TU Darmstadt / ULB / TUprints

Understanding Humidity‐Enhanced Adhesion of Geckos: Deep Neural Network‐Assisted Multi‐Scale Molecular Modeling

Materzok, Tobias ; Eslami, Hossein ; Gorb, Stanislav N. ; Müller‐Plathe, Florian (2023)
Understanding Humidity‐Enhanced Adhesion of Geckos: Deep Neural Network‐Assisted Multi‐Scale Molecular Modeling.
In: Small : nano micro, 2023, 19 (22)
doi: 10.26083/tuprints-00024319
Article, Secondary publication, Publisher's Version

[img] Text
SMLL_SMLL202206085.pdf
Copyright Information: CC BY 4.0 International - Creative Commons, Attribution.

Download (1MB)
[img] Text (Supplement)
smll202206085-sup-0001-suppmat.pdf
Copyright Information: CC BY 4.0 International - Creative Commons, Attribution.

Download (12MB)
Item Type: Article
Type of entry: Secondary publication
Title: Understanding Humidity‐Enhanced Adhesion of Geckos: Deep Neural Network‐Assisted Multi‐Scale Molecular Modeling
Language: English
Date: 18 July 2023
Place of Publication: Darmstadt
Year of primary publication: 2023
Publisher: Wiley-VCH
Journal or Publication Title: Small : nano micro
Volume of the journal: 19
Issue Number: 22
Collation: 9 Seiten
DOI: 10.26083/tuprints-00024319
Corresponding Links:
Origin: Secondary publication DeepGreen
Abstract:

A higher relative humidity leads to an increased sticking power of gecko feet to surfaces. The molecular mechanism responsible for this increase, however, is not clear. Capillary forces, water mediating keratin‐surface contacts and water‐induced softening of the keratin are proposed as candidates. In previous work, strong evidence for water mediation is found but indirect effects via increased flexibility are not completely ruled out. This article studies the latter hypothesis by a bottom‐up coarse‐grained mesoscale model of an entire gecko spatula designed without explicit water particles, so that capillary action and water‐mediation are excluded. The elasticity of this model is adjusted with a deep neural network to atomistic elastic constants, including water at different concentrations. Our results show clearly that on nanoscopic flat surfaces, the softening of keratin by water uptake cannot nearly account for the experimentally observed increase in gecko sticking power. Here, the dominant mechanism is the mediation of keratin‐surface contacts by intervening water molecules. This mechanism remains important on nanostructured surfaces. Here, however, a water‐induced increase of the keratin flexibility may enable the spatula to follow surface features smaller than itself and thereby increase the number of contacts with the surface. This leads to an appreciable but not dominant contribution to the humidity‐increased adhesion. Recently, by atomistic grand‐canonical molecular dynamics simulation, the room‐temperature isotherm is obtained for the sorption of water into gecko keratin, to the authors’ knowledge, the first such relation for any beta‐keratin. In this work, it relates the equilibrium water content of the keratin to the ambient relative humidity.

Uncontrolled Keywords: deep neural networks, gecko adhesion, humidity, molecular dynamics, multiscale molecular model, pull‐off, spatula
Identification Number: 2206085
Status: Publisher's Version
URN: urn:nbn:de:tuda-tuprints-243198
Classification DDC: 500 Science and mathematics > 540 Chemistry
500 Science and mathematics > 590 Animals (zoology)
Divisions: 07 Department of Chemistry > Theoretische Chemie (am 07.02.2024 umbenannt in Quantenchemie)
Date Deposited: 18 Jul 2023 12:47
Last Modified: 17 Oct 2023 07:35
SWORD Depositor: Deep Green
URI: https://tuprints.ulb.tu-darmstadt.de/id/eprint/24319
PPN: 512234515
Export:
Actions (login required)
View Item View Item