Hygroscopic compounds in spider aggregate glue remove interfacial water to maintain adhesion in humid conditions

Adhesion in humid environments is fundamentally challenging because of the presence of interfacial bound water. Spiders often hunt in wet habitats and overcome this challenge using sticky aggregate glue droplets whose adhesion is resistant to interfacial failure under humid conditions. The mechanism by which spider aggregate glue avoids interfacial failure in humid environments is still unknown. Here, we investigate the mechanism of aggregate glue adhesion by using interface-sensitive spectroscopy in conjunction with infrared spectroscopy. We demonstrate that glycoproteins act as primary binding agents at the interface. As humidity increases, we observe reversible changes in the interfacial secondary structure of glycoproteins. Surprisingly, we do not observe liquid-like water at the interface, even though liquid-like water increases inside the bulk with increasing humidity. We hypothesize that the hygroscopic compounds in aggregate glue sequester interfacial water. Using hygroscopic compounds to sequester interfacial water provides a novel design principle for developing water-resistant synthetic adhesives.


Supplementary Note 2 Raman Spectrum of Pristine and Washed Aggregate Glue
The sapphire prism with pristine aggregate glue is placed in contact with water for ∼1 h to remove the water-soluble low molecular mass compounds (LMMCs). The washing process is expected to irreversibly remove the LMMCs and leave washed aggregate glue containing only the glycoproteins as previously shown using NMR. 1 The efficacy of the washing process is tested by collecting Raman spectra of aggregate glue before (pristine) and after washing (washed) using a LabRam HR Micro Raman Spectrometer (Horiba) coupled to a Olympus BX41 motorized stage microscope ( Supplementary Fig. 2). 2 The washed aggregate glue Raman spectrum (black curve) does not show the LMMC-specific peak (indicated by blue dashed box), which is present in the pristine aggregate glue Raman spectrum (red curve), confirming our claim that only glycoproteins are left behind after washing.

Supplementary Figure 2:
Raman spectra of pristine (red curve) and washed (black curve) aggregate glue. The spectra have been vertically offset for clarity. The absence of LMMCspecific peak (indicated by blue dashed box) in the washed aggregate glue spectrum indicates that washed aggregate glue consists of only glycoproteins.  Fig. 3). The reduction in SFG intensity at 90% RH could be attributed to a change in the refractive index (RI) of pristine aggregate glue at 90% RH (Supplementary Table 2). All the SFG spectra include dominant peaks at ∼3560 cm −1 and ∼3660 cm −1 , attributed to the loosely coordinated water (or less hydrogen bonded water), [3][4][5][6] and the shifted sapphire O-H due to strong interactions between pristine aggregate glue and sapphire substrate. 7,8 The interaction energy between pristine aggregate glue and sapphire is calculated as a function of RH using the Badger-Bauer equation (equation 1) and tabulated in Supplementary Table 1. 7 A minimal change in the interaction energy with humidity suggests that the adhesive bonds between the pristine aggregate glue and sapphire stay intact at high humidity. In addition, no liquid-like water peak is observed at 90% RH even though pristine aggregate glue shows 100% water uptake inside the bulk. 9

Supplementary
Supplementary Figure 3: SFG spectra collected in PPP polarization for pristine aggregate glue/sapphire interface at 10% (red empty circles), 50% (green empty squares) and 90% (blue empty triangles) RH. Solid lines are fits obtained using a Lorentzian equation. The raw SFG spectra are presented in (A). The SFG spectrum collected at 90% RH has been multiplied by a factor of 3 for comparison purposes and the SFG spectra at different humidities have been vertically offset for clarity in (B).

Badger-Bauer Equation
In equation 1, m and C represent the slope and intercept of enthalpy, ΔH (kcal/mol) vs. frequency shift, Δν (cm -1 ) plot. We used literature reported values of m and C for calculating the interaction energies. 7

Supplementary Note 4 SFG Spectra for LMMC-extract
The LMMC-extract obtained by washing the pristine aggregate glue with water is used for dropcasting a LMMC film on the sapphire prism. The sapphire prism with LMMC-extract film is vacuumed for 15 min before collecting SFG spectra. SFG spectra are collected for the LMMCextract film at 42° (LMMC-extract/air, red curve) and 10° (LMMC-extract/sapphire, black curve) incident angles, as shown in Supplementary Fig. 5. The SFG spectra collected at two different incident angles look identical, except for overall changes in SFG intensity, indicating the LMMC-film on the sapphire prism is too thin to differentiate between the two interfaces. The LMMC-extract/sapphire SFG spectra in both PPP and SSP polarizations consist of hydrocarbon peaks in the C-H stretching vibration region along with shifted sapphire-OH peak at ~3600 cm -1 . 7,8 Unlike the pristine and washed aggregate glue SFG spectra ( Supplementary Fig. 1), the LMMC-extract/sapphire SFG spectrum in SSP polarization does not show a N-H peak, again suggesting that the N-H peak observed for both pristine and washed aggregate glue could be assigned to the glycoproteins.

Supplementary Note 5 SFG Spectra for Polyacrylic Acid
Supplementary Figure 6. SFG spectra collected in PPP polarization for polyacrylic acid/sapphire at 10% RH (red empty circles) and 90% RH (blue empty squares) using D 2 O vapors.
A polyacrylic acid (PAA) film spin-coated on sapphire prism is used as a control and subjected to varying humidity conditions using D 2 O vapors to determine the water structure at PAA/sapphire interface ( Supplementary Fig. 6). The SFG spectra collected at 10% RH predominantly shows the presence of low coordination water, whereas at 90% RH additional liquid-like water peak at ~2400 cm -1 is observed. This is in stark contrast to pristine aggregate glue/sapphire, where no liquid-like water peak is observed at 90% RH, although pristine aggregate glue shows a 100% water uptake in the bulk. 9 Supplementary Note 6 ATR-IR Spectra of Pristine Aggregate Glue Supplementary Figure 7. ATR-IR spectra of Larinioides cornutus pristine aggregate glue at 10% RH (red dotted curve), 50% RH (green dashed curve) and 90% RH (blue solid curve) collected using D 2 O vapors during the cyclic measurements (where the humidity is cycled from 10% to 50% to 90% RH and then returned to 10% RH and so on).
Supplementary Fig. 7 shows the ATR-IR spectra of pristine aggregate glue at 10%, 50% and 90% RH collected using D 2 O vapors. The ATR-IR spectra at different relative humidities show minor changes in the amide I/II, C-H stretch and N-H/O-H stretch regions, except for changes in the D 2 O (2200-2800 cm -1 ) region which are shown in Fig. 5A in the main text. The minimal change in the amide I/II region suggests that no significant changes occur in the protein conformation in the bulk.
We also cycled the humidity from 10% to 50% to 90% RH and then returned to 10% RH to see if any changes in ATR-IR spectra occur with different exposure time to D 2 O vapors. As shown in  Fig. 7, the spectra look very similar during the different cycles 1, 2 and 3 suggesting that the presented spectra are collected after allowing complete H/D exchange.

Supplementary Note 7 SFG Spectra for Pristine and Washed Aggregate Glue in PSP Polarization
SFG spectra are collected for pristine and washed aggregate glue at 10% and 90% RH in PSP polarization, which is selective to the presence of chiral structures at the interface ( Supplementary Fig. 8). The SFG spectrum for both pristine aggregate glue/sapphire and washed aggregate glue/sapphire interface at 10% RH shows a N-H stretch peak at ~3250 cm -1 , assigned to the presence of α-helices or β-sheets in pristine and washed aggregate glue. The N-H peak disappears at 90% RH for pristine aggregate glue/sapphire, whereas the N-H peak stays intact for washed aggregate glue/sapphire interface. The disappearance of N-H peak in PSP is reversible (black curve) as the humidity is reduced again to 10% RH.
Supplementary Figure 8. SFG spectra collected for pristine aggregate glue/sapphire interface at 10% RH (red empty triangles), 90% RH (blue solid triangles) in PSP polarization and again at 10% RH (black empty triangles) obtained using D 2 O vapors (A). Similar SFG spectra collected for washed aggregate glue/sapphire interface at 10% RH (black empty triangles) and 90% RH (blue solid triangles) in PSP polarization (B).

SFG Spectra for Pristine Aggregate Glue in SSP Polarization
We cycled the humidity from 10% RH to 90% RH and returned to 10% RH to see if any changes occur in the SFG spectra with different exposure time to D 2 O vapors. Supplementary Fig. 9 shows the SFG spectra collected for pristine aggregate glue/sapphire interface at 10% RH, 90% RH and again at 10% RH in SSP polarization. The SFG spectra at 10% RH collected before and after cycling to 90% RH show similar peaks (the N-H stretch peak at ~3270 cm -1 along with C-H stretch peaks) suggesting that the spectra shown in the present study are collected after allowing complete H/D exchange. Figure 9. SFG spectra for pristine aggregate glue collected at 10% (red empty squares), 90% (blue filled squares) and again at 10% RH (black filled squares) in SSP polarization. The SFG spectrum collected at 10% RH after cycling to 90% RH has been vertically offset for clarity.

Supplementary Note 8 Linear Reflectivity Measurements
Helium-neon reflectivity measurements are used to estimate the RI of pristine and washed aggregate glue, where the reflected intensity is measured as a function of incident angle. Supplementary Fig. 10 shows the normalized intensity plotted as a function of incident angle for pristine aggregate glue at 10% (red empty circles), 50% (black empty triangles) and 90% (blue empty squares) RH. The incident angle at which the intensity changes dramatically corresponds to the critical angle for total internal reflection. The critical angle increases with increasing RH due to a consistent increase in water content inside pristine aggregate glue with increasing RH. 9,15 The RI are calculated for pristine aggregate glue at varying RH using Snell's law with the help of a two-layer model (Supplementary Table 2 Figure 10. Normalized intensity plotted as a function of incident angle for pristine aggregate glue at 10% (red empty circles), 50% (black empty triangles) and 90% (blue empty squares) RH. The solid lines represent the fit obtained using a two-layer model.

Supplementary Note 9 Equilibration Time
Supplementary Figure 11. SFG intensity at 3650 cm -1 in PPP polarization plotted as a function of time as the humidity is increased from 10% to 90% RH (A) followed by decreasing the humidity from 90% back to 10% RH (B). The red arrow indicates the time instant at which the humidity is switched from high to low RH.
The SFG intensity at 3650 cm -1 (PPP, Supplementary Fig. 3) is monitored as a function of humidity to measure the time required to humidify (or dehumidify) the pristine aggregate glue on the sapphire prism. The SFG intensity at 3650 cm -1 attenuates quickly as the humidity is increased from 10% to 90% RH ( Supplementary Fig. 11A), indicating a quick water uptake by pristine aggregate glue due to the presence of hygroscopic LMMCs. However, as the humidity is subsequently reduced back to 10% RH, the SFG intensity at 3650 cm -1 gradually recovers over a period of 30-35 min ( Supplementary Fig. 11B). Thus, we use an equilibration time of 30 min before collecting SFG spectra at a particular RH.

Supplementary Note 10 Comparison of Buried vs. Air Interface SFG Spectra
In addition to the SFG spectra for aggregate glue/sapphire interface, we also collected SFG spectra for aggregate glue/air interface at different relative humidities in PPP and SSP polarizations using 42° incident angle (w.r.t. sapphire surface normal). Supplementary Fig. 12 shows a comparison of buried and air interface SFG spectra for pristine aggregate glue at 10% RH obtained using D 2 O vapors in SSP polarization. Clearly, the N-H peak observed at the buried aggregate glue/sapphire interface is absent at the aggregate glue/air interface, suggesting the interfacial structure at the buried interface is different from the air interface.
Supplementary Figure 12. Comparison of SFG spectra for pristine aggregate glue collected at 10% RH (obtained using D 2 O vapors) in SSP polarization for buried (red empty circles) and air (blue solid squares) interface.