Adhesion is the attachment of two bodies or surfaces through interactions acting across their interface. It includes the attraction of liquids to solids, the bonding of an adhesive to a substrate, the attachment of particles to surfaces, and the regulated binding of living cells. Adhesion is conventionally distinguished from cohesion, which holds a material together internally. The distinction concerns the location of the interaction rather than necessarily its chemical nature: the same intermolecular forces can contribute to both.
Adhesion connects molecular-scale interactions with large-scale mechanical behavior. Forces acting over nanometres can support a climbing animal, retain a coating on a bridge, or allow cells to assemble into tissues. Nevertheless, strong molecular attraction does not automatically produce a strong joint. Surface contamination, incomplete contact, deformation, internal defects, and environmental exposure can determine whether attachment forms and how it fails.
The subject therefore spans surface chemistry, thermodynamics, contact mechanics, fracture mechanics, materials engineering, and cell biology. Its central questions concern how interfaces form, what holds them together, how their strength is measured, and how attachment changes under load or with time.
Molecular origins of attachment
At sufficiently short separations, almost all ordinary materials attract one another. An important contribution comes from van der Waals forces, including dispersion interactions arising from correlated fluctuations in electronic charge. These operate even between electrically neutral, nonpolar substances. Individually weak interactions can produce substantial adhesion when many atoms approach closely over a large area. At still shorter distances, repulsion associated with overlapping electron distributions prevents unlimited approach and establishes an equilibrium separation.
Other interactions depend more strongly on surface chemistry. A hydrogen bond can form between suitable donor and acceptor groups, contributing to the attachment of water, polymers, and biological molecules to polar surfaces. Acid–base interactions also occur between electron-donating and electron-accepting groups. Under appropriate conditions, an interface develops a covalent bond or other strong chemical linkage. Silane coupling agents, for example, can connect hydroxylated inorganic surfaces with compatible organic resins. Such bonding requires suitable reactive groups and processing conditions; it is not a universal feature of adhesive joints.
Electrostatic interactions provide another route to attachment. Oppositely charged surfaces attract, and charge redistribution can accompany contact between dissimilar materials. In liquids containing dissolved ions, however, mobile charges screen electrostatic fields. The range and magnitude of attraction or repulsion then depend on ion concentration, surface charge, and the surrounding medium. For colloidal particles, the balance between electrostatic repulsion and dispersion attraction can determine whether particles remain separated or aggregate. A surface that adheres strongly in dry air may behave quite differently in an electrolyte solution.
Attachment can also involve the interpenetration of molecular chains. When compatible polymers contact while their chains have sufficient mobility, diffusion across the original boundary can produce entanglements. This mechanism is important in polymer welding and some solvent-assisted bonds. It is limited by chain mobility, compatibility, molecular mass, and contact time. Mechanical interlocking, in contrast, occurs when an adhesive enters pores, grooves, or other surface features and becomes geometrically constrained after solidifying. It need not involve chain interdiffusion or strong chemical bonding.
These mechanisms commonly act together rather than defining mutually exclusive classes. An epoxy bonded to an oxidized metal may benefit from polar interactions, chemical reactions, and penetration into microscopic roughness. Conversely, a strongly adsorbed contaminant can prevent the adhesive from reaching the underlying substrate. Adsorption describes accumulation at an interface; adhesion describes attachment between bodies. Adsorption may promote adhesion, as with a coupling agent, or undermine it, as with an oily boundary layer.
Interfacial energy and wetting
The thermodynamic description of adhesion compares the free energies of separated and contacting surfaces. For two materials, 1 and 2, separated initially in a surrounding medium, the reversible work of adhesion per unit area is
where and are their surface free energies relative to that medium, and is the free energy of their mutual interface. This relation is commonly called the Dupré equation, associated with nineteenth-century work by Athanase Dupré. The surrounding medium must be specified: replacing air with water changes the interfaces being compared and therefore changes the work of adhesion.
The quantity has units of joules per square metre and describes an ideal reversible process. It is not generally identical to the measured energy required to break a real bond. That measurement may include polymer deformation, plastic yielding, friction, or other irreversible processes. The thermodynamic work is closely connected to changes in Gibbs free energy under appropriate conditions, whereas engineering fracture measurements describe a particular separation process. Confusing these quantities obscures why two joints with similar surface chemistry can have very different resistance to peeling.
For a liquid on a solid, wetting provides an accessible expression of interfacial energetics. On an ideal smooth, chemically uniform surface at equilibrium, Young’s equation is
where the subscripts denote solid–vapor, solid–liquid, and liquid–vapor interfaces, and is the contact angle measured through the liquid. Combining this expression with the Dupré relation gives the Young–Dupré equation:
A smaller equilibrium contact angle therefore corresponds to a greater reversible work of adhesion for that liquid–solid pair under the stated assumptions.
The liquid–vapor term is its surface tension. For clean water near room temperature, it is about . If its ideal equilibrium contact angle on a surface is , the Young–Dupré relation gives approximately . This calculation characterizes liquid–solid adhesion, not the strength of a subsequently cured adhesive joint. A liquid may wet well yet solidify into a brittle material, while a reactive adhesive may change the substrate during curing and no longer correspond to its initial contact angle.
Real surfaces complicate this framework. Roughness, chemical patches, contamination, and molecular rearrangement can produce different advancing and receding contact angles, a phenomenon called contact-angle hysteresis. A droplet may remain pinned in a metastable configuration rather than attaining global equilibrium. Young’s relation also requires modification for sufficiently deformable substrates, where surface stresses create a ridge near the contact line. Consequently, a single static contact-angle measurement rarely determines a solid’s surface free energy uniquely.
Real contact, elasticity, and surface roughness
Adhesive interactions are short-ranged, but engineering surfaces are rarely flat at molecular scales. Two apparently smooth objects initially touch at scattered high points, or asperities. Their true area of contact can be much smaller than the area visible from their overall dimensions. Adhesion therefore depends not only on intermolecular attraction but also on whether the bodies can deform enough to bring a substantial area into intimate contact. This is a central problem of contact mechanics.
The classical Hertz theory describes the nonadhesive elastic contact of curved bodies. Adhesive theories extend that framework by including attraction between the surfaces. In 1971, Kenneth Johnson, Kevin Kendall, and A. D. Roberts introduced the JKR theory, which is especially useful for compliant bodies, relatively large contact sizes, and strong, short-ranged adhesion. It predicts that a contact can remain finite without an externally applied compressive load and can persist under a tensile load before becoming unstable.
For an ideal elastic sphere contacting a flat surface, JKR theory gives the magnitude of the pull-off force as
where is the effective radius of curvature. A different limiting theory, developed by Boris Derjaguin, Vladimir Muller, and Yuri Toporov in 1975, gives
The DMT description is more appropriate when deformation caused by adhesion is relatively small and attractive forces act substantially outside the contact. These are limiting models rather than competing universal laws. Transition models connect them, and applicability depends on a dimensionless combination of size, stiffness, adhesion energy, and interaction range.
Surface roughness can suppress adhesion because conforming to asperities stores elastic energy. If the energetic cost of deformation exceeds the gain from making contact, valleys remain uncontacted. Soft materials often adhere more readily than stiff materials because they conform under modest loads. Nevertheless, roughening does not always reduce practical bond strength. It can enlarge the accessible area, remove weak surface layers, or enable mechanical interlocking when a liquid adhesive fills the texture. The outcome depends on roughness wavelength and amplitude as well as the adhesive’s ability to penetrate and deform.
Real materials also depart from ideal elasticity. A viscoelastic solid responds differently at different loading rates: it may conform slowly during attachment yet resist rapid detachment through dissipative deformation. Such behavior helps explain why rubbery adhesives can combine easy contact formation with substantial removal forces. Adhesion also interacts with friction, since interfacial junctions must often be sheared during sliding. Normal pull-off force and sliding resistance nevertheless remain distinct observables; one cannot generally be calculated from the other without a model of the contact.
Detachment, fracture, and time-dependent strength
Separation usually begins at an edge, flaw, or local region of weak bonding rather than by simultaneous rupture across an entire interface. Fracture mechanics treats detachment as the growth of a crack and asks whether the mechanical energy released by its advance can supply the energy required to create new separated surfaces and deform the surrounding material. This approach explains why specimen geometry and defect size strongly influence measured strength even when the interfacial chemistry is unchanged.
The energy release rate is the decrease in available mechanical potential energy per unit increase in crack area. Crack growth becomes possible when reaches the relevant fracture resistance, often denoted . For an ideal reversible interface, that resistance may approach the thermodynamic work of adhesion. In practical joints it frequently exceeds it because energy is consumed in plastic deformation, viscoelastic loss, cavitation, molecular chain stretching, and other processes around the crack tip. Fracture energy can consequently depend on speed, temperature, and previous loading.
Peeling illustrates how geometry controls removal. For a sufficiently flexible, inextensible tape peeled steadily from a rigid substrate, neglecting bending and other corrections,
where is peel force, is tape width, and is the peel angle relative to the substrate. At , ; at , . Stretching of the backing and deformation within the adhesive can change the interpretation substantially. A tape that resists a direct pull can therefore be removed much more easily by peeling from an edge.
Failure location is equally important. Adhesive failure follows the boundary between the adhesive and substrate; cohesive failure occurs within one of the materials. A substrate can also fracture before its interface separates. Many apparently clean interfacial failures actually occur inside a thin damaged or contaminated surface layer. Examination of both fracture surfaces is therefore necessary to distinguish a truly weak interface from a weak boundary layer. A cohesive failure demonstrates that another region was weaker under that test, not that the interface has unlimited strength.
Time introduces further differences between initial attachment and durable bonding. Under sustained load, creep can progressively deform an adhesive and drive delayed crack growth. Repeated loading can produce fatigue, allowing cracks to advance below the resistance measured in a single monotonic test. Water uptake, thermal cycling, and chemical reactions can alter both the interface and the adhesive bulk. Durability is therefore a property of a material–interface–environment system, not a number intrinsic to the adhesive alone.
Measuring adhesion across scales
No single test measures every meaning of adhesion. A useful experiment specifies the contacting materials, surface preparation, loading mode, contact history, environment, and quantity being reported. Common outputs include maximum tensile force, apparent shear stress, peel force per unit width, fracture energy, and single-bond lifetime. These quantities have different dimensions and answer different questions. Comparing them requires attention to the physical process that each test actually probes.
Macroscopic tests include lap shear, pull-off, peel, and cleavage configurations. In a lap-shear specimen, overlapping adherends are loaded in tension, transferring load through the adhesive. Dividing failure force by overlap area gives a nominal stress, but stresses are not uniformly distributed: they concentrate near overlap ends, and bending can introduce opening forces. Peel tests are useful for flexible materials, whereas double-cantilever-beam and related specimens can provide a controlled geometry for studying crack propagation. Standards improve reproducibility but do not turn a geometry-dependent result into a universal interface constant.
Probe-tack experiments bring a defined probe into contact with a soft adhesive under controlled conditions and then withdraw it. The force–displacement curve reveals attachment, deformation, and detachment processes; integrating the curve gives a separation work for that experiment. Contact pressure, dwell time, withdrawal speed, and probe cleanliness can all alter the result. In soft adhesives, the observed response may include cavities and stretched fibrils, so the peak force alone omits much of the relevant behavior.
At smaller scales, atomic force microscopy measures cantilever deflection as a tip approaches and retracts from a surface. A colloidal particle attached to the cantilever can provide a better-defined contact geometry than a sharp tip. Conversion of a force curve into a work of adhesion requires an appropriate contact model and knowledge of probe radius and material properties. Capillary bridges in humid air, electrical charging, and cantilever calibration can strongly influence measurements. Surface force apparatus experiments offer another approach, measuring interaction forces between carefully prepared surfaces as their separation changes.
Biological experiments extend from whole-cell detachment to single-molecule force spectroscopy. Micropipette aspiration, optical trapping, and cantilever-based methods can resolve receptor-mediated attachment under controlled loads. Even here, rupture force is not simply a fixed bond strength: it depends on loading rate, bond number, molecular geometry, and rebinding. Reliable interpretation therefore requires calibration, repeated measurements, suitable controls, and an explicit treatment of measurement uncertainty. Combining mechanical measurements with microscopy and surface-sensitive spectroscopy is often necessary to identify why a particular interface behaves as it does.
Adhesives and engineered interfaces
An engineered adhesive must first establish contact and then retain an appropriate combination of strength, stiffness, flexibility, and environmental resistance. These demands can conflict. Low viscosity promotes spreading into narrow gaps, but excessive flow can drain adhesive from the joint. A highly cross-linked material may carry large loads but tolerate little deformation. A softer material may accommodate differential expansion yet deform excessively under sustained loading. Adhesive selection and joint design therefore concern the whole bonded assembly rather than interfacial attraction alone.
Structural adhesives include epoxies, polyurethanes, acrylic systems, and other reactive formulations. Epoxies commonly cure through reactions of epoxide groups with hardeners such as amines, creating a cross-linked network. Cyanoacrylate adhesives polymerize rapidly in the presence of suitable nucleophilic initiators, including those associated with surface moisture. Hot-melt adhesives instead gain strength primarily by cooling and solidification. Waterborne and solvent-borne systems form bonds as their carrier leaves, sometimes with additional chemical curing. Each route imposes different constraints on gap size, processing time, heat exposure, and substrate compatibility.
Pressure-sensitive adhesives are designed to attach under modest pressure without requiring a curing reaction during use. Their polymers must be compliant enough to establish real contact while sufficiently cohesive to resist flow and leave little residue. Acrylics, rubber-based formulations, and silicones serve different combinations of tack, aging resistance, and temperature performance. Their removal energy often derives largely from bulk viscoelastic dissipation, not simply from unusually strong molecular attraction. The response can change markedly with peel speed and temperature.
Surface preparation addresses obstacles to intimate and durable contact. Cleaning removes oils and particulates; abrasion can remove weak layers and modify texture; plasma, flame, or chemical treatment can introduce polar groups. Polyethylene and polypropylene are difficult to bond in their untreated states because of their low surface energy and limited chemical reactivity. Fluorinated surfaces present additional challenges. On metals, the interface generally involves an oxide or other surface film rather than bare bulk metal, making the condition and stability of that layer especially important.
Applications range from packaging and bookbinding to vehicle structures, electronic assemblies, coatings, and fiber-reinforced composites. Bonded joints distribute loads over an area and can join dissimilar materials, but differences in thermal expansion and stiffness create residual stresses. Moisture can plasticize a polymer or weaken an interface; curing shrinkage can introduce stress before service begins. For composite materials, the fiber–matrix interface must transfer load while supporting the intended fracture behavior. Maximum possible adhesion is not always optimal: controlled interfacial debonding can sometimes dissipate energy and prevent a brittle crack from crossing the entire structure.
Cell adhesion and tissue organization
Cell adhesion is the regulated attachment of a cell to another cell or to its surroundings. It shares the physical requirements of close contact and favorable interactions with nonliving adhesion, but living systems add molecular recognition, active force generation, and continuous renewal. Adhesion molecules are not merely passive glue. Their binding can organize the cytoskeleton, alter intracellular signaling, and influence migration, survival, proliferation, and tissue development. Attachment strength depends on the number, arrangement, activation state, and mechanical loading of many molecular bonds.
Cadherins are major mediators of cell–cell adhesion, with many classical cadherins requiring extracellular calcium for their adhesive function. Through associated proteins, they connect adhesion complexes to the actin cytoskeleton in adherens junctions. Desmosomal cadherins participate in junctions linked to intermediate filaments, contributing to the mechanical integrity of tissues such as skin and heart muscle. These are chemically specific interactions, yet they need not occur between different materials: neighboring cells can bind through the same cadherin type. Thus, biological usage does not restrict adhesion to unlike chemical substances.
Integrins are heterodimeric transmembrane receptors central to attachment to the extracellular matrix. They bind ligands including fibronectin, laminins, and collagens, depending on the integrin combination. On the cytoplasmic side, proteins such as talin and vinculin help connect integrins to actin and organize focal adhesions. Integrins can transmit signals in both directions: intracellular events alter their ligand-binding properties, while ligand binding and force transmission influence intracellular pathways. The same interface can therefore be mechanically reinforced or disassembled as cellular needs change.
Adhesion is especially dynamic during immune-cell movement. In the classical leukocyte recruitment cascade, selectin-mediated interactions support initial tethering and rolling along activated vascular endothelium. Local signals promote integrin activation, followed by firmer attachment and passage across the vessel wall. Leukocytes must withstand forces from flowing blood while retaining the ability to release contacts and migrate. Platelets similarly use specialized receptors to attach to damaged vascular surfaces and to one another during the formation of a hemostatic plug. These processes involve coordinated receptor binding rather than a single universal adhesive substance.
Mechanical force can alter receptor–ligand lifetimes in different ways. Many bonds behave as slip bonds, whose lifetimes decrease as tensile load rises. Some receptor systems exhibit catch-bond behavior over a particular force range, in which moderate force prolongs binding. At larger loads, their behavior can again become destabilizing. At the cellular scale, clustering, rebinding, and cytoskeletal reinforcement further reshape the response. Consequently, the force required to detach an entire cell cannot be obtained simply by multiplying a fixed single-bond rupture force by the number of receptors on its surface.
Biological attachment beyond cell junctions
Organisms have evolved attachment systems suited to dry surfaces, wet rocks, moving hosts, and other demanding environments. These systems differ in whether they rely primarily on secreted adhesives, microscopic contact structures, suction, or mechanical hooks. Their performance often depends on architecture as much as chemistry. A material that performs well as a thin secretion or at the tip of a flexible fiber may behave poorly as a bulk glue. Natural attachment also frequently includes a mechanism for controlled release.
Gecko feet provide a well-studied dry adhesive system. Their toe pads bear arrays of microscopic setae that branch into still smaller terminal structures, often called spatulae. This hierarchical organization allows many small contacts to conform to a surface. Experiments reported by Kellar Autumn and colleagues in 2000 demonstrated strong adhesion from individual setae, helping establish the mechanical basis of the system. Dispersion interactions make an important contribution, while humidity and surface chemistry can modify performance under particular conditions. Gecko attachment does not require a continuously secreted liquid glue.
The direction of loading is crucial to gecko adhesion. Appropriate engagement and shear promote attachment, whereas changes in the orientation of the fibers and peeling of the toes permit rapid release. Splitting a contact into many compliant elements can improve conformity and control local detachment. Similar principles motivate synthetic fibrillar adhesives, although natural toe pads combine geometry, material properties, and active movement in ways that are difficult to reproduce. The biological system is therefore not explained by surface chemistry alone.
Mussels use byssal threads ending in adhesive plaques to attach underwater. Proteins in these structures contain chemically functional residues, including the catechol-bearing amino acid derivative DOPA. Catechol groups can interact with mineral surfaces through several mechanisms, including coordination and hydrogen bonding, while oxidation and cross-linking help regulate the cohesion of the deposited material. Underwater attachment is particularly demanding because water already occupies the substrate and can compete for bonding sites. The secretion’s composition and local chemical environment help govern how contact is established before the structure matures.
Microbial attachment operates on another scale. Bacteria can approach a surface through transport and motility, make initial reversible contacts, and subsequently develop stronger attachment through adhesins and extracellular material. A mature biofilm contains cells embedded in a matrix that can include polysaccharides, proteins, and extracellular DNA. Its resistance to removal depends on both substrate attachment and cohesion within that matrix. Flow, nutrient conditions, surface chemistry, and microbial identity influence development. Attachment to surfaces is therefore an ecological process as well as a molecular one, affecting colonization in streams, industrial equipment, and biological environments.
Capillary adhesion, particles, and environmental interfaces
Liquids can join solid bodies through capillary action. When a liquid bridge forms between two surfaces, its curved interface creates a pressure difference described by the Young–Laplace equation:
where and are signed principal radii of curvature. The resulting force combines pressure acting over the wetted area with surface-tension forces at the contact line. Depending on bridge geometry and wetting conditions, this can pull the bodies together. A water film between glass plates is a familiar example.
Capillary bridges can arise even when no liquid has deliberately been added. In humid air, water vapor may condense in sufficiently narrow gaps, particularly between hydrophilic surfaces. The Kelvin relation connects equilibrium vapor pressure with interface curvature and explains why condensation can occur below bulk saturation. This is important in force measurements, powder handling, and microscale devices. The resulting adhesion can change strongly with relative humidity and contact time, although real surfaces often show complex behavior because roughness and adsorbed contaminants control where bridges form.
Particle size determines the relative importance of adhesive forces. For many idealized sphere–surface contacts, adhesive pull-off force scales approximately with particle radius, whereas particle weight scales with the cube of radius. Adhesion therefore becomes increasingly important relative to gravity as particles become smaller. Fine powders may clump, adhere to containers, and resist free flow even when larger grains of the same material pour readily. Electrostatic charging and moisture can amplify these effects, while surface treatments and particle shape alter the contact geometry.
In soils and sediments, capillary bridges contribute to the apparent cohesion of partially saturated granular material. Damp sand can sustain structures that dry loose sand cannot because menisci transmit attractive forces between grains. Fully saturating the pore space removes those air–water menisci, so the same capillary contribution does not persist unchanged. Grain-scale surface forces also influence fine clays and colloids, where ionic conditions and mineral surface chemistry become particularly important. These mechanisms must be distinguished from cementation by precipitated minerals, which creates more permanent solid connections.
At very small device dimensions, adhesion can exceed the elastic restoring force available to separate components. This produces stiction in microelectromechanical systems, including during drying after wet processing or during repeated contact in service. Related problems occur in particle deposition, dust retention, and fouling of surfaces. Predicting attachment in these settings requires coupling surface interactions with fluid flow, deformation, and environmental history. Roughness measured over several length scales, humidity-dependent chemistry, and contact aging remain major challenges because each can change the effective interface without changing the nominal bulk materials.