Morphological transformation from fibers to sheets in embiopteran silk

Embioptera (webspinners) are insects that construct domiciles using silk produced from their front feet. This silk is the finest known with measured single fiber diameters in the 30–140 nm range. In the wild, some webspinner silk on trees is observed to have a clothlike or shiny sheetlike appearance. Both forms of silk shield the occupants from rain water effectively: presumably valuable in tropical environments. In this article we elucidate the mechanism by which silk fibers are transformed into these structures through interaction with water. We quantify the evaporation rates of single water droplets which have been suspended on unmodified as-spun silk for two Trinidadian arboreal species: Antipaluria urichi (Clothodidae) and Pararhagadochir trinitatis (Scelembiidae). These rates are compared to those of droplets suspended on rose petals due to similar wetting properties (both hydrophobicity and pinning). We observe that on sufficiently thick silk, droplet evaporation rates decrease with time. This behavior is a result of a thin soluble film developing on the drop surface that later becomes a solid residual film. Experimentally verified theoretical models are invoked to support the results.

Figure 1

Comparison images for the two species studied: (a) Antipaluria urichi (typical body length, 1.6 cm), (b) previously wetted A. urichi silk as it appears on a tree, (c) SEM image of previously wetted A. urichi silk, (d) Pararhagadochir trinitatis (typical body length, 0.98 cm), (e) previously wetted P. trinitatis silk as it appears on a tree, and (f) SEM image of previously wetted P. trinitatis silk; inset from the same sample at 5× relative magnification with an underlying fiber highlighted in a black rectangle. Note the large front tarsi (feet) that contain the silk glands. The white scale bar is 20 μm for both SEM images.

Figure 2

Graphite substrate for Pararhagadochir trinitatis. Grooves shown are 3.2 mm wide and 4.8 mm deep. Circular films from dried water droplets are apparent near the center of the image. In order to encourage the silk placement to be higher, later substrates used a shallower (2.4 mm deep) groove with the same width.

Figure 3

Asymmetrical 2 μl water drop suspended on Antipaluria urichi silk showing (a) wrinkled surface film during drying and (b) final dried film.

Figure 4

2 μl water drops on silk of Pararhagadochir trinitatis and graphite: (a) $t=0$, (b) frame when drop on graphite dries: 28.1 min, and (c) frame when drop on silk dries: 59.8 min.

Figure 5

Graph of total drying time for drops of de-ionized water of various nominal volumes on Pararhagadochir trinitatis silk, Antipaluria urichi silk, and graphite. The crossed symbols refer to later samples which were also measured with a side-view camera (see Fig. 8).

Figure 6

Side-view images of 2 μl water drops on (a) silk spun by Antipaluria urichi, (b) graphite, and (c) a rose petal. The scales are roughly 7 mm across for each image.

Figure 7

(a) Top view (Zeta-20) and (b) auxiliary side view (Celestron) of a 9.2 μl water drop suspended on Antipaluria urichi silk. The diameter scale bars are 2.76 mm across. The hexagonal array of lights at the center of the side view (from the Celestron camera) correspond to the six blue reflection and refraction points on the upper right of the top view. This alignment indicates the location of the diameter measurements.

Figure 8

Droplet volume as a function of time. (a) Results for droplets suspended on Pararhagadochir trinitatis silk (squares) and Antipaluria urichi silk (diamonds). The solid line is derived from (b) to show the average behavior of droplets on rose petals. Approximate error bars are shown. (b) Results for water on rose petals. The inset to panel b displays the same data but with the times offset so that the 4 μl times coincide. The inset supports the notion that the evaporation rate on rose petals is roughly constant regardless of volume.

Figure 9

Modeling of flattening ellipsoid evaporation rates. (a) Calculated ratio of the evaporation rate of an oblate ellipsoid of rotation to that of a sphere (adapted from Fuchs [20]). (b) Data adapted from Houghton [25] of evaporating water droplets (solid black curve). The dashed red curve is our model calculated from assuming the droplets are flattening ellipsoids with a fixed semimajor axis, and the evaporation rate is increased using the factor from (a). Although the red (dashed) line is not exactly linear, the increase in evaporation rate is consistent with our control data for droplets on rose petals.