Adsorption of Chitosan on PET Films Monitored by Quartz Crystal

Jun 28, 2008 - Chitosan as well as PET were chosen for this study due to their promising biocompatible properties and numerous possibilities to be use...
0 downloads 10 Views 4MB Size
Download PDF
Biomacromolecules 2008, 9, 2207–2214

2207

Adsorption of Chitosan on PET Films Monitored by Quartz Crystal Microbalance Tea Indest,*,† Janne Laine,‡ Volker Ribitsch,§ Leena-Sisko Johansson,‡ Karin Stana-Kleinschek,† and Simona Strnad† Laboratory for Characterization and Processing of Polymers, Faculty of Mechanical Engineering, University of Maribor, ⊥ Smetanova 17, SI-2000 Maribor, Slovenia, Laboratory of Forest Products Chemistry, Helsinki University of Technology, Post Office Box 6300, FI-02015 HUT, Finland, and Institute of Chemistry, Rheology, and Colloid Science, Karl Franzens University, ⊥ Heinrichstrasse 28, A-8010 Graz, Austria Received March 31, 2008; Revised Manuscript Received May 13, 2008

The adsorption behavior of chitosan on poly(ethylene terephthalate) (PET) model film surface was studied using the quartz crystal microbalance (QCM) technique. QCM with a dissipation unit (QCM-D) represents a very sensitive technique for adsorption studies at the solid/liquid interface in situ, with capability of detecting a submonolayer of adsorbate on the quartz crystal surface. Chitosan as well as PET were chosen for this study due to their promising biocompatible properties and numerous possibilities to be used in biomedical applications. As a first step, PET foils were activated by alkaline hydrolysis in order to increase their hydrophilicity. Model thin films were prepared from PET foils by the spin coating technique. The chemical composition of the obtained model PET films was analyzed using X-ray photoelectron spectroscopy (XPS) and their morphology was characterized by atomic force microscopy (AFM). Furthermore, the adsorption behavior of chitosan on these activated PET films and the influence of adsorption parameters (pH, ionic strength and chitosan solution concentration) were investigated in detail. Additionally, the surface chemistry and morphology of the PET films and the chitosan coated PET films were analyzed with XPS and AFM.

1. Introduction Adsorption of polysaccharides at a solid-liquid interface is the subject of intense investigation for technological as well as for biomedical applications. Therefore it is necessary to utilize special methods which provide direct information on the solid-liquid interactions. The quartz crystal microbalance (QCM) is a promising technique, which gives information about adsorption/desorption processes and is useful for sorption studies at the solid/liquid interface.1,2 It is based on the change of the oscillating frequency of the piezoelectric quartz crystal device upon mass loading, with the capability of detecting a submonolayer of adsorbate3 on the surface. Real-time measurements of the frequency shifts and energy dissipation due to the changes in mass and viscoelastic properties give information about the adsorbed layer. The chemical structure of PET with very few available polar groups (carboxyl, hydroxyl) on the surface results in low surface free energy and poor wettability. Although there are several other methods to increase PET hydrophilicity like radiation with plasma4,5 and corona treatment, alkaline hydrolysis6,7 is a standard and simple chemical pretreatment for PET surface activation. For the preparation of model surfaces of nonhydrolyzed (PET-N) and hydrolyzed PET (PET-H) as substrates for adsorption studies, the spin coating technique was used. Spin coating is a very simple and fast technique for obtaining model surfaces. As the most appropriate solvent, 1,1,2,2-tetrachloroethane was chosen, and the spin-coated films were formed on * To whom correspondence should be addressed. Tel.: +386-2-250-9646. Fax: +386-2-220-7990. E-mail: [email protected]. † University of Maribor. ‡ Helsinki University of Technology. § Karl Franzens University. ⊥ Member of the European Polysaccharide Network of Excellence (EPNOE).

silica quartz crystals. The surface chemistry of spin-coated PET-N and PET-H films was investigated with X-ray photoelectron spectroscopy (XPS) while the quality and smoothness of the PET-H film was characterized with atomic force microscopy (AFM). Furthermore, the adsorption of chitosan was investigated on such prepared model PET-N and PET-H films. Recently chitosan has attracted some attention due to its good biocompatibility8 and biodegradability.9 Chitosan macromolecules consist of glucosamine and N-acetylglucosamine units linked by 1-4 glycosidic bonds and is prepared from chitin by N-acetylation with alkali.8,10,11 It is an inexpensive, nontoxic biopolymer with numerous biological functions such as antimicrobial/antibacterial activity,12 antifungal activity,9 hemocompatibility, hypocholesterolemic activity,13 the ability to accelerate wound healing, the ability to suppress some leukemia processes,14 and it also has metal binding capacity.15,16 The purpose of this study was to investigate the adsorption equilibrium under different (pH, ionic strength, and concentration) conditions. The target was to learn if the adsorption process is driven by electrostatic or hydrophobic/entropic interactions and to prove if the adsorbed layer structure depends on the pH and the ionic strength of the solution. Additionally, the surface chemistry and morphology of the spin-coated PET-H films and the adsorbed chitosan layers (PET-HC) were analyzed with XPS and AFM.

2. Experimental Section 2.1. Materials. PET Foil. The original foil used for experiments was a Mylar polyethylene terephthalate (PET) foil, with a thickness of 175 µm. PET-N and PET-H Film Pretreatment. PET (Mylar, 175 µm) foil was immersed in 98% ethanol and cleaned in an ultrasonic bath for 10 min, washed thoroughly with demineralised water and air-dried.

10.1021/bm800333p CCC: $40.75  2008 American Chemical Society Published on Web 06/28/2008

2208

Biomacromolecules, Vol. 9, No. 8, 2008

Indest et al.

Hydrolysis Procedure. The PET foils were hydrolyzed in 4 M NaOH solution for 35 min at 70 °C. After that, the foils were immersed in 1 M HCl for 10 min to neutralize the NaOH and to stop the hydrolysis. The foils were air-dried after rinsing with a large amount of demineralized water. Spin Coating. Spin-coated nonhydrolyzed (PET-N) and hydrolyzed (PET-H) films were prepared by dissolving 1 wt % of PET foil in 1,1,2,2-tetrachloroethane (Fluka, 86960) and heated (T ≈ 150 °C) until the foil was dissolved. After the solution was cooled down, it was filtered through a 0.2 µm Acrodisc GHP filter. A total of 30 µL of solution was spread on a 14 mm silica quartz crystal and spin coated at a maximum of 2000 rpm for 60 s. QCM-D Crystals. The quartz crystals (supplied by Q Sense AB) were AT-cut quartz with gold plate electrodes and sputtered with silica on the active surface. The fundamental frequency of quartz crystals is f0 ≈ 5 MHz and the sensitivity constant is C ) 0.177 mg/m2 Hz. Chemical Compounds. All chemicals used were of analytical grade and used without further purification. Solutions were prepared at least 24 h before measurements using Milli-Q water. The pH was adjusted with NaOH (1 M, 2.5 M). Chitosan from crab shells with low molecular weight (Aldrich, 448869), 75-85% deacetylated was used. Chitosan was prepared as 0.03 g/L, 0.1 g/L, and 0.2 g/L solutions in 1 w/v % acetic acid and with different salt (0.001-0.5 M NaCl) concentrations. The rinsing of the chitosan adlayer was performed with phosphate buffer saline (PBS; 6.44 mM KH2PO4, 8 mM Na2HPO4, 140 mM NaCl; pH 7.4). 2.2. Analytical Methods. 2.2.1. Quartz Crystal Microbalance Technique. Adsorption of chitosan on PET films was studied using a quartz crystal microbalance QCM-D E4 instrument (Q-Sense AB, Gothenburg, Sweden), and the temperature was set constant to 23.34 °C within ( 0.02 °C. PET films were conditioned in aqueous solution with suitable salt concentration (0.001-0.5 M NaCl) for about 20 h prior to measurement to avoid the influence of swelling on the measurement. Two approaches were used for the performance of measurements, that is, at constant flow (0.1 mL/min) or as stepwise replacements of chitosan solutions (0.5 mL of solution every 20 min). The principle of the quartz crystal microbalance is based on the piezoelectric effect, and with this technique, frequency shifts and energy losses (dissipation) can be measured simultaneously. The resonant frequency (f0 ≈ 5 MHz) of the crystal decreases when additional mass is adsorbed on its surface. If the mass of adsorbed layer is rigid, evenly distributed, and is much smaller than the mass of the quartz crystal then the decrease in frequency (∆f) is proportional to the adsorbed mass. The adsorbed mass can be calculated from the Sauerbrey equation17

∆m ) -

C · ∆f n

(1)

where ∆m ) the adsorbed mass per unit surface (mg/cm2); C ) a constant that describes the sensitivity of device to changes in mass; n ) the overtone number; and ∆f is the frequency shift (Hz). For nonrigid overlayers, in addition to the measurement of frequency shift, it is necessary to measure the damping of the crystal oscillation: the dissipation factor D. Frictional losses occur in the crystal and the adsorbed material that lead to a damping of the oscillation, with decay in the amplitude of the piezolelectric resonator when the driving voltage is turned off, revealing dissipative properties of viscoelastic overlayers.18 The dissipation factor D is defined as

D)

Ediss 2πEstor

(2)

where Ediss is the total dissipated (lost) energy during one oscillation cycle and Estor is the total energy stored in the oscillator. The change in the dissipation factor (∆D ) D - Do) is measured when the material is adsorbed, while Do is the dissipation factor of the pure quartz crystal immersed in the solvent.19 Reproducibility of QCM Measurements. The repeatability of measurements of chitosan (0.2 g/L) adsorption at constant pH 5 and constant

Table 1. Average Values of Frequency Shift (Third Overtone, ∆f3 [Hz]) and Energy Dissipation Changes (Third Overtone, ∆D3) at Equilibrium for Chitosan (0.2 g/L) Adsorption Experiments and the Calculated Standard Deviations (SD) and Variation Coefficient (CV(%)) adsorption of chitosan on PET films

∆f3 [Hz]

∆D3 × 10-6

avg SD CV (%)

-34.5 1.7 5.6

10.0 1.3 13.0

Table 2. Results of Surface Elemental Concentration Obtained from XPS Survey Spectra of PET-N and PET-H Films sample

O 1s

C 1s

PET-N PET-H

26.23 71.91 26.97 72.55

N 1s S 2p Na 1s Si 2p Cl 2p K 2p 0.19

1.36 0.19

0.29 0.28

0.03 0.02

electrolyte concentration (0.14 M NaCl) are presented in Table 1. The values of average final frequency (third overtone, ∆f3 [Hz]) and dissipation changes (third overtone, ∆D3), calculations of standard deviation (SD), and variation coefficient (CV(%)) are shown for this sample. Very good repeatability was attained as the CV for the frequency response is about 5.6%. The CV is slightly higher for dissipation change, which amounts to 13%, but taking into consideration that the layer structure is very soft and thick, the repeatability is satisfying for this kind of experiment. 2.2.2. XPS Analysis. XPS measurements were recorded with an AXIS 165 electron spectrometer (by Kratos Analytical) using a monochromated Al KR X-ray source. Acquisition was performed at 100 W, using a low-resolution setting (1 eV step and 80 eV analyzer energy) for wide spectra, used in elemental quantification and a highresolution setting (0.1 eV step and 20 eV analyzer energy) for C 1s and O 1s regions, used in chemical analysis of carbon and oxygen species observed. The pressure in the analysis chamber was better than 1 × 10-8 Torr during the experiments. Each sample was measured on at least two locations. Surface elemental compositions were obtained from survey scan and high resolution spectra fittings were used for analysis of carbon (C1, C2, C3, C4) and oxygen (O1, O2). All spectra were collected with an electron takeoff angle of 90° from sample analysis areas less than 1 mm in diameter. 2.2.3. AFM InVestigation. Atomic force microscopy (AFM) was performed with a Nanoscope IIIa multimode scanning probe microscope (Digital Instruments, Santa Barbara, CA). The images were scanned in tapping mode using silicon cantilevers (Pointprobes, type )NCH) delivered by Nanosensors. At least three areas on each sample were scanned. No image processing except flattening was made. The scanned image size was 5 × 5 µm.

3. Results and Discussion 3.1. Characterization of Spin-Coated PET Films (PETN and PET-H). Surface Elemental Composition. XPS was performed on spin-coated PET-N and PET-H films to analyze the chemical composition of the nonhydrolyzed (PET-N) and of hydrolyzed (PET-H) film. The average atomic concentrations obtained from the XPS survey spectra for PET-N and PET-H samples are shown in Table 2. The XPS data show no significant difference between spin-coated nonhydrolyzed (PET-N) and hydrolyzed (PET-H) PET films and for both mostly carbon and oxygen were detected. Both samples exhibit similar carbon and oxygen content and contain some minor impurities of Si, Cl, K and additionally a small amount of nitrogen (0.19%) is observed only on the PET-N film. The results of XPS analysis obtained from component fittings of carbon (C1, C2, C3, C4) and oxygen (O1, O2) for nonhy-

Adsorption of Chitosan on Pet Films

Biomacromolecules, Vol. 9, No. 8, 2008

Table 3. Results of High Resolution Spectra Fittings of Carbon (C1, C2, C3, C4) and Oxygen (O1, O2) for Nonhydrolyzed (PET) and Hydrolyzed (PET-H) Films sample

C-C (C1)

C-O (C2)

PETa PET-N film PET-H film

59 63.2 62.1

22 19.5 20.2

a

CdO (C3)

CdO (C4)

(OdC-O) O1

(OdC-O) O2

19 17.4 17.7

47 46.7 34.8

53 53.3 65.2

Reference values for PET according to Beamson and Briggs.20

Figure 1. Height and phase contrast AFM images of spin-coated hydrolyzed PET film (PET-H).

drolyzed (PET-N) and hydrolyzed (PET-H) films are presented in Table 3. The C 1s of PET could be decomposed into three single peaks and as a reference the resulting values for PETa from Beamson and Briggs are represented in the first row. The obtained results are similar to reference values20 and to the theoretical ratio of 60:20:2016 for the different carbon atoms in PET films. The chemical shifts for carbon (C 1s) can be classified as unoxidized carbon (C-C, C1), carbon with one oxygen bond (C-O, C2), carbon with two oxygen bonds (O-C-O or CdO, C3) and carbon with three oxygen bonds (OdC-O, C4).21 In PET three single peaks can be decomposed from C 1s. C1 corresponds to phenyl carbon atoms (-C6H4-), C2 for methylene carbon atoms bound to one oxygen (-CH2-O-) and C4 for ester carbon atoms (OdC-O).16 The oxygen peak O 1s can be decomposed to two peaks. O1 where oxygen is bond to carbon with a double bond (OdC-O) and O2 where oxygen is bond with a single bond (OdC-O). Small differences can be observed between both PET films, although there is no increase of oxygen per se. Additionally only minor differences in the decrease of C1 and the increase of C2 and C4 are observed on PET-H samples in comparison to PET-N samples. However, when we compare the O 1s data from high resolution spectra of PET films shown in Table 3, the single bound oxygen (O2) to carbon rose from 53.3% for PET-N to 65.2% for PET-H sample while the oxygen (O1) double bound to carbon decreased from 46.7% to 34.8%. These changes indicate that hydrolysis occurred predominantly at ester groups rather than phenyl rings in the PET polymer chain and modified ester groups into C-O groups.3 This also explains why the O/C atomic ratio did not increase on the PET-H film surface. Morphology. The topography of hydrolyzed (PET-H) spincoated film was studied by atomic force microscopy (AFM). Figure 1 shows height and phase contrast AFM images of the hydrolyzed PET (PET-H) film spin coated on the quartz crystal; the surface is very smooth and uniform. 3.2. Adsorption of Chitosan on PET Films. 3.2.1. Influence of pH on Frequency and Dissipation Response. In the preliminary studies, the adsorption of chitosan onto PET-N films was monitored at different pH values (2, 3, 4, 5, 6), and it was

2209

evident that at pH 5 the relatively highest frequency decrease (the largest increase in adsorbed mass), indicating that at pH 5 the largest change in the structure occurs. To additionally investigate the influence of pH on the behavior of an adsorbed chitosan layer on the PET-N film, the pH was gradually changed. These results are presented in Figure 2, where the first part (until 300 min) represents the chitosan adsorption, followed by stepwise pH changes (gradually from 2 to 7.4). For the chitosan adsorption an aqueous solution of 0.2 g/L of chitosan in 0.001 M NaCl, 2% acetic acid at pH 2 was used. After the adsorption equilibrium was reached (∼300 min), the chitosan adlayer was rinsed stepwise with water of increasing pH value (pH 3, 4, 5, 6, 7.4). Interestingly a frequency decrease as well as a dissipation increase was observed at each pH step. Since no additional chitosan could be adsorbed, the changes indicate that the chitosan adlayer structure on the PET-N film was changed and became more dissipative. The frequency changed from -64 Hz (f3) at pH 2 to about -111 Hz (f3) at pH 3, then to about -149 Hz at pH 4 and to about -224 Hz at pH 5. Almost no frequency change is observed by changing the pH to 6 (f3 ∼ -228 Hz) and only slight increase in frequency at pH 7.4 (f3 ∼ -220 Hz). The dissipation changed similarly to frequency and, at pH 3, dissipation increased from ∼5 × 10-6 to ∼9 × 10-6, while the following dissipation changes at pH 4, 5, and 6 (∼11 × 10-6) were smaller, but at pH 7.4 a significant decrease of dissipation (∼7 × 10-6) is observed. At pH 7.4 there is a small increase in frequency accompanied by large decrease in dissipation (desorption of adsorbed mass) indicating that the chitosan layer became more compact (the layer contracts) and the water trapped in the layer must have been released, causing a decrease in dissipation. The reason for this behavior lies in the chitosan pKa value, which lies between 6.3-6.5.22,23 Increasing pH from 2 to 4 causes a thickening of chitosan adlayer due to the deprotonation of the chitosan amino groups and reduction of the electrostatic repulsion causing the rearrangements of chitosan adlayer. At pH 5, the layer is not getting thicker anymore although the frequency shift indicates additional rearrangement of chitosan molecules. At pH 6, no changes in the adlayer can be observed, but by increasing the pH to 7.4, which is above pKa value of chitosan, a significant decrease in solubility occurs. 3.2.2. Influence of Ionic Strength and Chitosan Concentration on Adsorption. To investigate the influence of ionic strength and chitosan concentration in solution on the adsorption kinetics, two different approaches were used. In the first set of experiments, the adsorption dependency on the ionic strength (0.01, 0.1, and 0.5 M NaCl) and the chitosan concentration (0.03 and 0.1 g/L) in solution at a constant pH (pH ) 5.0) was studied (Figure 3a,b). In this experiment, the effect of continuous flow conditions was tested, while other experiments were performed as sequential additions of chitosan solutions. In the second set of experiments, we investigated the influence of chitosan concentration (0.03 g/L and 0.2 g/L) on the adsorption on PET-H surface at constant ionic strength (0.14 M NaCl; Figure 4a,b). This ionic strength (0.14 M NaCl) was selected to minimize its influence in a further adsorption experiment (protein HSA), which will be presented in the subsequent publication. In the present experiments, only rinsing with PBS solution was performed as an initial step prior to protein adsorption. The resonance frequency and dissipation of the quartz crystal in contact with an aqueous medium depend on the density and viscosity of aqueous medium and, therefore, also on the

2210

Biomacromolecules, Vol. 9, No. 8, 2008

Indest et al.

Figure 2. Change in frequency (third overtone (15 MHz), black line) and dissipation (gray line) as a function of time for chitosan (0.2 g/L) adsorption on PET-N film at 0.001 M NaCl concentration (stepwise additions measurement). The adsorption of chitosan was monitored at pH ) 2, and afterward, the pH was stepwise changed up to pH 7.4.

Figure 3. Change in frequency (third overtone; Figure 2a) and dissipation changes (third overtone; Figure 2b) as a function of time for chitosan (0.03 g/L and followed by 0.1 g/L, pH 5) adsorption on PET-H film at various electrolyte concentrations (0.0 1, 0.1, and 0.5 M NaCl). The concentration of chitosan solution are quoted in brackets C(0.03 g/L) and C(0.1 g/L); constant flow measurement.

concentration of simple salt.24 The influence of the ionic strength (0.01, 0.1, and 0.5 M NaCl) on the frequency changes (Figure 3a) for two different chitosan concentrations (0.03 and 0.1 g/L) was monitored at constant pH 5. The ionic strength effect is

very important since addition of simple electrolyte screens the intramolecular repulsion between polymer segments and polymer/ surface interactions.19 In the first part of the experiment the adsorption from the lower concentrated chitosan solution (0.03

Adsorption of Chitosan on Pet Films

Biomacromolecules, Vol. 9, No. 8, 2008

2211

Figure 4. Change in frequency (∆f3, third overtone; Figure 3a) and change in dissipation (∆D3; Figure 3b) as a function of time for chitosan (0.03 g/L and 0.2 g/L, pH 5) adsorption on PET-H film at constant electrolyte concentration (0.14 M NaCl). The concentration of chitosan solution are quoted in brackets (C(0.2 g/L) and C(0.03 g/L)). The points where the rinsing with phosphate buffer (0.154 M salts, pH ) 7.4) has started are marked with “PBS” (stepwise additions measurement).

g/L) was monitored at different ionic strength until the adsorption equilibrium was reached or until only very small changes were observed. In this first step the frequency changes did not differ significantly, however the frequency decrease was the smallest for 0.01 M NaCl sample (∼-23 Hz) and the largest for 0.5 M NaCl (∼-34 Hz). After that first adsorption step the chitosan concentration was increased to 0.1 g/L at the same pH 5 and ionic strengths (after approximately 290 min.). This caused a further frequency decrease; in the case of high ionic strengths samples (0.5 M NaCl) to ∼-47 Hz, followed by 0.1 M NaCl to ∼-40 Hz. The 0.01 M NaCl sample gained only a small change to ∼-25 Hz. Figure 3b presents the changes in energy dissipation (third overtone, ∆D3) as a function of time caused by chitosan (0.03 and 0.1 g/L) adsorption at pH 5 and at different electrolyte concentration (0.01-0.5 M NaCl). In the lower chitosan concentration region (0.03 g/L), the dissipation change follows the ionic strength; it was the smallest for 0.01 M NaCl sample (∆D3 ) 4.7 × 10-6), followed by the 0.1 M NaCl sample (∆D3 ) 8.4 × 10-6), and for the 0.5 M NaCl sample, it amounts to ∆D3 ) 11 × 10-6. When the chitosan solution was changed to higher chitosan concentration (0.1 g/L), the dissipation values were significantly

increased in the case of the solutions in 0.5 and 0.1 M NaCl. This happened also with the 0.01 M NaCl sample immediately after chitosan concentration increased, but then dissipation slowly decreased even though the frequency was not changed much by increasing the concentration of chitosan. The adsorbed amount increases with increasing ionic strength and increases again with increased chitosan concentration. The increasing ionic strength compresses the electrokinetic double layer and reduces the repulsive forces between cationic charges of dissociated chitosan molecules, causing increased polymer coiling. This leads to the transformation of a loose film (at 0.01 M NaCl) to a much denser packed film at 0.5 M NaCl, as reduced repulsion between the charged segments allows additional adsorption of chitosan. Furthermore, the influence of ionic strength is more pronounced at higher chitosan solution concentration. This indicates that the nonelectrostatic interactions contribute to the adsorption of the densely coiled polymer molecules on free PET sites. The influence of multiple (16) subsequent replacement steps with chitosan solutions (concentration 0.03 and 0.2 g/L) was additionally investigated at constant electrolyte concentration (0.14 M NaCl). The results are shown in Figure 4a (frequency

2212

Biomacromolecules, Vol. 9, No. 8, 2008

Table 4. Average Values of Frequency Shift (Third Overtone, ∆f3 [Hz]) and Energy Dissipation Changes (Third Overtone, ∆D3)a sample

∆f3 (Hz)

∆D3 × 10-6

PET-HC (0.03) PET-HC (0.20)

-20.6 ( 1.3 -34.5 ( 1.9

5.5 ( 0.2 10.0 ( 1.3

a At adsorption equilibrium with standard deviation value (SD) for PETHC (0.03 g/L) and PET-HC (0.2 g/L) samples.

change as a function of time) and Figure 4b (dissipation changes as a function of time). One observes a successive chitosan adsorption until equilibrium after 14 to 16 rinsing addition steps is reached. Each rinsing step creates changes in frequency and damping caused by an additional adsorption. This process is more pronounced with the higher polymer concentration. The final steps in the equilibrium indicate always an initial adsorption increase followed by desorption to the equilibrium conditions. The chitosan adsorption (concentration 0.03 g/L and 0.2 g/L) on the bare PET surface was repeated nine times and therefore the average values of frequency changes and dissipation changes at adsorption equilibrium are summarized in Table 4. In Table 4 the average values for the changes in frequency (third overtone, ∆f3 [Hz]) and dissipation (∆D3) at chitosan adsorption equilibrium on PET-H film are presented. For low chitosan concentration (0.03 g/L), the average frequency decrease is -20.6 Hz and dissipation increase is 5.5 × 10-6. For the high chitosan concentration (0.2 g/L), the average frequency decrease at adsorption equilibrium is -34.5 Hz and dissipation increase is 10.0 × 10-6. The frequency decrease is significantly larger in case of higher (0.2 g/L) concentration in comparison to lower chitosan solution concentration (0.03 g/L) on the PET-HC sample; therefore, the adsorption rate positively correlates to the chitosan concentration. After the chitosan adsorption steps, the samples were rinsed with phosphate buffer saline (PBS, 0.154 M salts, pH ) 7.4). This was performed to investigate the behavior of chitosan adlayers exposed to the conditions that would occur during exposure to biological fluids such as blood. For the PET-HC (0.03 g/L) samples, the frequency significantly increased from -19.7 Hz to -7 Hz and -5 Hz after a second rinsing step, indicating almost complete chitosan removal. In the case of PETHC (0.2 g/L) samples, the frequency increased from -37.5 Hz to -25 Hz and the adsorbed layer remains partially on the PET surface. This is after rinsing the remaining stable film. On Figure 4b, the corresponding changes in the energy dissipation of chitosan (0.03 g/L and 0.2 g/L) adsorption are shown. The adsorption of chitosan induces large energy dissipation changes, indicating a soft adlayer, with similar kinetics as the frequency shifts. The dissipation changes indicate that by more rapid adsorption of chitosan at higher concentration (0.2 g/L), a thicker and softer layer of chitosan is obtained compared to the 0.03 g/L solution. The PBS rinsing causes severe changes in the adsorbed polymer layer. The properties of the adsorbed layer from the low chitosan concentration are almost those of the pure PET-H. Those of the higher chitosan concentration are reduced to values obtained after 3-4 adsorption steps. From the ionic strength influence (Figure 3), one can conclude that these changes are not caused by the slight increase of the ionic strength, which would in any case cause a shift toward lower frequencies. The changes in pH are also not the reason for this shift; Figure 2 clearly shows that changes from pH 5 to pH 7.4 do not change the resonance frequency, only the damping. One has to conclude that these changes are caused by the reduction of the adsorbed layer due to the rinsing

Indest et al.

with a polymer-free solution. In the case of the higher concentration, the same amount of chitosan remains after PBS rinsing as is adsorbed after three to four adsorption steps. When plotting dissipation changes (∆D; Figure 4b) versus frequency changes (∆f; Figure 4a), the time dependence (the rate of binding) is avoided and the plot presents the change in damping for every new unit of mass adsorbed.17 Thus, the plot gives an estimation of how the new added mass affects the structure on the surface. In Figure 5 on which ∆D3 is plotted versus ∆f3 it is evident that the curves for PET-HC (0.03 g/L) and PET-HC (0.2 g/L) sample are very similar. However, in the case of higher chitosan concentration (0.2 g/L), the larger dissipation is observed (10.0 × 10-6) at equilibrium conditions than for lower chitosan concentration (5.5 × 10-6). Additional to that, in the case of higher concentration of chitosan, two breaks of the slope can be seen, while for low concentration, only one break is observed and the slope above the break is slightly lower than in the case of higher chitosan concentration. The steepness of the slope and the changes in the slope during adsorption describe the packing of the layer structure, and structural changes are indicated by changed slopes. The steeper the ∆D/∆f curve, the more dissipative (softer) is the adsorbed layer, that is, more energy is dissipated.19 In this kind of plot, the spacing of data points becomes more distant when the kinetics is faster.17 Thus, very few points are observed initially for both low and high concentration of chitosan in solution due to the fast rate of adsorption. The curves of the PET-HC (0.03 g/L) and the PET-HC (0.2 g/L) samples (Figure 5) are rather flat at low frequency changes indicating a thin and rigid adsorbed layer until a frequency shift of about 10 Hz is reached (first adsorption step). The slope changes and indicates the formation of a softer, nonrigid layer (adsorption step 2 at 0.2 g/L corresponding to step 16 at 0.03 g/L) until a frequency shift of 20 Hz is reached. Above this frequency shift the slope becomes less steep again, indicating that the high concentrated chitosan solution builds a dense and more rigid outer adsorption layer (steps 3-16). Because QCM is in principle a microbalance, the mass of the adsorbed layer can be calculated by Sauerbrey equation (see section 2.2.1, eq 1). However, it is known that the quartz crystal resonators are sensitive to viscoelastic properties of tested materials,25 therefore, the Sauerbrey equation is valid only for thin films rigidly attached. Furthermore, the Sauerbrey equation is valid when the dissipation change is low (∆D < 1 × 10-6) and the frequency changes follow the order of the overtone.19 Large dissipation changes are observed for chitosan adsorption, indicating that the mass calculated according Sauerbrey equation is overestimated in case of chitosan adlayer obtained on PET-H film. Therefore, no mass balances are presented. 3.3. Characterization of the Adsorbed Chitosan Layer on PET Film (PET-HC). Surface Chemical Composition. The elemental composition of PET-H film and chitosan-covered film (PET-HC) was also compared (see sample spectra in Figure 5a). Nitrogen is an important marker to determine the presence of chitosan and it is not present on the hydrolyzed (PET-H) sample. In the case of chitosan (PET-HC) samples, there is about 1.4% of nitrogen N 1s observed (Table 5), which corresponds to chitosan amino groups on the sample’s surface. XPS high-resolution spectra on carbon and oxygen regions of the three samples (PET-N, PET-H, PET-HC) are presented in Figure 6b,c. If the PET-HC sample (Table 6) is compared with PET-H sample (Table 3), some differences in the C 1s peak deconvolution can be observed. For instance, the ester carbon peak, C4

Adsorption of Chitosan on Pet Films

Biomacromolecules, Vol. 9, No. 8, 2008

2213

Figure 5. Change in the dissipation factor as a function of change in frequency (third overtone) occurring at multiple adsorption steps of chitosan (0.03 g/L (black line) and 0.2 g/L (gray line), 0.14 M NaCl) on PET-H film. The concentration data of the chitosan solutions are listed in brackets. Table 5. Results of Surface Elemental Concentration Obtained from XPS Survey Spectra of PET-HC Films sample

O 1s

C 1s

PET-HC 25.15 71.96

N 1s S 2p Na 1s Si 2p Cl 2p K 2p 1.40

0.16

0.08

1.10

0.18

Table 6. Results of High Resolution Spectra Fittings of Carbon (C1, C2, C3, C4) and Oxygen (O1, O2) for Chitosan-Modified (PET-HC) Films sample

C-C C-O CdO C-dO (OdC-O) (OdC-O) (C1) (C2) (C3) (C4) O1 O2

PET-HC (0.2 g/L) 62.9 22.1

(C-)O), decreased from 17.7% for PET-H films to 13.8 atomic %, indicating that chitosan covered the PET-H surface to some extent. Additionally, a small amount of CdO (C3) is present on the surface of PET-HC sample, corresponding to the presence of carbonyl groups. Morphology. The topography of chitosan-modified (PETHC) film was studied by atomic force microscopy (AFM). The chitosan-coated (PET-HC (0.2 g/L)) sample shown in Figure 7

1.2

13.8

exhibits uneven coverage of the PET-H surface as chitosan formed some kind of blobs. However, it has to be taken into consideration that after chitosan adsorption, rinsing with PBS (pH 7.4) was performed (see Figure 4a). This could explain the contraction of the adsorbed chitosan to blobs on the PET surface. At pH above chitosan’s pKa, most of the amino groups are deprotonated, therefore, chitosan becomes insoluble and the solidification of chitosan is expected.

4. Discussion In this study, the adsorption behavior of chitosan on PET model films at equilibrium conditions was investigated using QCM-D. The PET model films were rigidly attached and stable also at high ionic strengths and, therefore, PET is a suitable substrate for adsorption studies with QCM-D. The purpose of this study was to investigate the influence of different parameters (pH, ionic strength, chitosan concentration in solution) on the chitosan-adsorbed layer structure at equi-

Figure 6. XPS spectra of the PET samples: (a) a low-resolution wide spectrum of PET-HC; (b) high resolution carbon C 1s regions for PETN, PET-H, and PET-HC; and (c) high resolution oxygen O 1s regions for PET-N, PET-H, and PET-HC.

Figure 7. Height and phase contrast AFM images of PET-HC sample, with adsorbed chitosan (02 g/L).

2214

Biomacromolecules, Vol. 9, No. 8, 2008

librium conditions. The influence of pH on already adsorbed chitosan layer was investigated by stepwise increase of the pH (from 2 to 7.4), which caused a decrease in frequency although no additional mass was added. The decrease in frequency and increase in dissipation during the rinsing step with water by gradually increasing pH corresponds to deprotonation and, hence, weakening of the polymer/surface interaction. During the last rinsing step at pH 7.4, the chitosan layer is compressed due to the deprotonation of amino groups. A large dissipation decrease in comparison to minor change in frequency shift indicates the release of water from the chitosan adlayer as the layer is compressed. Altogether, these effects are related to a chitosan solubility decrease. The influence of the ionic strength on the chitosan adsorption was monitored at different electrolyte concentrations (0.01, 0.1, 0.5 M NaCl) and at constant pH 5. By increasing ionic strength, chitosan molecules adopt a more coiled conformation in solution due to the reduced repulsion between the charged segments. The persistence length of the polymer coil decreases and more polymer molecules can adsorb on the surface, which is seen as a higher adsorbed amount.24 The results show that even at low ionic strengths chitosan is adsorbed, and with increasing ionic strength, the attraction between the chitosan and the PET-H surface is not weakened, meaning that nonelectrostatic interactions are mainly responsible for the adsorption. Besides the influence of ionic strength it was of interest to investigate whether additional chitosan would be adsorbed from a higher concentrated solution (0.1 g/L) after the adsorption equilibrium was reached with a lower concentrated chitosan solution (0.03 g/L). This was observed in the case of higher ionic strength. It was clearly seen that the adsorbed amount (frequency decrease) differs between all ionic strengths, less at lower chitosan concentration, and more pronounced (more adsorbed amount) at higher chitosan concentration (0.1 g/L). From these results it can be concluded that more chitosan is adsorbed and that the obtained layer is thicker at higher ionic strength and higher chitosan solution concentration. From the adsorption experiments performed (Figures 2–4), it can be concluded that under both (stepwise additions and permanent flow) equilibrium is attained and a stable film forms. The collapse of the film and transfer into a very thin and dense layer is a reason why in the case of 0.2 g/L chitosan concentration in solution (Figure 3a), there is a frequency shift of about -25 Hz and a dissipation value of about 4.2 × 10-6 that remain after the PBS rinsing. This is after rinsing a stable remaining film.

5. Summary and Conclusion To our knowledge, the chitosan adsorption onto PET model surface has not been monitored by QCM-D until now. The great advantage of the QCM-D technique is that measurements are performed in the liquid environment. In evaluating these results, it has to be taken into consideration that the mass corresponding to the frequency shift is the total adsorbed mass, which also contains water trapped in the adsorbed layer; a mass calculation can only be obtained for rigid thin films and was, therefore, not performed. It was also shown that the chitosan adsorption is measured with good reproducibility using QCM-D technique. Based on the measurements performed in this study, it can be concluded that the QCM-D technique is a very sensitive

Indest et al.

method to monitor the adsorption process. The results here are influenced by various parameters of the liquid phase (pH, ionic strength, chitosan concentration in solution). Analytical techniques as AFM and XPS were found to be complementary and confirmed the QCM results, enabling conclusions about the state of the adsorbed layer. Future investigations will be performed to study the difference in the adsorption kinetics of chitosan between PET-N and PET-H films. Acknowledgment. We gratefully acknowledge the support ¨ sterberg regarding interpretation of AFM of Dr. Monika O images, Ritva Kivela¨ and Marja Ka¨rkka¨inen are acknowledged for performing AFM measurements. Special thanks go to Petri Myllytie for the support regarding spin coating and to Dr. Joseph Campbell for performing XPS measurements.

References and Notes (1) Nakanishi, K.; Sakiyama, T.; Imamura, K. J. Biosci. Bioeng. 2001, 91, 233–244. (2) Kanazawa, K. K. In New characterization techniques for thin polymer films, 1st ed.; Tong, H. M., Nguyen, L. T., Eds.; Wiley Interscience Publication: New York, 1990; p 125. (3) Coffey, T.; Krim, J. Phys. ReV. B: Condens. Matter Mater. Phys. 2005, 72, 235414. (4) Inagaki, N.; Narushim, K.; Tuchida, N.; Miyazaki, K. J. Polym. Sci., Part B: Polym. Phys. 2004, 42, 3727–3740. (5) Riccardi, C.; Barni, R.; Selli, E.; Mazzone, G.; Massafra, M. R.; Marcandalli, B.; Poletti, G. Appl. Surf. Sci. 2003, 211, 386–397. (6) Zhang, X.; Bai, R. J. Colloid Interface Sci. 2003, 264, 30–38. (7) Liu, Y.; He, T.; Gao, C. Colloids Surf. B 2005, 46, 117–126. (8) Anand, S. C. In Medical textiles and biomaterials for healthcare, 2nd ed.; Anand, S. C., Kennedy, J. F., Miraftab, M., Rajendran, S., Eds.; Woodhead Publishing: Cambridge, England, 2006; p 329. (9) Vikhoreva, G.; Bannikova, G.; Stolbushkina, P.; Panov, A.; Drozd, N.; Makarov, V.; Varlamov, V.; Gal’braikh, L. Carbohydr. Polym. 2005, 62, 327–332. (10) Xing, R.; Liu, S.; Yu, H.; Guo, Z.; Li, Z.; Li, P. Carbohydr. Polym. 2005, 61, 148–154. (11) Gorochovceva, N.; Kulbokaite¨, R.; Juke¨nas, R.; Makusˇka, R. Chemija 2004, 15, 22–27. (12) Qi, L.; Xu, Z.; Jiang, X.; Hu, C.; Zou, X. Carbohydr. Res. 2004, 339, 2693–2700. (13) Khan, T. A.; Peh, K. K.; Ching, H. S. J. Pharm. Sci. 2002, 5, 205– 212. (14) Rosiak, J.; Ulan´ski, P.; Kucharska, M.; Dutkiewicz, J.; Judkiewicz, L. J. Radioanal. Nucl. Chem. 1992, 159, 87–96. (15) Burke, A.; Yilmaz, E.; Hasirci, N. Turk. J. Med. Sci. 2000, 30, 341– 348. (16) Zhang, X.; Bai, R. J. Colloid Interface Sci. 2003, 264, 30–38. (17) Ho¨o¨k, F.; Rodahl, M.; Brzezinski, P.; Kasemo, B. Langmuir 1998, 14, 729–734. (18) Rodahl, M.; Ho¨o¨k, F.; Fredriksson, C.; Keller, C. A.; Kozer, A.; Brzezinski, P.; Voinova, M.; Kasemo, B. Faraday Discuss. 1997, 107, 229–246. (19) Merta, J.; Tammelin, T.; Stenius, P. Colloids Surf. A 2004, 250, 103– 114. (20) Beamson, G.; Briggs, D. In High Resolution XPS of Organic Polymers, The Scienta ESCA300 Database; John Wiley & Sons: Chichester, 1992; p 174. (21) Laine, J. Doctoral dissertation, Department of Forest Products Technology, Helsinki University of Technology, Espoo, Finland, 1994. (22) Wu, L.-Q.; Gadre, A. P.; Yi, H.; Kastantin, M. J.; Rubloff, G. W.; Bentley, W. E.; Payne, G. F.; Ghodssi, R. Langmuir 2002, 18, 8620– 8625. (23) Qaqish, R. B.; Amiji, M. M. Carbohydr. Polym. 1999, 38, 99–107. ¨ sterberg, M.; Laine, J. Cellulose 2006, (24) Tammelin, T.; Saarinen, T.; O 13, 519–535. (25) Voinova, M. V.; Jonson, M.; Kasemo, B. Biosens. Bioelectron. 2002, 17, 835–841.

BM800333P