved imaging of evolution dynamics in the gas/liquid cell. e Schematic structure of the gas/liquid cell. f Typical TEM image of the electrospun cell containing the HgS NC. g TEM image and EDS spectrum (Area A) of the HgS NC after the e-beam irradiation. h TEM image and EDS spectrum (Area B) of the HgS NC after the long-term e-beam excitation. Source data are provided as a Source Data file.

Structural evolution of HgS NCs in gas cell

Figure 2 shows a series of TEM images depicting the nucleation and growth dynamics of Hg nanodroplets in the gas cell (see Supplementary Movie 1 for details). Herein, for the sake of convenience, the time zero (0 s) was defined as the moment of initiating the video recording (the initial TEM alignment was operated under a low dose rate of ≈50 e− Å−2 s−1 so that the structural evolution was not excited). At a high dose rate of 3240 e− Å−2 s−1 (Fig. 2a), abundant voids (white features marked by the rectangles) arose rapidly, accompanied with the appearance of Hg nanodroplets (marked by the circles) after e-beam excitation for 2 s. As the irradiation time elapsed (e.g., 15 and 19 s), the voids wriggled and coalesced when they met. The voids grew up after coalescence (Fig. 2b, a typical example given at 270 s), and later turned into a crack-like feature at 380 s (marked by the rectangle) through lengthening (310 s) and coalescence processes (365 s). Different from the phenomenon observed in vacuum27, the small Hg nanodroplets gradually grew up during the e-beam excitation. Upon comparison of the TEM images at 410 and 720 s (Fig. 2c, indicated by the circles), it showed that the Hg nanodroplets were further enlarged to a size of ≈10–15 nm due to the coalescence. At this stage, the droplets became relatively stable. When the dose rate was further increased to 7350 e− Å−2 s−1, however, the Hg nanodroplets moved again and became larger at 1150 s. Ultimately, a big and stable nanodroplet with a size of ≈70 nm appeared at 1250 s (the inset was a view of the whole HgS morphology). The dynamical evolutions of the voids and Hg nanodroplets are schematically illustrated in the rightmost column of the figure.

Fig. 2: Structural evolution of the voids and Hg nanodroplets in the gas cell.
Fig. 2: Structural evolution of the voids and Hg nanodroplets in the gas cell.
Full size image

a A series of time-dependent TEM images revealing the formation of voids and Hg nanodroplets as well as the coalescence of the voids. Scale bars, 20 nm. b The voids turned into the crack-like feature, and such crack was lengthened, coalesced and finally evolved into the long one at 380 s. Scale bars, 20 nm. c Hg nanodroplets gradually grew into a size of ≈10–15 nm from 410 to 720 s under a dose rate of 3240 e− Å−2 s−1. As the dose rate was increased to 7350 e− Å−2 s−1, the Hg nanodroplets was further enlarged through coalescence (1150 s) and finally became relatively stable with the size of ≈70 nm at (1250 s). Scale bars, 20 nm. The voids and Hg nanodroplets in (a–c) were marked by the dotted white rectangles and yellow circles, respectively.

Rather than the enlargement of Hg nanodroplets inside the gas cells, we also noted that the big nanodroplets could suddenly fade away if the cells were broken intentionally. As depicted in Supplementary Fig. 1, the Hg nanodroplets gradually grew as the irradiation time elapsed and had a size of ≈40 nm at 480 s. These nanodroplets suddenly began to fade away from 480 s, and almost disappeared at 515 s. Such evaporation and disappearance of the nanodroplets was similar to that observed in the vacuum condition27, which was due to the structural breakage of the gas cell that became vacuum environment in the TEM chamber.

Detailed Evolution of voids and Hg droplets in gas cell

To gain detailed insight into the dynamical behavior of voids and Hg droplets, the dose rate-dependent experiments were carried out, as depicted in Fig. 3. As the dose rate was increased from 2020 to 7310 e− Å−2 s−1, the stable size of Hg droplets increased from ≈1.6 to 40.7 nm (Fig. 3a, refer to the typical images in Supplementary Fig. 2). It indicates that Hg droplets would stabilize with an appropriate size at each specific dose rate and higher dose rate facilitated the formation of larger droplets in the closed gas cell. The growth speed of Hg nanodroplets was estimated to be from ≈0.004 to 0.162 nm s−1 when the dose rate was increased from 2020 to 4810 e− Å−2 s−1 (inset of Fig. 3a). Figure 3b shows the distribution characteristics of Hg and S elements as a function of the dose rate. Five different areas were selected to analyze the elemental distribution of the samples after the e-beam irradiation (an example was given in the inset of Fig. 3b), revealing the diffusion behavior of Hg nanodroplets. Upon comparison of Hg/S ratio at different dose rates, large variation was observed under the high dose rate, which was due to the random distribution of fewer Hg droplets with relatively big size (refer to the TEM image at 1150 or 1250 s in Fig. 2).

Fig. 3: Analyses of the behaviors from the voids and Hg nanodroplets.
Fig. 3: Analyses of the behaviors from the voids and Hg nanodroplets.
Full size image

a The relationship between the size of Hg nanodroplets and the dose rate. The inset shows the growth speed of Hg nanodroplets as a function of the dose rate. b The variation of the Hg/S ratio under different dose rates at five selected areas illustrated in the inset. c The movement speed of the voids and nanodroplets as a function of the dose rate. The inset presents the movement velocities at the dose rates lower than 4000 e− Å−2 s−1. Error bars in (a–c) represent the deviations of the mean (n = 3). d Time-dependent TEM images showing the coalescence process of the voids (indicated by the dotted white rectangles) through the bridges. The crack-like feature was formed after the coalescence. Scale bars, 20 nm. e The detailed coalescence process between two Hg droplets (marked by the dotted yellow circle) through the nanobridges. The inset at the 520 s image displays the location of such an example where the coalescence occurred. Scale bars, 10 nm. Source data are provided as a Source Data file.

The movement speeds of voids and Hg droplets at different dose rates were then quantitatively estimated, as presented in Fig. 3c. In the case of voids, the rate of movement increased from ≈0.19 to 5.04 nm s−1 with the enhancement of dose rate from 688 to 3520 e− Å−2 s−1 (inset of Fig. 3c). Similarly, the Hg nanodroplets exhibited the movement speed from ≈0.08 to 2.91 nm s−1 with the increment of dose rate from 1060 to 3520 e− Å−2 s−1. As the dose rate was further increased to a higher level (4340 e− Å−2 s−1), the Hg nanodroplets moved much faster and had a velocity of ≈55.94 nm s−1. Finally, they would reach the velocity of ≈65.14 nm s−1 under the dose rate of 7340 e− Å−2 s−1.

We noticed that the voids and Hg droplets grew after coalescence, which were detailed in Supplementary Movies 2 and 3. Figure 3d, e shows the sequential TEM images extracted from the videos, showing the coalescence processes of the voids and Hg droplets, respectively. At 259 s, several voids were seen in Fig. 3d (marked by the rectangles). After further e-beam irradiation for 10 s, a bridge began to appear between two neighboring voids with a distance of ≈18 nm (marked by the dotted arrows). Such bridges became clearly visible during the following period of 30 s. Subsequently, the large voids were formed at 324 s as a result of the coalescence. The large voids would finally turn into a long crack-like feature (349 s). On the other hand, the coalescence was also realized through the nanobridges between two Hg droplets (Fig. 3e). Initially (520 s), the two droplets (marked by the circles) were separated with a distance of ≈8 nm. When they approached at 6 nm, a bridge (marked by the dotted arrow) appeared at this stage (555 s). The bridge got wider with the time (565 s), which induced the subsequent coalescence of the two droplets (585 s). As the time elapsed (605 s), the coalesced droplet continued changing its shape to further reduce the surface energy.

Further observations at high resolution were performed to achieve the detailed understanding of the formation and coalescence behaviors of the voids and Hg nanodroplets. As illustrated in Fig. 4a (Supplementary Movie 4), the initial HgS NC exhibited regular crystalline lattices under the vacuum environment. After 12 s of e-beam exposure, a small portion of the lattice sites vanished (became disordering), resulting in the formation of the void. More voids appeared upon the continuous irradiation, e.g., two voids were observed at 17 s, but they were still separated by the lattice points. With the removal of such lattice sites, a bridge (indicated by the dotted arrow) was created between these two voids at 18 s, leading to the subsequent coalescence of the voids at 20 s. As the time elapsed (22 and 24 s), another surrounding void occurred and grew, with the formation of the bridge at 26 s. Those two voids were then coalesced through the bridge at 30 s and eventually evolved into a large one at 40 s, concurrent with the disappearance of the lattices. Figure 4b shows the structural evolution of the Hg nanodroplets in the carbon-film gas cell (see Supplementary Movie 5). Unlike the situation in vacuum, several Hg nanodroplets (marked by the dotted circles) were formed besides the appearance of the voids at 10 s. We noticed that the voids (e.g., at 30 s) between the Hg droplets acted as the diffusion channels for liquidus Hg, which resulted in the subsequent formation of the nanobridges (indicated by the dotted arrows at 60 s). Through these bridges, the Hg nanodroplets were gradually merged into the large ones (e.g., at 90 s). Similar processes were also repeated during the later stages (e.g., 120-300 s).

Fig. 4: High-resolution observations of the formation and coalescence behaviors of the voids and Hg nanodroplets.
Fig. 4: High-resolution observations of the formation and coalescence behaviors of the voids and Hg nanodroplets.
Full size image

a Structural evolution of the voids under the vacuum condition. The voids were marked with the dotted white rectangles, while the bridges were indicated by the dotted white arrows. Scale bars, 20 nm. b In-situ TEM images showing the evolution of the Hg nanodroplets in the carbon-film gas cell. The Hg droplets were marked with the dotted yellow circles. Scale bars, 20 nm.

Ink-jetting phenomenon of HgS NCs in liquid cell

To study the Hg behavior at the solid-liquid interface, the structural evolution of HgS NCs in liquids was conducted. Figure 5a shows the evolution of a single NC from a series of sequential TEM images (Supplementary Movie 6). The experiment was operated under a dose rate of ≈5200 e− Å−2 s−1. Under the e-beam excitation (20 s), the voids were formed rapidly with the appearance of some small Hg nanodroplets (see details in Supplementary Fig. 3). The voids were eventually turned into the large ones (marked by the rectangle) through the coalescence process (e.g., 312 s and Supplementary Fig. 4). Different from the phenomena observed in the gas cell, the structural evolution of HgS in water underwent periodic jetting of Hg liquid, hereafter named as the ink-jetting event (such like the “ink-jetting” from cuttlefish). Upon irradiation for a relatively long time, the first ink-jetting began to appear at 338 s (marked by the circle). It spread quickly during the following seconds and then finished at 346 s. Within the next tens of seconds, the ink-jetting additionally occurred five times (the commence and end processes similar as that displayed above) and finally ended at 470 s. The beginning ts and finish time tf of each ink-jetting event are summarized in Fig. 5b. From the inset, it can be seen that the period of spanning a whole ink-jetting process \(\Delta t={t}_{{\rm{f}}}-{t}_{{\rm{s}}}\) was in the range of several to tens of seconds. Note that the time needed to initiate the ink-jetting event was decreased with the increase of the dose rate (Supplementary Fig. 5). Figure 5c presents the dose rate-dependent jetting speed of the ink feature. The jetting rate of the ink increased from ≈2.7 to 16.4 nm s−1 with the increment of dose rate from 3620 to 8780 e− Å−2 s−1. The Hg/S ratios under the different dose rates are shown in Fig. 5d (the selected areas for EDS are given in the inset), revealing the distribution of Hg and S elements after the ink-jetting behavior. The variation of the Hg/S ratios was not significant under different dose rates, which was different from the overall trend observed in the gas environment (random distribution of large droplets displayed in Fig. 2).

Fig. 5: Jetting behavior of liquid Hg in the liquid cell.
Fig. 5: Jetting behavior of liquid Hg in the liquid cell.
Full size image

a Time-dependent TEM images showing the ink-jetting phenomenon upon e-beam excitation of the HgS NC. Six times of ink-jetting was observed, revealing the cycling of the beginning-spread-finish processes. The dotted white rectangles and yellow circles indicate the locations of the voids and ink-jetting, respectively. Scale bars, 20 nm. b The beginning and finish times for each ink-jetting event. The inset shows the total time of each ink-jetting process. c The jetting speed of the liquid Hg as a function of the dose rate. d The relationship between the Hg/S ratio in different areas and the dose rate. The variation of the Hg/S ratio was not obvious, compared to that in the gas cell. Error bars in (c, d) represent the deviations of the mean (n = 3). Source data are provided as a Source Data file.

When multiple HgS NCs existed in the liquid cell, similar continuous ink-jetting phenomenon also occurred (Supplementary Fig. 6). During the e-beam irradiation, the voids preferentially arose and grew up, e.g., at the stages of 65 and 107 s. The first ink-jetting began at 138 s, then spread and lasted for several seconds, and finally finished at 150 s. Afterwards, the second one was triggered at 169 s. The ink-like feature became larger in size at 170 and 172 s, and eventually stopped at 182 s. The last one (the seventh) was evolved in a similar way (began and finished at 282 and 294 s, respectively). Ultimately, the HgS NC became relatively stable and showed no obvious morphological change. Note that other neighboring NCs also experienced the similar ink-jetting, forming an agglomerated particle at the end (the inset image at 294 s).

Void nucleation, growth, and coalescence

In the experiments we observed the evolution of voids that was categorized into the nucleation, growth, and coalescence processes. It has been documented that the voids would be formed through the thermal heating28, local stress concentrators caused by extended defects (e.g., dislocations and grain boundaries)29, Kirkendall effect30 and atomic displacement induced vacancy zones31. The local stress concentrators and the Kirkendall effect might not mainly account for the void formation, because no grain boundaries and/or diffusion couple with different rates were involved during the process. The e-beam-induced temperature rise was estimated to be several degrees based on ref. 27, which was also in accordance to the typical range of the temperature rise reported in refs. 32,33. Meanwhile, the electrospun material for encapsulating the liquids is polyvinyl acetate (PVAc), which has the melting point of ≈60 oC. The PVAc shell remained intact throughout the whole in-situ TEM experiments, providing experimental evidence that the e-beam-induced temperature rise did not exceed ≈60 oC. To further experimentally validate the heating effect, we conducted the in-situ heating TEM observations (low electron dose rate of ≈20 e− Å−2 s−1 for almost neglecting the e-beam heating effect) on the associated dynamics (Supplementary Fig. 7 and Supplementary Movie 7). Through analyzing the TEM images and diffraction patterns, the morphology and crystal structure of the HgS NC remained unchanged after exposure to the temperature of 100 oC even higher than the melting point of PVAc.

In addition, e-beam on-off experiments were also conducted to verify the individual effect from the electron beam. Basically, the e-beam was deactivated for a specified duration before being reactivated, enabling a comparison of the size and position of the voids and Hg droplets. As illustrated in Supplementary Fig. 8 and Supplementary Movie 8, several voids appeared after 60 s of the e-beam excitation at a dose rate of ≈1100 e− Å−2 s−1. Subsequently, the e-beam was deactivated for 120 s. During the period without the e-beam, both the size and position of these voids remained unchanged; however, they exhibited movement and coalescence behaviors once the exposure of the e-beam started again. A similar phenomenon was observed when the e-beam was turned off at 240 s and reactivated at 360 s. Likewise, the Hg nanodroplets also moved and coalesced during the e-beam exposure, while their size and position remained stable in the absence of such irradiation (Supplementary Fig. 8 and Supplementary Movie 9). Based on the above control experiments, neither the e-beam-induced temperature rise nor the external heating was the main reason for the observed phenomena in this work.

In the absence of material discontinuities, the direct transfer of e-beam energy to atoms during collision may be large enough to knock them out of their lattice sites, creating the vacancies so that the condensation of vacancy clusters induces the voids. On the basis of both energy and momentum conservation, the maximum kinetic energy EA transferred from an electron to an atom through collision is estimated using the following equation,

$${E}_{{\rm{A}}}=561\varepsilon\, \left(\varepsilon+2\right)/{\rm{A}}$$
(1)

where \({\rm{A}}\) is the weight of atom and \({\varepsilon }\) = E/(mc2), E is the original energy of electron in TEM with 200 kV accelerating voltage, m is the mass of electron, and c is velocity of light. For the typical 200 keV e-beam in TEM, the EA values were calculated to be 2.62 and 16.37 eV for Hg and S, respectively. Upon comparison to the vacancy formation energies Ev of 4.5 and 7.8 eV for Hg and S, it indicates that it was easier for S atoms to be knocked out than that for Hg atoms. According to the EDS analysis, the loss of Hg was more than that of S element (Supplementary Fig. 9), which was owing to the occurrence, movement, and evaporation of Hg droplets (Fig. 3). Such evolution of Hg droplets would further offer localized reduction in displacement energy for S, energetically favoring the vacancy clustering to facilitate the evolution of voids.

In the subsequent stage, the dynamical evolution encompasses the simultaneous growth and coalescence of existing voids as well as the formation of few new ones. Compared to the nucleation of new ones, the small voids tended to add into the existing voids and to merge to form the large ones, which was probably driven by the reduction of the total energy of the system. Interestingly, unlike the direct coalescence of voids, e.g., the behavior reported in the Bi nanoparticles induced by heating, the coalescence between two neighboring voids in the HgS NCs was achieved by the bridge with ≈18 nm (Fig. 3). Ultimately, the large voids turned into a long crack-like feature to further lower the total energy. Such feature preferentially propagated along the <001> direction (refer to Supplementary Fig. 10 for details) of the crystal due to the fewer number of bonds bridging unit cells perpendicular to this direction. In this work, the structural evolution of the crystal tended to form the voids instead of the volume contraction so that the framework was maintained. This was probably due to the thermodynamic preference for reduction of the total surface energy of the voids greater than the crystal surface energy.

Formation, movement and coalescence of Hg nanodroplets

During the e-beam excitation of HgS semiconductor, a number of electron-hole (e–h) pairs are generated within an excitation volume. For a 200 kV e-beam with the typical current of 1.2 nA used here, the local rate of carrier pair generation was calculated to be ≈3.10 × 105 pairs per second in the HgS NC (Supplementary Fig. 11, Supplementary Tables 1, 2, and Supplementary Note 1). The utilization of such generated electrons was then roughly estimated. As presented in Fig. 2, eight discernible Hg nanodroplets with a diameter of ≈2 nm ( ≈ 1.46 × 103 atoms) formed at 2 s while ten bigger nanodroplets (diameter of ≈10 nm) appeared at 400 s which contained ≈2.27 × 105 Hg atoms. Within these e-beam excitation periods, the number of generated electrons were 6.20 × 105 and 1.24 × 108 for 2 s and 400 s, respectively. Despite the rough estimation, it provided a pictorial understanding for the formation of Hg nanodroplets (Hg2+ + 2e \(\to\) Hg), namely, the formation of the 2 nm ( ≈ 0.47%) and 10 nm ( ≈ 0.37%) nanodroplets only consumed less than one percent of the reductive electrons, whereas most of the carriers were recombined in the material.

After the formation, those Hg nanodroplets exhibited relatively large movement speed of ≈56 nm s−1 at the dose rate of 4340 e− Å−2 s−1, in comparison to that of ≈5-10 nm s−1 observed for the Bi nanodroplets34. Although the movement of metal nanodroplets has been rarely reported, several mechanisms about the driving forces originated from the gradient fields including temperature21, surface energy35, electric field36 and light field37 have been proposed for other common liquid droplets (e.g., water). Those factors might not be applicable or play the essential role for the phenomena observed in this work. It has also been shown that the ratchet surface leaded to the propulsion of the liquids, which was driven by the viscous force between the solid/liquid interface due to the Leidenfrost effect38. The HgS nanocrystals in our work adopted a saw-tooth morphological feature so that the Hg nanodroplets preferred to curve concavely near the tops of the ridges while presenting the convex characteristics elsewhere (Supplementary Fig. 12). In this scenario, we speculated that such ratchet surface may contribute to the fast movement of the Hg nanodroplets. During the e-beam-induced reduction of Hg, the nucleation of the nanodroplets occurred once the vapor pressure of Hg exceeded the saturated one. Such variation in the surface curvature would generate a pressure differential Δp between the concave ridge and the neighboring convex position, forming a net force to drive the motion of the nanodroplet (Supplementary Fig. 12). Such a pictorial understanding was again verified by the fact that the glide velocity of the Hg nanodroplets increased dramatically at the high e-beam dose rates since more Hg vapor was formed upon the irradiation (Fig. 3).

During the movement, the Hg nanodroplets would coalesce with a relatively large distance of ≈6 nm through the bridges (Fig. 3), which was different from that observed in other systems. For example, the Bi nanodroplets underwent a sudden coalescence on the SrBi2Ta2O9 platelet39, while the fusion or coalescence of Au nanocrystals was realized by the nanochannel with a critical spacing of less than 1 nm40,41,42. Moreover, the coalescence of nanobubbles in liquids was observed to occur within a distance of ≈2 nm between two bubbles43. In this work, the e-beam irradiation induced the reduction of Hg2+ to Hg on the HgS NCs. As the time elapsed, the accumulation of Hg atoms might create the atomic chains accordingly, resulting in the formation of relatively long bridges that finally leaded to the coalescence between the nanodroplets.

Mechanism for continuous ink-jetting behavior

It is known that the irradiation of water by the electron beam generally generates a variety of radiolysis products, including the common oxidative •OH, H2O2 and reductive solvated electron eh− species44,45,46,47,48,49. The steady-state concentrations of the radiolysis species were calculated and summarized (Supplementary Note 2 and Supplementary Tables 3, 4, 5), presenting that the concentration of the oxidative species was higher than that of the reductive ones. Meanwhile, the oxidative species possess very high reduction potentials (Supplementary Table 6), which have been reported to play the vital role for etching the nanostructures (especially metals) in liquids31,45,46,47,48. Despite the inevitable existence of slight etching, the HgS semiconductor studied here demonstrated the distinct behavior from that was shown in the metal nanostructures (almost pure etching).

Different types of scavengers were designed to suppress a certain number of specific species so that the dominated ones for the ink-jetting phenomenon would be clarified. For each experiment, a specific type of the scavenger was added into the solution during the liquid cell preparation. Afterwards, such as-prepared liquid cell sample was used for the in-situ TEM studies. Figure 6a shows the sequential TEM images in the electrospun liquid cells with the addition of H2O2 (scavenger for reductive species). In this case, the e-beam-induced oxidative species would predominate. The HgS NCs were gradually etched and more cavity feature was formed with the increment of time. No ink-jetting phenomenon was observed even during the long irradiation for ≈300 s. The above behavior was further validated in the thin carbon liquid cells with the same addition of H2O2 (Fig. 6b). Obvious contrast from the bubbles was observed, manifesting the liquid environment inside the carbon liquid cell (Supplementary Fig. 13). Similar etching phenomenon was also observed as the time elapsed to several hundreds of seconds. Besides the e-beam irradiation, the effect of the irradiation from other common source (i.e., photons) on the evolution dynamics was further investigated. As is well-known, upon photon excitation similar oxidative and reductive species are generated in liquids with the involvement of semiconductor materials50. For the liquid cell sample with the addition of H2O2, no ink-jetting behavior was observed either under the photon irradiation, except the gradual etching of the HgS NCs (Supplementary Fig. 14 and Movie 10). Additional ex-situ experiments were designed to distinguish the etching effect from either H2O2 or •OH (see Methods for details). Figure 6c depicts the TEM images and EDS spectra of the HgS NCs treated by •OH. Upon comparison to the initial intact structure of the NCs, the change in the morphology as well as the decrease in the EDS intensity was observed in the sample after etching for 60 min. When the etching was increased to 180 min, most parts of the NCs were etched off, resulting in the complete collapse of the initial bipyramid structure and the transformation of the crystalline phase into the amorphous one identified from the diffraction pattern. At this stage only slight remaining elements were detected in EDS. The associated EDS mapping images of the above products under different etching periods were illustrated in Supplementary Fig. 15. In contrast, the HgS NCs with the addition of pure Co2+ or H2O2 presented no obvious morphological and crystalline phase change at the same time scale (Supplementary Fig. 16). To further identify the oxidative species, additional electron paramagnetic resonance (EPR) experiments were conducted. In EPR, the common 5,5-dimethyl-1-pyrroline N-oxide (DMPO) was used as the scavenger for combining radicals to form the relatively stable radical adducts. As illustrated in Fig. 6d, characteristic signals from the DMPO − •OH adducts51 were detected at 15 min, and further enhanced at 30 min following the addition of Co2+ into H2O2, while no obvious signal was observed in the pure H2O2. This suggests that the hydroxyl radicals were produced through the reaction between Co2+ and H2O2. Thus, upon comparison of the above results under different irradiation sources, it indicates that the •OH plays the dominant role in etching the HgS NCs.

Fig. 6: Liquid environments with addition of reductive/oxidative scavengers.
Fig. 6: Liquid environments with addition of reductive/oxidative scavengers.
Full size image

a Sequential TEM images of the etching phenomenon in the electrospun liquid cell with the addition of H2O2. Scale bars, 20 nm. b Control experiment in the carbon liquid cell with the addition of H2O2. Scale bars, 20 nm. c TEM images, diffraction patterns (insets) and EDS spectra of the HgS NCs treated by the Fenton reaction for different periods. d EPR spectra of the hydroxyl associated radicals under different reaction conditions. e Time-dependent TEM images of the ink-jetting behavior in the electrospun liquid cell with the addition of methanol. The locations of the voids and ink-jetting were marked by the dotted white rectangles and yellow circles, respectively. Scale bars, 20 nm. f The e-beam irradiation time and g the interval time between two neighboring ink-jetting processes as a function of the jetting frequency in water and that with addition of methanol. Error bars represent the deviations of the mean (n = 3). Source data are provided as a Source Data file.

Figure 6e shows the time-dependent TEM images of the ink-jetting phenomenon in the electrospun liquid cells with the addition of methanol (scavenger for oxidative species, Supplementary Movie 11). Similar to the structural evolution of HgS NCs in water, the voids arose rapidly and the ink-jetting phenomenon occurred consecutively. Distinctly, the first ink-jetting occurred at ≈66 s, which was significantly shorter than that in the water condition ( ≈ 314 s). Meanwhile, the interval time \(\Delta t^{\prime}={t}_{{\rm{n}}}-{t}_{{\rm{n}}-1}\) between each ink-jetting event also became shorter, e.g., the fifth jetting behavior was finished within only ≈4 s. Given the nature of oxidative scavenger in the case of methanol, it suggests that the reductive electrons might play the key role for the continuous ink-jetting phenomenon. This was further evidenced by the fact that the Hg nanodroplets immediately nucleated under the reductive environment after the e-beam excitation for 2 s in the gas cells with the absence of •OH (Fig. 2). For the case of the liquid cells, the longer time needed to trigger the ink-jetting (Fig. 5, 338 s) indicates that part of the reductive electrons was consumed by the oxidative species and the long-time accumulation of the sufficient electrons therefore allowed the subsequent ink-jetting.

Based on the above discussion, a competition exists for the reduction of Hg2+ to the formation of Hg droplets, i.e., the reductive electrons promote such behavior whereas the oxidative species inhibit it. Once the accumulated electrons reach the threshold value during such competition process, the ink-jetting behavior occurs. We noticed that such behavior is irrespective of the initial nanodroplet size (Supplementary Fig. 17), which is ascribed to the fact that the oxidative species in the liquid environment retard the growth of the nanodroplets (for comparison, refer to the obvious growth illustrated in the gas cells). With the continuous irradiation, the small Hg nanodroplets tend to be coalesced together for reducing their surface energies, forming the ink-like feature at the droplet-water interface. At such interface, the surface tension is further decreased from the value of pure Hg52 to that of the interfacial Hg-H2O one53, leading to the spreading behavior of the ink-like membrane (ink-jetting). Since the reductive electrons are consumed in each reduction-involved ink-jetting process, a certain amount of time would be thereby required to accumulate enough electrons for the next jetting behavior. Figure 6f shows the relationship between the e-beam irradiation time and the ink-jetting frequency under different liquid conditions. At each frequency, the irradiation time needed to trigger the ink-jetting under the addition of methanol was shorter than that in the water. The interval time between two sequential ink-jetting processes is quantitatively displayed in Fig. 6g. Compared to the interval time of ≈11-22 s in water, the time span under the addition of methanol was significantly shortened ( ≈ 3-6 s). The phenomenon was ascribed to the consumption of a certain number of oxidative species through the addition of methanol. In this scenario, more electrons had the chance of being involved in the reduction process, leading to the shortened time for triggering the ink-jetting behavior.

Nanodroplet dynamics in gas/liquid cells

The overall evolution pictures of Hg Nanodroplets in the gas and liquid cells are schematically summarized in Fig. 7. In the gas cells (Fig.7a), the direct transfer of e-beam energy to the Hg and S atoms knocks them out of the lattice sites, creating a large number of the vacancies. The condensation of the vacancy clusters induces the voids, which is also accompanied by the appearance of Hg nanodroplets due to the reduction of Hg2+ to Hg. Subsequently, the voids gradually grow up after coalescence through the bridges, which further turn into the crack-like structure preferentially along the <001> long-axis direction. Meanwhile, the Hg nanodroplets move rapidly, and come into coalescence through the bridges when they are close each other, forming the bigger ones accordingly. With the continuous coalescence process, the large droplets are finally formed and stay relatively stable on the substrate.

Fig. 7: Schematics of overall evolutions of Hg Nanodroplets in the gas and liquid cells.
Fig. 7: Schematics of overall evolutions of Hg Nanodroplets in the gas and liquid cells.
Full size image

a Dynamics of the voids and Hg nanodroplets in the gas cell. Upon the e-beam excitation, the voids and Hg nanodroplets are formed, which are coalesced into the large ones through the nanobridges. The large voids are evolved into the crack-like feature along the long-axis direction of the HgS nanobipyramid. b Ink-jetting behavior in the liquid cell. The voids are formed rapidly with the occurrence of small Hg nanodroplets. Such voids are evolved into the big ones at the later stage. As the accumulated electrons reach the threshold, the ink-jetting suddenly happens. The ink-like feature spreads, which will be ended after a certain time. The ink-jetting occurs several times with the interval time depending on the competition between the reductive electrons and the oxidative species in the liquid cell.

In the case of liquid cells (Fig. 7b), the voids also arise rapidly in the preliminary stage upon the e-beam excitation, which are evolved into the large ones as the time elapses. Compared to that in the gas cells, distinctly, some smaller Hg nanodroplets appear in the early stage, but without the apparent growth. After the long-time excitation so that the accumulated electrons achieve the threshold, the ink-jetting behavior begins to occur. Such ink-like feature spread, and then becomes less obvious at the later stage. After a certain time of additional e-beam excitation, the second ink-jetting appears once the reaccumulation of the reductive electrons becomes sufficient for triggering such behavior. In this regard, the continuous the ink-jetting behaviors become feasible, with the interval time of two neighboring ones depending on the competition between the reductive electrons (accumulation) and the oxidative species (consumption of electrons). Ultimately, a stable liquid Hg layer is formed after the several ink-jetting processes at a certain dose rate.

In summary, the distinct evolution dynamics of the Hg nanodroplets mediated at the solid-gas and solid-liquid interfaces were directly visualized and statistically investigated by in-situ TEM. Upon exposure to e-beam in the gas cells, the voids and Hg nanodroplets nucleated and grew up, which were then coalesced into the large ones through the nanobridges. The voids could be further evolved into the crack-like structure preferentially along the <001> direction of the HgS solid substrate. The Hg droplets moved fast at the solid-gas interface and finally became relatively stable at each dose rate, which was distinct from the evaporation behavior observed at the solid-vacuum interface. In contrast to the typical behavior of the voids and Hg nanodroplets observed at either solid-gas or solid-vacuum interface, statistical results revealed the occurrence of the ink-jetting phenomenon in the liquid cells. Two competitive factors governed the interval time between the neighboring ink-jetting events, namely, the reductive electron species facilitated the shortening of such period while the oxidative species like •OH acted in the opposite manner. These phenomena were further verified by the control experiments with the addition of additional H2O2 or methanol. Our results contribute to advance a fundamental understanding of the full and unique evolution picture of liquid metal nanodroplets at nanoscale solid-gas and solid-liquid interfaces, and provide the practical implications in interface science and engineering. For instance, vermilion, commonly used in paintings, would experience degradation-induced darkening (the occurrence of liquid mercury spreading at the interface) under external illumination. Besides the approach of initiating the redox environment by electrons or photons, large quantity of metal nanodroplets can be achieved at desired locations of appropriate substrates through the similar strategy of supplying reductive electrons in the electrochemical route (Supplementary Fig. 18). Furthermore, the modulation of nanodroplets behavior at different interfaces enables the potential functionality in various fields including nanotransporters delivery system with controllable size, dosage and speed, and nanoscale cold-welding at precise targeted locations.

Methods

Synthesis of HgS seeds

The HgS seed NPs were synthesized based on a wet-chemical method. Specifically, 128 mg of Hg(NO3)2·H2O was added into 20.0 mL of deionized water in a round-bottom flask, followed by the dropwise addition of 4 mL of D-form penicillamine aqueous solution (0.09 M) under stirring to obtain a colorless solution. Then, 0.6 mL of NaOH aqueous solution (2 M) was added into the above colorless solution. After further stirring for 2 min, 2.0 mL of thioacetamide solution (0.18 M) was added into the mixture solution. The flask was sealed and placed in a 38 oC water bath for 15 h under stirring. After the reaction, the seed NPs were collected by centrifugation at 6,000× g for 10 min, washed three times with isopropanol, and re-dispersed in 5.0 mL of deionized water for further use.

Epitaxial growth of HgS NCs

Chiral α-HgS NCs were prepared by the seed-mediated epitaxial growth method using the above as-synthesized seeds. Before the epitaxy, a Hg precursor solution was prepared by the dissolution of 74.4 mg Hg(NO3)2·H2O into 7.2 mL of deionized water followed by the addition of 1.8 mL of D-penicillamine aqueous solution (0.09 M), whereas a S precursor solution was obtained by dissolving 19.8 mg of thioacetamide in 10 mL of deionized water. For a typical epitaxial growth, 50 μL of HgS seeds colloidal solution was dispersed into 5 mL of deionized water in a three-neck round-bottom flask submerged in the 38 oC water bath. 1.0 mL of D-penicillamine aqueous solution (0.09 M) and 0.6 mL of NaOH aqueous solution (2 M) were dropwise added into the flask under stirring. Subsequently, the as-prepared Hg and S precursor solutions were slowly injected into the flask by a syringe pump with an injection rate of 1.0 mL h−1 for 4 h. After the reaction, the orange precipitates were obtained by centrifugation at 6000× g for 10 min, washed three times with isopropanol and finally dispersed in 10 mL of deionized water for later studies.

Preparation of TEM samples and carbon-film gas cells

The TEM samples for observation in vacuum were prepared by air-drying of the α-HgS colloidal suspension on carbon-coated copper grids under ambient conditions. To prevent aggregation of the NCs, the colloidal suspension was sonicated for ≈10 min prior to deposition on the grids. To visualizing the evolution dynamics of the NCs with better resolution, the carbon-film gas cells were prepared. By using the film coating system (Leica EM ACE600), such gas cells were realized by depositing a thin layer of carbon ( ≈ 10 nm) onto the aforementioned samples for encapsulating the NCs.

Liquid and gas cell fabrication

The liquid cells were prepared by a coaxial electrospinning technique reported in our previous study. First, two kinds of liquid precursors (shell and core liquid precursors) were prepared, i.e., the shell liquid precursor was obtained by dissolving 3 g of polyvinyl acetate (PVAc) in 20 mL of dimethyl carbonate (DMC) under stirring for 8 h, while the core liquid precursor was prepared by dispersion of 1 mg of the as-prepared chiral α-HgS NCs into 10 mL of deionized water. Note that 0.05 g of glycerol was also added into the core liquid precursor to reduce the water volatilization during electrospinning. Then, the core and shell liquid precursors were pumped into two syringes for electrospinning with the flow rates of the core and shell liquids set at 2.0 μm s−1 and 7.8 μm s−1, respectively. The applied voltage and roller speed were set as 15 kV and 60 r min−1, respectively. Finally, the copper grids with ≈20 nm thick carbon films were directly used to collect the fibers for the ready-to-use liquid cell TEM (collection time was 45 s). For the gas cell fabrication, the procedures were almost the same but without the addition of glycerol so that the core water evaporated during electrospinning.

To investigate the effect of the scavengers on the reductive and oxidative species, the liquid cells with the additional addition of either H2O2 or methanol were also fabricated by electrospinning. All the electrospinning steps were the same as that shown above, except the preparation of the core liquid precursors. For this preparation, the extra addition of either 0.5 mL of H2O2 or 0.5 mL of methanol into 10 mL of HgS suspension (0.1 mg mL−1) was acted as the core precursor. Such as-prepared liquid cells with the addition of scavengers were used for later in-situ TEM studies, which were compared to that of the samples without the scavengers.

In addition, the carbon-film liquid cells were further prepared to verify the phenomena observed in the electrospun ones. Briefly, 0.5 mL of H2O2 was firstly added into 10 mL of HgS suspension (0.1 mg mL−1). Then, 1.5 μL of the above suspension was dropped onto one piece of the carbon film-coated grid and subsequently sandwiched by another one face to face. The liquid cells were dried naturally under ambient atmosphere for about 2 h. After the evaporation of excessive liquid solution, the liquid pockets containing the HgS NCs were finally formed between two thin carbon films owing to van der Waals force.

Electrochemical route of preparing Hg nanodroplets

The electrochemical experiments of preparing the Hg nanodroplets were conducted on an electrochemical workstation (CHI-760E, Chenhua Instrument) through a typical three-electrode setup in 0.1 M KOH electrolyte. In the three-electrode system, the glassy carbon electrode (3 mm in diameter), Ag/AgCl electrode and graphite rod were used as the working, reference and counter electrodes, respectively.

To prepare the ink for the working electrode, 2.5 mg of carbon black, 15 μL of 5 wt.% Nafion solution and 1 mL of HgS aqueous solution were mixed together and sonicated for 30 min to make the HgS NCs and Nafion uniformly dispersed. Then, 15 μL of the solution ink was evenly dropped on the surface of the glassy carbon electrode and dried at room temperature. The linear sweep voltammetry (LSV) measurements were performed with a potential window ranging from -0.6 to -1.9 V, with a scan rate of 50 mV s−1. The Hg nanodroplets were formed after the LSV experiments.

Ex-situ observation of etching behavior

To investigate the effect from the oxidative species, the Fenton reaction was performed to produce the •OH radicals. Briefly, 30 mg of CoCl2 power was ultrasonically dispersed into 10 mL of HgS suspension (0.1 mg mL−1). The Fenton reaction was triggered by the addition of 0.5 mL of H2O2 into the above solution at room temperature under dark environment with stirring. The Fenton reaction formula is shown as follows,

$${{{\rm{Co}}}}^{2+}+{{{\rm{H}}}}_{2}{{{\rm{O}}}}_{2}\to {{{\rm{Co}}}}^{3+}+{{{\rm{OH}}}}^{-}+• {{\rm{OH}}}$$
(2)

0.5 mL of suspension was collected at each given reaction time and added into a test tube containing 1 mL of methanol which was used to terminate the Fenton reaction. Finally, a drop of the above suspension was directly dropped onto the carbon film-coated grid for TEM imaging.

We also prepared the samples with the addition of only either H2O2 or CoCl2 to distinguish the effect from the above •OH radicals. 5 mL of H2O2 or 30 mg of CoCl2 power was added into 10 mL of HgS suspension (0.1 mg mL−1) at room temperature, which was placed under dark environment with stirring. The other procedures were the same to that presented in the above one. The as-prepared samples were finally used to the later TEM characterization.

Chirality and surface profile characterization

The chirality of HgS samples was characterized by circular dichroism spectroscopy. The circular dichroism spectra were obtained on a JASCO J-1500 spectropolarimeter with an optical length of 10 mm at 293.15 K. Atomic force microscopy (AFM) was used to reveal the surface profiles of the HgS NCs (FastScan Bio, Bruker, USA).

EPR characterization of radical associated species

Spin trapping allows the observation of transient free radicals by reacting such short-lived radicals with a spin-trapping agent. For the detection of hydroxyl (•OH) free radicals, 50 mM of DMPO was employed as the spin-trapping agent in an aqueous solution. The EPR spectra of the DMPO-treated samples were acquired using a Bruker EPR I200 spectrometer at room temperature, with a central field of 3320 G and a sweep width of 140 G.

In-situ TEM visualization of structural dynamics

Upon the e-beam irradiation, the structural evolution of HgS NCs in the liquid and gas cells under different conditions was in-situ visualized in an FEI Talos F200X TEM operating at 200 kV. To minimize the total electron dose during searching the target sample, the sample was observed at a low magnification ( ≤ 4300×, electron dose rate ≤ 10 e− Å−2 s−1). Once a suitable region was identified, the sample was moved to an adjacent area outside the interested region, and then the magnification was adjusted to a level appropriate for imaging the reaction dynamics (typically 100,000× or higher). Afterwards, the electron dose rate was changed to the desired level by adjusting the “brightness knob” from the instrument control panel. Finally, the sample was moved back to the target region for the subsequent recording of the dynamics. In this work, the electron dose rate was tuned in a range of 500-10000 e− Å−2 s−1 to ensure the initiation of the structural evolution. To reveal the compositional change during the evolution, the high-angle annular dark-field (HAADF) and EDS techniques in the scanning transmission electron microscopy (STEM) mode were utilized to analyze the elemental composition and distribution under different dynamical stages.

Besides the e-beam irradiation, the effect from another source on the liquid cell samples, i.e., the photon irradiation, was also investigated through the in-situ ultrafast TEM setup54,55,56. A green laser (wavelength of 532 nm) was introduced into the TEM chamber for exciting the sample dynamics, while the e-beam was utilized to probe the evolution processes. The laser fluence used in this work was ≈40 mJ cm−2. To neglect the impact from the e-beam, a low dose rate of ≈2 e− Å−2 s−1 was applied for imaging. The in-situ observation procedures were the same as that shown above.

Calculation of the carrier pair generation

The local carrier pair generation rate was calculated using Eq. (3)57:

$$g= < g > \frac{{EI}(1-f)}{q{E}_{{\rm{eh}}}}({\rm{carrier}}\, {{\rm{pairs}}}\,{{\rm{s}}}^{-1})$$
(3)

where <g> is the normalized distribution, E is the energy of e-beam, I is the beam current, f is the fraction of e-beam reflected by the sample, q is the electron charge, and Eeh is the energy cost for the formation of a single electron-hole (e–h) pair. These parameters were listed in Supplementary Tables 1 and 2. Based on the sample geometry and electron range shown in Supplementary Fig. 11, the generation of the carrier pairs in the HgS NC was then estimated. The detailed estimation procedure was presented in Supplementary Note 1.

Calculation of the number of Hg atoms in one Hg nanodroplet

Based on the basic parameters of Hg (one unit-cell volume is 69.22 Å3, which contains 3 Hg atoms), the number of Hg atoms were calculated to be ≈1.82 × 102 and 2.27 × 104 for the 2 nm-sized and 10 nm-sized Hg, respectively.

Estimation of the steady-state concentration of radiolysis species

The steady-state concentration of the radiolysis species in a liquid cell produced by a 200 kV e-beam was estimated by the kinetic model of Elliot and McCracken44,58. Briefly, it was supposed that a certain amount of the incident e-beam energy was transferred to the water molecule, leading to the formation of a few oxidative and reductive species. For simplicity, the mass of the water was assumed to be constant throughout the radiolysis process. The basic parameters used for the calculation, such as the beam radius and current, were summarized in Supplementary Tables 3, 4. The estimation procedure was detailed in Supplementary Note 2. The concentrations of common radiolysis species including the oxidative •OH and reductive solvated electrons in liquid cells were listed in Supplementary Table 5.