Fig. 2: Influence of capillary walls’ material on evaporation.

a Schematic of the devices with different wall materials’ combination. A hyphen is placed in the acronym to represent spacer (which is always graphene) present in between the bottom and top walls. As illustrated, device G-G (also referred as graphite channel) has both the bottom and top graphite walls. Similarly, hBN channel (B-B) has both the bottom and top walls made from hBN crystals. Asymmetric capillaries contain bottom and top walls made from different 2D materials (e.g., G-B has bottom graphite and top hBN). Hybrid capillaries refer to those where the bottom is contoured with another 2D material (GB-B has bottom graphite contoured with monolayer hBN and top hBN). b Water flow through graphite (grey triangles), hBN (green squares), and asymmetric G-B devices (cyan circles) with different channel heights. Results of hBN channels are compared with those for graphite channels reported in ref. ²². The weight loss is normalized for L ≈ 1 µm and per channel. The symbols represent the experimental data; the solid curves are the best fits using Eq. (1); the driving force is a combination of capillary and disjoining pressures. The fits yield the slip lengths δ ≈ 1 and δ ≈ 60 nm, for hBN and graphite, respectively. Dotted line is a guide to the eye, showing the deviation from the Eq. (1) for the sub-2 nm water flow rates. The anomalous peak at N ∼ 4, which deviates from the fitting, observed for graphite capillaries, is due to increased structural ordering of water. c Water flow rates for capillaries made of graphite (gray bar), roughened graphite G*-G (light grey bar), hBN (green bar), MoS₂ (pink bar), and hybrid channels GB-B (cyan bar); spacers heights N = 3 and 9. Error bars in b and c indicate the data scatter for two devices measured for that particular channel height.