Reversible Adsorption and Interfacial Photoisomerization of Azobenzene Surfactants Studied by QCM

TL;DR

This study explores how light can control the behavior of azobenzene surfactants at liquid interfaces, revealing how their structure changes under UV and blue light.

Contribution

The study provides new insights into the dynamic adsorption and photoisomerization mechanisms of azobenzene surfactants at interfaces.

Findings

Trans isomers of AzoC6 show strong adsorption at the interface, while cis isomers are less surface-active.

Photoisomerization kinetics at the interface depend on trans isomer concentration and interfacial structure.

Light-driven isomer gradients generate diffusio-osmotic flows, demonstrating remote control of interfacial processes.

Abstract

Photoresponsive surfactants
offer a versatile approach
for remotely
controlling interfacial properties through light-triggered isomerization.
Among them, azobenzene-based surfactants are particularly attractive
due to their structural reversibility and stability under repeated
irradiation. In this study, we investigate the dynamic adsorption
and desorption behavior of the azobenzene-containing surfactant AzoC6 at a glass–water interface under controlled UV and
blue-light illumination. Using quartz crystal microbalance (QCM) measurements,
we show that the interfacial mass change is governed by the isomeric
composition in the bulk solution: the trans isomer
exhibits strong adsorption, while the cis isomer
is significantly less surface-active. We further quantify the photoisomerization
kinetics at the interface, revealing that the isomerization rate constant
decreases with a lower trans isomer concentration
due to a transition from a diffuse multilayer to a confined double-layer
structure. At higher concentrations, the rapid exchange between trans and cis isomers sustains dynamic
interfacial rearrangements, facilitating the formation of spatial
isomer gradients. These gradients generate light-driven diffusio-osmotic
flows, with a time evolution that reflects the interfacial photoresponse.
Our findings provide mechanistic insight into light-induced interfacial
processes and highlight the potential of azobenzene surfactants for
designing stimuli-responsive systems and soft materials with remote,
reversible control.

Figures

(a) Chemical structure of the photosensitive surfactant. The photoisomerization between trans and cis is depicted in the figure. (b) Adsorption spectra of the azobenzene-containing surfactant in the dark (black line), after exposure to UV (λ = 365 nm, red dashed line) and blue (λ = 455 nm, blue dotted line) light. (c) Scheme of the dynamic process of both isomers at the solid–liquid interface; the indexes of k correspond to the adsorption (a) or desorption (d) constants of the trans (T) and cis (C) isomers. m is the adsorbed mass of the trans and cis molecules on the surface. θ means the surface coverage resulting from the adsorbed mass.

Results of QCM-D measurements during the adsorption/desorption of a photosensitive surfactant (c = 0.3 mM) on a borosilicate glass surface. (a) Scheme of surfactant adsorption and desorption on a surface. (b) Frequency shift Δf and (c) dissipation shift ΔD as functions of time t. Region I indicates the water baseline, Region II corresponds to the adsorption of surfactant in the trans conformation, and Region III depicts the change in signal during exposure of the sample to UV (λ = 365 nm, I = 0.2 mW·cm–2, purple line) and blue light (λ = 455 nm, I = 0.5 mW·cm–2, blue line).

Surface coverage, θ, of the borosilicate glass surface as a function of time at different irradiations plotted for surfactant concentrations ranging between 0.01 and 2 mM (see the legend in the plots). Region I corresponds to the water baseline (a, b), Region II depicts surface coverage at different concentrations (a, b), and Region III shows surface coverage under irradiation with blue light (λ = 455 nm, I = 0.5 mW·cm–2) (a) and UV light (λ = 365 nm, I = 0.2 mW·cm–2) (b) for all surfactant concentrations, in situ measurements. (c) Adsorption isotherms of AzoC6 surfactants on a hydrophilic borosilicate surface under ex situ conditions. Black and pink lines show the dependence on the dark-state solution and the solution preirradiated with a UV LED (365 nm), respectively. The gradient lines describe the theoretical adsorption for different fractions of trans isomers in the solution.

COMSOL simulation of the flow velocity of the liquid in the QCM chamber. (a) Steady-state flow velocity in the chamber at the volumetric flow rate of 100 μL·min–1. (b) Adsorbed surface coverage θ over time t for c = 0.1 mM (see also the corresponding Video S1). (c) Calculated dependencies of adsorbed surface coverage θ on time for different AzoC6 surfactant concentrations.

Surface coverage θ change as a function of irradiation wavelength over time t for different surfactant concentrations (a) 0.1 mM, (b) 0.5 mM, and (c) 1 mM AzoC6. Region I: water baseline. In Region II, two experiments are shown: the adsorption of isomers starts from the state when the majority of the molecules are in the trans conformation (black and green lines), followed by irradiation with UV and blue light (region III), respectively. The red line depicts the experiment where the initial state is a cis-enriched solution, followed by irradiation with UV light. The blue line depicts the situation where the solution is first preirradiated with blue light, and in Region III, the irradiation again with blue light is switched on. The intensities used are UV (λ = 365 nm) I = 0.2 mW·cm–2 and blue (λ = 455 nm) I = 0.5 mW·cm–2.

Shift of surface coverage θ over time t for (a) 0.3 mM and (b) 2 mM surfactant solutions during irradiation with the UV (λ = 365 nm, I = 0.2 mW·cm–2) and blue (λ = 455 nm, I = 0.5 mW·cm–2) light shown for 3 cycles.

(a) Change of surface coverage θT over time for a 0.1 mM AzoC6 surfactant solution on a borosilicate surface. The black line shows the experimental data obtained during UV light irradiation (λ = 365 nm, I = 0.2mW·cm–2). The red line is the surface coverage calculated using eq VI. The green line depicts the change in the bulk concentration, c T, of the trans molecules as calculated using eq VIII. (b) Fitting of the surface coverage by different photoisomerization rate constant k TC,I values.

(a) Change of the photoisomerization rate constants as a function of surfactant concentrations for molecules in the bulk (k TC, blue and purple dots) and on the surface (k TC,I, blue and purple squares). (b) Comparison of rate constant ratios k TC,I·(k TC)−1 for different surfactant concentrations under exposure to blue (λ = 455 nm, blue squares) and UV (λ = 365 nm, purple squares) light. (c) Molecule density β as a function of trans surfactant concentration c T in the trans state (dark state, green squares) on a borosilicate glass surface. (d) k TC,I as a function of trans surfactant concentration irradiated with blue light (λ = 455 nm, blue squares) and irradiated with UV light (λ = 365 nm, purple squares). (e) Scheme of the process.

(a) Scheme of the irradiated bulk solution from the above with UV light. The rainbow color codec illustrates the considered finite layers at a specific distance from the light source. The time evolution of the concentration of trans isomers (b) and incident light (c) at different distances from the top surface over time in the case of λ = 365 nm and I = 0.2 mW·cm–2.

Desorption under the irradiation by UV LED with λ = 365 nm: (a) dependence of the concentration of trans isomers within the bottom layer; (b) theoretical θtheory and (c) experimental θexperimental surface coverage over time t in the case of different irradiation intensities I. The bulk surfactant concentration is c azo= 0.75 mM. (d) Scheme of the photoisomerization of surfactants in the bulk and on the surface, with the latter determining the number of adsorbed molecules over time.

Full text

Introduction

The design of smart materials relies on their ability to respond dynamically to external stimuli. Among various types of stimuli such as changes in ionic strength, pH, electric fields, or temperature ?−? ? light offers particularly attractive advantages. It can be applied remotely, with high spatial and temporal precision, and introduced into closed systems in a noninvasive and contactless manner.

Over the past decades, numerous light-responsive systems have been developed that encompass a wide range of applications. These include reversible nanoparticle precipitation and redispersion, ?,? photoresponsive polymer networks and microgels, ?,? surface relief gratings, ?,? light-guided fluid transport, manipulation of liquid crystals along planar interfaces,? reversible wetting and dewetting of surfaces,? emulsion destabilization, ?−? ? photoswitchable foams, ?,? controlled mixing of fluids,? light-induced particle motion in quiescent liquids,? selective separation of microparticles based on interfacial properties, ?,? and even biomimetic blood clotting using inorganic microparticles in microfluidic channels.?

One common approach to achieving light responsiveness in materials is the irreversible incorporation of photoactive molecular moieties into the chemical structure or matrix of the material itself. ?−? ? While effective, this strategy often involves complex synthesis procedures and limited adaptability, as any change in light-responsive behavior typically requires the design and synthesis of new compounds. An alternative and more versatile strategy is invasive triggering, in which inherently photoresponsive substances are introduced into a system to interact with the material and thereby transfer their light responsiveness. Photoresponsive surfactants are especially promising in this context, as they combine amphiphilic behavior with light-switchable properties within a single molecule. Their dynamic and often reversible interactions with interfaces allow light-triggered modulation of interfacial properties and, consequently, remote control over material behavior. ?−? ? As a result, such systems have attracted growing attention for their simplicity, adaptability, and potential in the development of responsive materials. ?−? ? ?

Among the various classes of photoresponsive surfactants incorporating light-sensitive moieties such as stilbenes, spiropyrene, ?,? azopyrazole,? azobenzene-based surfactants remain the most widely used. This is largely due to the favorable balance between long-term photostability and moderate quantum yields, resulting in minimal photodegradation over repeated switching cycles. As such, azobenzene-containing surfactants have found broad application in systems in which interfacial behavior can be modulated by light. When a solution of azobenzene-based surfactants is exposed to light, dynamic exchange processes between the bulk and interfacial regions (e.g., solid–liquid or air–liquid interfaces) are triggered. These exchanges are governed by the interplay of competing trans and cis isomers, both in solution and at the interface. ?,?,? For example, we previously demonstrated that the trans isomer of the surfactant 6-[4-(4-Hexylphenylazo)-phenoxy]-butyl-trimethylammonium bromide (C_4_-Azo-OC_6_-TAB, hereafter AzoC_6_; Figurea) exhibits significantly stronger adsorption at a glass–water interface compared to the cis isomer, with an adsorbed ratio of approximately 80:1 (trans:cis).?

(a) Chemical structure of the photosensitive surfactant. The photoisomerization between trans and cis is depicted in the figure. (b) Adsorption spectra of the azobenzene-containing surfactant in the dark (black line), after exposure to UV (λ = 365 nm, red dashed line) and blue (λ = 455 nm, blue dotted line) light. (c) Scheme of the dynamic process of both isomers at the solid–liquid interface; the indexes of k correspond to the adsorption (a) or desorption (d) constants of the trans (T) and cis (C) isomers. m is the adsorbed mass of the trans and cis molecules on the surface. θ means the surface coverage resulting from the adsorbed mass.

These interfacial interactions were previously investigated by using ex situ measurements under static conditions. In these studies, the isomer ratio in solution was predefined by controlled bulk illumination, and no further light was applied to the interface during measurement. Under these conditions, we investigated the adsorption kinetics of AzoC_6_ and determined the adsorption and desorption rate constants for each isomer (Figurec).

However, the kinetics of photoisomerization and the corresponding rate constants for AzoC_6_ at the interface under in situ illumination remain unknown. In this study, we address this gap by analyzing the dynamic interfacial response under controlled UV and blue-light irradiation at varying bulk concentrations of the surfactant. From these dynamic measurements, we extracted the isomerization kinetics of AzoC_6_ at the glass–water interface.

The overall aim of this research is to gain a fundamental understanding of mass transport and interfacial dynamics at solid–liquid boundaries in aqueous solutions of photoresponsive surfactants. These findings may also serve as a model system for more complex light-driven phenomena, such as diffusioosmosis? and localized light-induced interfacial transport.?

Experimental Section

Materials

Light-Responsive

Surfactant

Azobenzene-containing trimethylammonium bromide surfactant, full name 6-[4-(4-Hexylphenylazo)-phenoxy]-butyl-trimethylammonium bromide (C_4_-Azo-OC_6_-TAB, Figurea), is synthesized as described elsewhere.? The surfactant consists of a spacer of six methylene groups between the positively charged trimethylammonium bromide headgroup and the azobenzene unit with a butyl tail attached. A stock aqueous solution (Milli-Q) of the surfactant is adjusted to a concentration of 10 mM and diluted to the required concentrations. The trans isomer has a characteristic absorption band (π–π* transition) with a maximum at 351 nm (Figureb). The spectrum of the cis isomer has two absorption bands with maxima at 313 nm (π–π* transition) and at 437 nm (n-π* transition). The lifetime of the cis isomers in aqueous solution stored in dark or under illumination with red light of λ = 600 nm is 48 h (at room temperature of 20 °C), while the photoisomerization from the cis to trans state under irradiation with blue light (λ = 455 nm) takes place within seconds (light intensity dependent), with a fraction of trans and cis isomers of 66% and 34% for 1 mM aqueous solution at the photostationary state, respectively (Figureb), we used the dark state as an idealized reference case of 100% trans isomers.? Under UV illumination (λ = 365 nm) at the photostationary state, the surfactant molecules are predominantly in the cis state, with a fraction of 12% and 88% for 1 mM surfactant solution.?

A borosilicate glass sensor (QS-QSX336, LOT Quantum Design GmbH) is used for the QCM-D measurements. Before each measurement, the sensor is cleaned according to a cleaning protocol from BiolinScientific. In short, the borosilicate glass sensors are placed for 10 min in the plasma cleaner (UV/Ozon ProCleaner, BioForce Nanoscience), followed by exposure to a bath with 2% sodium dodecyl sulfate (SDS) solution for 30 min and washing with Milli-Q water. Afterward, the sensors are dried in a nitrogen atmosphere and etched once more in a plasma cleaner for 10 min to eliminate the last possibly remaining residues.

Light Source

The samples are irradiated with a UV lamp of λ = 365 nm wavelength (M365L2 THORLABS Inc., Germany) and blue light of λ = 455 nm (M455L3C1 THORLABS Inc., Germany). For the in situ measurements, the intensities between I = 50 μW·cm^–2^ and 0.7 mW·cm^–2^ are applied to avoid the effect of light-induced detuning (LID).?

Method

Quartz crystal microbalance with dissipation monitoring (QCM-D) measurements are performed on a four-chamber Q-Sense E4 instrument (BiolinScientific). The chamber is equipped with a window module, which allows irradiation of the sensor surface with light of different wavelengths in situ, i.e., during adsorption/desorption of the photosensitive surfactant. This enables us to observe the real-time adsorption kinetics under illumination.

Adsorption is monitored by collecting frequency and dissipation shift, Δf and ΔD, from the third to the ninth overtone.? Before each measurement, the resonance frequency of the crystal is determined in Milli-Q water. Values of Δf and ΔD are recorded as long as the equilibrium of the signal is achieved.

Typical measurement is conducted as follows: first, Milli-Q water is pumped over the sensor to record a water baseline. After that, the liquid is replaced by the surfactant solution of the desired concentration, and the measurements are performed until saturation is reached. At this point, the pump is switched off, and the alternating irradiation with UV and blue light in both cases during the time needed to reach the saturation of the frequency and dissipation signal is conducted.

The flow rate is kept constant at 100 μL·min^–1^ for all measurements. When the solution is changed, the flow is stopped to prevent the formation of air bubbles in the tubing system. The temperature in the chamber is kept at 23 °C.

QCM Data

Analysis

The adsorbed mass m ads is determined by the modified Sauerbrey equation for unknown adsorbed films that considers so-called fluid effects to a first small load approximation: ?,?

$$[eqn]$$

where C is the crystal constant (C ∼ 17.7 ng·cm^–2^), n is the overtone number, and f 0 is the resonance frequency (f 0 = 4.95 MHz). Because the data exhibit rather low dissipation, the mass is calculated from Δf and ΔD values from the third to the ninth overtone after reaching a steady-state value and averaged over all overtone numbers. The surface coverage θ is calculated using the adsorbed mass m _ ads _ obtained from the QCM-D measurement and the saturation mass m sat corresponding to the complete surfactant adsorption (m sat = 391.06 ± 7.88 ng·cm^–2^)? according to the following relation:

$$[eqn]$$

Theoretical Model of the

Dynamic Exchange of Trans Isomers

Given that the surfactant exists in two isomeric forms, we consider a binary mixture of trans and cis isomers undergoing dynamic exchange at the solid–liquid interface. The overall surface coverage is therefore governed by the competitive adsorption of both isomer species. Consequentially, the total surface coverage, θ, is the sum of the surface coverages of trans (θ_T_) and cis isomers (θ_C_):

$$[eqn]$$

Previously, we have shown that the surface coverage on a borosilicate glass surface is dominated by molecules in the trans state? (see also short explanation in Supporting Information, Section 2), so the total surface coverage is simplified from θ = θ_T_ + θ_C_ to θ = θ_T_, and the Langmuir equation can be expressed as follows:

$$[eqn]$$

where $[eqn]$ is the equilibrium constant of the trans molecules, k a and k d are the rate constants of adsorption and desorption, respectively, and c T is the bulk concentration of the trans- isomers.

The rate of the surface coverage by trans isomers can be expressed as follows: ?,?

$$[eqn]$$

where k TC,I is the photoisomerization rate constant from the trans to cis state at the interface and I is the intensity of irradiation. The first term of the equation, k TC,I_θ_T I, describes the time-dependent change in the surface coverage of trans molecules θ_T_ due to the dynamic process during irradiation, k a c T(1 – θ_T_), resulting from the change of trans isomers between surface and bulk. The rate constant of desorption of trans molecules from the surface is defined as k d_θ_T.

The solution of eq V (see Section 2, Supporting Information) describes the surface coverage after irradiation for the time period of (t 2 – t 1) with coverage θ_T,1_ at t 1:

$$[eqn]$$

where the concentration, c T, of trans isomers in the bulk at any time t during irradiation is given?

$$[eqn]$$

where k TC and k CT are the photoisomerization rate constants from the trans to cis and cis to trans state in the bulk, respectively, and c _ T,0_ is the surfactant concentration in the bulk at time t = 0. Using eq VI, one gets the photoisomerization rate constant k TC,I on the surface from of the measured adsorption curves.

COMSOL Multiphysics 6.3 (COMSOL Inc., Burlington, MA) is used to simulate fluid flow, the surfactant concentration distribution, and their adsorption onto the sensor surface. The model is implemented in 3D using a fine, physics-controlled mesh consisting of 372,433 elements (for more details, see Section 3 and Figure S2, Supporting Information).

Results and Discussion

The dynamic adsorption and desorption behavior of a photosensitive surfactant at the solid–liquid interface is investigated using a quartz crystal microbalance with dissipation monitoring (QCM-D), equipped with a window module to allow in situ measurements under controlled light illumination. A typical data are displayed in Figure, which shows the frequency (Δf) and dissipation (ΔD) change of the sensor coated with a borosilicate glass during adsorption of a photosensitive surfactant (c azo= 0.3 mM) under illumination with light of two different wavelength, i.e., UV (λ = 365 nm, I = 0.2 mW·cm^–2^) and blue (λ = 455 nm, I = 0.5 mW·cm^–2^). The concentration of 0.3 mM was chosen as a representative example to demonstrate the experimental procedure; this concentration does not have a specific physical significance but was selected because it clearly illustrates the difference between UV- and blue-light irradiation. The same procedure applies to other concentrations, as well. The measurements start with pure water (region I in Figure) set as the baseline, followed by adding the surfactant predominantly in trans conformation (dark state) at the 10th minute of measurements (region II in Figure). After equilibration of the signal, the pumping of the flow is stopped, and the illumination with either UV (red line) or blue light (blue line) is switched on (region III in Figure). As can be seen from Figure (see also scheme in Figurea), the trans isomers adsorb to the glass surface, causing a decrease in the frequency and dissipation change (region II), while irradiation and consequent photoisomerization from the trans to cis state force the surfactant to desorb. The frequency increase is more pronounced during UV irradiation (from −15 Hz up to −8 Hz) (Figureb), while blue-light irradiation alters it only slightly (Figurec).

Results of QCM-D measurements during the adsorption/desorption of a photosensitive surfactant (c = 0.3 mM) on a borosilicate glass surface. (a) Scheme of surfactant adsorption and desorption on a surface. (b) Frequency shift Δf and (c) dissipation shift ΔD as functions of time t. Region I indicates the water baseline, Region II corresponds to the adsorption of surfactant in the trans conformation, and Region III depicts the change in signal during exposure of the sample to UV (λ = 365 nm, I = 0.2 mW·cm–2, purple line) and blue light (λ = 455 nm, I = 0.5 mW·cm–2, blue line).

The adsorption and desorption behavior of trans isomers at the glass surface depends on the initial concentration of the surfactant solution (Figure). At lower concentrations, fewer molecules are available for adsorption, resulting in reduced surface coverage. Starting from a surfactant concentration of 0.5 mM (Figurec, dashed line), no further increase in the adsorbed amount is observed with increasing concentration, indicating that surface saturation has been reached.? The COMSOL simulations of the adsorption of surfactants in the presence of flow accurately reproduce the experimental observations in Region II of the QCM experiment. The adsorption rate depends on the bulk concentration of the surfactants and the flow velocity of the liquid in the QCM chamber. In the current experiment, the volumetric flow rate is kept constant, V = 100 μL·min^–1^. Figurea shows the steady-state velocity distribution in the chamber, obtained using COMSOL. To calculate the adsorbed mass over time (Figureb), the experimentally obtained velocities, diffusion coefficient for the surfactant, D = 2.4 × 10^2^ μm^2^·s^–1^, and the Langmuir model for the adsorption are employed (Supporting Information for the details, Section 3). The circular region in the center marks the surface over which the adsorbed mass is averaged. As shown in Figurec, the equilibrium adsorbed mass increases with higher bulk concentration, and at higher concentrations, saturation is reached more rapidly.

Surface coverage, θ, of the borosilicate glass surface as a function of time at different irradiations plotted for surfactant concentrations ranging between 0.01 and 2 mM (see the legend in the plots). Region I corresponds to the water baseline (a, b), Region II depicts surface coverage at different concentrations (a, b), and Region III shows surface coverage under irradiation with blue light (λ = 455 nm, I = 0.5 mW·cm–2) (a) and UV light (λ = 365 nm, I = 0.2 mW·cm–2) (b) for all surfactant concentrations, in situ measurements. (c) Adsorption isotherms of AzoC6 surfactants on a hydrophilic borosilicate surface under ex situ conditions. Black and pink lines show the dependence on the dark-state solution and the solution preirradiated with a UV LED (365 nm), respectively. The gradient lines describe the theoretical adsorption for different fractions of trans isomers in the solution.

COMSOL simulation of the flow velocity of the liquid in the QCM chamber. (a) Steady-state flow velocity in the chamber at the volumetric flow rate of 100 μL·min–1. (b) Adsorbed surface coverage θ over time t for c = 0.1 mM (see also the corresponding Video S1). (c) Calculated dependencies of adsorbed surface coverage θ on time for different AzoC6 surfactant concentrations.

After the adsorption is complete, the sample is irradiated with light of different wavelengths. Upon light exposure (Region III, blue light (λ = 455 nm*)* in Figurea) and UV light (λ = 365 nm) in Figureb, the desorption of the molecules depends on concentration and wavelength. Indeed, in the case of the exposure to blue light at concentrations below the bulk CMC (c CMC = 0.5 mM), a slight decrease in surface coverage is detected (Figurea), and irradiation of saturated surfaces (above the 0.5 mM) does not cause a detectable change in surface coverage (Figure S1). The situation is different for exposure to UV light. For the investigated concentration range of 0.01–2 mM, there is a pronounced decrease in adsorbed molecules during irradiation (Figuresb and S1). At surfactant concentrations below the critical micelle concentration (CMC), the difference in surface coverage between the dark state and the UV-irradiated state increases progressively (Figureb). This difference reaches a maximum at the CMC (c = 0.5 mM) (Δθ_UV_ = 0.6). Beyond this point, the effect diminishes, and the difference in surface coverage decreases with further increasing concentration.

The desorption kinetics also exhibit a concentration-dependent behavior. Notably, at concentrations above the CMC, a delayed onset of desorption under UV irradiation is observed. In contrast to sub-CMC concentrations, where a decrease in surface coverage begins immediately upon illumination, the onset of desorption above the CMC occurs only after a certain period of exposure (see the curves starting from c = 0.75 mM in Figureb). This phenomenon is addressed and interpreted within the framework of a physical model introduced in the following section.

As previous studies indicate that adsorption above 0.005 mM is mainly due to trans isomers,? the total adsorbed mass can be assumed to depend on their fraction in solution. Figurec shows the adsorption isotherm constructed for different trans isomer fractions, based on the given assumption and using previously obtained data.? This enables us to estimate the adsorbed mass as a function of the surfactant concentration and the trans-to-cis isomer ratio over time and to interpret the newly observed dependencies presented in Figures–?. At a fixed AzoC_6_ concentration in bulk, c 0, irradiation reduces the trans isomer content, leading to measurable differences in the desorbed mass under UV and blue light at various bulk concentrations. When started with the preirradiated solution (either UV or blue light), the irradiation itself does not result in the change of the coverage as shown in Figure for three different surfactant concentrations ((a) c azo = 0.1 mM, (b) c azo = 0.5 mM, and (c) c azo = 1 mM).

Surface coverage θ change as a function of irradiation wavelength over time t for different surfactant concentrations (a) 0.1 mM, (b) 0.5 mM, and (c) 1 mM AzoC6. Region I: water baseline. In Region II, two experiments are shown: the adsorption of isomers starts from the state when the majority of the molecules are in the trans conformation (black and green lines), followed by irradiation with UV and blue light (region III), respectively. The red line depicts the experiment where the initial state is a cis-enriched solution, followed by irradiation with UV light. The blue line depicts the situation where the solution is first preirradiated with blue light, and in Region III, the irradiation again with blue light is switched on. The intensities used are UV (λ = 365 nm) I = 0.2 mW·cm–2 and blue (λ = 455 nm) I = 0.5 mW·cm–2.

The adsorption–desorption process is reversible across all concentrations studied (Figure). However, at concentrations below 0.5 mM (the CMC), photoisomerization induced by alternating UV and blue light does not fully restore the original surface coverage observed in the dark state. For example, at 0.3 mM (Figurea), the surface coverage decreases from 0.75 to 0.3 (40% reduction) upon UV irradiation. Subsequent blue-light irradiation increases the coverage again, but the final value remains approximately 20% below that of the initial dark state.

Shift of surface coverage θ over time t for (a) 0.3 mM and (b) 2 mM surfactant solutions during irradiation with the UV (λ = 365 nm, I = 0.2 mW·cm–2) and blue (λ = 455 nm, I = 0.5 mW·cm–2) light shown for 3 cycles.

At surfactant concentrations above the critical micelle concentration (CMC), the adsorption/desorption behavior exhibits a distinct pattern (Figureb). In the dark state, the surface is fully saturated with surfactant molecules. Exposure to UV light induces a reduction in surface coverage of approximately 25%, as observed at a concentration of 2 mM. Irradiation with blue light subsequently restores the surface coverage to its initial value. This reversible behavior can be attributed to the persistence of a fraction of trans isomers in the system following UV irradiation (Figurec). Specifically, after UV exposure, ca. 12% of the trans molecules remain in the system. At initial concentrations of 0.3 mM and 2 mM, this results in trans isomer concentrations of 0.04 and 0.24 mM, respectively, both of which are below the CMC. This reduction in trans isomer concentration contributes to the observed decrease in surface coverage. When irradiated with blue light, ∼66% of the trans isomers are converted back to their original state. For the 0.3 mM solution, this results in a trans isomer concentration of ∼0.2 mM, which remains below the CMC, preventing the surface from reaching full saturation. In contrast, at 2 mM, the trans isomer concentration rises to ca. 1.3 mM, a value well above the CMC, leading to complete surface saturation.

Further, it can be explained by considering the adsorption rate of only the remaining trans isomer in the solution during illumination with the blue and UV light. During illumination, the imbalance between two competing rates determines the surface coverage: the desorption rate of the cis isomer, which is proportional to the photoisomerization rate of the trans isomer at the interface, k TC,I, and the adsorption rate, k a, which is proportional to ∼c T (1 – θ_T_) (see the Theory section for more details). Under blue-light illumination, a higher fraction of trans isomers remains in the solution (∼66%) compared to that under UV light (∼12%). Consequently, the adsorption rate is consistently higher under blue light than under UV irradiation, and this effect becomes more pronounced with an increasing bulk surfactant concentration.

At low concentrations, the lower availability of trans isomers limits adsorption. As a result, the surface coverage during illumination is reduced compared to the dark state, with a more significant decrease observed under UV light due to the lower trans isomer content in the photostationary state. In this regime, the surface coverage is adsorption rate-limited. In contrast, at higher surfactant concentrations, the rate imbalance shifts toward photoisomerization. The surface coverage during illumination remains high, following the same wavelength-dependent trend. Under blue light, the system clearly exceeds the CMC, leading to no significant change in surface coverage with or without illumination. In this regime, the surface coverage becomes limited by photoisomerization kinetics. Only minor differences are observed under UV light, consistent with its approximately 4-fold higher photoisomerization rate constant compared to blue light, k TC,365 nm > k TC,455 nm ∼ 0.03 > 0.008 cm^2^·(mW·s)^−1^.?

In other words, the dynamic exchange shifts the imbalance of adsorption (I) to the photoisomerization (II) rate-limited regime with increasing concentration from (I) to (II) relative to the applied wavelength.

Kinetic

Interpretation

To analyze the dynamic exchange kinetics, we assume that only trans isomers adsorb at the interface and contribute to the signal detected by the QCM-D technique. As discussed earlier, this has been experimentally confirmed: the trans-to-cis isomer ratio adsorbed at the interface is 80:1.? During illumination, the concentration of trans isomers changes both at the interface and in the bulk solution. In the following analysis, we focus on the time interval during which the adsorbed mass undergoes measurable changes, as shown in the experimental data in Figurea (black line). Initially (time set to t = 0s), the surface of the borosilicate glass is saturated with molecules in the trans state. As a result of irradiation-induced isomerization from the trans to the cis state, the surface coverage decreases over time until a new equilibrium is reached. Further, we approximate the theoretical surface coverage using eq VI, and by setting parameters fixed to experimental conditions [c 0 = 0.1 mM, I = 0.79 mW·cm^2^, k TC of bulk is 0.05 cm^2^·(mW·s)^−1^]. The theoretical surface coverage is depicted as the red curve in Figurea, which shows that the photoisomerization rate constant at the interface, k TC,I, is 0.03 cm^2^·(mW·s)^−1^. This value is smaller in comparison to the bulk one (k TC = 0.05 cm^2^·(mW·s)^−1^), k TC > k TC,I (see green curve in Figurea).

(a) Change of surface coverage θT over time for a 0.1 mM AzoC6 surfactant solution on a borosilicate surface. The black line shows the experimental data obtained during UV light irradiation (λ = 365 nm, I = 0.2mW·cm–2). The red line is the surface coverage calculated using eq VI. The green line depicts the change in the bulk concentration, c T, of the trans molecules as calculated using eq VIII. (b) Fitting of the surface coverage by different photoisomerization rate constant k TC,I values.

To support this finding, the surface coverage data are fitted using different values of the interfacial photoisomerization rate constant, k TC,I, keeping all other parameters fixed (Figureb). When k TC,I is set significantly lower than the bulk rate constant k TC (blue curve), desorption is slower, resulting in a delayed and modest decrease in surface coverage. As k TC,I increases (red and green curves), desorption accelerates, and a new equilibrium under UV irradiation is reached more rapidly. When k TC,I > k TC (violet curve), a sharp onset of desorption is observed, leading to a prompt decrease in surface coverage. These results indicate that to match the experimental data, k TC,I must be smaller than the bulk rate constant at the given surfactant concentration.

As a next iteration, we fit the surface coverage change under illumination by varying the bulk surfactant concentration in the range between 0.05 and 2 mM.

During irradiation with blue light (λ = 455 nm, Figurea, blue dots and squares), the value for k TC remains constant for all bulk surfactant concentrations below CMC. The same trend for k TC is shown for UV light irradiation (λ = 365 nm, Figurea, purple dots and squares).

(a) Change of the photoisomerization rate constants as a function of surfactant concentrations for molecules in the bulk (k TC, blue and purple dots) and on the surface (k TC,I, blue and purple squares). (b) Comparison of rate constant ratios k TC,I·(k TC)−1 for different surfactant concentrations under exposure to blue (λ = 455 nm, blue squares) and UV (λ = 365 nm, purple squares) light. (c) Molecule density β as a function of trans surfactant concentration c T in the trans state (dark state, green squares) on a borosilicate glass surface. (d) k TC,I as a function of trans surfactant concentration irradiated with blue light (λ = 455 nm, blue squares) and irradiated with UV light (λ = 365 nm, purple squares). (e) Scheme of the process.

For both wavelengths, the value of k TC,I shows the same trend as a function of the total surfactant concentration c 0 = c T + c C. The value of k TC,I increases above a certain bulk concentration, which is c 0 ∼ 0.03 mM for UV (λ = 365 nm) and c 0 ∼ 0.4 mM for blue (λ = 455 nm) light. A similar trend is observed, although the initial concentration (c 0) belongs to two different regimes. Since primarily trans isomers adsorb at the glass–water interface, the dynamic exchange during illumination is governed by the trans isomer concentration c T, rather than the total surfactant concentration, c 0 = c T + c C. During illumination, the value of c T in the photostationary state changes as a function of applied wavelength and is c T ∼ c 0 0.12 for UV (λ = 365 nm) and c T ∼ c 0 0.66 for blue (λ = 455 nm) light illumination.? With this correction, the onset of k TC,I increase occurs at the same concentration threshold (c T ∼ 0.04 mM) (Figured). This increase correlates with structural surface coverage change of the adsorbed trans isomers beginning at c T of 0.04 mM, where the surface undergoes a transition from a bilayer formation into a more diffuse multilayer-like formation (Figuree). A plot of the number of adsorbed molecules per nm^2^ (Figurec) reveals a clear transition in the adsorption behavior. At low concentrations, the surface coverage approaches ∼2 molecules·nm^–2^, consistent with the formation of a bilayer structure. At higher concentrations, however, the number of adsorbed molecules increases substantially beyond this value. Such high surface densities cannot be explained by a simple monolayer or bilayer arrangement and, therefore, strongly indicate the formation of additional adsorption layers (multilayers) at the interface.

In the following, a potential explanation for this phenomenon is explored. Molecules located within the bilayers are under strong confinement and are essentially locked in place. Furthermore, the benzene rings may undergo π–π stacking, leading to highly aligned molecular arrangements within the bilayers. Both factors can reduce the probability of the trans–cis photoisomerization, as this process requires lateral space for molecular rearrangement.? Tightly aligned azobenzene molecules are therefore unable to switch to the cis form, a phenomenon confirmed by numerical studies on pure azobenzene molecules, ?−? ? as well as on azobenzene-based surfactants, in reasonable agreement with experimental findings. ?,?,? Upon exceeding a critical trans-isomer concentration, the adsorption probability at the solid–liquid interface increases, promoting the formation of a diffuse, less ordered secondary adsorption layer. This transition to a more diffuse interfacial structure at elevated concentrations is reflected by increased dissipation values recorded via QCM-D (Section 4 in the Supporting Information). The intricate dependence of dissipation on the surfactant concentration will be comprehensively analyzed in a subsequent study. In this sterically less constrained environment, the trans-to-cis photoisomerization rate is enhanced, resulting in an elevated effective photoisomerization rate constant, k TC,I (Figuree). This behavior likely underpins the observed wavelength-independent trend, where the interfacial dynamics are predominantly controlled by the trans-isomer concentration, governing the isomeric exchange kinetics.

Explanation of Desorption Delay: Model of

Layered Bulk Concentration

The observed delay in surfactant desorption under ultraviolet (UV) irradiation represents a notable and nonintuitive characteristic of the system (Figureb). This phenomenon is absent under blue-light illumination. As a plausible explanation, a model is proposed that accounts for a layered concentration gradient of trans isomers within the bulk solution as well as the progressive penetration of UV light into deeper solution layers, leading to spatially heterogeneous photoisomerization dynamics. The Beer–Lambert law is commonly used to analyze the switching behavior of photosensitive compounds:

$$[eqn]$$

where I abs is the fraction of the initial light intensity I 0. The parameter A′ is the absorbance by solution at the specified wavelength, and it is defined by the thickness of the sample d, as well as the extinction coefficients (Figureb) and concentrations of the trans and cis isomers, c _ T _ and c C, respectively, as follows:

$$[eqn]$$

When the absorption coefficient is low (A′ < 0.1), the relationship between intensity and concentration remains linear.? However, at high absorption coefficients, nonlinear effects arise; it occurs at high concentrations and/or large layer thicknesses. In this case, we propose to consider a solution of fixed concentration in the form of thin layers, within which the linear relationship is maintained (I abs·I 0 ^–1^) < 0.2. Figurea shows the scheme, where the 0.5 mM AzoC_6_ solution is irradiated with a UV LED from the top, λ = 365 nm, I = 0.2 mW·cm^–2^. Rainbow-colored lines indicate considered layers, where the ratio between the concentration of trans isomers and absorbed light is linear. It can be observed that the concentration varies within each layer; it decreases more rapidly in layers near the top surface and experiences a delay in deeper layers (Figureb). This is due to the amount of incident light reaching each layer (Figurec). It turns out that the upper layers absorb the majority of the photons at the beginning of irradiation. Over time, the concentration of cis isomers in the system increases, the absorption parameter decreases by an order of magnitude (ε T ^365 nm^ = 2.03 × 10^4^ L·mol^–1^ cm^–1^ and ε C ^365 nm^ = 1.9 × 10^3^ L·mol^–1^ cm^–1^ in Figureb, these values correspond to rate constants in eq VII), and the layers begin to absorb an equal amount of photons. It is worth noting that the system is considered on time scales shorter than those required for mixing solely by diffusion, t < τ ∼ H ^2^·D ^–1^ ∼ 4200 s = 70 min, where D = 2.4 × 10^2^ μm^2^·s^–1^ is the diffusion coefficient for azobenzene surfactants? and H = 1 mm is the height of the chamber. Thus, the observed delay in surfactant desorption under UV light irradiation can be explained by the physical picture presented above.

(a) Scheme of the irradiated bulk solution from the above with UV light. The rainbow color codec illustrates the considered finite layers at a specific distance from the light source. The time evolution of the concentration of trans isomers (b) and incident light (c) at different distances from the top surface over time in the case of λ = 365 nm and I = 0.2 mW·cm–2.

To gain further insight, we performed intensity-dependent measurements and the corresponding theoretical calculations for a solution with a surfactant concentration of c azo= 0.75 mM. The calculated concentration of trans isomers in the near-surface layer under irradiation at different intensities is shown in Figurea. Using the equilibrium adsorption isotherms from Figurec, the theoretical surface coverage was obtained and is presented in Figureb. A clear correlation is observed between the experimental surface coverage as a function of time and the theoretically calculated trans isomer concentration in the deeper layer, located 1 mm below the surface (the height of the QCM chamber). Notably, desorption occurs when the trans isomer concentration drops to approximately 0.6 mM, which is in agreement with the equilibrium adsorption isotherm. A comparison of the experimental and theoretical surface coverage curves confirms that the desorption delays under varying light intensities are consistent and the equilibrium values of θ are also in good agreement.

Desorption under the irradiation by UV LED with λ = 365 nm: (a) dependence of the concentration of trans isomers within the bottom layer; (b) theoretical θtheory and (c) experimental θexperimental surface coverage over time t in the case of different irradiation intensities I. The bulk surfactant concentration is c azo= 0.75 mM. (d) Scheme of the photoisomerization of surfactants in the bulk and on the surface, with the latter determining the number of adsorbed molecules over time.

Similar calculations for the system under irradiation with blue light are shown in Figures S3 and S4. Since the extinction coefficients (ε T ^455 nm^ = 1.7 × 10^3^ L·mol^–1^ cm^–1^ and ε C ^455 nm^ = 3.1 × 10^3^ L·mol^–1^ cm^–1^ in Figureb), and consequently absorption parameters, are smaller, there is no noticeable delay in the photoisomerization of molecules in the solution over the height under the current conditions. Moreover, using isotherm adsorption, we show that there is no significant desorption of molecules in the case of high initial concentrations. Returning to UV irradiation, it is important to note that the observed decrease in surface coverage in the initial moments of irradiation with high-intensity light is not an equilibrium effect, as the number of adsorbed surfactants eventually stabilizes at a constant value. This effect is of particular interest in future studies. We can speculate about possible causes, such as light interaction with the substrate, potential local heating, reorganization of surfactant molecular aggregates on the adsorption surface, and so on.

Conclusion

In this study, we demonstrated that azobenzene-containing surfactants can be reversibly adsorbed to and desorbed from a glass–water interface under remote control by light irradiation; to reveal the processes involved, we performed in situ experiments under different wavelengths. The interfacial mass change is highly sensitive to the isomeric composition in the bulk solution: the trans isomer exhibits significantly higher surface affinity compared to the cis isomer, whose increased molecular polarity reduces its interfacial activity.?

Our experiments show that the light-controlled modulation of interfacial mass is most pronounced at low surfactant concentrations and under ultraviolet (UV) irradiation, whereas the effect is considerably diminished under blue light. This can be attributed to the dominant adsorption of the trans isomer; thus, the extent of mass change at the interface is governed by its absolute concentration. Under UV irradiation, the photostationary state favors the cis isomer, resulting in a low trans isomer concentration and, consequently, a strong desorption response. In contrast, blue light produces a nearly equal distribution of isomers, maintaining a relatively high trans isomer concentration that leads to nearly saturated interfacial coverage regardless of illumination.

To further elucidate the light-induced interfacial behavior, we investigated the photoisomerization kinetics at the interface by analyzing the temporal evolution of the interfacial mass during irradiation. The experimental data reveal that the photoisomerization rate constant decreases with decreasing trans isomer concentration in the bulk. This trend can be explained by a two-layer adsorption model: (i) a tightly packed, strongly adsorbed layer with highly aligned molecules and (ii) a more diffuse, loosely bound layer with less structural order. In the diffuse layer, molecular motion is less restricted, allowing efficient trans–cis isomerization. In contrast, steric hindrance in the densely packed layer suppresses isomerization, resulting in lower conversion rates.

Upon irradiation, the interface undergoes a structural transition from a diffuse multilayer to a more compact double-layer configuration as the limited cis isomer formation within the confined layer impedes further desorption-driven reorganization. Consequently, fast photoisomerization and dynamic interfacial exchange occur only at higher bulk surfactant concentrations (e.g., ≥1 mM), where rapid replenishment of the trans isomer sustains a multilayer adsorption regime.

This dynamic behavior enables the formation of spatial cis isomer concentration gradients near the interface driven by light. These gradients are sufficiently strong to induce repulsive diffusio-osmotic flows near microscopic objects. We demonstrate that the temporal evolution of these flows correlates with both the interfacial photoisomerization rate and the associated changes in interfacial mass. Furthermore, a delayed flow response under UV irradiation can be attributed to the high absorbance of the trans isomer at 365 nm, which attenuates the light intensity through the bulk before reaching the interface.

Supplementary Material