Article pubs.acs.org/Langmuir
Dynamics of Layer-by-Layer Growth of a Polyelectrolyte Multilayer Studied in Situ Using Attenuated Total Reflectance Infrared Spectroscopy Silas Owusu-Nkwantabisah, Madhira Gammana, and Carl P. Tripp* Department of Chemistry, and Laboratory for Surface Science and Technology, University of Maine, Orono, Maine 04469, United States S Supporting Information *
ABSTRACT: Attenuated total reflectance infrared spectroscopy (ATR-IR) was used to study the dynamic layer-by-layer (LBL) growth of a sodium polyacrylate (NaPA)/poly(diallydimethylammonium) chloride (PDADMAC) multilayer on TiO2 particles. Molecular weights (Mw) used were 30 and 60 kDa for NaPA and 8.5 and 150 kDa for PDADMAC. IR spectra were recorded in situ as a function of time and were used to obtain the dynamic mass adsorbed and bound fraction of the polymers during each deposition step. For 30 kDa NaPA layers, the dynamics of adsorption show an initial rapid rise in mass followed by a slow increase toward a plateau value upon LBL with 150 kDa PDADMAC. In contrast, the 60 kDa NaPA layers achieve a plateau quickly and do not show a slow increase toward a plateau. In the case of LBL with 150 kDa PDADMAC, the dynamics of the bound fraction of polymer per layer suggest that polymer diffusion and conformational rearrangement occur for the layers of 30 kDa NaPA but not for the 60 kDa NaPA layers. Furthermore, PDADMAC adsorption profiles show that there is no diffusion of the PDADMAC layers and that PDADMAC flattens onto the underlying layer. A linear growth in the mass adsorbed per layer was observed for 150 kDa PDADMAC with both molecular weights of NaPA. In the case of 8.5 kDa PDADMAC, smaller growth increments and the desorption of underlying layers were observed. This work demonstrates the use of ATR-IR in obtaining the dynamics of LBL multilayer formation. Furthermore, it provides an example in which polymer diffusion during LBL film formation does not lead to exponential growth.
■
diffusion of polymers within the film, which is known as the in/ out diffusion model.15 According to this model, polyanions from solution diffuse into the growing film to form a reservoir of free polymers. These free polymers subsequently diffuse out of the film to contribute to the complexation of the next layer of incoming polycations. Evidence to support this diffusion model has mainly come from CLSM in which a fluorescently tagged polymer layer was found to diffuse throughout the entire structure.15 Salomaki et al.24 also proposed that all LBL systems follow exponential buildup and become linear when diffusion is slower than the deposition time. Recently, other models involving island and dendritic growth have also been proposed to explain the exponential growth in films.22
INTRODUCTION Layer-by-layer (LBL) multilayer formation has become a facile and versatile method for preparing thin films on solid surfaces.1−8 A common LBL approach is the formation of a polyelectrolyte multilayer (PEM) by successively adsorbing polyanions and polycations onto a charged surface.1,3,7−10 The nature of PEM films has been analyzed using techniques such as atomic force microscopy (AFM),11 quartz crystal microbalance (QCM),9,12,13 X-ray reflectivity,14 ellipsometry,6 confocal laser scanning microscopy (CLSM),15−17 and the zeta potential.6,12,18 Generally, the PEM films show linear or exponential growth in thickness depending on several factors such as the nature of the polymers, the pH,9 and the ionic strength.19 Understanding the molecular structure and processes giving rise to linear and exponential films has been of interest in many studies.6,15,20−23 Exponential growth is usually attributed to the © 2014 American Chemical Society
Received: August 7, 2014 Revised: September 6, 2014 Published: September 9, 2014 11696
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
Figure 1. IR spectra for NaPA recorded at the end of depositing each polymer layer. The inset shows the 2500−3050 cm−1 region for spectra recorded at the end of the NaPA#1 (black trace) and PDADMAC#1 (red trace) cycles. The Mw values are 30 and 150 kDa for NaPA and PDADMAC, respectively.
smaller growth increments, and desorption of the underlying layer was observed during the third bilayer deposition. For 150 kDa PDADMAC, linear film growth occurred as well as rearrangement and polymer diffusion for the lower-molecularweight NaPA. However, PDADMAC (150 kDa) lies flat on the surface of the underlying layer during multilayer formation. Therefore, our work provides an example in which polymer rearrangement and diffusion are observed in a linear PEM film.
It is well known that the properties of polymers adsorbed on surfaces are often dictated by kinetic factors.25,26 In many cases, the final polymer conformation depends on the sample history. Kinetic traps can exist in which the polymer resides in metastable nonequilibrium conformations. Furthermore, polymer adsorption from solution typically requires anywhere from several minutes to a few hours to achieve a maximum in the amount adsorbed.23,26,27 This is due to the length of time required for the rearrangement of the polymer on the substrate, which is much longer than the typical incubation times used in LBL deposition. However, experimental studies on the dynamics of polymer adsorption in LBL deposition have been reported only recently. Guzman et al.14,21 measured the change in the adsorbed amount as a function of time using Xray reflectivity, the dissipative quartz crystal microbalance, and ellipsometry. They showed that interdiffusion occurs in linearly grown PEMs and thus polymer diffusion is not exclusive to exponential growth. Although this work showed the importance of the dynamics of mass adsorbed to the LBL film structure and properties, further insight into the mechanism of film growth in LBL processes would benefit from the development of methods that can measure both the dynamic change in the amount adsorbed and the dynamic changes in the polymer conformation. We report for the first time a method that measures the dynamic change in mass and the bound fraction of an LBL multilayer system comprising sodium polyacrylate (NaPA) and poly(diallyldimethylammonium) chloride (PDADMAC) of different molecular weights on a TiO2 surface. We selected these polymers because they are commonly used in LBL systems. Furthermore, the approach builds on a method we developed to follow the dynamic conformation of NaPA adsorbed on TiO2.27 In particular, LBL films were fabricated using either 30 or 60 kDa NaPA with 150 or 8.5 kDa PDADMAC. Experiments were conducted at pH 3.5 because the amount of NaPA adsorbed on TiO2 passes through a maximum value at this pH.27 The method is based on attenuated total reflectance infrared (ATR-IR) spectroscopy in which spectra are recorded as a function of time during each LBL deposition cycle. The 8.5 kDa PDADMAC results in
■
EXPERIMENTAL SECTION
Materials and Solution Preparation. A flow-through ATR cell was purchased from Harrick and used with a 45° ZnSe internal reflection element (IRE) of dimensions 50 × 10 × 2 mm3. A description of the flow-through ATR cell and its use in measuring polymer adsorption on TiO2 coated on the IRE is described elsewhere.27 All spectra were recorded using GRAMS AI software with an ABB-Bomem FTLA 2000 spectrometer. Typically, 100 scans at 8 cm−1 requiring about 2 min were coadded for each spectrum recorded. Fumed TiO2 powder (P25) was obtained from Degussa and had a BET (N2) surface area of 50 m2/g. The measured isoelectric point of the P25 powder was pH 6.5. Sodium polyacrylate (NaPA, Mw = 30 000 (30 kDa), Mw/Mn = 1.4) and poly(diallyldimethylammonium) chloride (PDADMAC, Mw = 100 000−200 000 (150 kDa), Mw/Mn = 1.6) were obtained from Sigma-Aldrich and used as received. NaPA with Mw = 60 000 (60 kDa) and PDADMAC with Mw = 8500 (8.5 kDa) were obtained from Polysciences. Solutions of NaPA (20 ppm) and PDADMAC (30 ppm) were prepared by adding known quantities of the polymer to deionized water (Milli-Q, 18.2 MΩ·cm). At these solution concentrations, bands due to NaPA or PDADMAC above a bare ZnSe IRE are not detected.27 Hence, all bands in the spectra are due to the adsorbed NaPA or PDADMAC on TiO2. The pH values of these solutions and deionized water were adjusted using either dilute hydrochloric acid (HCl) or sodium hydroxide (NaOH) solutions. TiO2 Coating of the IRE. The ZnSe IRE was coated with a layer of TiO2 using an established procedure.27,28 Specifically, the TiO2 powder (30 mg) was dispersed in 25 mL of methanol and ultrasonicated for 30 min. A 200 μL aliquot of the suspension was then evenly deposited on one side of the IRE using a pipet. Evaporation of the methanol under ambient conditions resulted in a uniform TiO2 film of about 500 nm in thickness on the IRE.28 This thickness is less than the 0.7 μm penetration depth of the ATR evanescent wave. Typically, the methanol evaporated within 10 min but the TiO2 film was left at ambient conditions for at least 1 h to 11697
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
ensure thorough drying. The TiO2-coated IRE was then mounted in the flow-through cell. Polymer Deposition Method. The flow-through cell containing the TiO2-coated IRE was initially flushed at a flow rate of 5.8 mL/min with deionized water adjusted to pH 3.5. A reference spectrum was recorded, followed by the recording of spectra (100% baseline) at various time intervals. During the initial flow period, the 100% baseline spectra usually showed changes in the amount of water probed by the IRE. This was primarily due to the removal of air bubbles from the cell cavity and associated tubing. The procedure for recording a reference spectrum, followed 15 min later by a 100% baseline spectrum, was repeated until no changes in a 100% baseline spectra were observed. This typically required about 30−45 min from the start of flushing the cell with water. During this initial flow of water and in subsequent additions of polymer solutions, no evidence of removal of the TiO2 coating was observed (no decrease in TiO2 bulk modes near 720 cm−1). Once this initial break-in period was established, a reference spectrum was recorded and used for the remainder of the experiment. A 20 ppm NaPA solution at pH 3.5 was then allowed to flow through the cell at 5.8 mL/min, and spectra were recorded as a function of time for approximately 3.5 h. This deposition cycle is referred to as NaPA#1. The cell was then flushed with pH 3.5 water for 5 min to remove excess NaPA from the cell cavity. Next, a 30 ppm PDADMAC solution at pH 3.5 was allowed to flow through the cell for 3.5 h at 5.8 mL/min, and spectra were recorded at specified time intervals. This cycle is referred to as PDADMAC#1. The cell was then again flushed with water at pH 3.5 for 5 min to remove excess PDADMAC from the tubing and cell cavity. No change in the IR spectrum was observed during this flushing step. The sequential flow of NaPA, rinsing, PDADMAC, and rinsing was repeated twice using the above procedures, denoted as NaPA#2, PDADMAC#2, NaPA#3, and PDADMAC#3, respectively. IR spectra were collected every 5 min during each deposition cycle. All experiments were repeated at least three times.
calculated using this band gave the same value as obtained using the band at 2930 cm−1. This confirms that the subtraction procedure to account for the contribution from NaPA to the intensity of the band at 2930 cm−1 was valid. It is noted that a peak near 1637 cm−1 of 0.002 to 0.01 absorbance units in intensity appeared in about 3−5% of the spectra recorded. This band is due to the deformation mode of water. The water deformation mode is typically on the order of 1 absorbance unit when a reference spectrum of a dry TiO2coated IRE is used. Thus, the appearance of a band at 1637 cm−1 from time to time corresponds to less than a 1% fluctuation in the amount of water passing through the cell. This band was removed from the spectra by subtraction using a prerecorded spectrum of water. It is also noted that the signalto-noise level in the 1640 cm−1 region is about 10 times lower than in the surrounding spectral region, reflecting the 90% reduction in signal due to adsorption by the bulk water in the cell. Amount of Polymer Adsorbed. The procedure to determine the amount of adsorbed NaPA from the intensity of the band at 1455 cm−1 is described in our earlier work,27 and details are provided for NaPA and PDADMAC in the Supporting Information section. In brief, a final spectrum was recorded at the end of the experiment while the cell was rinsed with DI water. The IRE was then removed and dried, and a transmission IR spectrum was measured through the coated IRE. The amount of NaPA adsorbed was determined from the intensity of the 1455 cm−1 band in the transmission spectrum. This value was used to calibrate the peak intensity in the final ATR spectrum. The same procedure was used to calibrate the intensity of the 2930 cm−1 band for PDADMAC and the 720 cm−1 peak for TiO2. Figure 2 is a plot of the ratio of the
■
RESULTS AND DISCUSSION IR Spectra of Adsorbed Polymers. Figure 1 shows the IR spectra recorded as a function of time at the end of each deposition cycle. A 3D plot of spectra showing the time evolution of these bands during the adsorption of NaPA#1 and PDADMAC#1 is provided in the Supporting Information section (Figure S1). The inset in this figure shows the region between 2500 and 3050 cm−1 for the last spectrum recorded in NaPA#1 and PDADMAC#1, respectively. The key bands of interest for NaPA occur at 1713 cm−1 for a CO stretching mode of COOH, 1545 and 1414 cm−1 for the asymmetric and symmetric stretching modes of COO−, and 1455 cm−1 for a CH2 bending mode. The CH2 band at 1455 cm−1 was used to determine the amount of NaPA. This band overlaps with the COO− mode at 1414 cm−1 and had a shoulder on the high-wavenumber side at 1473 cm−1 due to a C−H mode of PDADMAC. The region was curve fitted with a Gaussian−Lorentzian function using GRAMS AI software over the range of 1490 to 1370 cm−1 to remove the contributions from the COO− and PDADMAC bands. For PDADMAC, there was a band due to the C−H stretching mode at 2930 cm−1 (Figure 1 inset), and the intensity of this band was used to calculate the amount of PDADMAC. Bands in this region, due to NaPA, are weak in intensity (inset in Figure 1). In calculating the mass of PDADMAC, we subtracted the band intensity of the final NaPA#1 spectrum from the overall band intensity in this region. For PDADMAC#2 and PDADMAC#3 deposition cycles, a C−H band of PDADMAC at 1473 cm−1 was of sufficient intensity to calculate the amount of PDADMAC (for example, see PDADMAC#3 in Figure 1). The amount
Figure 2. Ratio of the intensity of the band at 1455 cm−1 recorded in transmission IR to ATR-IR as a function of the number of layers. The Mw values were 30 kDa for NaPA and 150 kDa for PDADMAC.
intensity of the band at 1455 cm−1 in the transmission spectrum to the final ATR spectrum. The data points shown in Figure 2 were obtained using separate LBL experiments that we stopped at the end of a different number of NaPA deposition cycles. (For example, in one experiment we stopped at the end of NaPA#2, and then in a separate experiment, we stopped at NaPA#3.) In each experiment, after the rinsing step with DI water, the IRE was removed and dried and a transmission spectrum was recorded. The band intensity measured in the transmission spectrum showed a linear increase in its value (Figure S3 in the Supporting Information). Thus, the nonlinearity observed in Figure 2 originated with the evanescent wave dependence on 11698
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
Figure 3. Amount of polymer (mg per m2 of TiO2 surface) adsorbed as a function of time for LBL using 30 kDa NaPA and 150 kDa PDADMAC.
Table 1. Final Amount of Polymer (mg per m2 of TiO2 Surface) per Layer for LBL Films Prepared Using 30 kDa NaPA and 150 kDa PDADMACa
intensity found in the ATR spectra. In ATR, the evanescent wave decays exponentially from the surface; therefore, the band intensity will decrease with distance away from the IRE. Figure 2 shows that the dependence on the evanescent wave was about 10% for the first three cycles, increasing to more than 25% in the fourth cycle. The calculated penetration depth at 1455 cm−1 was 0.7 μm (equation S1 in Supporting Information), exceeding the 0.5μm-thick TiO2 layer. Thus, at first glance, the evanescent wave dependence as a function of intensity would occur when the amount of polymer adsorbed is not uniform throughout the TiO2 film and increases in the upper section of the TiO2 film with each deposition cycle. However, several control experiments showed that the adsorption of the polymers occurs throughout the entire TiO2 layer. In one control experiment, LBL deposition was performed using TiO2 powder stirred in beakers containing solutions of NaPA or PDADMAC. The amount of polymer adsorbed in each cycle was the same as that measured at the end of each cycle in the ATR experiment. In a second control experiment, an IRE was coated with a TiO2 film that was 50% thinner than normally used. In this case, the mass of polymer adsorbed per cycle was 50% of the value obtained with the thicker TiO2 layer. This shows that the polymers are adsorbed throughout the entire TiO2 layer and that the effective surface area available on the film is the same as for a dispersed suspension of TiO2 particles in a beaker. Because the deposition of the polymer is uniform throughout the TiO2 film, the only possibility to account for the decreasing absorbance value with each cycle is a refractive index change in the rarer medium. It is noted that the refractive index of TiO2 is about 2.5, whereas the refractive indexes of NaPA and PDADMAC are 1.43 and 1.38, respectively. Coating the TiO2 with successive polymer layers would tune the refractive index to increasingly lower values for the TiO2/water layer and hence reduce the penetration depth and the index matching. A reduction in both of these factors led to a lower intensity of the bands in the ATR spectrum.29 The refractive index tuning of materials via the formation of thin films is well known.30−32 Thus, the curve in Figure 2 provided the correction value to be applied to the intensity of bands in calculating the adsorbed amount. Figure 3 is a plot of the mass of polymer adsorbed per layer as a function of contact time for LBL using 30 kDa NaPA and 150 kDa PDADMAC. Table 1 gives the average amount of polymer at the end of each deposition cycle. There are clear differences in the amounts obtained for the first layer,
amount/mg·m−2
layer NaPA#1 PDADMAC#1 NaPA#2 PDADMAC#2 NaPA#3 PDADMAC#3
7.8 2.5 6.5 3.9 6.3 3.9
± ± ± ± ± ±
0.4 0.3 0.5 0.4 0.6 0.5
a
The errors are at the 95% confidence level based on a minimum of three measurements.
compared to those for the second and third layers. The total amount adsorbed for NaPA#1 is about 20% higher than for NaPA#2 and NaPA#3. However, the adsorbed amount of PDADMAC#1 is 40% lower than for the succeeding PDADMAC layers. Clearly, the adsorption of NAPA#1 and PDADMAC#1 represent a transition layer. This transition is well known to occur in LBL-based depositions, but typically the transition occurs over three to four layers.6,10,17,33 In our experiment, the transition occurs in the first cycle. This shorter transition is likely due to the long incubation time (3.5 h). Typically, the longest incubation times in LBL depositions are 10 to 30 min per layer.11,17 Figure 4 shows the cumulative amount of adsorbed polymer at the end of each deposition and is the most common way of plotting the LBL growth of PEM films.5,9,11,34 The plot (Figure
Figure 4. Cumulative amount of polymer (mg per m2 of TiO2 surface) deposited for each layer for LBL using 30 kDa NaPA and 150 kDa PDADMAC. 11699
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
4) follows a typical sawtooth pattern5,10 and a linear growth regime with a calculated R2 value of 0.98. Dynamics of Polymer Adsorption. Figure 3 also shows that the dynamics for NaPA and PDADMAC adsorption follow different profiles. For NaPA, the curves have a sharp initial rise followed by a slow increase in amount that tends toward a plateau value. In contrast, the curves due to PDADMAC adsorption reach a plateau rapidly (PDADMAC#1 curve rises rapidly to a plateau value within 20 min). Revisiting Figure 1, we see that the addition of PDADMAC#1 leads to a decrease in the COOH band at 1713 cm−1, which is mirrored by an increase in intensity of the COO− bands at 1545 and 1414 cm−1. There is no loss of NaPA because the band at 1455 cm−1 remains constant in intensity. Li and Tripp27 have shown that the change in intensity of the COOH and COO− bands for NaPA adsorbed on TiO2 can be used to measure the dynamic bound fraction of charged sites (COO− bound) to charged sites on the TiO2. We have extended this approach to the alternating addition of PDADMAC and NaPA. Details are provided in the Supporting Information section. The decrease in the COOH band upon exposure of NaPA#1 to PDADMAC#1 shown in Figure 1 can be attributed to two sources. First, it is primarily due to the adsorption of the positively charged sites on PDADMAC with the COO− groups located in the loops and tails of NaPA#1. Second, the rearrangement of the polymer molecules in NaPA#1 that occurs with PDADMAC deposition could lead to a change in the fraction of NaPA bound to the underlying TiO2 substrate. The combination of the two effects leads to a decrease in the total number of COOH groups because the ratio of COOH/ COO− groups in the loops and tails remains constant at their solution equilibrium values. Figure 5 is a plot of the total bound %COO− recorded as a function of time during the deposition of each layer in the LBL
rearrangement of NaPA on top of the PDADMAC layer with little, if any, diffusion of the NaPA into the PDADMAC layer. However, there are differences in the trend in the total bound fraction for NaPA#1 and other deposition cycles (Figure 5). The value for the bound fraction reaches a constant value of 48% within 40 min for NaPA#1, whereas in all other NaPA cycles the bound fraction continues to vary, albeit slowly at the end of the 3.5 h incubation. Therefore, NaPA diffuses into the underlying PDADMAC layer. Furthermore, while NaPA#1 passes through a maximum in bound fraction at 80% before reaching a 48% constant value, NaPA#2 and NaPA#3 do not pass through such a maximum (Figure 5). This suggests that NaPA#2 and NaPA#3 do not flatten on the underlying PDADMAC layers. The total bound fraction at the end of NaPA#2 and NAPA#3 are lower in value (about 32 and 28%, respectively) than the value obtained at the end of NaPA#1 (48%). This shows that the number of bonds between PDADMAC and NaPA is smaller than that for NaPA with TiO2. Furthermore, there is a stronger interaction between the COO− groups on NaPA and the positively charged sites on TiO2 than for the positively charged sites on PDADMAC. The difference in wavenumbers (Δν = νas − νs) between the asymmetric (νas) and symmetric (νs) stretching frequencies of COO− bands for NaPA#1 was Δν = 132 cm−1, whereas for NaPA#2 and NaPA#3 we obtain Δν values of 134 and 142 cm−1, respectively. The Δν values for NaPA#2 and NaPA#3 are composite values for the combined NaPA in the PEM film. It is noted that for NaPA#2 the Δν represents the overall value for both NaPA#1 and NaPA#2. Therefore, the similarity of Δν for NaPA#2 and NaPA#1 is indicative of the higher percent contribution of NaPA#1. Nevertheless, this shows that the Δν obtained for NaPA adsorbing with TiO2 sites is smaller than when NaPA binds with charged sites on PDADMAC. A smaller Δν implies a stronger interaction with the COO− functionalities.35 Thus, the combination of a higher bound fraction and stronger interaction shows that NaPA#1 is more strongly bound to TiO2 than to the PDADMAC layers. Bound Fraction of Individual Layers. The deposition of PDADMAC onto the underlying NaPA layer always increases the bound fraction of NaPA (Figure 5). From the knowledge of the amount of PDADMAC adsorbed and assuming intrinsic compensation involving a 1:1 interaction between the positively charged sites on PDADMAC and the negatively charged sites on NaPA, an estimate of the bound fraction of PDADMAC was determined. (Details of this calculation are provided in the Supporting Information section.) Figure 6 displays the bound fraction for each layer. Here, the bound fraction of PDADMAC refers to the fraction of NR4+ moieties binding to the COO− groups of NaPA. In generating these values, it was assumed that the number of groups already bound to the underlying layer does not change in value. In other words, the adsorption of PDADMAC#1 onto NaPA#1 does not result in any change in the fraction of NaPA#1 bound to the underlying TiO2. This is a reasonable assumption for PDADMAC#1 given the strong interaction between COO− groups and the charged sites on TiO2. From Figure 5, the addition of PDADMAC#1 leads to a 12% increase in the COO− bound fraction. Using the mass of adsorbed PDADMAC, we calculate that these COO− groups bind with 68% of the charged sites on the PDADMAC. A value of 68% bound fraction could suggest a high level of interpenetration of PDADMAC into the underlying NaPA#1
Figure 5. Total bound fraction of NaPA as a function of time for LBL using 30 kDa NaPA and 150 kDa NaPA.
process. The NaPA#1 layer forms on TiO2 with an initial bound fraction greater than 80% that drops to a plateau value of 48% at 40 min. This is because the initial NaPA in solution arriving at a bare TiO2 flattens onto the surface. The adsorbed NaPA subsequently rearranges to a conformation containing more loops and tails to accommodate more polymer adsorption on TiO2.27 The curves obtained for the dynamic amount adsorbed in NaPA#2 and NaPA#3 (Figure 3) are similar in shape to that of NaPA#1. This could be evidence that there is a slow 11700
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
ment of the adsorbed NaPA to accommodate more NaPA, as shown by the slow increase in adsorption toward a plateau during NaPA#1 deposition (Figure 3) and the drop in value for the bound fraction with time. Figure 6 shows that there is a general trend of a decrease in the bound fraction for each NaPA layer, and in fact the bound fraction for NaPA#3 is negative in value. While a decrease in the bound fraction may imply an increase in the coil-like conformation of the adsorbed layer,27 a negative bound fraction could indicate the loss of PDADMAC from the underlying layer to the solution phase. However, the IR spectra show no loss of PDADMAC during any of the NAPA cycles. Furthermore, the mass adsorbed for NaPA#3 was 6.3 mg·m−2 (Table 1), and this layer must have some positive value for the bound fraction to the underlying PDADMAC layer. Thus, a negative bound fraction means that the assumption that the bound fraction of the underlying layer does not change is invalid for cycle #2 and higher. This shows that the adsorption of PDADMAC leads to a reduction of charged sites on NaPA bound to the underlying PDADMAC layer. Hence, the adsorption of the next PDADMAC layer leads to a rearrangement of the entire underlying NaPA layer. This is also consistent with the slow change in the bound fraction for NaPA#2 and NAPA#3, which was attributed to the interdiffusion of NAPA into the underlying layer. LBL Using NaPA/PDADMAC of 60/150, 60/8.5, and 30/ 8.5 kDa Mw Ratios. The dynamics of LBL using 60 kDa NaPA and 150 kDa (and separately with 8.5 kDa) PDADMAC are provided in the Supporting Information section (Figures S3 and S4, respectively). The dynamics of LBL using 30 kDa NaPA and 8.5 kDa PDADMAC are provided in Figure S6. Compared to the dynamics of LBL using 30 kDa NaPA (Figure 3), the 60 kDa NaPA layers rise quickly initially and do not show a slow continual increase in mass. Thus, the 60 kDa NaPA layers do not show polymer diffusion into the growing PEM film. The plot of cumulative amounts of adsorbed polymer shows that the PDADMAC/NaPA multilayer grows linearly independent of the molecular weight of NaPA (Figure S7a in the Supporting Information section). Therefore, 60 kDa NaPA provides an example in which there is no polymer diffusion in a PEM film that shows linear LBL growth. In the case of 8.5 kDa PDADMAC (Figure S7b), there is a decrease in bilayer increments in the amount of adsorbed polymer. The LBL bilayer increments for 60 kDa NaPA are 14 and 3 mg·m−2 with 150 and 8.5 kDa PDADMAC, respectively. This is consistent with a report by Kujawa et al.32 on the molecular weight dependence of the hyaluronic acid/chitosan (HA/CH) multilayer system. Although the film growth rate did not change for the molecular weights (Mw) studied, Kujawa et al. indicated that an increase in Mw of HA or CH resulted in thicker films. In the case of 30 kDa NaPA and 8.5 kDa PDADMAC, the dynamic amount of NaPA layers continues to increase, never reaching steady state during the entire 3.5 h incubation time for each layer. This is consistent with a slow diffusion of NaPA. There is almost no growth during the deposition of NaPA#3. This is likely due to few binding sites of the 8.5 kDa PDADMAC resulting in fewer electrostatic attraction sites with NaPA during LBL. The NaPA#3 that is further extended from the TiO2 surface can easily be washed into the solution phase, thus preventing NaPA#3 growth. As was observed for 150 kDa PDADMAC (Figure 3), the 8.5 kDa PDADMAC layers rapidly reach a plateau in the adsorbed
Figure 6. Fraction of individual layers bound to the respective underlying layer. Data points refer to the fraction of COO− in NaPA or NR4+ in PDADMAC from spectra at the end of deposition for each layer. The Mw values were 30 kDa for NaPA and 150 kDa for PDADMAC.
layer. However, this is unlikely given the rapid plateau observed in the dynamic amount for PDADMAC#1 in Figure 3 and the rapid plateau in the dynamic bound fraction shown in Figure 5. Thus, a value of 68% bound fraction indicates that PDADMAC lies flat on top of the NaPA layer (Scheme 1). Decher1 Scheme 1. Molecular Picture of NaPA#1 (Mw = 30 kDa) and PDADMAC#1 (Mw = 150 kDa) Adsorbed on TiO2
described that the self-regulation of LBL is achieved because when charge reversal occurs, the incoming equally charged molecules are repelled from the growing film surface. Therefore, because each segment of PDADMAC carries a charge, the incoming molecules will be repelled upon approaching the initial flatly arranged PDADMAC molecules in the layer. As a result, the flat PDADMAC layer does not rearrange to accommodate the adsorption of more PDADMAC. Figure 3 shows that PDADMAC#2 and PDADMAC#3 have similar profiles in that both show a rapid plateau in the adsorbed amount. Guzman et al.14 have studied the kinetics of formation of poly(allylamine hydrochloride)/poly(sodium 4styrenesulfonate) (PAH/PSS) and PDADMAC/PSS multilayers as a function of ionic strength. For PDADMAC adsorption at low ionic strength, adsorption was rapid, whereas longer times were necessary to reach steady state at high ionic strength. It was concluded that there was limited interdiffusion for the PDADMAC/PSS multilayer at low ionic strength. This is consistent with our results in DI water where PDADMAC is the polycation. Thus, we conclude that the diffusion of PDADMAC into the underlying NaPA layer does not occur to any significant extent. In contrast to PDADMAC#1, there is a rearrangement of the initial flat layer of NaPA on the TiO2. In this case, there will be minimal repulsion of incoming NaPA molecules because 25% of the segments are charged, much lower than the 100% charged segments on PDADMAC. Hence, there is a rearrange11701
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
PAH Multilayer Films and on Their Permeability to an Electroactive Probe. Langmuir 2009, 25, 2282−2289. (6) Bragaru, A.; Kusko, M.; Radoi, A.; Danila, M.; Simion, M.; Craciunoiu, F.; Pascu, R.; Mihalache, I.; Ignat, T. Microstructures and growth characteristics of polyelectrolytes on silicon using layer-by-layer assembly. Cent. Eur. J. Chem. 2013, 11, 205−214. (7) Priya, A. J.; Vijayalakshmi, S. P.; Raichui, A. M. Enhanced Survival of Probiotic Lactobacillus acidophilus by Encapsulation with Nanostructured Polyelectrolyte Layers through Layer-by-Layer Approach. J. Agric. Food Chem. 2011, 59, 11838−11845. (8) Ma, Y. D.; Dong, J. L.; Bhattacharjee, S.; Wijeratne, S.; Bruening, M. L.; Baker, G. L. Increased Protein Sorption in Poly(acrylic acid)Containing Films through Incorporation of Comb-Like Polymers and Film Adsorption at Low pH and High Ionic Strength. Langmuir 2013, 29, 2946−2954. (9) Barrantes, A.; Santos, O.; Sotres, J.; Arnebrant, T. Influence of pH on the build-up of poly-L-lysine/heparin multilayers. J. Colloid Interface Sci. 2012, 388, 191−200. (10) Paloniemi, H.; Lukkarinen, M.; Aaritalo, T.; Areva, S.; Leiro, J.; Heinonen, M.; Haapakka, K.; Lukkari, J. Layer-by-layer electrostatic self-assembly of single-wall carbon nanotube polyelectrolytes. Langmuir 2006, 22, 74−83. (11) McAloney, R. A.; Sinyor, M.; Dudnik, V.; Goh, M. C. Atomic Force Microscopy Studies of Salt Effects on Polyelectrolyte Multilayer Film Morphology. Langmuir 2001, 17, 6655−6663. (12) Han, L. L.; Mao, Z. W.; Wuliyasu, H.; Wu, J. D.; Gong, X.; Yang, Y. G.; Gao, C. Y. Modulating the Structure and Properties of Poly(sodium 4-styrenesulfonate)/Poly(diallyldimethylammonium chloride) Multilayers with Concentrated Salt Solutions. Langmuir 2012, 28, 193−199. (13) Abdelkebir, K.; Gaudiere, F.; Morin-Grognet, S.; Coquerel, G.; Labat, B.; Atmani, H.; Ladam, G. Evidence of different growth regimes coexisting within biomimetic Layer-by-Layer films. Soft Matter 2011, 7, 9197−9205. (14) Guzman, E.; Ritacco, H.; Ortega, F.; Rubio, R. G. Evidence of the influence of adsorption kinetics on the internal reorganization of polyelectrolyte multilayers. Colloids Surf., A 2011, 384, 274−281. (15) Lavalle, P.; Picart, C.; Mutterer, J.; Gergely, C.; Reiss, H.; Voegel, J. C.; Senger, B.; Schaaf, P. Modeling the buildup of polyelectrolyte multilayer films having exponential growth. J. Phys. Chem. B 2004, 108, 635−648. (16) Porcel, C.; Lavalle, P.; Ball, V.; Decher, G.; Senger, B.; Voegel, J. C.; Schaaf, P. From exponential to linear growth in polyelectrolyte multilayers. Langmuir 2006, 22, 4376−4383. (17) Porcel, C.; Lavalle, P.; Decher, G.; Senger, B.; Voegel, J. C.; Schaaf, P. Influence of the polyelectrolyte molecular weight on exponentially growing multilayer films in the linear regime. Langmuir 2007, 23, 1898−1904. (18) Ladam, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F. In situ determination of the structural properties of initially deposited polyelectrolyte multilayers. Langmuir 2000, 16, 1249−1255. (19) von Klitzing, R. Internal structure of polyelectrolyte multilayer assemblies. Phys. Chem. Chem. Phys. 2006, 8, 5012−5033. (20) Xu, L.; Pristinski, D.; Zhuk, A.; Stoddart, C.; Ankner, J. F.; Sukhishvili, S. A. Linear versus Exponential Growth of Weak Polyelectrolyte Multilayers: Correlation with Polyelectrolyte Complexes. Macromolecules 2012, 45, 3892−3901. (21) Guzman, E.; Ritacco, H. A.; Ortega, F.; Rubio, R. G. Growth of Polyelectrolyte Layers Formed by Poly(4-styrenesulfonate sodium salt) and Two Different Polycations: New insights from Study of Adsorption Kinetics. J. Phys. Chem. C 2012, 116, 15474−15483. (22) Haynie, D. T.; Cho, E. H.; Waduge, P. ″In and Out Diffusion″ Hypothesis of Exponential Multilayer Film Buildup Revisited. Langmuir 2011, 27, 5700−5704. (23) Dubas, S. T.; Schlenoff, J. B. Factors controlling the growth of polyelectrolyte multilayers. Macromolecules 1999, 32, 8153−8160.
amount (Figure S7) and hence do not undergo polymer diffusion. Porcel et al.17 showed that for a hyaluronic acid/ poly(L-lysine) (HA/PLL) multilayer the diffusion of highmolecular-weight (but not low-molecular-weight) PLL was limited to the upper layer of the PEM film. In our case, when we used a low-molecular-weight PDADMAC (8.5 kDa), the dynamics of PDADMAC were no different from those of the higher-molecular-weight PDADMAC (150 kDa). Thus, the absence of diffusion in 150 kDa PDADMAC during LBL with 30 kDa NaPA is not due to the large size of PDADMAC. Rather, PDADMAC flattens onto the underlying NaPA layer during the LBL deposition because of its high charge density. Because linear growth and the polymer diffusion of NaPA are observed for the multilayer, polymer diffusion in PEM films does not necessarily indicate exponential growth.
■
CONCLUSIONS We demonstrate for the first time an ATR-IR method that simultaneously determines the dynamics of the mass adsorbed and polymer conformation during the LBL process. This work provides evidence that linear growth can occur in systems in which polymer diffusion occurs. For the NaPA/PDADMAC system (Mw = 30 and 150 kDa, respectively), we observe linear growth despite the slow adsorption and interdiffusion of NaPA into the underlying PDADMAC layer. In contrast, PDADMAC adsorbs rapidly and flattens on the underlying NaPA layer, showing no evidence of diffusion. However, the underlying NaPA layer responds to PDADMAC by rearranging and reducing the number of bonds to the underlying PDADMAC layer. In addition to rearrangement, there is diffusion into the underlying film by NaPA but not PDADMAC. Therefore, our ATR-IR method shows polymer diffusion occurring in a PEM film that shows linear growth. Our study shows that, contrary to the in/out diffusion model, polymer diffusion is not exclusive to exponential growth in PEM films.
■
ASSOCIATED CONTENT
S Supporting Information *
Mass (amount) of polymer adsorbed. Depth of penetration. Bound fraction of adsorbed polymers. Dynamics of LBL. LBL growth as a function of molecular weight. This material is available free of charge via the Internet at http://pubs.acs.org.
■ ■
AUTHOR INFORMATION
Notes
The authors declare no competing financial interest.
REFERENCES
(1) Decher, G. Fuzzy nanoassemblies: Toward layered polymeric multicomposites. Science 1997, 277, 1232−1237. (2) Yin, M. J.; Gu, B. B.; Zhao, Q. A.; Qian, J. W.; Zhang, A.; An, Q. F.; He, S. L. Highly sensitive and fast responsive fiber-optic modal interferometric pH sensor based on polyelectrolyte complex and polyelectrolyte self-assembled nanocoating. Anal. Bioanal. Chem. 2011, 399, 3623−3631. (3) Chen, W.; McCarthy, T. J. Layer-by-layer deposition: A tool for polymer surface modification. Macromolecules 1997, 30, 78−86. (4) Zhang, P.; Qian, J. W.; Yang, Y.; An, Q. F.; Liu, X. Q.; Gui, Z. L. Polyelectrolyte layer-by-layer self-assembly enhanced by electric field and their multilayer membranes for separating isopropanol-water mixtures. J. Membr. Sci. 2008, 320, 73−77. (5) El Haitami, A. E.; Martel, D.; Ball, V.; Nguyen, H. C.; Gonthier, E.; Labbe, P.; Voegel, J. C.; Schaaf, P.; Senger, B.; Boulmedais, F. Effect of the Supporting Electrolyte Anion on the Thickness of PSS/ 11702
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
(24) Salomaki, M.; Vinokurov, I. A.; Kankare, J. Effect of temperature on the buildup of polyelectrolyte multilayers. Langmuir 2005, 21, 11232−11240. (25) Tripp, C. P.; Hair, M. L. Kinetics of the adsorption of a polystyrene-poly(ethylene oxide) block copolymer on silica: A study of the time dependence in surface/segment interactions. Langmuir 1996, 12, 3952−3956. (26) Guzman, E.; Ritacco, H.; Rubio, J. E. F.; Rubio, R. G.; Ortega, F. Salt-induced changes in the growth of polyelectrolyte layers of poly(diallyl-dimethylammonium chloride) and poly(4-styrene sulfonate of sodium). Soft Matter 2009, 5, 2130−2142. (27) Li, H. Y.; Tripp, C. P. Interaction of sodium polyacrylate adsorbed on TiO2 with cationic and anionic surfactants. Langmuir 2004, 20, 10526−10533. (28) Ninness, B. J.; Bousfield, D. W.; Tripp, C. P. The importance of adsorbed cationic surfactant structure in dictating the subsequent interaction of anionic surfactants and polyelectrolytes with pigment surfaces. Colloids Surf., A 2002, 203, 21−36. (29) Harrick, N. J. Internal Reflection Spectroscopy; Harrick Scientific Corporation: New York, 1987; pp 42−44. (30) Ninness, B. J.; Bousfield, D. W.; Tripp, C. P. Formation of a thin TiO2 layer on the surfaces of silica and kaolin pigments through atomic layer deposition. Colloids Surf., A 2003, 214, 195−204. (31) Hiller, J.; Mendelsohn, J. D.; Rubner, M. F. Reversibly erasable nanoporous anti-reflection coatings from polyelectrolyte multilayers. Nat. Mater. 2002, 1, 59−63. (32) Kujawa, P.; Moraille, P.; Sanchez, J.; Badia, A.; Winnik, F. M. Effect of molecular weight on the exponential growth and morphology of hyaluronan/chitosan multilayers: A surface plasmon resonance spectroscopy and atomic force microscopy investigation. J. Am. Chem. Soc. 2005, 127, 9224−9234. (33) Sun, Y.-x.; Ren, K.-f.; Zhao, Y.-x.; Liu, X.-s.; Chang, G.-x.; Ji, J. Construction of Redox-Active Multilayer Film for Electrochemically Controlled Release. Langmuir 2013, 29, 11163−11168. (34) Ruths, J.; Essler, F.; Decher, G.; Riegler, H. Polyelectrolytes I: Polyanion/polycation multilayers at the air/monolayer/water interface as elements for quantitative polymer adsorption studies and preparation of hetero-superlattices on solid surfaces. Langmuir 2000, 16, 8871−8878. (35) Jones, F.; Farrow, J. B.; van Bronswijk, W. An infrared study of a polyacrylate flocculant adsorbed on hematite. Langmuir 1998, 14, 6512−6517.
11703
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Dynamics of Layer-by-Layer Growth of a Polyelectrolyte Multilayer Studied in Situ Using Attenuated Total Reflectance Infrared Spectroscopy Silas Owusu-Nkwantabisah, Madhira Gammana, and Carl P. Tripp* Department of Chemistry, and Laboratory for Surface Science and Technology, University of Maine, Orono, Maine 04469, United States S Supporting Information *
ABSTRACT: Attenuated total reflectance infrared spectroscopy (ATR-IR) was used to study the dynamic layer-by-layer (LBL) growth of a sodium polyacrylate (NaPA)/poly(diallydimethylammonium) chloride (PDADMAC) multilayer on TiO2 particles. Molecular weights (Mw) used were 30 and 60 kDa for NaPA and 8.5 and 150 kDa for PDADMAC. IR spectra were recorded in situ as a function of time and were used to obtain the dynamic mass adsorbed and bound fraction of the polymers during each deposition step. For 30 kDa NaPA layers, the dynamics of adsorption show an initial rapid rise in mass followed by a slow increase toward a plateau value upon LBL with 150 kDa PDADMAC. In contrast, the 60 kDa NaPA layers achieve a plateau quickly and do not show a slow increase toward a plateau. In the case of LBL with 150 kDa PDADMAC, the dynamics of the bound fraction of polymer per layer suggest that polymer diffusion and conformational rearrangement occur for the layers of 30 kDa NaPA but not for the 60 kDa NaPA layers. Furthermore, PDADMAC adsorption profiles show that there is no diffusion of the PDADMAC layers and that PDADMAC flattens onto the underlying layer. A linear growth in the mass adsorbed per layer was observed for 150 kDa PDADMAC with both molecular weights of NaPA. In the case of 8.5 kDa PDADMAC, smaller growth increments and the desorption of underlying layers were observed. This work demonstrates the use of ATR-IR in obtaining the dynamics of LBL multilayer formation. Furthermore, it provides an example in which polymer diffusion during LBL film formation does not lead to exponential growth.
■
diffusion of polymers within the film, which is known as the in/ out diffusion model.15 According to this model, polyanions from solution diffuse into the growing film to form a reservoir of free polymers. These free polymers subsequently diffuse out of the film to contribute to the complexation of the next layer of incoming polycations. Evidence to support this diffusion model has mainly come from CLSM in which a fluorescently tagged polymer layer was found to diffuse throughout the entire structure.15 Salomaki et al.24 also proposed that all LBL systems follow exponential buildup and become linear when diffusion is slower than the deposition time. Recently, other models involving island and dendritic growth have also been proposed to explain the exponential growth in films.22
INTRODUCTION Layer-by-layer (LBL) multilayer formation has become a facile and versatile method for preparing thin films on solid surfaces.1−8 A common LBL approach is the formation of a polyelectrolyte multilayer (PEM) by successively adsorbing polyanions and polycations onto a charged surface.1,3,7−10 The nature of PEM films has been analyzed using techniques such as atomic force microscopy (AFM),11 quartz crystal microbalance (QCM),9,12,13 X-ray reflectivity,14 ellipsometry,6 confocal laser scanning microscopy (CLSM),15−17 and the zeta potential.6,12,18 Generally, the PEM films show linear or exponential growth in thickness depending on several factors such as the nature of the polymers, the pH,9 and the ionic strength.19 Understanding the molecular structure and processes giving rise to linear and exponential films has been of interest in many studies.6,15,20−23 Exponential growth is usually attributed to the © 2014 American Chemical Society
Received: August 7, 2014 Revised: September 6, 2014 Published: September 9, 2014 11696
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
Figure 1. IR spectra for NaPA recorded at the end of depositing each polymer layer. The inset shows the 2500−3050 cm−1 region for spectra recorded at the end of the NaPA#1 (black trace) and PDADMAC#1 (red trace) cycles. The Mw values are 30 and 150 kDa for NaPA and PDADMAC, respectively.
smaller growth increments, and desorption of the underlying layer was observed during the third bilayer deposition. For 150 kDa PDADMAC, linear film growth occurred as well as rearrangement and polymer diffusion for the lower-molecularweight NaPA. However, PDADMAC (150 kDa) lies flat on the surface of the underlying layer during multilayer formation. Therefore, our work provides an example in which polymer rearrangement and diffusion are observed in a linear PEM film.
It is well known that the properties of polymers adsorbed on surfaces are often dictated by kinetic factors.25,26 In many cases, the final polymer conformation depends on the sample history. Kinetic traps can exist in which the polymer resides in metastable nonequilibrium conformations. Furthermore, polymer adsorption from solution typically requires anywhere from several minutes to a few hours to achieve a maximum in the amount adsorbed.23,26,27 This is due to the length of time required for the rearrangement of the polymer on the substrate, which is much longer than the typical incubation times used in LBL deposition. However, experimental studies on the dynamics of polymer adsorption in LBL deposition have been reported only recently. Guzman et al.14,21 measured the change in the adsorbed amount as a function of time using Xray reflectivity, the dissipative quartz crystal microbalance, and ellipsometry. They showed that interdiffusion occurs in linearly grown PEMs and thus polymer diffusion is not exclusive to exponential growth. Although this work showed the importance of the dynamics of mass adsorbed to the LBL film structure and properties, further insight into the mechanism of film growth in LBL processes would benefit from the development of methods that can measure both the dynamic change in the amount adsorbed and the dynamic changes in the polymer conformation. We report for the first time a method that measures the dynamic change in mass and the bound fraction of an LBL multilayer system comprising sodium polyacrylate (NaPA) and poly(diallyldimethylammonium) chloride (PDADMAC) of different molecular weights on a TiO2 surface. We selected these polymers because they are commonly used in LBL systems. Furthermore, the approach builds on a method we developed to follow the dynamic conformation of NaPA adsorbed on TiO2.27 In particular, LBL films were fabricated using either 30 or 60 kDa NaPA with 150 or 8.5 kDa PDADMAC. Experiments were conducted at pH 3.5 because the amount of NaPA adsorbed on TiO2 passes through a maximum value at this pH.27 The method is based on attenuated total reflectance infrared (ATR-IR) spectroscopy in which spectra are recorded as a function of time during each LBL deposition cycle. The 8.5 kDa PDADMAC results in
■
EXPERIMENTAL SECTION
Materials and Solution Preparation. A flow-through ATR cell was purchased from Harrick and used with a 45° ZnSe internal reflection element (IRE) of dimensions 50 × 10 × 2 mm3. A description of the flow-through ATR cell and its use in measuring polymer adsorption on TiO2 coated on the IRE is described elsewhere.27 All spectra were recorded using GRAMS AI software with an ABB-Bomem FTLA 2000 spectrometer. Typically, 100 scans at 8 cm−1 requiring about 2 min were coadded for each spectrum recorded. Fumed TiO2 powder (P25) was obtained from Degussa and had a BET (N2) surface area of 50 m2/g. The measured isoelectric point of the P25 powder was pH 6.5. Sodium polyacrylate (NaPA, Mw = 30 000 (30 kDa), Mw/Mn = 1.4) and poly(diallyldimethylammonium) chloride (PDADMAC, Mw = 100 000−200 000 (150 kDa), Mw/Mn = 1.6) were obtained from Sigma-Aldrich and used as received. NaPA with Mw = 60 000 (60 kDa) and PDADMAC with Mw = 8500 (8.5 kDa) were obtained from Polysciences. Solutions of NaPA (20 ppm) and PDADMAC (30 ppm) were prepared by adding known quantities of the polymer to deionized water (Milli-Q, 18.2 MΩ·cm). At these solution concentrations, bands due to NaPA or PDADMAC above a bare ZnSe IRE are not detected.27 Hence, all bands in the spectra are due to the adsorbed NaPA or PDADMAC on TiO2. The pH values of these solutions and deionized water were adjusted using either dilute hydrochloric acid (HCl) or sodium hydroxide (NaOH) solutions. TiO2 Coating of the IRE. The ZnSe IRE was coated with a layer of TiO2 using an established procedure.27,28 Specifically, the TiO2 powder (30 mg) was dispersed in 25 mL of methanol and ultrasonicated for 30 min. A 200 μL aliquot of the suspension was then evenly deposited on one side of the IRE using a pipet. Evaporation of the methanol under ambient conditions resulted in a uniform TiO2 film of about 500 nm in thickness on the IRE.28 This thickness is less than the 0.7 μm penetration depth of the ATR evanescent wave. Typically, the methanol evaporated within 10 min but the TiO2 film was left at ambient conditions for at least 1 h to 11697
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
ensure thorough drying. The TiO2-coated IRE was then mounted in the flow-through cell. Polymer Deposition Method. The flow-through cell containing the TiO2-coated IRE was initially flushed at a flow rate of 5.8 mL/min with deionized water adjusted to pH 3.5. A reference spectrum was recorded, followed by the recording of spectra (100% baseline) at various time intervals. During the initial flow period, the 100% baseline spectra usually showed changes in the amount of water probed by the IRE. This was primarily due to the removal of air bubbles from the cell cavity and associated tubing. The procedure for recording a reference spectrum, followed 15 min later by a 100% baseline spectrum, was repeated until no changes in a 100% baseline spectra were observed. This typically required about 30−45 min from the start of flushing the cell with water. During this initial flow of water and in subsequent additions of polymer solutions, no evidence of removal of the TiO2 coating was observed (no decrease in TiO2 bulk modes near 720 cm−1). Once this initial break-in period was established, a reference spectrum was recorded and used for the remainder of the experiment. A 20 ppm NaPA solution at pH 3.5 was then allowed to flow through the cell at 5.8 mL/min, and spectra were recorded as a function of time for approximately 3.5 h. This deposition cycle is referred to as NaPA#1. The cell was then flushed with pH 3.5 water for 5 min to remove excess NaPA from the cell cavity. Next, a 30 ppm PDADMAC solution at pH 3.5 was allowed to flow through the cell for 3.5 h at 5.8 mL/min, and spectra were recorded at specified time intervals. This cycle is referred to as PDADMAC#1. The cell was then again flushed with water at pH 3.5 for 5 min to remove excess PDADMAC from the tubing and cell cavity. No change in the IR spectrum was observed during this flushing step. The sequential flow of NaPA, rinsing, PDADMAC, and rinsing was repeated twice using the above procedures, denoted as NaPA#2, PDADMAC#2, NaPA#3, and PDADMAC#3, respectively. IR spectra were collected every 5 min during each deposition cycle. All experiments were repeated at least three times.
calculated using this band gave the same value as obtained using the band at 2930 cm−1. This confirms that the subtraction procedure to account for the contribution from NaPA to the intensity of the band at 2930 cm−1 was valid. It is noted that a peak near 1637 cm−1 of 0.002 to 0.01 absorbance units in intensity appeared in about 3−5% of the spectra recorded. This band is due to the deformation mode of water. The water deformation mode is typically on the order of 1 absorbance unit when a reference spectrum of a dry TiO2coated IRE is used. Thus, the appearance of a band at 1637 cm−1 from time to time corresponds to less than a 1% fluctuation in the amount of water passing through the cell. This band was removed from the spectra by subtraction using a prerecorded spectrum of water. It is also noted that the signalto-noise level in the 1640 cm−1 region is about 10 times lower than in the surrounding spectral region, reflecting the 90% reduction in signal due to adsorption by the bulk water in the cell. Amount of Polymer Adsorbed. The procedure to determine the amount of adsorbed NaPA from the intensity of the band at 1455 cm−1 is described in our earlier work,27 and details are provided for NaPA and PDADMAC in the Supporting Information section. In brief, a final spectrum was recorded at the end of the experiment while the cell was rinsed with DI water. The IRE was then removed and dried, and a transmission IR spectrum was measured through the coated IRE. The amount of NaPA adsorbed was determined from the intensity of the 1455 cm−1 band in the transmission spectrum. This value was used to calibrate the peak intensity in the final ATR spectrum. The same procedure was used to calibrate the intensity of the 2930 cm−1 band for PDADMAC and the 720 cm−1 peak for TiO2. Figure 2 is a plot of the ratio of the
■
RESULTS AND DISCUSSION IR Spectra of Adsorbed Polymers. Figure 1 shows the IR spectra recorded as a function of time at the end of each deposition cycle. A 3D plot of spectra showing the time evolution of these bands during the adsorption of NaPA#1 and PDADMAC#1 is provided in the Supporting Information section (Figure S1). The inset in this figure shows the region between 2500 and 3050 cm−1 for the last spectrum recorded in NaPA#1 and PDADMAC#1, respectively. The key bands of interest for NaPA occur at 1713 cm−1 for a CO stretching mode of COOH, 1545 and 1414 cm−1 for the asymmetric and symmetric stretching modes of COO−, and 1455 cm−1 for a CH2 bending mode. The CH2 band at 1455 cm−1 was used to determine the amount of NaPA. This band overlaps with the COO− mode at 1414 cm−1 and had a shoulder on the high-wavenumber side at 1473 cm−1 due to a C−H mode of PDADMAC. The region was curve fitted with a Gaussian−Lorentzian function using GRAMS AI software over the range of 1490 to 1370 cm−1 to remove the contributions from the COO− and PDADMAC bands. For PDADMAC, there was a band due to the C−H stretching mode at 2930 cm−1 (Figure 1 inset), and the intensity of this band was used to calculate the amount of PDADMAC. Bands in this region, due to NaPA, are weak in intensity (inset in Figure 1). In calculating the mass of PDADMAC, we subtracted the band intensity of the final NaPA#1 spectrum from the overall band intensity in this region. For PDADMAC#2 and PDADMAC#3 deposition cycles, a C−H band of PDADMAC at 1473 cm−1 was of sufficient intensity to calculate the amount of PDADMAC (for example, see PDADMAC#3 in Figure 1). The amount
Figure 2. Ratio of the intensity of the band at 1455 cm−1 recorded in transmission IR to ATR-IR as a function of the number of layers. The Mw values were 30 kDa for NaPA and 150 kDa for PDADMAC.
intensity of the band at 1455 cm−1 in the transmission spectrum to the final ATR spectrum. The data points shown in Figure 2 were obtained using separate LBL experiments that we stopped at the end of a different number of NaPA deposition cycles. (For example, in one experiment we stopped at the end of NaPA#2, and then in a separate experiment, we stopped at NaPA#3.) In each experiment, after the rinsing step with DI water, the IRE was removed and dried and a transmission spectrum was recorded. The band intensity measured in the transmission spectrum showed a linear increase in its value (Figure S3 in the Supporting Information). Thus, the nonlinearity observed in Figure 2 originated with the evanescent wave dependence on 11698
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
Figure 3. Amount of polymer (mg per m2 of TiO2 surface) adsorbed as a function of time for LBL using 30 kDa NaPA and 150 kDa PDADMAC.
Table 1. Final Amount of Polymer (mg per m2 of TiO2 Surface) per Layer for LBL Films Prepared Using 30 kDa NaPA and 150 kDa PDADMACa
intensity found in the ATR spectra. In ATR, the evanescent wave decays exponentially from the surface; therefore, the band intensity will decrease with distance away from the IRE. Figure 2 shows that the dependence on the evanescent wave was about 10% for the first three cycles, increasing to more than 25% in the fourth cycle. The calculated penetration depth at 1455 cm−1 was 0.7 μm (equation S1 in Supporting Information), exceeding the 0.5μm-thick TiO2 layer. Thus, at first glance, the evanescent wave dependence as a function of intensity would occur when the amount of polymer adsorbed is not uniform throughout the TiO2 film and increases in the upper section of the TiO2 film with each deposition cycle. However, several control experiments showed that the adsorption of the polymers occurs throughout the entire TiO2 layer. In one control experiment, LBL deposition was performed using TiO2 powder stirred in beakers containing solutions of NaPA or PDADMAC. The amount of polymer adsorbed in each cycle was the same as that measured at the end of each cycle in the ATR experiment. In a second control experiment, an IRE was coated with a TiO2 film that was 50% thinner than normally used. In this case, the mass of polymer adsorbed per cycle was 50% of the value obtained with the thicker TiO2 layer. This shows that the polymers are adsorbed throughout the entire TiO2 layer and that the effective surface area available on the film is the same as for a dispersed suspension of TiO2 particles in a beaker. Because the deposition of the polymer is uniform throughout the TiO2 film, the only possibility to account for the decreasing absorbance value with each cycle is a refractive index change in the rarer medium. It is noted that the refractive index of TiO2 is about 2.5, whereas the refractive indexes of NaPA and PDADMAC are 1.43 and 1.38, respectively. Coating the TiO2 with successive polymer layers would tune the refractive index to increasingly lower values for the TiO2/water layer and hence reduce the penetration depth and the index matching. A reduction in both of these factors led to a lower intensity of the bands in the ATR spectrum.29 The refractive index tuning of materials via the formation of thin films is well known.30−32 Thus, the curve in Figure 2 provided the correction value to be applied to the intensity of bands in calculating the adsorbed amount. Figure 3 is a plot of the mass of polymer adsorbed per layer as a function of contact time for LBL using 30 kDa NaPA and 150 kDa PDADMAC. Table 1 gives the average amount of polymer at the end of each deposition cycle. There are clear differences in the amounts obtained for the first layer,
amount/mg·m−2
layer NaPA#1 PDADMAC#1 NaPA#2 PDADMAC#2 NaPA#3 PDADMAC#3
7.8 2.5 6.5 3.9 6.3 3.9
± ± ± ± ± ±
0.4 0.3 0.5 0.4 0.6 0.5
a
The errors are at the 95% confidence level based on a minimum of three measurements.
compared to those for the second and third layers. The total amount adsorbed for NaPA#1 is about 20% higher than for NaPA#2 and NaPA#3. However, the adsorbed amount of PDADMAC#1 is 40% lower than for the succeeding PDADMAC layers. Clearly, the adsorption of NAPA#1 and PDADMAC#1 represent a transition layer. This transition is well known to occur in LBL-based depositions, but typically the transition occurs over three to four layers.6,10,17,33 In our experiment, the transition occurs in the first cycle. This shorter transition is likely due to the long incubation time (3.5 h). Typically, the longest incubation times in LBL depositions are 10 to 30 min per layer.11,17 Figure 4 shows the cumulative amount of adsorbed polymer at the end of each deposition and is the most common way of plotting the LBL growth of PEM films.5,9,11,34 The plot (Figure
Figure 4. Cumulative amount of polymer (mg per m2 of TiO2 surface) deposited for each layer for LBL using 30 kDa NaPA and 150 kDa PDADMAC. 11699
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
4) follows a typical sawtooth pattern5,10 and a linear growth regime with a calculated R2 value of 0.98. Dynamics of Polymer Adsorption. Figure 3 also shows that the dynamics for NaPA and PDADMAC adsorption follow different profiles. For NaPA, the curves have a sharp initial rise followed by a slow increase in amount that tends toward a plateau value. In contrast, the curves due to PDADMAC adsorption reach a plateau rapidly (PDADMAC#1 curve rises rapidly to a plateau value within 20 min). Revisiting Figure 1, we see that the addition of PDADMAC#1 leads to a decrease in the COOH band at 1713 cm−1, which is mirrored by an increase in intensity of the COO− bands at 1545 and 1414 cm−1. There is no loss of NaPA because the band at 1455 cm−1 remains constant in intensity. Li and Tripp27 have shown that the change in intensity of the COOH and COO− bands for NaPA adsorbed on TiO2 can be used to measure the dynamic bound fraction of charged sites (COO− bound) to charged sites on the TiO2. We have extended this approach to the alternating addition of PDADMAC and NaPA. Details are provided in the Supporting Information section. The decrease in the COOH band upon exposure of NaPA#1 to PDADMAC#1 shown in Figure 1 can be attributed to two sources. First, it is primarily due to the adsorption of the positively charged sites on PDADMAC with the COO− groups located in the loops and tails of NaPA#1. Second, the rearrangement of the polymer molecules in NaPA#1 that occurs with PDADMAC deposition could lead to a change in the fraction of NaPA bound to the underlying TiO2 substrate. The combination of the two effects leads to a decrease in the total number of COOH groups because the ratio of COOH/ COO− groups in the loops and tails remains constant at their solution equilibrium values. Figure 5 is a plot of the total bound %COO− recorded as a function of time during the deposition of each layer in the LBL
rearrangement of NaPA on top of the PDADMAC layer with little, if any, diffusion of the NaPA into the PDADMAC layer. However, there are differences in the trend in the total bound fraction for NaPA#1 and other deposition cycles (Figure 5). The value for the bound fraction reaches a constant value of 48% within 40 min for NaPA#1, whereas in all other NaPA cycles the bound fraction continues to vary, albeit slowly at the end of the 3.5 h incubation. Therefore, NaPA diffuses into the underlying PDADMAC layer. Furthermore, while NaPA#1 passes through a maximum in bound fraction at 80% before reaching a 48% constant value, NaPA#2 and NaPA#3 do not pass through such a maximum (Figure 5). This suggests that NaPA#2 and NaPA#3 do not flatten on the underlying PDADMAC layers. The total bound fraction at the end of NaPA#2 and NAPA#3 are lower in value (about 32 and 28%, respectively) than the value obtained at the end of NaPA#1 (48%). This shows that the number of bonds between PDADMAC and NaPA is smaller than that for NaPA with TiO2. Furthermore, there is a stronger interaction between the COO− groups on NaPA and the positively charged sites on TiO2 than for the positively charged sites on PDADMAC. The difference in wavenumbers (Δν = νas − νs) between the asymmetric (νas) and symmetric (νs) stretching frequencies of COO− bands for NaPA#1 was Δν = 132 cm−1, whereas for NaPA#2 and NaPA#3 we obtain Δν values of 134 and 142 cm−1, respectively. The Δν values for NaPA#2 and NaPA#3 are composite values for the combined NaPA in the PEM film. It is noted that for NaPA#2 the Δν represents the overall value for both NaPA#1 and NaPA#2. Therefore, the similarity of Δν for NaPA#2 and NaPA#1 is indicative of the higher percent contribution of NaPA#1. Nevertheless, this shows that the Δν obtained for NaPA adsorbing with TiO2 sites is smaller than when NaPA binds with charged sites on PDADMAC. A smaller Δν implies a stronger interaction with the COO− functionalities.35 Thus, the combination of a higher bound fraction and stronger interaction shows that NaPA#1 is more strongly bound to TiO2 than to the PDADMAC layers. Bound Fraction of Individual Layers. The deposition of PDADMAC onto the underlying NaPA layer always increases the bound fraction of NaPA (Figure 5). From the knowledge of the amount of PDADMAC adsorbed and assuming intrinsic compensation involving a 1:1 interaction between the positively charged sites on PDADMAC and the negatively charged sites on NaPA, an estimate of the bound fraction of PDADMAC was determined. (Details of this calculation are provided in the Supporting Information section.) Figure 6 displays the bound fraction for each layer. Here, the bound fraction of PDADMAC refers to the fraction of NR4+ moieties binding to the COO− groups of NaPA. In generating these values, it was assumed that the number of groups already bound to the underlying layer does not change in value. In other words, the adsorption of PDADMAC#1 onto NaPA#1 does not result in any change in the fraction of NaPA#1 bound to the underlying TiO2. This is a reasonable assumption for PDADMAC#1 given the strong interaction between COO− groups and the charged sites on TiO2. From Figure 5, the addition of PDADMAC#1 leads to a 12% increase in the COO− bound fraction. Using the mass of adsorbed PDADMAC, we calculate that these COO− groups bind with 68% of the charged sites on the PDADMAC. A value of 68% bound fraction could suggest a high level of interpenetration of PDADMAC into the underlying NaPA#1
Figure 5. Total bound fraction of NaPA as a function of time for LBL using 30 kDa NaPA and 150 kDa NaPA.
process. The NaPA#1 layer forms on TiO2 with an initial bound fraction greater than 80% that drops to a plateau value of 48% at 40 min. This is because the initial NaPA in solution arriving at a bare TiO2 flattens onto the surface. The adsorbed NaPA subsequently rearranges to a conformation containing more loops and tails to accommodate more polymer adsorption on TiO2.27 The curves obtained for the dynamic amount adsorbed in NaPA#2 and NaPA#3 (Figure 3) are similar in shape to that of NaPA#1. This could be evidence that there is a slow 11700
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
ment of the adsorbed NaPA to accommodate more NaPA, as shown by the slow increase in adsorption toward a plateau during NaPA#1 deposition (Figure 3) and the drop in value for the bound fraction with time. Figure 6 shows that there is a general trend of a decrease in the bound fraction for each NaPA layer, and in fact the bound fraction for NaPA#3 is negative in value. While a decrease in the bound fraction may imply an increase in the coil-like conformation of the adsorbed layer,27 a negative bound fraction could indicate the loss of PDADMAC from the underlying layer to the solution phase. However, the IR spectra show no loss of PDADMAC during any of the NAPA cycles. Furthermore, the mass adsorbed for NaPA#3 was 6.3 mg·m−2 (Table 1), and this layer must have some positive value for the bound fraction to the underlying PDADMAC layer. Thus, a negative bound fraction means that the assumption that the bound fraction of the underlying layer does not change is invalid for cycle #2 and higher. This shows that the adsorption of PDADMAC leads to a reduction of charged sites on NaPA bound to the underlying PDADMAC layer. Hence, the adsorption of the next PDADMAC layer leads to a rearrangement of the entire underlying NaPA layer. This is also consistent with the slow change in the bound fraction for NaPA#2 and NAPA#3, which was attributed to the interdiffusion of NAPA into the underlying layer. LBL Using NaPA/PDADMAC of 60/150, 60/8.5, and 30/ 8.5 kDa Mw Ratios. The dynamics of LBL using 60 kDa NaPA and 150 kDa (and separately with 8.5 kDa) PDADMAC are provided in the Supporting Information section (Figures S3 and S4, respectively). The dynamics of LBL using 30 kDa NaPA and 8.5 kDa PDADMAC are provided in Figure S6. Compared to the dynamics of LBL using 30 kDa NaPA (Figure 3), the 60 kDa NaPA layers rise quickly initially and do not show a slow continual increase in mass. Thus, the 60 kDa NaPA layers do not show polymer diffusion into the growing PEM film. The plot of cumulative amounts of adsorbed polymer shows that the PDADMAC/NaPA multilayer grows linearly independent of the molecular weight of NaPA (Figure S7a in the Supporting Information section). Therefore, 60 kDa NaPA provides an example in which there is no polymer diffusion in a PEM film that shows linear LBL growth. In the case of 8.5 kDa PDADMAC (Figure S7b), there is a decrease in bilayer increments in the amount of adsorbed polymer. The LBL bilayer increments for 60 kDa NaPA are 14 and 3 mg·m−2 with 150 and 8.5 kDa PDADMAC, respectively. This is consistent with a report by Kujawa et al.32 on the molecular weight dependence of the hyaluronic acid/chitosan (HA/CH) multilayer system. Although the film growth rate did not change for the molecular weights (Mw) studied, Kujawa et al. indicated that an increase in Mw of HA or CH resulted in thicker films. In the case of 30 kDa NaPA and 8.5 kDa PDADMAC, the dynamic amount of NaPA layers continues to increase, never reaching steady state during the entire 3.5 h incubation time for each layer. This is consistent with a slow diffusion of NaPA. There is almost no growth during the deposition of NaPA#3. This is likely due to few binding sites of the 8.5 kDa PDADMAC resulting in fewer electrostatic attraction sites with NaPA during LBL. The NaPA#3 that is further extended from the TiO2 surface can easily be washed into the solution phase, thus preventing NaPA#3 growth. As was observed for 150 kDa PDADMAC (Figure 3), the 8.5 kDa PDADMAC layers rapidly reach a plateau in the adsorbed
Figure 6. Fraction of individual layers bound to the respective underlying layer. Data points refer to the fraction of COO− in NaPA or NR4+ in PDADMAC from spectra at the end of deposition for each layer. The Mw values were 30 kDa for NaPA and 150 kDa for PDADMAC.
layer. However, this is unlikely given the rapid plateau observed in the dynamic amount for PDADMAC#1 in Figure 3 and the rapid plateau in the dynamic bound fraction shown in Figure 5. Thus, a value of 68% bound fraction indicates that PDADMAC lies flat on top of the NaPA layer (Scheme 1). Decher1 Scheme 1. Molecular Picture of NaPA#1 (Mw = 30 kDa) and PDADMAC#1 (Mw = 150 kDa) Adsorbed on TiO2
described that the self-regulation of LBL is achieved because when charge reversal occurs, the incoming equally charged molecules are repelled from the growing film surface. Therefore, because each segment of PDADMAC carries a charge, the incoming molecules will be repelled upon approaching the initial flatly arranged PDADMAC molecules in the layer. As a result, the flat PDADMAC layer does not rearrange to accommodate the adsorption of more PDADMAC. Figure 3 shows that PDADMAC#2 and PDADMAC#3 have similar profiles in that both show a rapid plateau in the adsorbed amount. Guzman et al.14 have studied the kinetics of formation of poly(allylamine hydrochloride)/poly(sodium 4styrenesulfonate) (PAH/PSS) and PDADMAC/PSS multilayers as a function of ionic strength. For PDADMAC adsorption at low ionic strength, adsorption was rapid, whereas longer times were necessary to reach steady state at high ionic strength. It was concluded that there was limited interdiffusion for the PDADMAC/PSS multilayer at low ionic strength. This is consistent with our results in DI water where PDADMAC is the polycation. Thus, we conclude that the diffusion of PDADMAC into the underlying NaPA layer does not occur to any significant extent. In contrast to PDADMAC#1, there is a rearrangement of the initial flat layer of NaPA on the TiO2. In this case, there will be minimal repulsion of incoming NaPA molecules because 25% of the segments are charged, much lower than the 100% charged segments on PDADMAC. Hence, there is a rearrange11701
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
PAH Multilayer Films and on Their Permeability to an Electroactive Probe. Langmuir 2009, 25, 2282−2289. (6) Bragaru, A.; Kusko, M.; Radoi, A.; Danila, M.; Simion, M.; Craciunoiu, F.; Pascu, R.; Mihalache, I.; Ignat, T. Microstructures and growth characteristics of polyelectrolytes on silicon using layer-by-layer assembly. Cent. Eur. J. Chem. 2013, 11, 205−214. (7) Priya, A. J.; Vijayalakshmi, S. P.; Raichui, A. M. Enhanced Survival of Probiotic Lactobacillus acidophilus by Encapsulation with Nanostructured Polyelectrolyte Layers through Layer-by-Layer Approach. J. Agric. Food Chem. 2011, 59, 11838−11845. (8) Ma, Y. D.; Dong, J. L.; Bhattacharjee, S.; Wijeratne, S.; Bruening, M. L.; Baker, G. L. Increased Protein Sorption in Poly(acrylic acid)Containing Films through Incorporation of Comb-Like Polymers and Film Adsorption at Low pH and High Ionic Strength. Langmuir 2013, 29, 2946−2954. (9) Barrantes, A.; Santos, O.; Sotres, J.; Arnebrant, T. Influence of pH on the build-up of poly-L-lysine/heparin multilayers. J. Colloid Interface Sci. 2012, 388, 191−200. (10) Paloniemi, H.; Lukkarinen, M.; Aaritalo, T.; Areva, S.; Leiro, J.; Heinonen, M.; Haapakka, K.; Lukkari, J. Layer-by-layer electrostatic self-assembly of single-wall carbon nanotube polyelectrolytes. Langmuir 2006, 22, 74−83. (11) McAloney, R. A.; Sinyor, M.; Dudnik, V.; Goh, M. C. Atomic Force Microscopy Studies of Salt Effects on Polyelectrolyte Multilayer Film Morphology. Langmuir 2001, 17, 6655−6663. (12) Han, L. L.; Mao, Z. W.; Wuliyasu, H.; Wu, J. D.; Gong, X.; Yang, Y. G.; Gao, C. Y. Modulating the Structure and Properties of Poly(sodium 4-styrenesulfonate)/Poly(diallyldimethylammonium chloride) Multilayers with Concentrated Salt Solutions. Langmuir 2012, 28, 193−199. (13) Abdelkebir, K.; Gaudiere, F.; Morin-Grognet, S.; Coquerel, G.; Labat, B.; Atmani, H.; Ladam, G. Evidence of different growth regimes coexisting within biomimetic Layer-by-Layer films. Soft Matter 2011, 7, 9197−9205. (14) Guzman, E.; Ritacco, H.; Ortega, F.; Rubio, R. G. Evidence of the influence of adsorption kinetics on the internal reorganization of polyelectrolyte multilayers. Colloids Surf., A 2011, 384, 274−281. (15) Lavalle, P.; Picart, C.; Mutterer, J.; Gergely, C.; Reiss, H.; Voegel, J. C.; Senger, B.; Schaaf, P. Modeling the buildup of polyelectrolyte multilayer films having exponential growth. J. Phys. Chem. B 2004, 108, 635−648. (16) Porcel, C.; Lavalle, P.; Ball, V.; Decher, G.; Senger, B.; Voegel, J. C.; Schaaf, P. From exponential to linear growth in polyelectrolyte multilayers. Langmuir 2006, 22, 4376−4383. (17) Porcel, C.; Lavalle, P.; Decher, G.; Senger, B.; Voegel, J. C.; Schaaf, P. Influence of the polyelectrolyte molecular weight on exponentially growing multilayer films in the linear regime. Langmuir 2007, 23, 1898−1904. (18) Ladam, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F. In situ determination of the structural properties of initially deposited polyelectrolyte multilayers. Langmuir 2000, 16, 1249−1255. (19) von Klitzing, R. Internal structure of polyelectrolyte multilayer assemblies. Phys. Chem. Chem. Phys. 2006, 8, 5012−5033. (20) Xu, L.; Pristinski, D.; Zhuk, A.; Stoddart, C.; Ankner, J. F.; Sukhishvili, S. A. Linear versus Exponential Growth of Weak Polyelectrolyte Multilayers: Correlation with Polyelectrolyte Complexes. Macromolecules 2012, 45, 3892−3901. (21) Guzman, E.; Ritacco, H. A.; Ortega, F.; Rubio, R. G. Growth of Polyelectrolyte Layers Formed by Poly(4-styrenesulfonate sodium salt) and Two Different Polycations: New insights from Study of Adsorption Kinetics. J. Phys. Chem. C 2012, 116, 15474−15483. (22) Haynie, D. T.; Cho, E. H.; Waduge, P. ″In and Out Diffusion″ Hypothesis of Exponential Multilayer Film Buildup Revisited. Langmuir 2011, 27, 5700−5704. (23) Dubas, S. T.; Schlenoff, J. B. Factors controlling the growth of polyelectrolyte multilayers. Macromolecules 1999, 32, 8153−8160.
amount (Figure S7) and hence do not undergo polymer diffusion. Porcel et al.17 showed that for a hyaluronic acid/ poly(L-lysine) (HA/PLL) multilayer the diffusion of highmolecular-weight (but not low-molecular-weight) PLL was limited to the upper layer of the PEM film. In our case, when we used a low-molecular-weight PDADMAC (8.5 kDa), the dynamics of PDADMAC were no different from those of the higher-molecular-weight PDADMAC (150 kDa). Thus, the absence of diffusion in 150 kDa PDADMAC during LBL with 30 kDa NaPA is not due to the large size of PDADMAC. Rather, PDADMAC flattens onto the underlying NaPA layer during the LBL deposition because of its high charge density. Because linear growth and the polymer diffusion of NaPA are observed for the multilayer, polymer diffusion in PEM films does not necessarily indicate exponential growth.
■
CONCLUSIONS We demonstrate for the first time an ATR-IR method that simultaneously determines the dynamics of the mass adsorbed and polymer conformation during the LBL process. This work provides evidence that linear growth can occur in systems in which polymer diffusion occurs. For the NaPA/PDADMAC system (Mw = 30 and 150 kDa, respectively), we observe linear growth despite the slow adsorption and interdiffusion of NaPA into the underlying PDADMAC layer. In contrast, PDADMAC adsorbs rapidly and flattens on the underlying NaPA layer, showing no evidence of diffusion. However, the underlying NaPA layer responds to PDADMAC by rearranging and reducing the number of bonds to the underlying PDADMAC layer. In addition to rearrangement, there is diffusion into the underlying film by NaPA but not PDADMAC. Therefore, our ATR-IR method shows polymer diffusion occurring in a PEM film that shows linear growth. Our study shows that, contrary to the in/out diffusion model, polymer diffusion is not exclusive to exponential growth in PEM films.
■
ASSOCIATED CONTENT
S Supporting Information *
Mass (amount) of polymer adsorbed. Depth of penetration. Bound fraction of adsorbed polymers. Dynamics of LBL. LBL growth as a function of molecular weight. This material is available free of charge via the Internet at http://pubs.acs.org.
■ ■
AUTHOR INFORMATION
Notes
The authors declare no competing financial interest.
REFERENCES
(1) Decher, G. Fuzzy nanoassemblies: Toward layered polymeric multicomposites. Science 1997, 277, 1232−1237. (2) Yin, M. J.; Gu, B. B.; Zhao, Q. A.; Qian, J. W.; Zhang, A.; An, Q. F.; He, S. L. Highly sensitive and fast responsive fiber-optic modal interferometric pH sensor based on polyelectrolyte complex and polyelectrolyte self-assembled nanocoating. Anal. Bioanal. Chem. 2011, 399, 3623−3631. (3) Chen, W.; McCarthy, T. J. Layer-by-layer deposition: A tool for polymer surface modification. Macromolecules 1997, 30, 78−86. (4) Zhang, P.; Qian, J. W.; Yang, Y.; An, Q. F.; Liu, X. Q.; Gui, Z. L. Polyelectrolyte layer-by-layer self-assembly enhanced by electric field and their multilayer membranes for separating isopropanol-water mixtures. J. Membr. Sci. 2008, 320, 73−77. (5) El Haitami, A. E.; Martel, D.; Ball, V.; Nguyen, H. C.; Gonthier, E.; Labbe, P.; Voegel, J. C.; Schaaf, P.; Senger, B.; Boulmedais, F. Effect of the Supporting Electrolyte Anion on the Thickness of PSS/ 11702
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
Langmuir
Article
(24) Salomaki, M.; Vinokurov, I. A.; Kankare, J. Effect of temperature on the buildup of polyelectrolyte multilayers. Langmuir 2005, 21, 11232−11240. (25) Tripp, C. P.; Hair, M. L. Kinetics of the adsorption of a polystyrene-poly(ethylene oxide) block copolymer on silica: A study of the time dependence in surface/segment interactions. Langmuir 1996, 12, 3952−3956. (26) Guzman, E.; Ritacco, H.; Rubio, J. E. F.; Rubio, R. G.; Ortega, F. Salt-induced changes in the growth of polyelectrolyte layers of poly(diallyl-dimethylammonium chloride) and poly(4-styrene sulfonate of sodium). Soft Matter 2009, 5, 2130−2142. (27) Li, H. Y.; Tripp, C. P. Interaction of sodium polyacrylate adsorbed on TiO2 with cationic and anionic surfactants. Langmuir 2004, 20, 10526−10533. (28) Ninness, B. J.; Bousfield, D. W.; Tripp, C. P. The importance of adsorbed cationic surfactant structure in dictating the subsequent interaction of anionic surfactants and polyelectrolytes with pigment surfaces. Colloids Surf., A 2002, 203, 21−36. (29) Harrick, N. J. Internal Reflection Spectroscopy; Harrick Scientific Corporation: New York, 1987; pp 42−44. (30) Ninness, B. J.; Bousfield, D. W.; Tripp, C. P. Formation of a thin TiO2 layer on the surfaces of silica and kaolin pigments through atomic layer deposition. Colloids Surf., A 2003, 214, 195−204. (31) Hiller, J.; Mendelsohn, J. D.; Rubner, M. F. Reversibly erasable nanoporous anti-reflection coatings from polyelectrolyte multilayers. Nat. Mater. 2002, 1, 59−63. (32) Kujawa, P.; Moraille, P.; Sanchez, J.; Badia, A.; Winnik, F. M. Effect of molecular weight on the exponential growth and morphology of hyaluronan/chitosan multilayers: A surface plasmon resonance spectroscopy and atomic force microscopy investigation. J. Am. Chem. Soc. 2005, 127, 9224−9234. (33) Sun, Y.-x.; Ren, K.-f.; Zhao, Y.-x.; Liu, X.-s.; Chang, G.-x.; Ji, J. Construction of Redox-Active Multilayer Film for Electrochemically Controlled Release. Langmuir 2013, 29, 11163−11168. (34) Ruths, J.; Essler, F.; Decher, G.; Riegler, H. Polyelectrolytes I: Polyanion/polycation multilayers at the air/monolayer/water interface as elements for quantitative polymer adsorption studies and preparation of hetero-superlattices on solid surfaces. Langmuir 2000, 16, 8871−8878. (35) Jones, F.; Farrow, J. B.; van Bronswijk, W. An infrared study of a polyacrylate flocculant adsorbed on hematite. Langmuir 1998, 14, 6512−6517.
11703
dx.doi.org/10.1021/la5031558 | Langmuir 2014, 30, 11696−11703
