Langmuir isotherm model assumes that sorption takes place at specific homogeneous sites within an adsorbent where all sorption sites are identical and energetically equivalent and there is no interaction between molecules adsorbed on neighboring sites. Therefore, a saturation value is reached beyond which no further sorption could take place, meaning that once a site is occupied no further adsorption can take place in this site. This theory considers that the adsorbent behaves as an ideal solid surface, where adsorbate can be bonded either by physisorption or chemisorption. Intermolecular forces decrease abruptly with distance and the model predicts only monolayer coverage [16] .

The Langmuir model equation is:

where q_e is the equilibrium concentration of adsorbate per unit weight of dry adsorbent (mmol・kg⁻¹ or mg・g⁻¹); C_e is the equilibrium concentration of adsorbate in the solution (mmol・dm⁻³ or mg・dm⁻³); X_m is the theoretical monolayer capacity (mmol・kg⁻¹ or mg・g⁻¹) and a_L is a constant related to the affinity binding sites (dm³・mmol⁻¹ or dm³・mg⁻¹). Langmuir constant can be defined as. In the present work in order to compare the interaction between molecules of RBBR and molecules of CHX in the adsorption process, every concentration of adsorbate has been expressed as mmol instead of mg.

The constants have been determined from linear fit of Equation (1) in the following form:

This model assumes some modifications of Langmuir model such as surface roughness, inhomogeneity, and adsorbate-adsorbate interactions. It is the most important multisite adsorption isotherm for rough surfaces and this empirical model has been shown to be consistent with exponential distribution of active centers characteristic of heterogeneous surface. The empirical equation is:

where K_F is the Freundlich constant (dm³・mmol⁻¹ or dm³・mg⁻¹) and 1/n_F is the heterogeneity factor. If n > 1, then the adsorption is favorable. From a linear fit of Equation (4), these empirical values were obtained.

Values of thermodynamic functions could be obtained from thermodynamic sorption curves. The standard free energy of adsorption, at a temperature T, can be evaluated from the following equation assuming an ideal behavior of the solution:

where the term is the process constant, in which V is the specific volume of the fiber (0.87

- 1.15 dm³/kg) data obtained from bibliography [17] and for the calculation has been assumed as 1 dm³/kg.

The changes of the enthalpy of the sorption process, can be evaluated at various temperatures intervals by means of the Clausius-Clapeyron equation, whose integrated form is:

Such evaluation assumes that the solution behaves as an ideal one and that the heat of dilution is negligible compared with the heat of adsorption, at this low concentration.

The change of entropy, could be determined using the equation:

Integrating for the boundary conditions t = 0 to t = t and q_t = 0 to q_t = q_e gives:

where q_e is the amount of dye sorbed at equilibrium (mmol・kg⁻¹) or (mg・g⁻¹); q_t is the amount of dye sorbed at time t (mmol・kg⁻¹) or (mg・g⁻¹) and k₁ is the equilibrium rate constant of first order sorption (min⁻¹). The intercept of the straight line plots of against t should equal. However, if the intercept does not equal q_e then the reaction is not likely to be first order reaction, and even this plot has high correlation coefficient with experimental data [19] .

The rate constant is described by:

Integrating for the boundary conditions t = 0 to t = t and q_t = 0 to q_t = q_e gives:

From the fit of against t it can be obtained k₂ and q_e.

The initial sorption rate h (g・mg⁻¹・min⁻¹) or (kg・mmol⁻¹・min⁻¹):

Fick’s equation, should be applied to estimate the approximated diffusion coefficient, D, at the cylindrical walls of the fiber

where a is the fiber radio which has been determined with a microscope as 5 × 10⁻⁴ cm in accordance with the

value determined by Espinosa-Jiménez [17] . The apparent diffusion coefficient can be obtained plotting

versus under assumption that the surface concentration is constant.

Fibers of Nylon 6.6, used in the present work is 100% polyamide 6.6 yarn, style 361, was supplied by Testfabrics (Spain). Its density is about 1 g/cm³, fusion temperature 250˚C, water absorption 10% - 20% when dipped in water and water content is ca. 3.8% - 4.1% at 20˚C and 65% of relative humidity. The fiber was repeatedly rinsed with deionizer water until water conductivity remained constant, and them they were dried in an oven and stored. Nylon 6.6 structure is represented in Scheme 1.

Reactive dye used in the present work is Remazol Brilliant Blue R (RBBR), Colour Index Number 61,200 Reactive Blue 19) (Scheme 2) was A.R grade was provided by Sigma Company. Its molecular formula is: C₂₂H₁₆N₂Na₂O₁₁S₃ and has a molecular weight of 626.55 g・mol⁻¹, and its solutions present a maximum of adsorbance at 593 nm. RBBR structure is depicted in Scheme 2.

The antiseptic used, Chlorhexidine (CHX) is presented as a salt solution, Chlorhexidine Digluconate, with initial concentration 20% w/v aqueous solution, and has been supplied by Dentaid S.A (Spain). Its molecular formula is C₃₄H₃₀N_10.2 (C₆H₁₂O₇) and has a molecular weight is 897.8 g・mol⁻¹. The maximum adsorbance of its solutions corresponds to 231 nm. CHX structure is depicted in Scheme 3.

Scheme 1. Nylon 6.6.

Scheme 2. Remazol brilliant blue R molecule.

Scheme 3. Chlorhexidine molecule.

Pellets of 1 g of dried fiber were used for sorption and electrokinetic experiments. The samples were dipped in 100 ml of RBBR or CHX water solutions of different concentration and pH and temperature conditions were adapted depending on the experiment performed. The sorption was performed in a thermostated water bath (WNB Memmert) with constant stirring and temperature accuracy of ±0.1 K during 48 hours, time enough to achieve equilibrium. The kinetic determinations were performed in the same way, using 250 ml of initial solution. The amount of adsorbate that has been uptaken by the fiber was determined from the difference between the initial and final concentration of the solution, estimated by optical absorbance at both wavelength 581 nm for the dye solutions and 231 for CHX solutions, with a Hitachi U-2000 spectrophotometer. The amount of adsorbate uptaken by the fiber has been expressed as mmol・kg⁻¹ of dry fiber, and the concentration of solution as mmol・l⁻¹. Adsorption data isotherms have been analyzed considering Langmuir and Freundlich models, as well as to investigate about kinetics mechanism involved we have tested a pseudo-first order, pseudo- second order an intraparticle diffusion models.

The evolution of zeta potential of the fiber plug has been determined by streaming potential method, with Fairbrother-Mastin [20] equation:

where streaming potential, pressure difference, conductivity of the electrolyte solution, dyamic viscosity of electrolyte solution. Zeta potential has been determined by streaming potential with an Electrokinetic Analyzer, EKA (Anton Paar, KG). Samples of 1 g of fiber were placed between two electrodes at a constant distance of 1 cm of a cylindrical cell of 20 mm diameter, which correspond to a bulk sample density in the cell of around 0.35 to 0.40 g/cm⁻³. An electrolyte solution of 1 mM KCl is forced to stream along the solid surface by a pressure difference, and the flow should change its direction periodically. The streaming potential is measured with two electrodes Ag/AgCl placed at both sides of the sample. Tritiation measures were carried out starting at pH 10 and finishing at pH 2.5 - 3 adjusting pH with NaOH or HCl respectively. Operating temperature should be close to standard room temperature (25˚C ± 3˚C).

Most of dyes present an ionic character in solution being the reactive dye RBBR an anionic one. The sorption of these charged dyes onto the adsorbent surface is primarily influenced by the surface charge on the adsorbent, which in turns is influenced by the solution pH. Thus the knowledge of the electrical potential of the fiber surface as function of the pH of the medium gives information about the electrostatic interactions that could take place between the dye and the fiber surface. Due to the amphoteric character of the fiber the concentration of the four possible fiber terminal groups, Ny-NH₂, Ny-, Ny-COOH AND Ny-COO⁻ (where Ny―represents the Nylon 6.6 polymer chain) will be determined not only by the concentration of amino or carboxyl terminal group but mainly by the pH of the medium. The equilibrium between the four species depends on the pH of the medium as follow [21] .

In order to know the charge of the fiber surface, in Figure 1 is presented the evolution of the zeta potential of Nylon 6.6 as function of the pH of aqueous medium and 1 mM NaCl solution. The similarity in the evolution of zeta potential indicates that NaCl behaves as an indifferent electrolyte for Nylon 6.6 fibers.

As it can be observed, Nylon 6.6 is highly influenced by the pH of the medium. After the fiber is immersed in an ionic medium the presence of H⁺ or OH⁻ should be responsible of the ionization to a greater or lesser extend of the amino and carboxylic acid groups of the fiber. An isoelectric point, IP, is presented at pH ranging from 4 to 4.5 in water and 1 mM NaCl solution respectively. Espinosa et al., [18] using a microelectrophoretic technique for zeta potential determination, have obtained a value of 3.5 units pH. The difference could be due to some alteration in the number of polar groups onto the fiber surface causes by morphological changes in fiber structure due to the samples treatment. The zeta potential of the fiber is negative for pH higher than 4 - 4.5 units and it increases up to ca −50 mV for pH 10. From pH below 4, the fiber presents positive charge. The change in the sign of zeta potential from an IP depends on the ionization of both amine and carboxylic groups. At pH lower than 4 units, ionization of prevails on the COOH one, while for pH above IP, COO⁻ is the most important one. Nevertheless, at both very acidic and highly alkaline pH, ζ is determined by adsorption of H⁺ and OH⁻ ions in the double layer.

The influence of the pH of the medium in the adsorption of RBBR is presented in Figure 2 where it is shown

the amount of dye adsorbed, in mmol・kg⁻¹ of dried fiber, for an initial concentration of the treatment of 0.5 mM of RBBR at two different temperature 303 K and 293 K. As it can be seen the maximum absorption takes place at the lowest pH, pH 4, and at the highest temperature tested. It is also relevant to realize, that the influence of the pH is much accurate at the lower temperature tested. In our opinion, as chemical interactions takes place at high temperature, in this situation, two mechanisms would be responsible of sorption process. Besides covalent bonding between vinyl sulfone group of RBBR and amine group of the fiber [22] , we must consider electrostatic interactions between ionized amino group, and the sulfonate terminal of RBBR,. But at higher pH, when the fiber presents negative charge on its surface, the lack of this last mechanism will justified the drop in the amount of RBBR uptaken by the fiber. At low temperature chemical interactions diminish, and this situation besides electrostatic repulsion give rises to the scarce adsorption observed at high pH and low temperature.

The results of adsorption of RBBR onto Nylon 6.6 at four different temperatures and at natural pH of solution that is about 6 units pH, are plotted in Figure 3.

As it can be observed sorption rises abruptly with temperature, being its value low and almost constant at low temperature, which is 293 K and 303 K. For temperature higher than this last one, there is an abrupt increase in the amount adsorbed, and it seems, because of the sharp of the curve, that at the highest temperature no saturation is reached in the conditions tested. The effect of the initial concentration is more pronounced the higher the temperature is. As we have point before, some reactive dyes, can be covalently fixed to Nylon because of the covalent bonds formed between the amino group of the fiber and the vinyl sulfone group of the dye, being the adsorption of dye by polyamide fibers highly influenced by temperature under transition temperature that is about 90˚C. As many researchers have observed [13] [22] [23] working with reactive dyes, initial dye uptake increases with temperature, and the influence of it is much marked at low pH, Figure 2, as we have just observed.

Thermodynamic data have been fitted to Langmuir and Freundlich theoretical model of adsorption. The results are depicted in Figure 4.

The evolution of the data obtained is similar to that presented by Ho and Mc. Kay working with Basic Blue 69 onto peat [23] . The values of X_m and K_L are favored by temperature, correlation coefficient keeps higher than 0.9 in all the range tested. The values of the theoretical monolayer capacity X_m (mmol・kg⁻¹) obtained from Langmuir model present good correlation with experimental data presented in Figure 3. This result seems to point to specific chemisorptions of dye onto the fiber and monolayer distribution.

In order to complete this information thermodynamic functions in the same conditions have been tested. The Standard Free Energy of sorption process obtained from Equation (5), is negative and increases its absolute value with temperature. These results confirm that the process is spontaneous, and favored by temperature. Furthermore, the change in Heat of Adsorption, Equation (6), determined at different intervals of temperature is always positive because the reaction is endothermic and temperature favors the process. As Minguand

suggested [11] , when dyeing activated carbon with cationic dyes, the electrostatic repulsion weakens the adsorption forces and then heat is necessary for it to take place. Finally the negative value of change of Entropy, Equation (7) seems to suggest a formation of ordered structures. As it could be appreciated in Figure 4, there is no possibility of fitting data to Freundlich model and therefore only Langmuir constant values were presented in Table 1.

The results of the kinetic experiences carried out at two temperatures are represented in Figure 5. These experiments have been performed with 1 g of clean and dry fiber and 250 ml of 0.5 mM RBBR as initial concentration of dye solution. Kinetic data have been fitted to pseudo-first order, pseudo-second order and diffusion model. Results are depicted respectively in Figure 6, Figure 7 and Figure 8, besides, in Table 2 there are listed the kinetic parameters.

Although correlation coefficient, R², of both, first and second order are high, the calculated q_e (amount adsorbed at equilibrium) values obtained for the first-order kinetic model are very similar to the experimental ones. In Table 2 is observed that the rate constant, k₁, as well as the amount adsorbed at equilibrium, q_e, increases with temperature. As Espinosa and cols. [17] have observed with Acid Blue 45 over Nylon 6.6, it can be observed that for the first stage of adsorption (up to 100 min) the process seems to respond to first order kinetic model.

Furthermore, when comparing the Langmuir results in Table 1 with equilibrium sorption capacity obtained from the kinetic analysis, it is possible to appreciate that at least at high temperature values are very similar, that is, Xm = 46.10 mmol・kg⁻¹ and q_e1 = 49.77 mmol・kg⁻¹ even they are similar to the experimental one that is 50.40 mmol・kg⁻¹. These observations can reveal that the reaction could be considered pseudo-first order.

Determination of the activation energy E^* of dye sorption process evaluated form the equation of Arrhenius-type:

gives rise to a value of 30.10 kJ・mol⁻¹, that means that the process needs from temperature to take place although the results are slightly significant because the fit has been done only with two points.

Fitting kinetics data to Fick equation in the form mentioned above, lead us to determination of the apparent diffusion coefficients which are listed in Table 2. This equation follows a model that is only valid when adsorption takes place with constant concentration of the adsorbate in the adsorbent surfaces and therefore the results

discussed are only relatives. The values obtained are similar to those from Espinosa-Jiménez [12] and we can point that diffusion is favored by temperature. In this situation, the value of the activation energy E^* for the diffusion process is 34.36 kJ・mol⁻¹, being this value quite similar to that obtained from adsorption process.

Sorption experiments of CHX over Nylon 6.6-RBBR system have been carried out, considering initial conditions of pH, temperature and concentration of the pretreatment of the fiber. The pH of the CHX aqueous medium has been kept between the interval 6.5 - 7, that correspond to the pH of CHX solutions as function of its concentration, because if it diminishes from this value the protonation of the terminal groups of the fiber could hinder the posterior adsorption of the amino groups of CHX. Besides, the competition between the amino groups of the fiber and the amino groups of the CHX molecule for the anionic RBBR molecule, leads to a release of the dye by the fiber, and the solution gets slightly blue. This effect is more pronounced the more acid the CHX solution is. This undesirable situation has been observed in other circumstances. Because of its importance we will describe it here.

When adsorption of CHX as function of temperature is analyzed, we have observed that at low temperature, and low initial concentration of CHX solution, desorption of dye molecules takes place. This effect is more noticeable the lower the concentration and the lower temperature is, as we have been able to verify by measuring RBBR concentration in supernatants solution, although the previously adsorption of dye has been carried at the optimum conditions tested (50 mmol・kg⁻¹ of RBBR over the dry fiber). When the fiber is previously treated with low concentration of dye, desorption is then not observed. From this observation we can infer that there could be described two possibility of sorption of dye onto the fiber, a strongly one possible due to chemical interactions and another one that leads to a weaker linkage fiber-dye molecule, maybe electrostatic interactions. The molecules adsorbed by this last mechanism would be the ones first desorbed in less favored sorption requirements for the reactive dye. In Figure 9, it is represented the amount of CHX adsorbed (mmol・kg⁻¹) at equilibrium as function of final CHX concentration solution, at three temperatures and at constant concentration of the pretreatment of 0.5 mM RBBR. It can be notice that the temperature favored the process and that a low temperature the amount adsorbed is scarce.

Results of the fitting of the adsorption data from Figure 9 Langmuir isotherms models, besides the values obtained for thermodynamic functions are shown in Table 3. As linear fits to Freudlich model exhibits lower correlation coefficients, in the following section only Langmuir results will be analyzed.

Negative value of which exhibits a small increase for a slight raise in temperature from 293 to 313 or 323, indicates that the process is spontaneous. Besides, it could be observed that the process is more endothermic the higher the temperature of the interval tested is, however, the values obtained are lower than those for RBBR sorption listed in Table 1. Finally, values of are negatives and similar to those obtained for sorption of

RBBR, in our opinion the influence of CXH in the structure of the adsorbent surface is very little. Values of Langmuir parameters obtained from adsorption data of Figure 9 listed in Table 3 shown that in this situation the influence of temperature is only appreciated for the lower one, although a slight increase of Langmuir constant, K_L, adsorption capacity, X_M, and a_L can be observed, however these values are lower than those of RBBR. We have also estimated the proportion of adsorption capacity of CHX respect of that of RBBR used in the pretreatment as function of the temperature of CHX adsorption, and it is remarkable that the value at 293 K represents half the value obtained at 323 K.

The influence that the RBBR concentration used in the pretreatment of the fiber exerts in the posterior adsorption of CHX is presented in Figure 10. Experiments have been performed at the higher temperature tested, 323K. It is possible to realize that at low concentration of RBBR in the pretreatment, the amount of CHX adsorbed is low, and this value experiments a marked increase for higher concentration of the pretreatment.

Dynamic parameters of the adsorption of 0.5 mM of CHX over Nylon pretreated with 0.5 mM of RBBR at 323 K, as function of two temperatures, have been determined following the same methods as has been done previously for the uptaken of RBBR over the fiber. The results are plotted in Figure 11 as well as in Figure 12 and Figure 13. The kinetic data obtained are shown in Table 4, these data are important in order to obtain information about the reaction involved in the process.

In this situation correlation coefficients for the pseudo-second order fitting, are higher than those of first order. Else, the coincidence between q_e values from pseudo-second order and the experimental ones are slightly higher. This fact could point to an activated sorption mechanism characteristic of chemical interaction, as has been suggested by Ho and Mc Kay [23] . Finally the fitting of data to diffusion model gives a low correlation coefficient possible because the linkage of the molecules to the surface avoids the diffusion of CHX into the fiber. Nevertheless the order of magnitude of the diffusion coefficient is similar to the one obtained by Espinosa-Jiménez, and cols. working with Nylon 6.6 [17] . Also, in this situation we appreciate that this factor increases with temperature.