• Aucun résultat trouvé

Adsorption of Dendronized Polymers on Planar Water-Silica Interface Investigated by

Microscopy

6.1 Introduction

In recent years, dendronized polymers (DPs) have attracted considerable interest for their potential applications in biology and material sciences, including the development of optical devices [73, 74], biosensors [75], supports for enzymes or nucleic acids [76, 77], and drug delivery systems [78]. DPs comprise linear backbones carrying repeatedly branched dendrons of varying generation as the side chains. The homologous series of DPs relies on the same chemistry but differs in the number of monomers in the side chain [81]. Therefore, DPs allow studying the scaling behavior, i.e. variation in thickness, persistence length and other physicochemical properties as a function of generation. Numerous studies on DPs have addressed their responsive behavior [51, 82-84], the dimensions of adsorbed chains [81], self-folding of single charged polymer chain [86], and their interaction with surfaces [37]. The adsorption of DPs on various surfaces was investigated by atomic force microscopy (AFM) imaging and AFM-based single molecule force spectroscopy in a mildly acidic electrolyte solution. AFM images showed a worm-like structure of the adsorbed chains up to generation 4. Single molecule experiments revealed a substantial effect of hydrophobicity of the substrate, polymer generation and ionic strength on the adhesion of DPs [37].

Numerous studies on polymer adsorption have been published [2, 11, 13, 15, 37, 110, 230-233] and are directed at determining the influence of polymer and surface charge densities, solution of variable composition, polymer conformation at the surface, surface characteristics, and the interaction force of adsorbed polymers at surfaces. When a polymer attaches to an interface, it may undergo conformational changes to a certain extent depending on the force exerted by the surface. Moreover, these changes may be influenced by the presence of neighboring molecules as the competition for available surface comes into play. There are several factors influencing how a polymer is spreading on the surface, namely surface chemistry [232], adsorption kinetics [234] and polymer concentration [235]. Various experimental techniques can be employed in the investigation of the polymer adsorption on solid substrates, including optical reflectivity [109, 110], the quartz crystal microbalance [235, 236], direct force measurements [237, 238] and AFM imaging [144, 239]

In the present paper, we study the adsorption of DPs of different generations and at different salt levels with reflectometry and AFM. Combination of these two techniques allows us to follow the adsorption process of DPs in greater detail, to obtain the adsorbed amount, and to explore conformational changes of

96

adsorbed DPs. For a selected sample, effect of polymer concentration on the kinetics of adsorption is illustrated. We compare the adsorbed amount calculated from reflectometry signal with values directly obtained from corresponding AFM images. Finally, the AFM images are further analyzed to obtain persistence length of adsorbed polymer chains and to demonstrate the effect of surface saturation on polymer conformation.

6.2 Experimental section

Dendronized polymers. Amino-terminated polymethacrylate-based dendronized polymers (DPs) of different generations (PGn, n = 1–4) were synthesized by attach-to route as described elsewhere [154]. The molecular mass of these cationic polymers is in the range (0.3 – 5.8) × 106 g/mol depending on the generation of DPs. Polymers were investigated in aqueous electrolyte solution at pH 4 where all terminal amine groups are positively charged. DP of each generation were dissolved in electrolyte solution adjusted with HCl to pH 4.0 and then, depending on the experimental conditions used, diluted in 0.1, 1, 10, 100 and 200 mM NaCl solution at pH 4.0 to a final concentration of 5 mg/L, unless stated otherwise.

Reflectometry. The reflectometric measurements were performed on the native oxidized silicon wafer surfaces (Silchem, Germany). Prior to the experiments, the surfaces were cleaned for 20 min with hot piranha solution (H2SO4/H2O 3:1 v/v), rinsed with water and dried in a stream of nitrogen. The typical thickness of grown silica layer was in the range 1.3 – 1.5 nm as determined for each surface by null-ellipsometry (Multiskop, Optrel, Germany) in air.

The dry mass of the adsorbed polymers was measured with a home-build fixed-angle reflectometer with a stagnation-point flow cell. The reflectometer is equipped with two rotating arms with a light source and a detector. The light source is a stabilized He-Ne laser with a wavelength of 632.8 nm and the light beam is focused with an incidence angle of 60°. As a result of light refraction into the prism, the final angle of incidence on the surface was 71°. The silica surface is installed in the cell and covered with capped equilateral prism made with a borehole with a radius of 0.5 mm perpendicular to its base. The adsorption and mass transfer rate should be maximal in the stagnation point of the flow [110]. The position of the cell is therefore adjusted parallel to the surface through spacers. Two spacers keep the horizontal gap between the surface and the prism at the distance of 0.85 mm ensuring the stagnation point condition, where the collision with the surface is diffusion limited. The solutions are pumped with a peristaltic pump through the borehole in the prism. The reflected beam is split into its parallel (p) and perpendicular (s) components with respect to the plane of reflection by a polarizing beam-splitter and their intensities are then measured with lock-in amplification detection scheme. Further detail on the reflectometry setup can be found elsewhere [112].

97

The intensities are proportional to the reflectances Rp and Rs, respectively. The ratio represents the reflectometry signal R. The signal R t

 

is normalized to its initial value R

 

0 as

This reflectometry signal S t

 

is directly proportional to the adsorbed dry mass per unit area according to

 

t S t( )

  A (6.2)

The sensitivity constant A is calculated with a homogeneous slab model with the Abelès matrix method when the refractive index increment dn/dc of the adsorbed polymers in solution is known [107, 115]. The refractive index at 632.8 nm of silica is 1.547 and that of the bulk silicon is 4.146 + 0.0444i. The layer thickness of silica was taken from ellipsometric measurements and the thickness of the polymer layer was estimated. The values 0.152 mL/g for dn/dc of dendronized polymers, calculated from interferometric measurements in DMF [240], and 0.1318 mL/mol for dn/dc of NaCl salt were used.

At first, the cell was filled with the pure electrolyte solution at pH 4 to obtain a stable baseline, which was used as reference for the subsequent measurements. At time zero, polymer solution in the same electrolyte solution and pH as the background solution was injected. Once the adsorption reached a stable plateau the surface was rinsed again with the pure electrolyte solution. All measurements were performed with polymer solutions of concentration 5 mg/L, except the adsorption kinetics study where the concentration was varied. All experiments were performed at room temperature (21± 2°C).

AFM Imaging. Polymers adsorbed on silica surfaces during the reflectometry measurements were imaged in air and also in corresponding electrolyte solution with a Cypher AFM device (Asylum Research, Santa Barbara, CA) in amplitude-modulation (AC) mode. The imaging was performed close to the stagnation point, as was ensured by marking the surface through the bore hole after the reflectometry experiment with a needle and finding the mark with AFM. For imaging in air AC mode cantilevers (AC240TS, Olympus, Japan) of a nominal tip radius below 10 nm and resonance frequency around 70 kHz in air were used. The spring constants were in the range of 1.9–2.1 N/m as determined by monitoring the thermal fluctuations in air. Images were acquired in air with a scan rate of 1.5 Hz, scan size of 1 μm  1 μm, free oscillation amplitude (FOA) of about 60 nm, and a set point corresponding to around 70% of the FOA were used.

98

Biolever mini cantilevers (BL-AC40TS, Olympus, Japan) with a nominal tip radius smaller than 9 nm and a resonance frequency of 25-35 kHz in water were used for imaging in AC-mode in electrolyte solution of pH 4.0 with ionic strength adjusted with NaCl to the desired value. The spring constants were in the range of 0.06–0.1 N/m as determined by monitoring the thermal fluctuations in air. Scan rate of 3 Hz, scan size of 1 μm  1 μm, free oscillation amplitude (FOA) of about 20 nm, and a set point corresponding to around 70% of the FOA were used. All cantilevers were ozone cleaned with an UV-ozone cleaner (PSD Pro, Novascan, Ames, USA) in an oxygen enriched atmosphere for 20 minutes prior to use. The experiments were carried out at the room temperature of 25 ± 1 °C.

Measurements in liquid and in air were compared in order to verify any eventual change in conformation of adsorbed polymers induced by surface drying. It was not possible to image higher generations of DPs in liquid due to the presence of nanobubbles (Figure 6.1). However, there was a good agreement between images in air and liquid obtained for lower generations of DPs, therefore the imaging in air was generally preferred.

Figure 6.1 AFM height images of PG4 in 0.1 mM (a) and 1 mM (b) electrolyte solution at pH4 adsorbed on bare silica after reflectometry measurement. Both images indicate presence of nanobubbles formed on the adsorbed molecules. The density of nanobubbles increases with ionic strength, i.e. polymer surface coverage.

Image analysis. Images recorded at the low salt conditions were used to determine the mass of DPs from AFM images. This mass can be then directly compared to the mass obtained from reflectometric measurements. Directly after reflectometric measurements the reflection spot of laser beam on the surface is marked with a needle. The area nearby the mark is then imaged ensuring that the AFM images will illustrate the situation on the surface reported by reflectometry.

99

The molecular mass of PGn molecules was calculated from the length of the molecule using the DNA trace program [224]. All molecules adsorbed on the surface were traced to obtain the length of each molecule. Knowing the length of one monomer, the length of the molecule, the number of monomers and molecular weight, one could obtain the molecular mass.

The DNA trace program was also used to determine the persistence length of adsorbed DPs from 0.1 mM electrolyte solution at pH4. The contour of 100-160 molecules was traced and only molecules fully contained in the image frame where included in the analysis. The bond-bond correlation function was used to calculate the persistence length (Eq. (6.3)).

 

Where s is the contour length, is the angle between the tangents to two points on the molecule separated by the distance and P is the persistence length.

6.3 Results and discussion

A comprehensive set of data on the adsorption of dendronized polymers onto oppositely charged silica surfaces is presented. The adsorbed dry mass per unit area is obtained by reflectometry. The adsorbed polymer layer is subsequently imaged with atomic force microscopy. AFM images can provide information on the structure of adsorbed polymer layer, persistence length of dendronized polymers and also adsorbed mass per unit area. The combination of these two techniques provides new information on adsorption of dendronized polymers as a function of generation and different salt conditions.

Amino-terminated polymethacrylate-based dendronized polymers of different generations have been investigated on bare silica surface by reflectometry and AFM. The experiments have been carried out for different generations of DPs at pH 4 and various salt concentrations. Subsequently, the effect of polymer concentration on the adsorption process was studied. The structure of dendronized polymers is shown in Figure 6.2. Dendrons of progressive generations are attached to the polymethacrylate backbone leading to an increase in the height and the width with increasing generation of DPs. One can then easily distinguish between different generations of DPs even when co-adsorbed on the same surface as illustrated on the AFM images. Details on dimensions and other characteristics of DP adsorbed on solid surfaces can be found elsewhere [37, 81, 154].

100

Figure 6.2 Dendronized polymers of different generations adsorbed on bare silica. (a) Schemes and chemical structure of the dendronized polymers. (b) AFM height image in air of a mixture of PG1, PG2, PG3, and PG4 coadsorbed on the same surface from electrolyte solution at pH 4.

Reflectometry. Typical experimental results for the adsorption of DPs of different generations on bare silica surface in 200 mM NaCl solution at pH 4 are shown in Figure 6.3. The adsorption of DPs on the surface is characterized by a change in the reflectivity signal (Figure 6.3a) which is directly proportional to the adsorbed mass (Figure 6.3b).

Figure 6.3 Reflectometry data of DPs of different generations adsorbing on bare silica at pH 4 with 200 mM NaCl.

(a) Reflectometry signal and (b) corresponding adsorbed mass per unit area.

101

The adsorption of DPs of different generations exhibits a similar trend. At first the adsorbed mass increases linearly with time and the adsorption is not reduced by previously deposited polymer. This part of the adsorption curve shows the initial adsorption rate

d dt

t0. After a certain time period, the surface is saturated with polymer, the adsorption rate decreases and the adsorbed mass attains a plateau.

When the surface is rinsed with the pure electrolyte solution the adsorbed mass decreases slightly due to desorption of weakly bound polymers. Only minor desorption is observed, most of the polymer chains is strongly attached to the surface and therefore the adsorption of DPs on silica can be described as irreversible. The initial slope of adsorption curve changes with the generations of DPs. The final plateau value depends strongly on generations of DPs, with increasing generation of DPs there is an increase in the adsorbed mass.

Figure 6.4 Adsorbed mass obtained from reflectometry measurements of DPs of different generations and various ionic strengths in solution with NaCl at pH 4.

Figure 6.4 illustrates the dependence of the adsorbed mass per unit area of different generations of DPs on the ionic strength in NaCl solution at pH 4 and polymer concentrations of 5 mg/L. The adsorbed amount increases with increasing ionic strength and generation of DPs. The increase of adsorbed mass per unit area with increasing ionic strength can be explained by screening of electrostatic repulsive interactions between adsorbed polymer chains [241-245]. When a polymer arrives to the surface it has certain probability to stick to it depending on the total interaction, under sticky conditions this probability equals unity. In case of random sequential adsorption (RSA), polymers that cannot adhere to the surface because

102

the area where they arrive is already occupied by previously adsorbed polymers are removed from the system. In stable aqueous solutions, double-layer repulsion between polymers will reduce the maximum surface coverage substantially [107, 241].

At low ionic strengths, each polymer chain is surrounded by an extensive diffuse layer. As the adsorption to the surface continues, the diffuse layers of neighboring polymer chains repeal each other leading to the saturation of the surface at low polymer coverage, i.e. low adsorbed mass. At higher ionic strengths, the diffuse layer is less extended which leads to higher surface coverage and higher adsorbed mass per unit area, provided that the non-electrostatic polymer-surface interaction is strong enough to keep the screened polymer chain adsorbed.

The adsorption of polymers and other polyelectrolytes is controlled by several kinetic process including the mass transport to the surface, reconformation at the surface and desorption of weakly attached chains. The adsorption kinetics as well as the reconformation process will influence both the adsorbed amount and the structure of the polymer layers [236].

The effect of polymer concentration on the kinetics of adsorption of DPs is illustrated for PG4 and bare silica surface. The adsorbed amount (i.e., dry mass) dependence on the polymer concentration of PG4 in 50 mM NaCl solution at pH 4 is shown in Figure 6.5a. The initial slope increases with increasing polymer concentration, since the flux of the polymer to the surface increases accordingly. The plateau value varies with the changes in the polymer concentration. At lower polymer concentrations, the plateau is not visible in Figure 6.5a as it is only reached for longer times. The adsorbed amount at low polymer concentrations is lower than at high polymer concentrations. As the initial adsorption rate is slower at low polymer concentrations, the polymer chains have enough time to spread on the surface and block more polymer chains from coming to the surface. If the initial adsorption rate is higher, the filling time of the surface is shorter, the polymer molecules have little time to relax on the surface and more molecules can be then accommodated on the same surface area. The initial slope of the adsorption curve corresponds to the initial adsorption rate (d/dt)t=0. This slope was estimated as a plateau in the time derivative plot of the adsorbed mass (Figure 6.5b) and its value is shown in Figure 6.5c. The obtained data suggest that the adsorption rate increases monotonically with the polymer concentration. These findings are in good agreement with similar measurements performed with proteins [109] and other polyelectrolytes [244].

103

Figure 6.5 Adsorption of PG4 on bare silica at different polymer concentrations in 50 mM NaCl solution at pH4.

(a) Adsorption curves for time dependence of the adsorbed mass per unit area. (b) Corresponding time derivative plots of the adsorbed mass. (c) The initial adsorption rates as a function of polymer concentration. These values were obtained from the plateaus in the time derivative plots of the adsorbed mass.

Atomic force microscopy. DPs of different generations adsorbed on bare silica after reflectometry are imaged with AFM in the amplitude modulation mode. The AFM imaging was performed both in air and liquid. No significant difference between these images was observed; therefore imaging in air was preferred to imaging in liquid where the AFM tip interacted highly with the dense layer of adsorbed polymers. Figure 6.6 shows AFM height images of PG3 adsorbed on the silica surface during the

104

reflectometry experiments at four different levels of ionic strength at pH4. Images represent qualitatively well the adsorption of all generations of DPs at different solution conditions. As evident from presented images, the adsorption process of DPs can be well documented by AFM imaging. At low ionic strength, the maximum surface coverage is considerably limited by double-layer repulsion between adsorbed DPs chains. With increasing ionic strength, the screening of electrostatic repulsion between the adsorbed polymer chains increases and leads then to higher adsorbed amount of DPs. The AFM imaging shows increased surface coverage with increasing ionic strength which is in good agreement with earlier reflectometry measurements.

Figure 6.6 AFM height images in air of adsorbed PG3 on bare silica after reflectometry measurements in electrolyte solution of various ionic strength at pH 4.

105

Figure 6.7 shows typical AFM height images of saturated polymer layers of different generations recorded in air after reflectometry measurements in 0.1mM electrolyte solution at pH4. Each generation of DPs is also imaged shortly after polymer deposition in order to record a diluted polymer layer on the surface as a reference. In order to quantitatively characterize the adsorbed DPs of different generations, the AFM images are analyzed with DNA trace program [224]. Typically 150-200 molecules for each generation of DPs are traced and the coordinates of the polymer backbone are recorded. One can then obtain the persistence length P of adsorbed DPs using the bond-bond correlation function and fitting to the worm-like chain model (WLC) (Eq. (6.3)).

Figure 6.7 AFM height images in air of adsorbed PG1-4 on bare silica after reflectometry measurements in 0.1 mM electrolyte solution at pH 4.

106

The persistence lengths of different generations of DPs in 0.1 mM electrolyte solution at pH4 obtained from the WLC fitting are summarized in Figure 6.8. Surprisingly, in saturated layers adsorbed on the surface (cross symbol) the persistence length of DPs shows no generation dependence in contrast to the diluted case (circles). In the diluted regime, polymer molecules adsorb freely with certain conformation without restrictions induced by proximity of neighboring chains. As shown here experimentally and reported previously [80], DPs become stiffer with increasing generation which leads to an increase of the

The persistence lengths of different generations of DPs in 0.1 mM electrolyte solution at pH4 obtained from the WLC fitting are summarized in Figure 6.8. Surprisingly, in saturated layers adsorbed on the surface (cross symbol) the persistence length of DPs shows no generation dependence in contrast to the diluted case (circles). In the diluted regime, polymer molecules adsorb freely with certain conformation without restrictions induced by proximity of neighboring chains. As shown here experimentally and reported previously [80], DPs become stiffer with increasing generation which leads to an increase of the