A tree in a gale is bent far beyond the point at which a steel beam of similar slenderness would be permanently damaged. When the wind drops it stands up again, apparently undiminished. The explanation lies not in the shape of the trunk but in a molecular mechanism inside the cell wall: one that researchers named, aptly, molecular Velcro.
The problem a tree has to solve
A tree cannot move away from load. It must accommodate wind and snow where it stands, repeatedly, for decades or centuries, without the option of repair between events. It must also be different things in different places at once: flexible in twigs and in the stem of a sapling, rigid in the outer part of a mature stem. It builds a single material (wood) and varies its properties enormously to meet those conflicting demands.
Engineering materials do not usually manage this. Beyond the yield point, most materials that deform permanently also become weaker: the deformation is a record of damage. Wood does something stranger.
The observation that started it
In 2003, a group led by Jozef Keckes and Peter Fratzl combined tensile tests on individual wood cells and on thin wood foils with simultaneous synchrotron X-ray diffraction, which allowed them to separate what happens inside a cell wall from what happens between cells.
The result was counter-intuitive. Tensile deformation beyond the yield point did not deteriorate the stiffness of either individual cells or foils. The material had been permanently deformed, yet it had not been weakened. Something inside the wall was re-forming the amorphous matrix between the cellulose microfibrils and restoring its mechanical properties.
Why the name
The authors described a stick–slip mechanism, “rather like Velcro operating at the nanometre level”, which provides a plastic response similar to that produced by moving dislocations in metals. The analogy was needed because wood has no crystalline structure of the kind that would allow dislocations. It achieves a comparable behaviour by entirely different means: bonds that let go and take hold again.
The structure that makes it possible
Wood is a biocomposite of cellulose microfibrils, several micrometres or more in length, embedded in a matrix of hemicelluloses and lignin. The microfibrils are not aligned with the cell axis but wound around it at the microfibril angle, and the longitudinal tensile properties of wood depend strongly on that angle. A small angle gives stiffness; a large angle gives extensibility. This single geometric variable is much of how a tree tunes twig against trunk.
Because the fibrils are helical, pulling on the cell wall does not simply stretch them. It shears the matrix between them. The question the molecular Velcro model answers is what that matrix does under shear.
The mechanism
In the original formulation, hemicellulose chains attached to adjacent cellulose microfibrils interact with one another through non-covalent bonds. These interactions hold until a threshold level of interfibrillar shear is reached, at which point they are disrupted, the fibrils slip past one another, and then re-form.
The bonds are weak and numerous, which is precisely the point. Breaking one costs little energy and endangers nothing. Breaking many, sequentially, absorbs a great deal of energy. And because each broken bond can find a new partner a short distance away, the wall arrives at a new configuration that is as sound as the old one.
A refinement, and why it matters
Models in biomechanics are arguments, not photographs, and this one was revised. In 2008, Clemens Altaner and Michael Jarvis pointed out that other evidence did not confirm the importance of hemicellulose–hemicellulose association in holding the interfibrillar matrix together. They proposed an alternative in which hemicellulose chains bridge continuously from one microfibril aggregate (a macrofibril) to the next, and provide most of the cohesion that way.
Their description of the resulting motion is worth dwelling on. When one macrofibril slides past another, a domain of the cell wall extends but simultaneously twists, until the spacing between macrofibrils is reduced again and contact through hemicellulose bridges is restored. Deformation therefore proceeds as a series of local stick–slip events involving temporary twisting of small domains within the wall.
The modelled load–deformation curves are similar to those of the original model, though not identical. The mechanism, however, is different, and more consistent with current understanding of cell wall structure. Both models agree on the essential behaviour; they disagree about which molecular contact does the work.
Where lignin comes in
Neither version works without something to stop the macrofibrils simply separating. Altaner and Jarvis note that other polymers, lignin in particular, must restrain that separation. This is the same lignin that gives the secondary cell wall its compressive rigidity and hydrophobicity. It is not an inert filler around the cellulose; it is part of the mechanism that lets the wall yield without failing.
Water is part of the design
These models describe water-saturated wood, which is the condition of wood in a living tree: free water is present and the cell walls hold roughly their saturation value of moisture. Plastic deformation is much easier at high moisture content, and the recovery mechanism depends on it. Weak bonds can only detach and re-form freely in a hydrated, mobile matrix.
This has a practical corollary that anyone working with fibre knows in their hands: dry wood and wet wood are mechanically different materials. The same is true of paper. Much of what we call the behaviour of a fibre-based material is really the behaviour of a hydrogen-bonded network at a particular moisture content.
What this means for the tree
For the tree, the consequence is survival across repeated loading. A gust that carries the stem past its elastic limit does not spend that event weakening the material for the next one. The wall deforms, dissipates energy, and re-establishes cohesion. Damage is not cumulative in the way it would be in a material whose yielding meant fracture at the molecular scale.
It also explains how one material can serve contradictory functions. By varying microfibril angle, density and matrix composition, the tree produces flexible twigs, springy saplings and rigid mature stems from the same molecular toolkit. The stick–slip mechanism, described as a universal phenomenon dominating tensile deformation across different wood tissue types, underlies all of them.
How nature arrives at this
It is tempting to describe this as elegant design. It is more accurate, and more interesting, to describe it as an accumulation of constraints. The tree cannot select exotic elements; it builds from carbon, hydrogen and oxygen fixed from air and water. It cannot anneal, forge or heat-treat; assembly happens at ambient temperature and pressure in water. It cannot replace a failed component; it can only add new material outside the old.
Under those constraints, weak reversible bonds are not a compromise. They are the only route to a tough material. Strong covalent bonds everywhere would give a stiff, brittle wall. Weak bonds, numerous and re-formable, give toughness (the ability to absorb energy without breaking) and they give it repeatedly. The same principle appears in bone and in nacre, which is why those materials were the comparison when this work began.
What we can take from it
Toughness is not strength. A material that yields and recovers may outlast a stronger one that does not. In packaging, the relevant property is often energy absorption under a real distribution of loads, not peak stress in a test.
Sacrificial bonds are a design strategy. Building in interactions that are meant to break, and to re-form, is now an established approach in self-healing polymers and tough hydrogels. Wood has been doing it since vascular plants colonised land.
Hierarchy does the work. The performance is not in the chemistry alone but in how structure is organised across scales: molecule, microfibril, macrofibril, cell wall layer, cell, tissue, organ. This is precisely why converting biomass to defined chemicals is so hard, and why fibre products behave as they do on recycling.
Moisture is a design variable, not a nuisance. In a living tree, water is a functional component of the mechanism. In fibre-based packaging, humidity dependence is often treated as a weakness to be engineered out. It is worth asking, occasionally, whether it could be used instead.
References
- Altaner, C. M. & Jarvis, M. C. (2008). Modelling polymer interactions of the ‘molecular Velcro’ type in wood under mechanical stress. Journal of Theoretical Biology, 253(3), 434–445. doi:10.1016/j.jtbi.2008.03.010
- Keckes, J., Burgert, I., Frühmann, K., Müller, M., Kölln, K., Hamilton, M., Burghammer, M., Roth, S. V., Stanzl-Tschegg, S. & Fratzl, P. (2003). Cell-wall recovery after irreversible deformation of wood. Nature Materials, 2(12), 810–813. doi:10.1038/nmat1019
- Fratzl, P., Burgert, I. & Keckes, J. (2004). Mechanical model for the deformation of the wood cell wall. Zeitschrift für Metallkunde, 95, 579–584.
- Farber, J., Lichtenegger, H. C., Reiterer, A., Stanzl-Tschegg, S. & Fratzl, P. (2001). Cellulose microfibril angles in a spruce branch and mechanical implications. Journal of Materials Science, 36, 5087–5092.
Related reading
This article reflects independent professional analysis and is provided for informational purposes. The mechanisms described are models under active discussion in the literature, and the two accounts summarised here differ in their molecular detail; readers should consult the primary papers. The diagram is a schematic reconstruction for explanatory purposes and is not to scale. Last reviewed July 2026.