Adsorption Isotherms of Cellulose-Based Polymers onto Cotton Fibers

Article pubs.acs.org/Langmuir

Adsorption Isotherms of Cellulose-Based Polymers onto Cotton Fibers Determined by Means of a Direct Method of Fluorescence Spectroscopy Ingo Hoffmann,† Claudia Oppel,† Ulrich Gernert,‡ Paula Barreleiro,§ Wolfgang von Rybinski,§ and Michael Gradzielski*,† †

Stranski Laboratorium für Physikalische und Theoretische Chemie, Technische Universität Berlin, Strasse des 17, Juni 124, 10623 Berlin, Germany ‡ Zentraleinrichtung Elektronenmikroskopie, Technische Universität Berlin, Strasse des 17, Juni 135, 10623 Berlin, Germany § Global R&D Chemistry, Henkel AG & Co. KGaA, Laundry & Home Care Advanced Materials/Technology Brokerage, Düsseldorf, Germany S Supporting Information *

ABSTRACT: We present a novel method for the measurement of polymer adsorption on fibers by employing fluorescently labeled polymers. The method itself can be used for any compound that either shows fluorescence or can be labeled with a fluorescent dye, which renders it ubiquitously applicable for adsorption studies. The main advantage of the method is that the choice of adsorbent is not limited to flat surfaces, thereby allowing the investigation of fibrous and porous systems. As an example of high interest for application we determined the adsorption isotherms of various polysaccharide-based polymers with different charges and different substituents on cotton fibers. These experiments show that the extent of adsorption depends not only on the charge conditions but also very much on the specific interactions between the polymer and fiber. For instance, the cationic hydroxyethyl cellulose can become bound to an extent similar to that of the anionic alginate, while the anionic carboxymethyl cellulose of similar charge density adsorbs much less under these conditions. This shows that the adsorption of polymers depends subtly on the details of the interaction between the polymer and fiber but can be determined with good precision with our direct fluorescence method.

■

INTRODUCTION The measurement of adsorption of polymers on surfaces or fibers is of significant commercial interest, since polymers are part of most formulations in detergency, for instance, for soil release and as dye transfer inhibitors, encrustation inhibitors, or antiredeposition agents.1−3 In particular, polymer adsorption on textile fibers plays a key role in detergency as it is the central aspect in their use for soil release. Also, the adsorption of polymers at the liquid−liquid interface can play an important role in the formation of microemulsions.4 While there are very sophisticated methods for the determination of polymer adsorption on flat model surfaces, such as measurements by means of ellipsometry, reflectometry, atomic force microscopy (AFM), or quartz crystal balance (QCB),5−10 the study of adsorption on not so well-defined materials such as fibers is much more complicated. Therefore, only relatively few publications can be found on this topic that address this question in a quantitative way, but some publications have been dealing with the adsorption of polymers and polyelectrolytes on cellulosic material.11−15 The adsorption on cotton is of high practical interest since it is one of the most important materials for fabrics.16−26 Polymer © 2012 American Chemical Society

adsorption plays an important role in the washing process since polymers are used to prevent the readsorption of soil on the fibers. Some data can be found concerning the performance of polymers in preventing readsorption of soil27−30 and on the adsorption of surfactants on cotton fibers,31−35 where typically synergistic effects for surfactant mixtures are quite pronounced.34 For the case of deposition of dyes (as a model soil), it has been observed that hydrophobic interactions play an important role and dye adsorption becomes less with increasing polymer concentration.36 To a much larger extent the adsorption of surfactant at the solid−liquid interface between an aqueous solution and a fiber or flat surface has been studied. A comprehensive overview of surfactant absorption is given by Somasundaran and Huang37 and Somasundaran and Zhang.38 As to be expected, the adsorption process depends to a large extent on the details of the solid substrate and is largely controlled by the electrostatic conditions between the substrate and adsorbing surfactant at low surfactant concentrations. At Received: January 20, 2012 Revised: April 30, 2012 Published: May 1, 2012 7695

dx.doi.org/10.1021/la300192q | Langmuir 2012, 28, 7695−7703

Langmuir

Article

the different sizes of the polymers. The chemical structures of all polymers are given in Figures 1 and 2; details are given in Table 1.

higher concentrations the formation of surface aggregates becomes increasingly relevant. In contrast, only a few studies deal with the adsorption of polymers on cotton fibers.30,39 Therefore, to the best of our knowledge, this is the first study providing quantitative adsorption isotherm data. It has also been observed recently by small-angle neutron scattering (SANS) measurements that the presence of water-soluble polymer or surfactant has only little effect on the mesoscopic structure of cotton fibers.40 The aim of this study is to establish a simple method to quantitatively measure polymer adsorption on arbitrarily shaped materials by taking advantage of the large selection of fluorescence labels available today. Of course, employing fluorescently labeled polymers for determining their adsorption isotherms is not really a new approach by itself. For instance, for the case of studying the adsorption of cationic polyacrylamide (C-PAM) onto cellulose fibers from kraft pulp, fluorescently labeled C-PAM was employed.14 However, in that work the adsorption isotherms were determined by measuring the filtrate, while we will show in this work that for many concentration regions of interest this method is rather imprecise and one can obtain values with smaller errors by the method introduced here, where the adsorbed amount is determined directly on the fiber. It might be added here that confocal fluorescence microscopy of kraft pulp cellulose fibers with adsorbed fluorescently labeled cationic dextran has successfully been employed to determine its location within the fiber structure.11 It should also be mentioned that total internal reflectance fluorescence (TIRF) has been used for studying the adsorption of poly(ethylene oxide) (PEO) and hydroxyethyl cellulose (HEC) on silica surfaces.41 The main difference from this previous work is that here we employ an approach by which the amount of adsorbed polymer is determined directly from the fluorescence of the fiber of arbitrary shape, i.e., the case that is typically not so easily studied by other methods, such as reflectivity or ellipsometry. With this novel method we then investigated the adsorption of cationic, nonionic, and anionic water-soluble cellulose-based polymers to see how the polymer charge and the molecular constitution of the polymer affect the adsorption properties for otherwise identical conditions. The obtained adsorption isotherms were correspondingly analyzed and show very pronounced differences with respect to the type of polymer employed.

■

Figure 1. Chemical structure of alginate.

Figure 2. Chemical structures of CHEC, MHEC, and CMC. The polymers were labeled with the fluorescein-based dye 5-[(4,6dichlorotriazin-2-yl)amino]fluorescein (DTAF; Sigma-Aldrich) (Figure 3) according to the procedure described by de Belder et al.43 The labeling density was chosen to yield about two dye molecules per polymer chain to limit its effect on modification of the chemical structure of the polymer. At such low labeling densities and for the given size of the polymers investigated, this modification should be so minor that substantial effects on the adsorption properties of the labeled polymers are not expected. Samples of unmodified cotton were placed in boiling water three times for at least 30 min to remove dust and chemicals from prior treatment. The surface area of dry cotton was 7 m2/g as measured by N2 adsorption and analysis by the Brunauer−Emmett−Teller (BET) method44 (see Figure S1 (Supporting Information) and ref 40 for further details). For a further characterization, SEM images were taken from samples of dried cotton after the boiling treatment (see Figure 4). It can be seen that the fibers are relatively flat and smooth, and their diameter is about 10−15 μm. One can also see well that the fibers are composed of helically arranged fibrils. Bean-shaped cross sections can be observed as they are characteristic for relatively young fibers.16,17,21 It should also be noted that cotton fibers are mostly composed of cellulose, a high molecular weight polymer, which is insoluble in most solvents and whose structure was first determined to be a long threadlike molecule by Staudinger.45 Methods. Sample Preparation. For the measurement of adsorption isotherms, small pieces of cotton with a mass of about 50 mg were exposed to solutions of labeled polymers with different polymer concentrations and at a constant cotton concentration of 25 g of cotton/kg of solution (resulting in about 2 mL of solution) for 24 h at 40 °C. Samples were stored in 8 mL test tubes closed with a screw cap. Afterward, the wet cotton specimens were taken out of their test tubes and shaken to remove excess solvent, and then their mass was recorded to obtain the swelling constant (mass of liquid per mass of adsorbent) of the cotton. Subsequently, the cotton samples were dried in a desiccator prior to measurement. The pH of the sample solutions was checked and always found to be somewhat less than 7, which corresponds to the value of deionized water. Fluorescence Measurements. Fluorescence intensities were recorded with a Hitachi F-4500 FL spectrophotometer. The excitation and emission wavelengths were 494 and 514 nm, respectively. Square pieces of cotton of about 1.5 cm2 were placed in a laboratory-made

EXPERIMENTAL SECTION

Materials. Cationic hydroxyethyl cellulose (CHEC; Dow Chemical, Midland, MI) is a cationically modified cellulose ether with a molar mass of about 500 000 g/mol and degrees of substitution (DSs) of 0.27 (N,N,N-trimethylammonium chloride) and 0.73 (hydroxyethyl), resulting in a charge density of 1000 g/mol of positive charges.42 Each monomeric unit is substituted with either a hydroxyethyl group or a charged group. Methyl hydroxyethyl cellulose (MHEC) (Clariant, Germany) is a nonionic cellulose ether with additional methyl and hydroxyethyl substituents; according to the supplier its DSs are 1.8 and 0.16, respectively, and the molar mass is 170 000 g/mol. Carboxymethyl cellulose (CMC; Hercules, Wilmington, DE) is an anionic cellulose ether, substituted with sodium carboxymethyl groups; according to the supplier the DS is 0.7 and the molar mass is about 250 000 g/mol. Sodium alginate is the sodium salt of alginic acid, a polysaccharide consisting of monomers of mannuronic and guluronic acids. All polymers were used as received. Furthermore, it should be noted that all of them have a rather broad size distribution, which makes it difficult to observe effects caused by 7696


Langmuir

Article

Table 1. Molar Mass, Degree of Substitution, Charge Density (Mass per Mole of Charges), z, Type of Charge of the Polymers, and Mean Contour Length between Charges, lc (Assuming a Length of 0.51 nm per Monomer Unit) polymer

Mw(g/mol)

DS

type

MHEC

170000

nonionic

CMC CHEC

250000 500000

1.8 (methyl) 0.16 (hydroxyethyl) 0.7 0.73 (hydroxyethyl) 0.27 (trimethylammonium)

z (g/mol)

lc (nm)

anionic cationic

310 1000

0.73 1.89

alginate

500000

anionic

198

0.51

Figure 3. Chemical structure of DTAF. Figure 6. Calibration curve for CHEC at 40 °C obtained with the calibration procedure described in the text. The solid line gives the fit with eq 1. I = mcsurf + n

(1)

with m the slope, n the intercept (accounting for the residual fluorescence signal of the pure cotton, which, however, was always very small and independent of the individual cotton sample employed), and csurf the amount of polymer per amount of cotton. BET Measurements. Nitrogen adsorption−desorption measurements at 77 K were performed using a Gemini 2375 volumetric gas adsorption analyzer (Micromeritics). The samples were dried and outgassed at 120 °C for about 1 h with a VacPrep 061 device from Micromeritics at about 0.5 mbar and reweighed before the gas sorption measurement to determine the net mass of the sample. The BET specific surface area was extracted from the isotherms in a range of relative pressures p/p0 from 0.01 to 0.31. For the cross-sectional area of nitrogen molecules, a value of 0.162 nm2 was used. Scanning Electron Microscopy (SEM). SEM micrographs were recorded using a Hitachi S-4000 in secondary electron imaging mode with an acceleration voltage of 20 kV and a probe current of 20 pA. Cotton samples (25 g/L of solvent) were kept in water or polymer solution (1.5 g/L) overnight and dried in vacuo. To improve conductivity, samples were sputtered with a 6 nm Au coating. Isotherms. The obtained isotherms were fitted to the isotherm equation obtained by Zhou et al.46−48

Figure 4. SEM image of dried cotton after the boiling treatment, scale bar 100 μm. sample holder (see Figure 5). Its outer dimensions are those of a regular 1 cm cuvette. The intensities were measured in four different

Figure 5. Scheme of the sample holder for the measurement of the fluorescence of fibers, top view.

cads =

( 1n + K 2csol n−1)

ΓmaxK1csol

1 + K1csol(1 + K 2csol n − 1)

(2)

which was originally derived to describe the adsorption of hemimicelle-forming surfactants. Γmax denotes the maximum amount that can be adsorbed on the surface, K1 and K2 are equilibrium constants for the adsorption of monomers and clusters of n molecules, respectively, and csol is the concentration of adsorptive in the bulk solution. The picture behind the equation is as follows: At low concentrations single molecules adsorb according to the Langmuir isotherm with equilibrium constant K1 up to a concentration of Γmax/n, and then clusters of n molecules are formed from n − 1 molecules in the solution and an adsorbed single molecule with equilibrium constant K2. Accordingly, a two-step adsorption process is described, where at lower concentration first molecular adsorption takes place, while at higher concentration micellar (or hemimicellar) aggregates become formed at the interface. Of course, in our case of polymer

places on the surface for at least 5 s to obtain well-averaged intensities. Measurements on one spot never varied by more than 1%. Typical deviations in such a series on different spots were about 1% and never exceeded 10%. To obtain quantitative data, calibration measurements were performed. In these measurements pieces of cotton were exposed to amounts of polymer solution with different concentrations. The amount of solution was kept below the swelling constant of cotton (about 2); thus, the amount of polymer contained in these cotton samples was known precisely. The obtained fluorescence intensities showed a linear dependence (for an example with CHEC, see Figure 6) on concentration and could be converted to the amounts of deposited polymer according to a calibration curve given as 7697


Langmuir

Article

Therefore, we refrained from adding another fit parameter into our model. The bulk concentration after adsorption, csol (employed in eq 2), is calculated according to

adsorption no micelle formation is expected, but polymers might adsorb in a cooperative fashion, e.g., by cluster formation, where the extent of cooperativity is described by the parameter n. This means that, after an initial adsorption of polymer directly onto the cotton fibers, subsequent further adsorption of polymer occurs at this adsorbed polymer layer, thereby leading to polymer aggregation at the surface.

■

csol =

RESULTS AND DISCUSSION Theoretical Considerations Concerning the Adsorption of Polymers on Cotton. Due to quite pronounced swelling of cotton, it is necessary to distinguish between the amount of polymer on the cotton due to adsorption onto the fiber, cads, and the total amount of polymer on the sample due to adsorption and swelling, csurf (both mg of polymer/g of cotton), which is the measured concentration; cads can be calculated according to cads = csurf − csolS (3)

c0 − 1−

csurf c B 1000 c BS 1000

(4)

with the bulk concentration, csol, initial polymer concentration, c0, and cotton concentration, cB, in grams of polymer per kilogram of solvent, surface concentration, csurf in milligrams of polymer per gram of cotton, and swelling constant, S, in grams of cotton per gram of adhered solution. Estimation of Errors. In general, there are two different approaches to measure adsorption from the bulk. It is either possible to measure the decrease in concentration in the bulk or to measure the concentration of adsorbate on the surface directly. The former approach has already been employed to measure the adsorption of polymers by means of fluorescence spectroscopy,11 but the latter approach is more accurate for the large majority of measurements presented here, as shall be outlined in detail in the Appendix. Figure S2−S4 (Supporting Information) provide an overview of under which conditions which type of measurement yields better results. These calculations have been done assuming realistic experimental conditions and errors. In general, it can be said that measurements directly on the surface are superior for weak adsorption and somewhat higher concentrations in the bulk. At high Γmax the precision of the measurement in solution would be better for not too high solution concentrations, but this case of high adsorption (and correspondingly large effects observed) is easily and precisely measured by both methods; i.e., the absolute error in this range is small in any case. Adsorption Isotherms. After having established this novel method for determining adsorption isotherms, we were now interested in its application to a problem of interest. For this purpose we chose cotton fabric as the substrate and investigated the adsorption of different cellulose-based polymers onto it. All of them were labeled similarly with DTAF to be detectable by their fluorescence signal. They vary, however, with respect to their charge and the details of their structure (Figures 1 and 2). The adsorption isotherms of fluorescently labeled CHEC, alginate, MHEC, and CMC were measured at 40 °C at a cotton concentration, cB, of 25 g/kg. The obtained results are shown in Figure 8. The isotherms were fitted with eq 2, and from these fits we determined Γmax,

with S the swelling constant (g of swelling solution/g of cotton) and csol the bulk concentration of the polymer in the solution (g of polymer/kg of solution). The swelling constant is defined as the mass of solution taken up by the cotton sample divided by the mass of the cotton sample itself. The second term accounts for the amount of polymer simply contained in the solution that adheres to the fiber. To account further for excluded volume effects, it would be necessary to differentiate between the swelling in parts of the fabric accessible to the polymer, Sa, and that in parts that are inaccessible, Sb, where S = Sa + Sb (see Figure 7). Since it is

Figure 7. Schematic representation of a cross section through a cotton fiber. The green (Sa), outer area is accessible to the polymers. In contrast, the inner part (yellow, Sb) is inaccessible for the polymer but still accessible for the solvent, while the innermost, red part is inaccessible to both polymer and solvent and therefore does not contribute to the swelling at all.

difficult to determine Sa or Sb experimentally, it is assumed in the following that Sa = S, which is the equivalent of assuming that there are no excluded volume effects for the polymer. This is not entirely true as can be seen from isotherms that show a decline at higher concentrations. Therefore, the amounts of adsorbed polymer are somewhat underestimated. In principle, it is possible to estimate Sa from the decrease of adsorption at high bulk concentrations, since cads should become constant and therefore dcsurf/dcsol = Sa. Unfortunately, here the errors become quite large at sufficiently large concentrations.

Figure 8. Adsorption isotherms of CHEC (●), alginate (■), and MHEC (▲) at 40 °C. CMC does not display significant adsorption under these conditions. The solid lines give fits to eq 2. 7698


Langmuir

Article

i.e., the maximum amount of polymer that can be adsorbed. The obtained parameters together with c1/2, which is the concentration at which cads = Γmax/2, are summarized in Table 2. c1/2 is characteristic for how easily the respective polymer is adsorbed onto the cotton fibers.

That this is the case becomes readily clear when we consider the adsorption data for the anionic sodium alginate, for which we cannot expect an electrostatic driving force for adsorption. Nonetheless, the final plateau level for the adsorption of alginate is similar to that of CHEC (Figure 8), which demonstrates that there can be a very strong attraction for adsorption without having electrostatic attraction between the fiber and the polymer to be adsorbed. Unlike for CHEC, the isotherm is of the S2 type and shows a sudden increase in adsorption at a bulk concentration of about 0.5 g/kg, indicating strong cooperativity in the adsorption process. This strong cooperativity is also evidenced by the relatively high value of n = 8.7 for the adsorption isotherm (eq 2), while the cooperativity is much less for CHEC (n = 2.7). It is interesting to note that the adsorption constant K1 for the individual polymers is rather similar for CHEC and alginate but much less for the less strongly adsorbing MHEC. In contrast, the equilibrium constant K2 for the cooperative binding is rather similar for these three polymers, being the highest for CHEC. As the CHEC does not bind much in a cooperative fashion, its amounts adsorbed are already relatively high for low concentrations (as evidenced by the low c1/2 value), while for alginate (and to a lesser extent for MHEC) much higher polymer concentrations are required to see substantial adsorption. However, alginate is known to interact strongly with divalent cations.55 Even though the prior treatment of the cotton itself should remove such contaminations, it cannot be ruled out that traces are still present on the cotton or in the alginate. Adsorbed divalent cations would then mediate the adsorption of the anionic polyelectrolyte (see Figure S5 (Supporting Information) for the influence of CaCl2 on the adsorption of alginate). In contrast, the other anionic polymer, CMC, does not seem to adsorb on cotton at all, as can be seen in Figure 9, where the

Table 2. Summary of Fit Parameters for Adsorption with Eq 2, Which Are the Maximum Amount of Polymer Adsorbed, Γmax, the Equilibrium Constants, K1 and K2, for the Adsorption of Monomers and Micelles, Respectively, the Cooperativity Constant, n, and the Polymer Concentration at Which 50% of the Maximum Adsorption Is Attained, c1/2 polymer

Γmax (mg/g)

K1 (kg/g)

K2[(kg/g)n−1]

n

c1/2 (g/kg)

alginate CHEC MHEC

5.6 5.5 1.6

15 11 1.8

15 44 19

8.7 2.7 4.3

0.7 0.15 0.4

The cationic CHEC, being oppositely charged compared to the cellulose fibers, shows strong adsorption and reaches a plateau of about 5.5 mg/g at quite low bulk concentrations of about 0.2−0.3 g/kg. It is interesting to note that for the adsorption of cationic dextran on kraft pulp cellulose the amount of adsorbed dextran has been found to be 50−500 mg/ g(fiber), being larger with increasing molar mass of the dextran. This means that in that case substantially larger amounts of polycation are adsorbed than on our cotton fibers.49 The isotherm is of the L type and does not show a plateau at intermediate concentrations. Since the surface charge of cotton has a negative sign in the relevant pH range, it is reasonable to assume that Coulombic interactions play a significant role in the adsorption of CHEC. The surface charge of cotton has been measured by Stana-Kleinschek et al.23 with streaming potential measurements.22,50 The values that were obtained for untreated cotton are about 0.075 μC/cm2. The specific surface area of the dry cotton is about 7 m2/g as has been determined by us with the BET method (see Figures S1 (Supporting Information) and ref 40; similar values for the specific surface area of cotton have been found previously,32 while significantly larger values have been obtained for cellulose51,52). SANS measurements on paper53 indicate that the value for wet cotton might be somewhat larger. The charge density of CHEC is 1000 g/mol of charges,42 so the resulting adsorption that would lead to charge neutralization would be (0.075 μC/cm2 × 7 m2/ g × 1000 g/mol)/(96500 C/mol) = 0.05 mg/g. It is interesting to note that the actual experimental values for saturation of adsorption are 2 orders of magnitude larger. This leads to the conclusion that simple Coulombic interactions between polymer and the charges on the surface are not the key element in the adsorption of CHEC. The actual driving force may well be entropic due to the release of counterions; also there is experimental evidence that the interactions of the polymer are not limited to the charges on the surface of the fiber but that the polymer penetrates more deeply into the fiber. A similar suggestion has been made for the interactions of CHEC with hair54 and has been corroborated by other experiments of ours that studied the change of the mesoscopic structure of cotton fibers upon polymer treatment by SANS.40 In summary, these findings show that adsorption in these cases is not just driven by simple electrostatics but in reality is a much more complex process (as already observed in general for adsorption on the liquid−solid interface).38

Figure 9. csurf of CHEC (●), alginate (■), CMC (◆), and MHEC (▲) as a function of csol at 40 °C. The dotted line displays the behavior of a compound that neither adsorbs nor shows any excluded volume effects with S = 2.

total amount of polymer on the cotton fabric is plotted as a function of the polymer concentration (compare eq 3). This is quite surprising given the fact that its molar mass and basic structure are not much different from those of the sodium alginate (at least if we assume that the adsorption of alginate is not due to contaminations with divalent cations). Its concentration on the cotton is constantly below the dotted line that indicates the behavior of a compound not showing any significant adsorption, i.e., the amount of polymer contained is just due to swelling, and the dotted line would be valid if all of the cotton fabric were accessible to the polymer. That csurf of 7699


Langmuir

Article

brushes, where the driving force is the release of counterions as the protein normally has positively and negatively charged patches.57 By appropriate combination with the polyelectrolyte one can release the corresponding counterions and thereby have a driving force for attachment even though the overall charges of polyelectrolyte and protein have the same sign. A partly similar mechanism might be envisioned here with corresponding charge inhomogeneities on the side of the cellulose fibers (having positively (e.g., remaining multivalent cations) and negatively charged patches). That this effect is not automatically governed by the overall sign of the charge and the electrostatic interactions becomes clear when looking at the case of the anionic CMC, which basically does not adsorb at all. Apparently the ability to interact strongly enough locally with the cellulose and to become incorporated into the fibers is at least as important. This aspect might be intimately linked to the structure and structural arrangement of the sugar groups in the polymer backbone as well as to the flexibility of the chains.

CMC is even less therefore can only be attributed to excluded volume effects. Under the assumption that CMC does not adsorb, we can estimate the upper limit of the extent of excluded volume effects from its measurement. If cads is given by eq 3 and the real amount adsorbed by cads,real = csurf − csolSa, then the difference is −csol(S − Sa) = −csolSb. Therefore, we can determine Sb from the slope of the isotherm (see Figure 10).

■

CONCLUSIONS We have demonstrated the use of fluorescently labeled polymers for the study of adsorption of polymers on cotton and the determination of the corresponding adsorption isotherms. The application of the method is not limited to such systems, though. It can be applied to the study of any compound as long as it either can be labeled with a fluorescent dye or is fluorescent itself. This method has not been applied before and allows for a determination of adsorption isotherms with much better precision than possible by most other methods. Its particular advantage is the applicability to arbitrarily shaped surfaces, such as fibers, which are of high practical importance. Furthermore, it is superior to classical methods of determining the adsorption isotherm from the solution concentration in the relevant concentration ranges for polymer adsorption on fibers. This newly established method was then applied to the investigation of four different cellulose-based polymers, the cationic CHEC, the anionic CMC and sodium alginate, and the neutral MHEC on cotton fabric, where the fibers possess a net negative charge. We were able to show that polymer adsorption on cotton is to some degree related to the charge of the polymer and the fiber, being very pronounced for the cationic CHEC, for which a Langmuir-type adsorption isotherm is observed (where, of course, one has to keep in mind that one could also have an S-type adsorption isotherm, where aggregation of polymer in the form of multilayers already takes place at very low concentrations, which are out of the observation window). However, adsorption is most likely not only due to Coulombic interactions, since the amount adsorbed is too large to be accounted for only by compensation of charges. It is more likely that the gain in entropy due to the release of counterions plays a more important role. For sodium alginate, at higher polymer concentrations, a similar maximum polymer adsorption is observed, which, however, occurs in a much more cooperative fashion only beyond a certain threshold concentration. The relatively strong adsorption of alginate takes place despite a lack of favorable overall Coulombic interactions, and the relatively complicated shapes of the adsorption isotherms show that the relevant interactions are more complicated. It might be speculated that the release of counterions from the cotton fibers in the adsorption process plays a prominent role in both cases. In that context it is also interesting to note

Figure 10. Adsorption isotherm of CMC at 40 °C. Its very weak adsorption together with excluded volume effects causes negative cads and allows for an estimation of Sb.

The slope has a value of 0.6, so withan S of about 1.8 we can conclude that at most about two-thirds of the internal volume of the fabric is accessible to the polymer. Unfortunately, these values should depend on the polymer, and we cannot be sure that this value is the same for all polymers used. However, it gives a first insight into and approximation of this relevant value which describes the internal structure of the fibers and which cannot be accessed easily by other experimental approaches. It should be kept in mind that this finding does not mean that there is no CMC present on the fibers. It just means that its presence is mostly due to swelling rather than adsorption. In fact, it has been shown that CMC-treated cotton fibers can bind larger amounts of cationic surfactant,56 which would be a clear indication of the presence of CMC on the fiber. Surprisingly, the neutral polymer MHEC adsorbs much less than CHEC and alginate with a Γmax of about 1.5 mg/g. Like that of alginate, the isotherm is of the S2 type, and adsorption shows a sudden rise after weak adsorption at low concentrations, indicating cooperativity. It should be noted that the adsorption isotherms of MHEC and alginate are quite similar at low concentrations and only differ at higher concentrations, where the isotherm of alginate shows strong cooperativity, while the cooperativity in the adsorption of MHEC is much less pronounced (n = 4.3 as compared to n = 8.7). Therefore, it can be assumed that the lower level of adsorption of MHEC compared to alginate is due to alginate's higher ability to form adsorbed clusters of polymers. However, in general, it has to be noted that the most interesting finding of this investigation is that adsorption is high for the two polyelectrolytes of opposite charge, CHEC and sodium alginate, while the neutral counterpart, MHEC, exhibits only little adsorption. One possible explanation for this finding is that, apart from strong hydrogen bonding and hydrophobic interactions, which certainly will also play an important role in the polymer adsorption, one additional driving force might be the release of the low-valent counterions (thereby gaining entropy). This has been described in detail for the case of adsorption of proteins into equally charged polyelectrolyte 7700


Langmuir

Article

that cotton has also been successfully employed as a substrate for the assembly of alternatively charged polyelectrolytes by the layer-by-layer technique,58 i.e., it can interact attractively with positively and negatively charged polyelectrolytes, which suggests the presence of both positive and negative charges on the surface. However, more specific interactions between polymer or fiber and polymer in solution and adsorbed polymers must also play a prominent role as the other anionic polymer, CMC, is basically not adsorbed at all. A similar conclusion has also been arrived at for the formation of multilayers of low charge density polyelectrolytes on thin films of cellulose.59 This means that presumably incorporation into the interior of the fibers plays an important role as much as the polymer's ability to form aggregates on the surface. This is in agreement with the observation that the amount of absorbed polyelectrolyte is more than sufficient to compensate the charges of the cotton completely (at the highest extent of adsorption, more than 100 times that of the cationic CHEC). Accordingly, the charge effect of the cotton then is compensated already at low polymer concentrations. The enhanced adsorption of CHEC and alginate then has to be due to interactions of polymer molecules in solution with adsorbed polymer, which then leads to the formation of polymer aggregates on the fibers, as evidenced by the cooperativity parameter, n, and a gain of counterion entropy. Apparently, for the CMC, the specific polymer−polymer interactions are not sufficient to allow for such aggregate formation, and hence, its adsorption is negligible. Finally, for the neutral MHEC about a factor 4 lower maximum adsorption is observed, which also occurs in a cooperative fashion. It is interesting to note that obviously it is not so important whether one adsorbs an oppositely or an equally charged polymer. For both, the adsorption can be high, provided the local interaction with the fiber and adsorbed polymer is favorable. Apparently, more complicated aspects such as hydrogen bonding, dipole−dipole interactions, hydrophobic interactions, and counterion entropy play an integral role as evidenced by the relatively strong adsorption of alginate despite the lack of favorable Coulombic interactions and the relatively complicated shapes of the isotherms. The application of our novel method for determining adsorption isotherms of polymers on fibers allows reliable information regarding the adsorption process to be obtained and is rather generally applicable. For the systems studied it shows that polymer adsorption on cotton fabric depends subtly on the detailed structure of the polymer employed and not just simply on the electrostatic conditions. This finding is of substantial relevance for all fabrics in which polymer treatment is relevant, such as detergency but also fabric care and fabric conditioning. In summary, we present an interesting new and widely useful method for measuring the adsorption isotherms on complex structured substrates and demonstrate how it can be applied to relevant questions from soft-matter research. This is of importance for further studies on even more complexly built systems.

■

Δcsurf =

⎛ ΔI ⎛ Δn Δm ⎞⎟ I Δm ⎟⎞ n ⎜ + +⎜ + ⎝ I ⎝ n m ⎠m m ⎠m

(5)

with ΔX being the error of X. Calculating cads according to eq 3 and assuming Sa = S, then Δcads = Δcsurf + Δ(csolS) = Δcsurf + (Δcsol/csol + ΔS/S)csolS. Calculating csol according to eq 4 and neglecting errors from cB, c0, and S yield Δcsol = (∂csol/ ∂csurf)Δcsurf = (cB/1000)/(1 − cBS/1000)Δcsurf. Using cB = 25 g/ kg and S = 2, i.e., (cB/1000)/(1 − cBS/1000) ≈ 0.026 (representing our experimental conditions), Δcads is larger than Δcsurf by less than 3%; therefore, it is safe to assume Δcads ≈ Δcsurf. It is also possible to calculate csurf by the decrease of csol as compared to c0 according to csurf =

c0 − csol(1 − c BS) cB

(6)

and csol =

i − isolvent c0 i0 − isolvent

(7)

with i and i0 being the bulk fluorescence intensities of the solution after and before adsorption, respectively, and isolvent being the bulk fluorescence intensity of the solvent (water). Assuming that the measurement of csol is the only significant source of error, we obtain Δcsurf =

⎛1 ⎞ ∂csurf Δcsol = ⎜ + S⎟Δcsol ∂csol ⎝ cB ⎠

(8)

If the error stems from the measured intensities, we obtain Δcsol Δi + Δisolvent Δi + Δisolvent = + 0 csol i − isolvent i0 − isolvent

(9)

Inserting eq 9 into eq 8 yields ⎛1 ⎞⎛ Δi + Δisolvent Δi + Δisolvent ⎞ Δcsurf = ⎜ + S⎟⎜ + 0 ⎟csol i0 − isolvent ⎠ ⎝ cB ⎠⎝ i − isolvent (10)

Since isolvent ≪ i and i0 and Δisolvent ≪ Δi and Δi0 and accounting for cB being in grams per kilogram and S in grams per gram, we obtain ⎛ 1000 ⎞⎛ Δi Δi ⎞ Δcsurf = ⎜ + S ⎟⎜ + 0 ⎟csol i0 ⎠ ⎝ cB ⎠⎝ i

(11)

Figures S3 and S4 (Supporting Information) compare the errors of both methods for typical experimental conditions as the ratio of error resulting from measurements in solution and at the surface as a function of the bulk concentration and maximum amount adsorbed, Γmax, and Langmuir constant, K, respectively, assuming adsorption according to the Langmuir equation. The ratio starts to exceed 1 at rather low concentrations in any case. This means that adsorption measurements from the surface become superior already at relatively low concentrations, in particular when Γmax is low and K is large, making them the method of choice for the determination of adsorption isotherms of polymers on fibers. Measuring the difference in bulk concentration remains superior up to relevant concentrations only if the adsorption is very strong, but for this situation in terms of the absolute error good results are obtained with either method.

APPENDIX

Error Estimate for the Fluorescence Method

Contributions to the error of csurf, if measured directly on the fiber, stem from the measurement itself and from the calibration (see eq 1); therefore, the error is given by 7701


Langmuir

Article

(9) Trabelsi, S.; Langevin, D. Co-adsorption of carboxymethylcellulose and cationic surfactants at the air−water interface. Langmuir 2007, 23, 1248−1252. (10) Claesson, P.; Makuska, R.; Varga, I.; Meszaros, R.; Titmuss, S.; Linse, P.; Pedersen, J. S.; Stubenrauch, C. Bottle-brush polymers: Adsorption at surfaces and interactions with surfactants. Adv. Colloid Interface Sci. 2010, 155, 50−57. (11) Horvath, A. T.; Horvath, A. E.; Lindström, T.; Wågberg, L. Adsorption of highly charged polyelectrolytes onto an oppositely charged porous substrate. Langmuir 2008, 24, 7857−7866. (12) Gimåker, M.; Wågberg, L. Adsorption of polyallylamine to lignocellulosic fibres: Effect of adsorption conditions on localisation of adsorbed polyelectrolyte and mechanical properties of resulting paper sheets. Cellulose 2009, 16, 87−101. (13) Solberg, D.; Wågberg, L. On the mechanism of cationicpolyacrylamide-induced flocculation and re-dispersion of a pulp fiber dispersion. Nord. Pulp Pap. Res. J. 2003, 18, 51−55. (14) Tanaka, H.; Ö dberg, L.; Wågberg, L.; Lindström, T. Adsorption of cationic polyacrylamides onto monodisperse polystyrene latices and cellulose fiber: Effect of molecular weight and charge density of cationic polyacrylamides. J. Colloid Interface Sci. 1990, 134, 219−228. (15) Winter, L.; Wåg berg, L.; Ö dberg, L.; Lindströ m , T. Polyelectrolytes adsorbed on the surface of cellulosic materials. J. Colloid Interface Sci. 1986, 111, 537−543. (16) Skau, E. L. The possible contribution of shape to the swelling and moisture-sorption behavior of cotton fiber at the saturation point. Text. Res. J. 1946, 16, 556−563. (17) Rollins, M. L. Microscopical methods in the measurement of swelling in cotton fibersA literature survey. Text. Res. J. 1947, 17, 19−26. (18) Jörgensen, L. Swelling and heterogeneous hydrolysis of cotton linters and wood pulp fibers related to their fine structure. Acta Chem. Scand. 1950, 4, 185−199. (19) Moore, A. T.; Scott, L. W.; Derguy, I. V.; Rollins, M. L. The swelling of cotton in waterA microscopical study. Text. Res. J. 1950, 20, 620−630. (20) Giuffria, R.; Tripp, V. W. Effects of tension on the swelling of cotton fibers. Text. Res. J. 1952, 22, 21−25. (21) Degruy, I. V.; Carra, J. H.; Goynes, W. R. In The Fine Structure of Cotton; O‘Connor, R. T., Ed.; Marcel Dekker, Inc.: New York, 1973. (22) Ribitsch, V.; Jorde, C.; Schurz, J.; Jacobasch, H. Measuring the zetapotential of fibers, films and granulates. Prog. Colloid Polym. Sci. 1988, 77, 49−54. (23) Stana-Kleinschek, K.; Ribitsch, V. Electrokinetic properties of processed cellulose fibers. Colloids Surf., A 1998, 140, 127−138. (24) Grundke, K.; Bogumil, T.; Werner, C.; Janke, A.; Pöschel, K.; Jacobasch, H.-J. Liquid-fluid contact angle measurements on hydrophilic cellulosic materials. Colloids Surf., A 1996, 116, 79−91. (25) Heinze, T.; Liebert, T. Unconventional methods in cellulose functionalization. Prog. Polym. Sci. 2001, 26, 1689−1762. (26) El Seoud, O. A.; Fidale, L. C.; Ruiz, N.; D’almeida, M. L. O.; Frollini, E. Cellulose swelling by protic solvents: Which properties of the biopolymer and the solvent matter? Cellulose 2008, 15, 371−392. (27) Fite, F. J. C. Behavior of cellulose ethers when used as antiredeposition agents during washingrelationship between the deposition and the zeta-potential of polyester-cotton using different finishes. J. Text. Inst. 1993, 84, 582−592. (28) von Stackelberg, M.; Kling, W.; Benzel, W.; Wilke, F. Die elektrokinetischen Potentiale von Textilfasern in Lösungen von normalen Elektrolyten und von grenzflächenaktiven Verbindungen. Kolloid Z. Z. Polym. 1954, 135, 67−80. (29) Kling, W.; Lange, H. Ü ber das elektrophoretische Verhalten von Pigmentteilchen in Waschmittel-Lösungen. Colloid Polym. Sci. 1952, 127, 19−27. (30) Jost, F.; Andree, H.; Schwuger, M. J. Mechanisms of polymeric antiredeposition agents in detergents. Colloid Polym. Sci. 1986, 264, 56−64. (31) Meader, A. L.; Fries, B. A. Adsorption in the detergent process. Ind. Eng. Chem. 1952, 44, 1636−1648.

Comparing the errors for both methods under realistic conditions shows that the error bars are significantly smaller if the adsorption is measured directly on the surface of the fabric, in particular when csol becomes larger, which can be understood intuitively because the relative differences of bulk concentration before and after adsorption become small at some point when the amount adsorbed remains constant. This means that for conditions of practical relevance the investigation of polymer adsorption on cotton fibers by our novel method of measuring the absorbed amount by means of fluorescence detection is clearly superior to techniques that would determine the adsorbed amount from the change of the solution concentration.

■

ASSOCIATED CONTENT

S Supporting Information *

Figures showing the nitrogen adsorption isotherm on cotton, critical concentration of polymer in solution as a function of Γmax, ratio between the error of csurf from bulk measurements and measurements on the surface as a function of K and csol, ratio between the error of csurf from bulk measurements and measurements on the surface as a function of Γmax and csol, and adsorption isotherms of alginate with CaCl2. This material is available free of charge via the Internet at http://pubs.acs.org/.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS For funding of this research project, we are grateful to Henkel AG & Co. KGaA. For the use of the BET instrument, we are grateful to Prof. Schomäcker. M.G. thanks the Institute LaueLangevin (ILL; Grenoble, France) and the Deutsche Forschungsgemeinschaft (Project GR1030/10) for hospitality and funding of his sabbatical stay, during which a large part of this manuscript was produced.

■

REFERENCES

(1) Kwak, J. C. T., Ed. Polymer-Surfactant Systems; Marcel Dekker, Inc.: New York, 1998; Vol. 77. (2) Bertleff, W.; Neumann, P.; Baur, R.; Kiessling, D. Aspects of polymer use in detergents. J. Surfactants Deterg. 1998, 1, 419−424. (3) Wagner, G. Waschmittel: Chemie, Umwelt, Nachhaltigkeit; WileyVCH: Weinheim, Germany, 2005. (4) Sottmann, T. Solubilization efficiency boosting by amphiphilic block co-polymers in microemulsions. Curr. Opin. Colloid Interface Sci. 2002, 7, 57−65. (5) Zimin, D.; Craig, V. S. J.; Kunz, W. Adsorption pattern of mixtures of trimethylammonium-modified hydroxyethylcellulose and sodium dodecyl sulfate at solid-liquid interfaces. Langmuir 2004, 20, 2282−2291. (6) Zimin, D.; Craig, V. S. J.; Kunz, W. Adsorption and desorption of polymer/surfactant mixtures at solid-liquid interfaces: Substitution experiments. Langmuir 2004, 20, 8114−8123. (7) Nylander, T.; Samoshina, Y.; Lindman, B. Formation of polyelectrolyte-surfactant complexes on surfaces. Adv. Colloid Interface Sci. 2006, 123, 105−123. (8) Nestor, J.; Esquena, J.; Solans, C.; Luckham, P. F.; Musoke, M.; Levecke, B.; Booten, K.; Tadros, T. F. Interaction forces between particles stabilized by a hydrophobically modified inulin surfactants. J. Colloid Interface Sci. 2007, 311, 430−437. 7702


Langmuir

Article

(32) Ginn, M.; Kinney, F.; Harris, J. Effect of cotton substrate characteristics upon surfactant adsorption. J. Am. Oil Chem. Soc. 1961, 38, 138−143. (33) Paria, S.; Manohar, C.; Khilar, K. C. Experimental studies on adsorption of surfactants onto cellulosic surface and its relevance to detergency. J. Inst. Chem. Eng., Singapore 2003, 43, 34−44. (34) Paria, S.; Khilar, K. C. A review on experimental studies of surfactant adsorption at the hydrophilic solid-water interface. Adv. Colloid Interface Sci. 2004, 110, 75−95. (35) Paria, S.; Manohar, C.; Khilar, K. C. Adsorption of anionic and non-ionic surfactants on a cellulosic surface. Colloids Surf., A 2005, 252, 221−229. (36) Oakes, J.; Dixon, S. Adsorption of dyes to cotton and inhibition by polymers. Color. Technol. 2003, 119, 140−149. (37) Somasundaran, P.; Huang, L. Adsorption/aggregation of surfactants and their mixtures at solid-liquid interfaces. Adv. Colloid Interface Sci. 2000, 88, 179−208. (38) Zhang, R.; Somasundaran, P. Advances in adsorption of surfactants and their mixtures at solid/solution interfaces. Adv. Colloid Interface Sci. 2006, 123, 213−229. (39) Kling, W. Theorie des Waschens. Z. Elektrochem. 1955, 59, 260−272. (40) Hoffmann, I.; Oppel, C.; Prévost, S.; Gernert, U.; Barreleiro, P.; von Rybinski, W.; Gradzielski, M. The influence of polymers, surfactants and salt on the fine structure of cotton revealed by SANS. Colloids Surf., B 2012, 91, 175−180. (41) Rebar, V. A.; Santore, M. M. History-dependent Isotherms and TIRF calibrations for homopolymer adsorption. Macromolecules 1996, 29, 6262−6272. (42) Thuresson, K.; Nilsson, S.; Lindman, B. Effect of hydrophobic modification on phase behavior and rheology in mixtures of oppositely charged polyelectrolytes. Langmuir 1996, 12, 530−537. (43) Debelder, A. N.; Granath, K. Preparation and properties of fluorescein-labelled dextrans. Carbohydr. Res. 1973, 30, 375−378. (44) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of gases in multimolecular layers. J. Am. Chem. Soc. 1938, 60, 309−319. (45) Staudinger, H. Zur Konstitution der Cellulose. Z. Elektrochem. Angew. Phys. Chem. 1934, 40, 729. (46) Zhu, B. Y.; Gu, T. R. General isotherm equation for adsorption of surfactants at solid liquid interfaces. J. Chem. Soc., Faraday Trans. 1 1989, 85, 3813−3817. (47) Zhu, B. Y.; Gu, T. R.; Zhao, X. L. General isotherm equation for adsorption of surfactants at solid liquid interfaces 0.2. Applications. J. Chem. Soc., Faraday Trans. 1 1989, 85, 3819−3824. (48) Zhu, B. Y.; Gu, T. Interfacial film of surfactant adsorbed at a solid liquid interface. J. Chem. Soc., Faraday Trans. 1991, 87, 2745− 2748. (49) Horvath, A. T.; Horvath, A. E.; Lindström, T.; Wågberg, L. Adsorption of low charge density polyelectrolytes to an oppositely charged porous substrate. Langmuir 2008, 24, 6585−6594. (50) Jacobasch, H. J.; Bobeth, W. Electrokinetic measurements for characterization of textile fibers. Papier 1975, 29, 555−561. (51) Thode, E. F.; Swanson, J. W.; Becher, J. J. Nitrogen adsorption of solvent-exchanged wood cellulose fibers: Indications of total surface area and pore size distribution. J. Phys. Chem. 1958, 62, 1036−1039. (52) Carles, J. E.; Scallan, A. M. The determination of the amount of bound water within cellulosic gels by NMR spectroscopy. J. Appl. Polym. Sci. 1973, 17, 1855−1865. (53) Garvey, C. J.; Parker, I. H.; Knott, R. B.; Simon, G. P. Small angle scattering in the Porod region from hydrated paper sheets at varying humidities. Holzforschung 2004, 58, 473−479. (54) Regismond, S. T. A.; Heng, Y.-M.; Goddard, E. D.; Winnik, F. M. Fluorescence microscopy observation of the adsorption onto hair of a fluorescently labeled cationic cellulose ether. Langmuir 1999, 15, 3007−3010. (55) Thiele, H.; Andersen, G. Ionotrope Gele von Polyuronsäuren. Colloid Polym. Sci. 1955, 140, 76−102.

(56) Fras-Zemljic, L.; Stenius, P.; Laine, J.; Stana-Kleinschek, K. The effect of adsorbed carboxymethyl cellulose on the cotton fibre adsorption capacity for surfactant. Cellulose 2006, 13, 655−663. (57) Henzler, K.; Haupt, B.; Lauterbach, K.; Wittemann, A.; Borisov, O.; Ballauff, M. Adsorption of β-lactoglobulin on spherical polyelectrolyte brushes: Direct proof of counterion release by isothermal titration calorimetry. J. Am. Chem. Soc. 2010, 132, 3159− 3163. (58) Ali, S. W.; Rajendran, S.; Joshi, M. Effect of process parameters on layer-by-layer self-assembly of polyelectrolytes on cotton substrate. Polym. Polym. Compos. 2010, 18, 175−187. (59) Wang, Z.; Hauser, P. J.; Laine, J.; Rojas, O. J. Multilayers of low charge density polyelectrolytes on thin films of carboxymethylated and cationic cellulose. J. Adhes. Sci. Technol. 2011, 25, 643−660.

7703


Adsorption Isotherms of Cellulose-Based Polymers onto Cotton Fibers

Recommend Documents