Templated Surfactant Readsorption on Polyelectrolyte-Induced

Laboratory for Surface Science and Technology, University of Maine, Orono, Maine 04469. Received August 27, 2004. In Final Form: December 20, 2004...
0 downloads 0 Views 75KB Size
2876

Langmuir 2005, 21, 2876-2880

Templated Surfactant Readsorption on Polyelectrolyte-Induced Depleted Surfactant Surfaces Jason S. Poirier,† Carl P. Tripp,‡ and David J. Neivandt*,† Department of Chemical and Biological Engineering, Department of Chemistry and Laboratory for Surface Science and Technology, University of Maine, Orono, Maine 04469 Received August 27, 2004. In Final Form: December 20, 2004 Changes in the structure of a surfactant adsorbed on oxidized silicon arising from interaction with a polyelectrolyte have been studied using polarized infrared attenuated total reflection spectroscopy. Specifically, the cationic surfactant cetyltrimethylammonium bromide (CTAB) was found to form a highly ordered layer on oxidized silicon at a concentration of 5.5 × 10-5 M and a pH of 9.6. Addition of a solution of the anionic polyelectrolyte poly(styrenesulfonate) to the ordered CTAB layer resulted in a rapid and dramatic decrease in the surface excess of CTAB. Interestingly however, the interfacial order of the residual surfactant was retained for a time period greater than 1 h, before decreasing. Reintroduction of a surfactant solution prior to destabilization of the residual interfacial CTAB resulted in the readsorption of the surfactant, the recovery of the initial equilibrium coverage, and the maintenance of an ordered CTAB conformation. This desorption/readsorption process may be subsequently repeated without destroying the order of the CTAB on the surface. If however sufficient time is allowed for the residual interfacial surfactant to destabilize prior to readdition of CTAB, the degree of surfactant order remains low, despite the rapid reobtainment of a surface excess equal to or greater than that initially measured. These results are interpreted in terms of polymer/surfactant interfacial complexation and the removal of adsorbed surfactant into solution. The ordering behavior of the residual surfactant suggests that CTAB is left on the surface in isolated patches of highly ordered species that maintain their order until two-dimensional diffusion leads to a more homogeneous surfactant surface distribution and hence the loss of conformational order. The degree of orientation order assumed by surfactant readsorbing on a depleted surface appears to be templated by the order of the residual interfacial surfactant, suggestive of a two-dimensional epitaxial growth mechanism for CTAB readsorption.

Introduction Polymers and surfactants are widely employed individually and in combination to modify the interfacial behavior of surfaces.1 Although a great deal of work has focused on the adsorption of each type of species individually at the solid/liquid interface,2,3 considerably less effort has been directed at studying coadsorption or sequential adsorption. Clearly, when considering adsorption in mixed systems of polymers and surfactants, interactions (whether attractive or repulsive) between the two entities, in addition to interactions between each entity and the surface, are critical. Indeed, a wide variety of systemdependent phenomena such as preferential, competitive, and synergistic adsorption and desorption have been observed.4-9 Desorption of one species by the addition of another without significant displacement due to prefer* To whom correspondence should be addressed. Phone: (207) 581-2288. Fax: (207) 581-2323. E-mail: dneivandt@umche. maine.edu. † Department of Chemical and Biological Engineering. ‡ Department of Chemistry and Laboratory for Surface Science and Technology. (1) Goddard, E. D. Colloids Surf. 1986, 19, 255. (2) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman & Hall: London, 1993. (3) Rosen, M. J. Surfactants and Interfacial Phenomena, 2nd ed.; Wiley: New York, 1989. (4) Velegol, S. B.; Tilton, R. D. J. Colloid Interface Sci. 2002, 2, 282. (5) Windsor, R.; Neivandt, D. J.; Davies, P. B. Langmuir 2001, 17, 7306. (6) Lauten, R. A.; Kjoniksen, A. L.; Nystrom, B. Langmuir 2000, 16, 4478. (7) Green, R. J.; Su, T. J.; Lu, J. R.; Webster, J. R. P. J. Phys. Chem. B 2001, 105, 9331. (8) Ninness, B. J.; Bousfield, D. W.; Tripp, C. P. Colloids Surf., A 2002, 203, 21. (9) Neivandt, D. J.; Gee, M. L.; Hair, M. L.; Tripp, C. P. J. Phys. Chem. B 1998, 102, 5107.

ential adsorption is perhaps the least investigated of these phenomena and one of the most interesting.6-9 Several workers have investigated desorption of adsorbed polymers induced by the addition of surfactants. For example, Lauten et al.6 studied the effect of the addition of both anionic and cationic surfactants (sodium dodecyl sulfate (SDS) and cetyltrimethylammonium bromide (CTAB), respectively) on the hydrodynamic thickness of ethyl(hydroxyethyl)cellulose (EHEC) adsorbed on anionic polystyrene latex particles. It was found that the addition of either SDS or CTAB resulted in desorption of the polymer due to the formation of a soluble polymer/ surfactant complex. Indeed, desorption was maximized at the respective critical aggregation concentrations of the two surfactant/EHEC systems. Green et al.7 investigated the surface excess of lysozyme adsorbed at the silica/solution interface as a function of the addition of varying concentrations of the cationic surfactant dodecyltrimethylammonium bromide (DOTAB). It was found that at low surfactant concentrations the protein was removed from the interface due to the formation of a highly soluble protein/surfactant complex with very little accompanying DOTAB interfacial adsorption. At higher surfactant concentrations, protein desorption was still observed; however, a more complex scenario of some polymer/surfactant complex adsorption and coadsorption of the surfactant was observed. Conversely, other workers have studied the removal of interfacial surfactant by the addition of polymeric species. For example, Ninness et al.8 showed that the addition of sodium polyacrylate (NaPA) to defective bilayers of a cationic surfactant (CTAB) on titania resulted in the effective transformation of the surfactant into a monolayer structure and indeed that it is the structure of the surfactant layer and not the adsorbed amount that

10.1021/la047861w CCC: $30.25 © 2005 American Chemical Society Published on Web 02/22/2005

Surfactact Readsorption on Depleted Surfaces

governed the interaction with anionic solution species. Previous work by the present authors9 demonstrated that the addition of an anionic polyelectrolyte, poly(styrenesulfonate) (PSS), to an equilibrium surface excess of cationic CTAB on a negatively charged oxidized silicon surface at high pH resulted in the desorption of the surfactant and the nonadsorption of the polyelectrolyte. This behavior was attributed to the interfacial formation of a PSS/CTAB complex10,11 which was subsequently removed from the surface via solubilization. In common with the work by Ninness et al., the latter study showed that the structure of the residual nonpolymer-complexed interfacial surfactant layer could be manipulated via the addition of the polymer, although the system employed in our earlier work displayed a time dependence for the structural changes not seen in the NaPA/CTAB/titania system. Specifically, using polarized infrared attenuated total reflection (IR-ATR), we showed that the residual interfacial surfactant (not complexed with PSS and removed into solution) maintained a high degree of orientational order, comparable to that observed at the initial equilibrium surface coverage, for several hours after the addition of PSS before gradually assuming a more random conformation at extended time periods. This behavior was attributed to complexation of interfacial CTAB with PSS in localized areas and hence the creation of patches or clusters of residual highly ordered CTAB. The present study probes the effect of the orientational order of the residual interfacial surfactant on that of CTAB readsorbed on the depleted surface. Experimental Section IR-ATR spectra were collected in a manner comparable to that outlined in ref 9. Briefly, a modified Bomem Michelson 110 FT-IR spectrometer12 and a twin parallel mirror reflection attachment from Harrick Scientific were employed. Two KRS-5 wire grid polarizers were mounted in two positions of a threeposition slide mechanism placed immediately in front of the MCT detector of the spectrometer. The polarizers were set to the s and p polarization orientations (with respect to the plane of incidence of the internal reflection element (IRE)), and polarization of the beam was achieved by sliding the appropriate polarizer into the optical path. The third slide position was employed to obtain nonpolarized spectra. A silicon trapezoidal IRE was mounted in a stainless steel flow-through cell and inserted into the reflection attachment. Solution was introduced to the cell from an external reservoir via a peristaltic pump and Teflon capillary tubing. Spectra were obtained via the coaddition of 100 scans (approximately 2 min) to provide a sufficient signal-to-noise ratio. All experiments were performed at an ambient temperature of 22 ( 2 °C. Conversion of absorbance to surface excess values was achieved via determination of the effective extinction coefficient of the relevant methylene mode and a waveguide calibration taking into account the number of internal reflections and the beam footprint. The native oxide layer on the IRE was employed as the adsorbate. Preparation of the IRE differed from that reported previously to remove the use of a potentially hazardous nitric acid reflux procedure.9 Specifically, the IRE was repeatedly wiped with a clean lint-free tissue soaked in CCl4, soaked in a Contrad 70 detergent solution for 24 h, rinsed copiously with water, soaked for 24 h in concentrated nitric acid, boiled in a 10 vol % concentrated NH3 in 30% H2O2 solution for 15 min, and finally copiously rinsed with water. The IRE was subsequently mounted in the IR-ATR flow cell, through which water flowed for 24 h prior to commencement of an experiment. The stainless steel cell and all glassware were soaked for 24 h in a Contrad 70 detergent solution, rinsed thoroughly with water, soaked for 24 (10) Hayakawa, K.; Kwak, J. C. T. J. Phys. Chem. 1982, 86, 3866. (11) Almgren, M.; Hansson, P.; Mukhtar, E.; van Stam, J. Langmuir 1992, 8, 2405. (12) Tripp, C. P.; Hair, M. L. Appl. Spectrosc. 1992, 46, 100.

Langmuir, Vol. 21, No. 7, 2005 2877 h in concentrated nitric acid, and finally rinsed thoroughly with water immediately prior to use. Teflon tubing and fittings were treated in a comparable manner with the exception of the nitric acid procedure. All water used for rinsing and sample preparation had a resistivity of 18.2 MΩ cm and a TOC of less than 5 ppb. Cetyltrimethylammonium bromide was obtained from SigmaAldrich and was doubly recrystallized from an acetone/ethanol mixture prior to use. Poly(styrenesulfonate) was obtained from Fluka and had an MP of 158000 and an MW of 152300. PSS was used as received. It has been shown previously9 that polarized IR-ATR data of CTAB adsorption on the oxide layer of a silicon IRE in water may be treated as a two-layer refractive index system (silicon/ water) in analyzing the relationship between the dichroic ratio (ratio of the absorbance in the s and p polarization states, As/Ap ratio) and the average orientation of the surfactant alkyl chain to the surface. Specifically, it was demonstrated that substitution of the refractive indices of silicon and water into the relevant electric field amplitude equations,13 followed by calculation of the predicted dichroic ratios, resulted in an As/Ap ratio of 0.50 for the methylene stretching modes of randomly distributed alkyl chains, and an As/Ap ratio of 1.22 for alkyl chains aligned normal to the surface. Although in principle it is possible to relate the measured dichroic ratio to a specific average tilt angle of an adsorbed surfactant molecule, in practice it requires a welldefined angle of incidence (not present for a noncollimated FTIR spectrometer) and a knowledge of the number of trans and gauche conformers of the alkyl chain.14 For this reason while changes in the dichroic ratio may be taken as an indication of the alteration of the average orientation of the surfactant to the surface, conversion to a specific angle of molecular orientation is avoided.

Results and Discussion CTAB Adsorption and Polymer-Induced Desorption. To study the readsorption of CTAB on a depleted surfactant surface, it was first necessary to create an ordered CTAB layer on the substrate and to induce surfactant removal through the addition of PSS. As demonstrated by Fleming et al.,15 the conformational order of CTAB adsorption on silica is highly pH and concentration dependent. Indeed, the critical surface aggregation concentration (CSAC) of CTAB at pH 9.6 was shown to occur at a significantly lower concentration than observed at neutral pH (∼5.5 × 10-5 M versus ∼4.5 × 10-4 M). This finding was attributed to the increased surface charge of the substrate at elevated pH and the passing of a threshold level of surfactant adsorption required for hydrophobic interactions to promote aggregation and adsorption beyond that required for charge neutralization. This observation is entirely consistent with the authors’ previous observation of ordered CTAB adsorption on oxidized silicon at pH 9.6 from 5.5 × 10-5 M CTAB solution. Consequently, identical solution conditions were employed in the present study to create an ordered surfactant layer. Specifically, a 5.5 × 10-5 M CTAB solution at pH 9.6 was introduced to the ATR flow cell containing the oxidized silicon IRE and allowed to equilibrate for 24 h. The surface excess was subsequently monitored for a period of 3 h and is plotted as open tilted squares, left ordinate, versus time in Figure 1. There is little if any change in the surface excess with time, indicating adsorption equilibrium of CTAB (likely present in its dissociated form) has been obtained, a finding consistent with previous studies by the authors and others.9,15 The dichroic ratio of the interfacial surfactant during the 3 h measuring period is (13) Harrick, N. J. Internal Reflection Spectroscopy; Interscience: New York, 1967. (14) Fringeli, U. P.; Gunthard, W. s. H. Infrared Membrane Spectroscopy. In Membrane Spectroscopy; Gell, E., Ed.; Springer: New York, 1981. (15) Fleming, B. D.; Biggs, S.; Wanless, E. J. J. Phys. Chem. B 2001, 105, 9537.

2878

Langmuir, Vol. 21, No. 7, 2005

Figure 1. Desorption of CTAB from an equilibrium surface excess due to the addition of 100 ppm PSS at a time of 3 h, indicated by the arrow. The surface excess (left ordinate, open tilted squares) and the As/Ap ratio (right ordinate, solid circles) are plotted as a function of time.

plotted as solid circles on the right ordinate of Figure 1 as a function of time. The As/Ap ratio shows little dependence on time, a finding consistent with the presence of an equilibrium interfacial surfactant conformation. This result concurs with the previous study in which the dichroic ratio was found to be essentially invariant at times greater than approximately 16 h after addition of the CTAB solution to the ATR flow cell. It is noted that a wide variety of dynamics are observed for the adsorption of CTAB and related surfactants on silica and oxidized silicon, with equilibration times ranging from seconds16,17 to hours/days.18,19 There has to date been no consensus reached in the literature as to the source of the variations between studies. The effect of the addition of 100 ppm PSS on the surface excess and dichroic ratio of the equilibrium surface excess of adsorbed CTAB is presented at times greater than 3 h in Figure 1. The experimental protocol employed was as follows: the CTAB solution in the reservoir was replaced with 18.2 MΩ cm water and the cell rinsed for 2 min to remove CTAB in solution from the system. The water in the reservoir was subsequently replaced with a 100 ppm PSS solution which was pumped through the cell commencing at the time indicated in Figure 1 by the arrow (3 h). It should be noted that separate experiments performed to monitor the rate of desorption of CTAB upon replacing the 5.5 × 10-5 M CTAB solution with 18.2 MΩ cm water showed a gradual loss of interfacial surfactant (20% over a period of 12 h) and no associated decrease in the As/Ap ratio (data not shown). The comparatively slow desorption of CTAB observed in water (in comparison to the very rapid removal of surfactant induced by the addition of PSS as seen in Figure 1, discussed below) is not surprising when, for example, the finding of Clark and Ducker16 that the rate of surface to solution exchange of C14TAB on an oxidized silicon surface in contact with a C14TAB solution was far faster than desorption of the surfactant in pure deuterated water (although both time scales were faster than observed for CTAB in the present study) is considered. In common with previous findings,9 the addition of 100 ppm PSS to the system resulted in extremely rapid desorption of CTAB from the interface, as indicated by (16) Clark, S. C.; Ducker, W. A. J. Phys. Chem B 2003, 107, 9011. (17) Atkin, R.; Craig, V. S. J.; Biggs, S. Langmuir 2000, 16, 9374. (18) Pagac, E. S.; Prieve, D. C.; Tilton R. D. Langmuir 1998, 14, 2333. (19) Biswas, S. C.; Chattoraj, D. K. J. Colloid Interface Sci. 1998, 205, 12.

Poirier et al.

the approximately 80% decrease in the surface excess over a period of 2 h (left ordinate, open tilted squares, of Figure 1). The corresponding dichroic ratio values however remained at their pre-PSS addition values for a period of approximately 1 h before starting to gradually monotonically decrease to a quasi plateau at a value of approximately 0.60, 31/2 h after addition of the polyelectrolyte (right ordinate, solid circles, of Figure 1). The rapid loss of interfacial CTAB is attributed to association of the surfactant with PSS and the removal of the PSS/CTAB complex from the surface into solution. Investigation of the aromatic C-H stretching region confirmed the absence of PSS at the interface. The maintenance of the dichroic ratio at the equilibrium value for a significant period of time upon the rapid removal of a large portion of the interfacial surfactant observed in Figure 1 and the previous study9 implies that the surfactant remaining on the surface maintained a high degree of order for an extended period before gradually relaxing into a less orientated state at low surface coverages. This behavior was previously attributed to the removal of CTAB by PSS in localized areas (consistent with the fact that the removal mechanism is via complexation with a macro-ion) and hence the existence of residual interfacial surfactant in isolated patches or clusters of highly ordered CTAB on a bare or sparsely populated surface. The surfactant molecules within such clusters would be expected to maintain a high degree of order since the local packing density is that of the initial equilibrium coverage. The gradual decay of the dichroic ratio was therefore attributed to relaxation of the surfactant orientation, a process likely coupled with twodimensional diffusion and an increase in the homogeneity of the surface, as per the mechanism subsequently described by Doudevski and Schwartz.20 It is important to note that the correlation between the CTAB adsorption and PSS-induced desorption data of the present study and that of the authors’ previous work9 suggests that the changes made in the oxide preparation procedure between the two studies did not have a dramatic effect on the adsorption characteristics of the system. The equilibrium surface excess and dichroic ratio values of the present work were however lower by factors of approximately 20% and 10%, respectively. Readsorption of CTAB. The nature of CTAB readsorption at the depleted CTAB/oxidized silicon interface was investigated via reintroduction of a CTAB solution to the ATR flow cell. Specifically, an equilibrium CTAB surface excess was created on a bare oxidized silicon IRE as described above, and 100 ppm PSS was added at the time indicated with an arrow in Figure 2 (employing the water rinsing procedure before addition of PSS) after a stable equilibrium baseline was recorded for approximately 40 min. The surfactant desorption process was monitored in terms of both the surface excess (open tilted squares, left ordinate) and the dichroic ratio (solid circles, right ordinate). As in Figure 1 the surface excess of CTAB was observed to decrease extremely rapidly, while no change in the dichroic ratio was observed. Once the surface excess of CTAB had dropped to approximately one-third of its equilibrium value, the cell was rinsed with water for 2 min, prior to the introduction of a 5.5 × 10-5 M, pH 9.6 CTAB solution (at the time marked with an arrow in Figure 2). The collection of nonpolarized IR-ATR spectra clearly showed that the surface excess of CTAB increased and in fact reobtained its initial equilibrium value within approximately 7 h. This is far faster than the time period (20) Doudevski, I.; Schwartz, D. K. Langmuir 2000, 16, 9381.

Surfactact Readsorption on Depleted Surfaces

Figure 2. PSS-induced desorption and subsequent readsorption of CTAB prior to the loss of conformational order. The respective addition times are indicated by the labeled arrows. The surface excess (left ordinate, open tilted squares) and the As/Ap ratio (right ordinate, solid circles) are plotted as a function of time.

taken for the same increase in surface excess in the original adsorption process commencing with a bare oxidized silicon surface of approximately 24 h. Polarized spectra of the CTAB readsorption process indicate that there is no change in the As/Ap ratio of the interfacial surfactant and that it remains at the equilibrium value of approximately 0.75. The results of Figure 2 clearly show that the initial equilibrium surface excess of CTAB may be regained if the surfactant is reintroduced to the depleted surface within a time frame in which the residual interfacial surfactant is highly ordered. This finding is consistent with the concept of readsorption of CTAB on a surfactantdepleted surface consisting of clusters of highly ordered CTAB on a sparsely populated oxidized silicon surface. Alkyl chain interactions would be expected to drive adsorption of surfactant at surface sites neighboring those already occupied, while electrostatic interactions would promote adsorption on unoccupied oxidized silicon sites. Such an adsorption mechanism would result in the adsorption of CTAB on the surface at the edges of residual surfactant clusters (whether initially or after twodimensional diffusion from a bare area to a cluster site). The clusters would consequently grow in two dimensions until equilibrium coverage was regained. Supporting evidence for this mechanism is found in the fact that no change in the dichroic ratio was observed upon surfactant readsorption. If CTAB were to adsorb on bare oxidized silicon in regions away from residual surfactant clusters, then it would be expected to do so with a degree of orientational order commensurate with its local surface density. That is, the dichroic ratio of the readsorbing CTAB would be expected to be on the order of 0.65 or less (the approximate value measured immediately upon introduction of CTAB to a bare oxidized silicon surface in the authors’ previous study and again in the present study, but not presented). Since the IRATR spectra are an average of those of all the interfacial CTAB, which consists of up to two-thirds readsorbed CTAB, the dichroic ratio would by necessity decrease if CTAB were to readsorb on bare oxidized silicon regions. Conversely, if CTAB readsorption occurs via a cooperative mechanism on the edges of highly ordered residual surfactant clusters, then readsorbing molecules would be expected to assume the same conformational order as their neighbors; consequently, no change in the measured dichroic ratio would be expected.

Langmuir, Vol. 21, No. 7, 2005 2879

Figure 3. Sequential desorption/readsorption cycles performed by the alternate addition of PSS and CTAB at the times indicated by the labeled arrows. The surface excess (left ordinate, open tilted squares) and the As/Ap ratio (right ordinate, solid circles) are plotted as a function of time.

A final piece of evidence in support of the proposed readsorption mechanism (which may be considered along the lines of two-dimensional epitaxial growth) is the fact that the reobtainment of the initial equilibrium coverage upon CTAB readsorption occurred in approximately onethird the time required to reach equilibrium coverage in the initial adsorption process. As stated earlier, it was demonstrated in the authors’ previous study that the increase in the measured dichroic ratio with the surface excess of CTAB in the initial adsorption process may be interpreted as reorientation of adsorbed surfactant in a direction more normal to the surface to facilitate further surfactant adsorption, and not initial nonordered adsorption followed by later highly orientated adsorption. In the readdition of CTAB procedure the surfactant on the surface is at its maximum orientational order, and consequently, no time-dependent reorientation is required to facilitate further surfactant adsorption. It would consequently be expected that a significantly shorter time period would be required to reach equilibrium coverage, as is in fact observed. Sequential Desorption/Readsorption Cycles. The data of Figure 2 imply that the adsorbed amount and order of the surfactant layer after the desorption/readsorption cycle are comparable to those after the initial adsorption process. As such it would be expected that the desorption/readsorption behavior should be repeatable through several cycles. Figure 3 presents the results of such an experiment. The data of Figure 2 (from 0 to approximately 7 h) are reproduced on the left of Figure 3. At 7 h and 20 min a 100 ppm PSS solution was introduced to the flow cell from the external reservoir following the rinsing protocol described above. In common with Figures 1 and 2, a very rapid decrease in the surface excess (open tilted squares, left ordinate) is observed, while the dichroic ratio is unaffected. Readdition of a 5.5 × 10-5 M CTAB solution at 7 h and 50 min, 30 min after PSS addition (again following the rinsing procedure), results in the readsorption of the surfactant and the regaining of the initial equilibrium surface excess value. No change in the dichroic ratio is observed upon the readdition of CTAB. Clearly, the data of Figure 3 indicate that the desorption/ readsorption cycle may be performed multiple times on the same substrate without major differences occurring in the behavior of the system. It is noted however that some subtle differences do exist between the first and second desorption/readsorption cycles. Specifically, ap-

2880

Langmuir, Vol. 21, No. 7, 2005

proximately 20% less surfactant was removed by PSS in the second cycle relative to the first, although the exposure time was the same (30 min). Second, the readsorption of CTAB leads to a quicker recovery of the initial equilibrium coverage. These findings suggest that the detailed structure of the second re-formed surfactant layer may differ somewhat from that initially re-formed. Alternatively, the differences may simply be representative of the variability in the amount of CTAB removed by PSS in the desorption process and a consequent difference in the readsorption process. The importance of the order of the residual interfacial CTAB in dictating the order of the readsorbing CTAB is shown in the final desorption/readsorption cycle of Figure 3, where surfactant readdition was not performed until the residual interfacial CTAB had been given sufficient time to disorder. At a time of 11 h a solution of 100 ppm PSS was introduced to the flow cell employing the standard rinsing procedure. The resulting surface excess is plotted as open tilted squares (left ordinate) at times from 11 to 15 h in Figure 3. Clearly, a very rapid desorption of CTAB is observed, with a loss of approximately 70% of the initial surface excess in a period of 2 h before a quasi plateau is reached. This behavior is comparable to that observed for the first two desorption cycles of Figure 3 and the analogous experiments presented in Figures 1 and 2. Likewise the dichroic ratio behavior (solid circles, right ordinate) mirrors that observed in Figure 1; that is, it remains stable for approximately 1 h (slightly longer in the latter experiment) before gradually decreasing to an equilibrium value of approximately 0.62. These findings suggest that the CTAB layer present on the oxidized silicon surface after multiple desorption/readsorption cycles is comparable in desorption behavior to that formed in the initial adsorption of CTAB on the bare substrate process. At a time of 15 h CTAB was reintroduced to the flow cell employing the rinsing procedure, and polarized and nonpolarized IR-ATR spectra were collected. The resulting surface excess (open tilted squares, left ordinate) and dichroic ratio (solid circles, right ordinate) are presented at times from 15 to 20 h in Figure 3. Clearly, a comparatively rapid increase in the surface excess of CTAB occurs upon its reintroduction to the system; indeed, it reaches the initial equilibrium value in approximately 5 h. This time period is much faster than that required for the adsorption process on a bare surface and is intermediate between the times required to regain the initial equilibrium value for the first and second desorption/ readsorption cycles. The dichroic ratio of the interfacial surfactant upon readdition of CTAB to the disordered surface is invariant with time and surface excess, maintaining a value of approximately 0.62. This finding is in stark contrast to the increasing As/Ap ratio observed for the adsorption

Poirier et al.

process on a bare oxidized silicon surface, although it is in agreement with the observation for the two desorption/ readsorption cycles of an invariant As/Ap ratio upon readdition of CTAB to the PSS-depleted surfactant surface (although in the latter case the As/Ap ratio and hence average orientational order of the surfactant were considerably larger). This observation is consistent with the two-dimensional epitaxial growth mechanism outlined earlier; i.e., the readdition of CTAB to a depleted surfactant surface results in the residual interfacial surfactant acting to template the orientation of the readsorbing surfactant. That is, a highly ordered partial surfactant layer results in the readsorption of surfactant in a similarly highly ordered state, while a disordered partial surfactant layer results in the readsorption of surfactant in a similarly disordered state. The rapidity of the of the reobtainment of the initial equilibrium surface excess value upon readdition of CTAB to a depleted surfactant surface (whether it be highly ordered or disordered) may consequently be explicable by the absence of a surfactant reorientation step in the CTAB readsorption process. Clearly, the fact that two significantly different CTAB conformations may be created on oxidized silicon at the same surface excess as a function of system history indicates the existence of nonequilibrium conditions and the presence of kinetic traps. Conclusion Polarized and nonpolarized IR-ATR spectroscopy has been employed to determine the surface excess and degree of conformational order of CTAB adsorbed on oxidized silicon substrates as a function of the adsorption and polyelectrolyte-induced desorption processes. It has been demonstrated that CTAB may be removed from the interface by complexation with the oppositely charged polyelectrolyte PSS and that the orientational order of the residual surfactant, after an initial period of stability, decreases as a function of time. Further, it has been shown that CTAB may readsorb on a surface that has been partially depleted and interestingly that the degree of orientational order of the readsorbed surfactant is highly dependent upon the order of the residual interfacial surfactant at the time of surfactant readdition. Indeed, it has been shown that the orientational order of the readsorbed surfactant is templated by that of the residual interfacial surfactant, suggesting a two-dimensional epitaxial growth mechanism for the readsorption process. Acknowledgment. J.S.P. gratefully acknowledges the University of Maine Pulp and Paper Foundation and the sponsors of the University of Maine Paper Surface Science Program for financial support. LA047861W