Quantitative Investigation of Layer-by-Layer Deposition and Dissolution Kinetics by New Label-Free Analytics Based on Low-Q-Whispering Gallery Modes

Abstract

A new instrument for label-free measurements based on optical Low-Q Whispering Gallery Modes (WGMs) for various applications is used for a detailed study of the deposition and release of Layer-by-Layer polymer coatings. The two selected coating pairs interact either via hydrogen bonding or electrostatic interactions. Their assembly was followed by common Quartz Crystal Microbalance (QCM) technology and the Low-Q WGMs. In contrast to planar QCM sensor chips of 1 cm, the WGM sensors are fluorescent spherical beads with diameters of 10.2 µm, enabling the detection of analyte quantities in the femtogram range in tiny volumes. The beads, with a very smooth surface and high refractive index, act as resonators for circular light waves that can revolve up to 10,000 times within the bead. The WGM frequencies are highly sensitive to changes in particle diameter and the refractive index of the surrounding medium. Hence, the adsorption of molecules shifts the resonance frequency, which is detected by a robust instrument with a high-resolution spectrometer. The results demonstrate the high potential of the new photonic measurement and its advantages over QCM technology, such as cheap sensors (billions in one Eppendorf tube), simple pre-functionalization, much higher statistic safety by hundreds of sensors for one measurement, 5–10 times faster analysis, and that approx. 25, 000 fewer analyte molecules are needed for one sensor. In addition, the deposited molecule amount is not superposed by hydrated water as for QCM. A connection between sensors and instruments does not exist, enabling application in any transparent environment, like microfluidics, drop-on slides, Petri dishes, well plates, cell culture vasculature, etc.

1. Introduction

In recent years, we have developed a novel label-free method for studying the kinetics of Layer-by-Layer (LbL) coatings on colloidal templates in real time and quantitatively evaluating the deposited materials [1,2]. The method is based on the Whispering Gallery Modes (WGMs) phenomenon [3] within spherical microparticles around 10 µm in diameter. WGMs are resonant light waves that circulate over 10,000 times within the microparticle due to total internal reflection at the interface between the particle and surrounding medium [4,5,6,7]. Adsorption of molecules onto the particle surface induces changes in both the diameter and the refractive index at the interface, resulting in a shift in resonance frequency that is directly proportional to the amount of adsorbed materials. To facilitate the generation and measurement of WGMs, we developed a compact instrument capable of exciting fluorescent dye molecules within the particle and measuring the WGM position with a wavelength accuracy of 10 pm. This allows for precise monitoring of the LbL film buildup, as well as its degradation or the release of immobilized compounds, as long as the film thickness remains within the range of the evanescent field of the traveling light wave.

The quality factor Q of WGMs describes the number of wave circulations in the resonator [8]. It can be calculated by means of the Full Widths at Half Maximum (FWHM) of the WGMs. The more circulations take place, the narrower the FWHM in the WGM spectra is, because only the resonant waves are increasingly selected. Besides a smooth surface and maximal sphericity of the resonators, the size and the refractive index difference between the resonator and medium determine the Q-factor [1]. With the decreasing diameter of the resonator, the waves are less effectively reflected because the necessary reflection angle becomes smaller. This leads to less wave circulations, lower Q-factor, and, as a consequence, broader WGM spectra and less accuracy in the determination of the peak position. However, simultaneously, the shift of the WGM increases in the case of adsorptions according to Equation (1) [9], leading to more sensitivity. Hence, there is an optimum between the accuracy of peak position determination and size of the peak shift.

For the well-designed 10.2 µm polystyrene beads with a refractive index of 1.56 in water, the FWHM is around 150 pm for 500 nm emission wavelength, resulting in a Q-factor of 10⁴. Due to limitations in construction, the spectrometer resolution is around 6–10 pm, and 150 pm FWHM is quite sufficient for the exact determination of the peak position. Hence, higher Q-factors would not yield better peak recognition but less WGM-shift or sensitivity. Therefore, we are working in the Low-Q-WGM range, while high-Q resonators can achieve factors of up 10¹⁰ [8].

Since the development of the Layer-by-Layer technology on planar substrates (Figure 1a) by G. Decher et al. [10], a huge number of investigations of such films have been performed. The technology relies on the alternating adsorption of nanometer thin films of charged polymers, such as polycations and polyanions, onto surfaces through electrostatic interactions [11]. In 1998, this technology was adapted to colloidal systems, enabling the fabrication of hollow capsules via the sacrificial dissolution of the colloidal templates (Figure 1b) [12]. This advancement significantly expanded the application potential for encapsulation and control of colloidal systems, spurring extensive research on spherical LbL systems [13,14]. A further significant advancement was the extension of the electrostatic assembling to hydrogen-bonded systems involving hydrogen donor and acceptor polymers, pioneered by Sukhishvili [15]. These films hold substantial application potential, as they can be deposited with precision and are easily dissolved by specific triggers, such as pH [16], temperature, ionic strength [17], or reductive environment [18]. This feature makes such capsule systems highly suitable for use in controllable drug delivery systems [18,19]. Notably, many biological processes, such as double-stranded DNA formation [20], RNA protein synthesis mechanisms [21], or antibody–antigen interactions [22], are also driven by hydrogen bonds or electrostatic interactions.

The analytical methods to study LbL films differ significantly between planar and spherical templates. This distinction is particularly evident in analytical techniques, such as UV/Vis, fluorescence spectroscopy, Zeta-potential determination, and reflection-based methods, like FTIR-ATR, ellipsometry, and neutron reflection, as well as high-resolution scanning methods, such as Atomic Force Microscopy (AFM), Scanning Tunneling Microscopy (STM), etc. To date, label-free detection of film growth kinetics has been feasible only on planar films by Surface Plasmon Resonance (SPR), Quartz Crystal Microbalance (QCM), or ellipsometry.

Interestingly, different analytical methods have often produced contradictory results, like when comparing LbL films on planar versus colloidal templates, even when the same material and assembling sequence were used. These discrepancies are especially pronounced in terms of assembling time, layer thickness, Zeta potential, and permeability. This leads to the question of whether colloidal coating processes result in more deviations compared with dip-coated planar films. Different results for coating planar and colloidal surfaces have been reported, especially for the more sensitive LbL coating based on hydrogen bonding [9]. The formation of hollow capsules and their subsequent analytical characterization have also revealed differences, often attributed to core dissolution and diffusion of dissolved material out of the capsule [14].

We applied the new WGM-based analytical technology to study hydrogen-bonded and electrostatically assembled LbL films, previously measured in planar setups. The results obtained from the WGM method are compared with QCM data for film assembling using the same sequences and solutions. The experiments show several clear advantages of the WGM measurements versus the QCM, such as faster analysis, higher statistic safety, and cheap and pre-functionalized sensors, as well as independence of film hydration.

2. Materials and Methods

2.1. Chemicals

The cationic polymer poly(allylamine hydrochloride) (PAH, 40,000 Da) was purchased from Beckmann-Kenko GmbH (Bassum, Germany). The anionic polymers poly(sodium-4-styrenesulfate) (PSS, 70,000 Da) and poly(methacrylic acid) (PMAA, 100,000 Da) were purchased from Sigma Aldrich (St. Louis, USA). The polyvinylpyrrolidone (PVP, 50,000 Da) was a gift from BASF (Ludwigshafen am Rhein, Germany). Polyelectrolyte solutions for LbL coating based on electrostatic attractions were prepared in 50 mM sodium acetate buffer (pH 5.6) with 200 mM NaCl and a polyelectrolyte concentration of 1 mg/mL. Polymer solutions for hydrogen-bonded LbL coating were prepared in salt-free solutions at pH 2 (adjusted with 1 M HCl) and polymer concentrations of 1 mg/mL. The buffer solutions for the film degradation were 50 mM acetate (pH 4–5.6) or 2-(N-Morpholino)ethane sulfonic acid (MES, pH 5.6–6.5) buffer solutions, freshly prepared and pH-controlled.

2.2. Sensor Particles

Monodisperse spherical sensor particles of 10.2 µm ± 0.4 µm were produced by Surflay Nanotec GmbH (Berlin, Germany) from polystyrene optimized for maximal spherical shape with smooth surface, equipped with a fluorescent, low bleaching coumarin dye. The surface was pre-coated by the common LbL coating process with polyallylamine hydrochloride (PAH), polystyrenesulfonat sodium salt (PSS), and polymethacrylate sodium salt (PMAA) with the sequence sensor bead/PAH/PSS/PAH/PMAA. The coating solutions contained 50 mM acetate buffer, pH 5.6, 200 mM sodium chloride, and 1 g/L polyelectrolyte.

2.3. WGM Measurement

The WhisperSense instrument (Figure 2a) is a compact device containing all necessary equipment for measurements, as follows:

• Microscopic objective and lenses for focusing on the 10.2 µm sensor beads, as well as a long-distance dry objective 20× (Nikon) with a numerical aperture of 0.75.
• Camera for microsensor visualization in the microfluidic device with red light illumination.
• Motorized XYZ stage with autofocus for scanning the selected beads.
• A 405 nm laser with 10 mW tunable intensity for excitation of the fluorescence dye in the beads. For measurements, 3.6 mW was set for particle excitation.
• Spectrometer for recording the emitted spectra with minimal (0.1 s) time resolution and 10 pm spatial resolution; the spectrometer was built inhouse together with the NanoBioAnalytics GmbH (Berlin, Germany).
• A microfluidic setup with a peristaltic pump, tubes, connectors, and microfluidic cells.
• The microfluidic cell consists of 8 channels of 500 µm width and 100 µm height.
• Each channel contains a microarray of 6000 conic wells of 14 µm depth and 12 µm (bottom) to 16 µm (top) diameter (Figure 2b) (Stratec Consumables GmbH, Anif, Austria); the lid (Chipshop GmbH, Jena, Germany) was laser welded with the microarray bottom.

A 100 µL aliquot of the pre-coated sensor beads in water (c = 0.5%) was introduced into one channel of the microfluidic device. The beads were allowed to sediment for 5 min before the channel was flushed with buffer solution to remove any excess particles. Up to 30 well positions containing beads were selected for statistic measurement, and their spectra were automatically recorded and stored. The small differences in bead diameter in the nanometer range lead to individual spectra of each bead. The diameter difference between wells (12–14 µm) and beads (10.2 µm) in connection with the peristaltic pump results in an undesired movement, which yields a negligible increase in noise.

For a kinetic experiment, a single bead is selected, and the measurement is initiated. At defined time intervals, the laser gives a pulse to the bead, and the spectrum is recorded (Figure 3a). The coating solution is then flushed through the channel, and the resulting changes in the emission spectrum are recorded. The difference between the recorded and the initial spectrum (Figure 3b) is converted into a wavelength shift, representing the adsorbed mass over time (Figure 3c).

During the deposition of polymer molecules onto the surface, a typical adsorption curve is obtained. Once equilibrium is reached, the analyte solution is replaced with a washing solution (identical to a coating solution). After thorough washing, the complementary coating solution is applied in the same manner. This process can be repeated as many times as desired. A limitation is given by the bleaching of the fluorescence dye inside the bead. Although this does not influence the peak position, it decreases the quality of the spectrum and the recognition of the exact position by the software, increasing the noise of the kinetic curve. Therefore, after 300–500 measuring points, the measured particle has to be exchanged for another one.

For statistical measurements, the kinetic measurement can be paused at any time (preferably during an equilibrium phase), and the spectra of all previously selected beads (e.g., 30) are then re-measured, allowing for the calculation of the adsorbed mass per bead and their standard deviations. The measurement protocol enables a high time resolution, sensitive determination of the adsorbed mass, and high statistical accuracy, which cannot be achieved by any of the other methods discussed above.

The thickness of the adsorbate layer ($∆R$) and adsorbed surface mass density ($\sigma$) were calculated using the equations proposed by A. François and M. Himmelhaus [23,24]:

$${\displaystyle \frac{∆\lambda }{\lambda }}={\displaystyle \frac{n_L}{n_s}} {\displaystyle \frac{∆R}{R}}$$

(1)

$$\sigma ={\displaystyle \frac{\rho }{3}}{\displaystyle \frac{{\left(R+∆R\right)}^3-R^3}{R^2}}$$

(2)

where $\lambda$ is the wavelength of the WGM, $n_s$ is the refractive index of the polystyrene bead, and $n_L$ is the refractive index of the adsorbed layer. $R$ is the radius, $∆R$ is the change in radius after adsorption of polymer, $∆\lambda$ is the shift in spectral position of a WGM resonance, and $\rho$ is the mass density of the adsorbate (for most proteins and polymers: 1.37 g/cm) [23].

2.4. QCM Measurement

QCM-D measurements were conducted using the Q-Sense E4 (four channels) device from Biolin Scientific (Västra Frölunda, Sweden) at 21 °C.

The gold-coated crystals were cleaned in the RCA cleaning solution consisting of H₂O:NH₃:H₂O₂ in a ratio of 5:1:1 at 70 °C for 20 min, followed by thorough rinsing with deionized water. To introduce carboxylic groups onto the gold surface, the chips were incubated overnight in 1 mM thiol glycolate solution dissolved in ultrapure ethanol. Polymer solutions were injected into the system using a peristaltic pump at a flow rate of 40 µL/min and rinsed with a buffer before introducing the subsequent polymer solution. Throughout the experiments, changes in frequency and dissipation were continuously recorded and converted to adsorbed mass using the Sauerbrey equation at the 5th harmonic (25 MHz) [25].

2.5. Hydrogen-Bonded LbL Coating

The two investigated polymers, poly(N-vinylpyrrolidone) (PVPon) and poly(methacrylic acid) (PMAA), were dissolved in salt-free aqueous solutions (MilliQ water) at a concentration of 1 g/L, and the pH was set with a calibrated pH meter to pH 2 with 1 M HCl. The washing solution consisted of diluted HCl, also set to pH 2. For the WGM measurement, the pre-coated beads were introduced into the WhisperSense device (Surflay Nanotec GmbH, Berlin, Germany) and initially equilibrated with the PMAA solution of pH 2, followed by a rinse with the washing solution. The measurement then began with the adsorption of PVPon. A coating time of 5 min per polymer was sufficient to reach equilibrium. The flow velocity over the sensor was set to 4.5 mm/s, achieved with a pumping volume of 80 µL/min. After each coating step, the sensor was flushed for 10 min with the washing buffer, which had the same composition as the coating solution without the polymer. This approach ensures that the refractive index of the medium remains nearly constant.

2.6. Degradation Experiment

The pre-coated sensor beads were further coated with two additional double layers of PVPon/PMAA following the procedure described for the measurement setup. For each experiment, the coated beads were immobilized in the WhisperSense instrument. After equilibration at pH 2, the beads were flushed with a buffer solution of 50 mM MES or acetate buffer solution, with the pH adjusted immediately before the experiment using a calibrated pH meter (Thermo Fischer Scientific, Waltham, MA, USA).

2.7. pH Calibration

For the pH measurement, the Orion VersaStar Pro pH-meter (Thermo Fischer Scientific, Waltham, MA, USA) (accuracy ± 0.002) equipped with a miniTrode microelectrode (Hamilton Company, Reno, NV, USA) was used. The pH meter was calibrated before each use with Hamilton buffer solutions of pH 4.01, 7.00, and 10.01.

3. Results

3.1. Coating Process

The coating of a single bilayer of PVPon/PMAA was studied in detail using the WGM sensor particle in a salt-free environment at pH 2 (Figure 4a). The polystyrene (PS) sensor bead surface was pre-coated with PAH/PSS/PAH/PMAA/PVPon/PMAA to ensure independence from the substrate surface. In parallel, the electrostatic assembling of a PAH/PSS bilayer at pH 5.6 and 0.2 M salt was investigated on a sensor bead pre-coated with (PAH/PSS)₂. Both coating processes were conducted under identical WGM measurement conditions, including polymer concentration, flow rate, etc. Despite the differing mechanisms of deposition, the kinetics and assembled amounts of electrostatic and hydrogen-bonded coatings were surprisingly similar. The time to reach equilibrium varied between the two types of coating. Specifically, the adsorption of PAH and PVPon reached equilibrium rapidly (less than 90 s), whereas the PSS and PMAA exhibited continued adsorption, reaching near equilibrium at around 300 s. Additionally, the deposited mass was smaller for PAH and PVPon compared with PSS and PMAA.

The same experiment was performed with the Quartz Crystal Microbalance (Figure 4b). The time required to reach equilibrium was notably longer compared with the WGM measurements, likely due to much slower streaming velocity in the QCM setup. In contrast to the WGM data, the assembled amounts differed considerably between the hydrogen-bonded polymers and the electrostatically assembled layers. The deposited masses were calculated using the equations above for the WGM and the Sauerbrey equation provided by the QCM software (QSoft 4.0.1.). The data are shown in

Table 1. Mass of the deposited polymers PVPON/PMAA layers by hydrogen bonding and PAH/PSS layers by electrostatic assembling, calculated from both WGM and QCM data (WGM: statistic of 25 sensors, QCM: statistic of 3 chips).

Layer	Adsorbed Mass of Polymer (mg/m2)
	WGM		QCM
	Avg.	SD	Avg.	SD
1. PAH	1.65	0.03	1.51	0.15
2. PSS	2.76	0.07	3.31	0.31
1. PVPon	1.83	0.20	3.21	0.33
2. PMAA	2.54	0.28	4.74	0.56

.

3.2. Multilayer Deposition

Eight double layers of PMAA/PVPon were assembled under identical conditions on spherical WGM sensor beads and on planar QCM sensor chips (Figure 5a,b). As previously noted, the kinetics of assembling were remarkably slower in the QCM measurement compared with the WGM measurements. The growth process was found to be linear for both methods, indicating that each additional layer was of equal thickness to the preceding layers. The total deposited mass of (PVPon/PMAA)₈ was determined to be 25.32 mg/m² using the WGM measurement. In contrast, the QCM measurements yielded a total deposited mass of 54.95 mg/m².

3.3. Release Process

Pre-coated WGM sensor particles were coated with two double layers of PVPon/PMAA and subsequently treated in the WGM system with buffer solutions of increasing pH values. The flushing experiments demonstrated a clear degradation of the coatings, which began to occur slowly at pH values above 4 (Figure 6). This observation contrasts sharply with the literature data [17]. Due to time constraints, the investigation was stopped after 30 min, resulting in only partial dissolution of the layers at lower pH values during this period. However, at pH values above 6, complete dissolution of the PVPon/PMAA layers was achieved within 30 min.

3.4. Stabilization of Hydrogen-Bonded Films by Top Coating Stable Electrostatically Bound LbL Films

3.4.1. WGM Data

In order to stabilize the hydrogen-bonded films, a top coating of electrostatically bound (PAH/PSS)_3.5 layers was deposited on the hydrogen-bonded (PVPon/PMAA)₃ film. This deposition was carried out in a salt-free environment at pH 2, adjusted by HCl (Figure 7a). As expected, the assembled mass of PAH/PSS was small due to the absence of salt. Flushing with pH 2 and pH 4 buffers demonstrated complete stability of the assembly. Subsequently, the assembly was flushed with a buffer solution of pH 5.8 without additional salt, which led to a slow degradation of the film. A more pronounced mass decrease was observed when the buffer was further increased to pH 7.4. After an additional 5 min, equilibrium was nearly reached under significant mass loss. Nevertheless, the dissolution process of the hydrogen-bonded film was remarkably slower compared with the assembly without a top layer, where the dissolution was completed within a few seconds at pH 7.4 (see Figure 6). Finally, the washing buffer of pH 2 was applied to accurately calculate the mass loss by compensating for the small differences in refractive indices caused by the different buffer solutions. The total mass loss was 11.60 mg/m², which corresponds to the measured deposited mass of 11.81 mg/m² by (PVPon/PMAA)₃.

In order to increase the thickness of the top layer, a similar assembly of (PAH/PSS)_3.5 was performed on top of (PVPon/PMAA)₃ in the presence of 0.2M salt (Figure 7b). While salt-free coating yielded only 1.99 mg/m² for the (PAH/PSS)_3.5 top layer, the coating with salt resulted in a remarkably thicker layer of 10.44 mg/m². Despite the increased thickness of the top film, subsequent treatment with buffers of pH 5.8 and 7.4 containing 0.2 M salt caused a mass decrease of 12.71 mg/m². This reduction is similar to the change caused by (PVPon/PMAA)₃ deposition (11.63 mg/m²), which indicates that the top coating of the PAH/PSS film on the sensor remained.

3.4.2. QCM Data

The same assembly process with salt was performed on planar QCM chips (Figure 7c). While the layer deposition on the QCM substrate did not especially deviate from the WGM data, the dissolution process exhibited notable differences. The application of buffer pH 5.6 did not result in substantial changes. However, at pH 5.8, the mass of the assembly increased rapidly within a few seconds to 150%. Nevertheless, after a few further seconds, the mass gets lost, except the mass of the base PAH/PSS/PAH/PMAA coating.

3.4.3. CLSM Data

To clarify the discrepancy in dissolution between WGM and QCM, the same polymer sequence was assembled on a glass substrate using rhodamine-labeled PAH for the top layers. The behavior of this assembly during the high pH treatment was monitored by Confocal Laser Scanning Microscopy (CLSM, Figure 7d). Upon the addition of a buffer solution of pH 5.8, increasing delamination of the (PAH/PSS)₃ film from the glass surface was observed (Figure 7d). The delaminated, flexible (PAH-Rho/PSS)₃ films of around 12 nm in thickness [26] were seen to be floating away from the surface.

The same experiment was performed for the sensor particles. However, in contrast to the planar case, the (PAH-Rho/PSS)₃ film remained on the particles, proving the assumption in Section 3.4.1.

4. Discussion

The study of the deposition of hydrogen and electrostatically bonded Layer-by-Layer (LbL) films on colloidal templates using Whispering Gallery Mode (WGM) analytics provides detailed insights into the deposited film’s quantity, the kinetics of the deposition process, and the degradation behavior of the films under varying environmental conditions.

The detailed analysis of the hydrogen- and electrostatically bonded LbL processes using the Whispering Gallery Modes (WGMs) method revealed a striking similarity in the kinetics (Figure 3a), despite the distinct mechanisms governing the assembly limitations of each process. In the case of PAH/PSS, the change in the Zeta potential leads to an electrostatic repulsion, preventing the attachment of further polymer molecules of the same polarity [27]. In contrast, the assembly of PVPon/PMAA relies solely on hydrogen bonding and is limited by the availability of free binding sites for further molecule attachment. Given the long-range nature of electrostatic forces and the short-range nature of hydrogen bonding, more pronounced differences in the kinetics between the two assembly processes were anticipated.

In both cases, the mass between the deposited complementary polymers is different. For electrostatic assembling, the mass ratio between PAH and PSS is expected, according to the molecular weight of the monomer units, to be 59 g/mol: 183 g/mol = 1:3.1. However, the WGM data reveal a ratio of 1:1.67, showing a remarkable excess of PAH. Such excess is already described in the literature and explained by the steric mismatch between PAH and PSS structure [28]. Another reason might be the limited protonation of the PAH units even at pH 5.6 [29].

For the PVPon/PMAA system, a mass ratio of 111 g/mol:86 g/mol = 1:0.78 was expected, but 1:1.48 was found. This means each PVPon unit interacts with 1.9 PMAA units. Obviously, the strong hydrogen acceptor strength of the PVPon is coordinated by almost two PMAA hydrogen donor sites.

In contrast to the WGM data, the QCM method showed a higher mass deposition for the same assembly process (Table 1). This discrepancy arises from different mechanisms underlying these techniques. In WGM measurements, the change in surface refractive index is measured, which is directly related to the mass of deposited polymers and does not depend on film hydration, porosity, or water-induced swelling [1]. In the case of QCMs, the total mass of the entire film is measured, which is the sum of adsorbed polymer and included water [25]. Hence, the mass difference between the QCM and WGM data should correspond to the amount of adsorbed water, as calculated in

Table 2. Calculation of hydration level of assembled polymer layers based on differences in adsorbed mass for QCM and WGM method (see Table 1).

Polymer Layer	Mass QCM	Mass WGM	Difference	Percentage of Hydration
PAH	1.51	1.65	−0.14	−9.2
PSS	3.31	2.76	0.55	16.6
PVPON	3.21	1.83	1.38	42.9
PMAA	4.74	2.54	2.2	46.4

. The water content in the PAH layer seems to be very low, as indicated by similar mass values obtained from both methods. However, the PAH mass in QCM was determined as the average of only two double layers, where a remarkable influence of the substrate (gold + thioglycolate) could influence the data and additionally reduce the adsorbed PAH amount. Furthermore, the three parallel QCM chips showed a high standard deviation (Table 1), making the data less reliable. For the PSS layer, a hydration level of 16% was calculated. A significantly higher water content of up to almost 50% was observed in hydrogen-bonded films, potentially due to stronger H-bond interactions with the water molecules.

When analyzing the adsorption kinetics of LbL films, it is notable that the assembly process observed in QCM measurements is significantly slower compared with WGM measurements. This discrepancy can be attributed to differences in streaming velocity and the size of the sensors, which lead to the depletion of polymer molecules on the QCM sensor surface. The streaming velocity on the QCM sensor, which has a diameter of 104 µm, is only 0.2 mm/s, whereas on the WGM sensor, it is 4.5 mm/s. Additionally, both methods revealed slightly faster adsorption kinetics for PVPon and PAH compared with PMAA and PSS (Figure 4). This can be attributed to the lower molecular weights of PVPon (50 kDa) and PAH (40 kDa) compared with PMAA (100 kDa) and PSS (70 kDa). Polymers with lower molecular weights can orient more rapidly and have a higher polymer concentration at the same mass concentration, facilitating quicker adsorption.

The assembly of eight double layers in the PVPon/PMAA system demonstrated a distinct linear growth mechanism in both measurement methods, indicating that each additional double layer has the same thickness as the previous one. This observation is consistent with findings reported in the literature [9,17].

The analysis of the degradation kinetics by WGM already showed a slow degradation at pH 4.2, which increases remarkably until pH 5.8, where the full PVPon/PMAA film is already dissolved within 30 min. This is in contrast to literature reports, where the PVPon/PMAA film was stable until pH 6.5 [9,19]. The reason for this difference is not quite clear, but the recalibration of our buffer solution and repetitions of the measurement showed the same result. In contrast to the experimental setup in the literature, our assembly is always flushed with a fresh buffer solution, always ensuring a constant pH value, while, in the literature, the assembly is treated with one buffer batch. Maybe, due to the large consumption of OH- ions for the deprotonation of the PMAA molecules, the original pH value was decreased during measurement in the literature experiments [9,16,17]. On the other hand, we used lower molecular weights of the polymers, which could also contribute to less stability of the assembly.

Several attempts to stabilize the assembly to physiological pH with further hydrogen bonding partners [30], crosslinking [19], or introduction of positive charges for electrostatic stabilization [31] were undertaken. We tried to use an additional stable electrostatic top coating of PAH/PSS on the hydrogen-bonded coating in order to stabilize the system and control the release time, especially for microencapsulation purposes for drug delivery systems (Figure 7). In the first attempt, we assembled (PAH/PSS)_3.5 (PAH was the outermost layer) in a salt-free manner on top of the PVPon/PMAA film. This resulted in a rather thin film that was not able to hold the PVPon/PMAA polymers inside the (PAH/PSS)_3.5 capsule. All of the PVPon/PMAA was diffusing out. Nevertheless, the decomposition time was decreased in comparison to dissolution without a top layer (Figure 7a). In a second experiment, thicker top layers of (PAH/PSS)_3.5 were assembled in the presence of 0.2 M salt. Although a five-times-higher polymer mass could be deposited and the release of PVPon/PMAA could furthermore be reduced, the polymers still diffused completely out at physiological pH (Figure 7b). This result was a bit surprising because such PAH/PSS layers should be impermeable for macromolecules [32]. However, the deprotonation of the PMAA results in a high osmotic pressure, which is probably able to create holes or large pores in the PAH/PSS capsule, leading to a release of the polymer molecules. Online investigation of the dissolution process by using PAH-Rho/PSS layers proved the almost unchanged position of the film on the sensor. Even a slight expected swelling of the PAH-Rho/PSS_3.5 capsule could not be observed.

The same experiment was repeated on the planar sensors of the QCM. The top LbL deposition was performed with salt, as for the WGM experiment in Figure 7c. However, the decomposition of the film showed quite different behavior than in the WGM. At pH 5.6, almost no change was observed, but by adding pH 5.8 buffer, the mass of the assembly increased tremendously, pointing to a strong uptake of water by swelling, but after a few seconds, almost the whole film was disassembled from the sensor surface.

An experiment with the same assembly on a glass surface using fluorescence-labeled PAH for the (PAH-Rho/PSS)_3.5 top layer showed in CLSM that the PVPon/PMAA layers underneath are dissolved, and free-floating films of the top layer moved away (Figure 7d). This was already reported in the literature, where the decomposition of hydrogen-bonded sacrificial layers has been used for the preparation of free-floating LbL membranes [33]. Similar free-standing LbL films were also produced by PAH/PAA (polyacrylic acid) as sacrificial film underneath, which dissolved at low pH [34], and by using zwitterionic polyelectrolytes dissolving at a high pH [35].

5. Conclusions

The reinvestigation of the kinetics and mass of LbL deposition of electrostatically and hydrogen-bonded multilayer films by the new method of Low-Q-Whispering Gallery Modes partly confirmed the literature data but, in addition, yielded further insights into the processes. This concerns the ratio between the monomer units of the polymer pairs, which is unequal in both cases, facilitating the electrostatic assembly of an excess of positive charges and, in hydrogen bonding, an excess of the hydrogen donor. The kinetics of assembling were quite similar, despite the different binding mechanisms. In both cases, the higher molecular weight polymers have slower assembling kinetics.

Due to the superposition of polymer mass with hydration in QCM but not in WGM, the degree of hydration in the films could be determined. As expected, the hydrogen-bonded films contained more water up to 50% mass. In contrast to the literature, the slow decomposition of hydrogen-bonded films has been found, already starting at pH 4.2 instead of 6.2, which originated either from the different molecular weights of the polymer combinations or by always floating fresh buffer solution in our experiments.

The study of protecting, stable top layers on hydrogen-bonded films showed different results in WGM and QCM. Due to the spherical sensors of the WGMsa, the dissolved polymers diffused out of the formed capsule in a homogeneous process, whereas, in QCM, the dissolved polymers, together with the top layer, were released in a fast process.

The new technology of Low-Q-WGM is well suited for the analysis of Layer-by-Layer coatings. Compared with Quartz Crystal Microbalance, which has been used up to now mainly for such adsorption studies, the following remarkable advantages can be shown:

• No planar and expensive sensor chips are required.
• Up to thousands of sensors can be measured in parallel, yielding much higher statistic safety than the four channels of QCM, often yielding deviating results for the same measurements.
• Up to billions of sensors can be pre-functionalized as desired, simultaneously with an identical base layer in one Eppendorf tube.
• The adsorption measurements are 5–10 times faster.
• The 10.2 µm WGM-bead surface requires 25,000 times less analyte molecules than the QCM sensor surface.
• The real mass of adsorbed material is measured and not superposed by water hydration.
• The WGM sensors are not connected to the optical signal transducer. Therefore, they can be detected in any environment (microfluidic, Petri dishes, etc.)