SciELO - Scientific Electronic Library Online

 
vol.49 número2TOWARDS A QUANTUM MOLECULAR MEASUREMENT THEORY: STERN-GERLACH THOUGHT EXPERIMENTS AT THE INTERFACE OF HILBERT AND REAL SPACESSENSING PROPERTIES OF HYBRID POLYMERIC FILMS OBTAINED BY SOL-GEL índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Journal of the Chilean Chemical Society

versión On-line ISSN 0717-9707

J. Chil. Chem. Soc. v.49 n.2 Concepción jun. 2004

http://dx.doi.org/10.4067/S0717-97072004000200003 

  J. Chil. Chem. Soc., 49, N 2 (2004), pags.:121-126

USE OF ULTRAFILTRATION ON THE EVALUATION AND QUANTIFICATION OF THE INTERACTIONS BETWEEN POLYMERS AND LOW MOLECULAR-WEIGHT MOLECULES IN AQUEOUS SOLUTIONS.

 

Ignacio Moreno-Villoslada* a), Víctor Miranda a), Felipe Oyarzún a), Susan Hess a), Maribel Luna b), and Bernabé L. Rivas b).

A) Facultad de Ciencias, Universidad Austral de Chile, Valdivia, Chile. B) Facultad de Ciencias Químicas, Universidad de Concepción, Concepción, Chile.

Dirección para correspondencia


Abstract

The theory and mathematical treatment of the evaluation of the interactions of water-soluble polymers with low molecular-weight molecules is presented. The interaction of the water-soluble polymer poly(sodium 4-styrenesulfonate) with L-tryptophan (Try), L-phenylalanine (Phe), chlorpheniramine maleate (CPM), and 2,3,5-triphenyl-2H-tetrazolium chloride (TTC) is studied by this technique. Ultrafiltration experiments at pH 7.5, and different ionic strength values show weak interactions of the low molecular weight molecules TTC and CPM with the polymer with apparent dissociation constants of 0.26 and 0.32 respectively in the absence of NaNO3, and 2.02 and 1.2 in the presence of 0.10 of NaNO3 and 0.13 M of NaCl respectively. Negligible interactions are found for Try or Phe at pHs ranging between 4 and 10 and NaNO3 concentrations ranging between 0.0 and 0.15 M. This suggests that the interactions are mainly due to both long-range electrostatic forces and short-range interactions that may include hydrophobic forces or charge transfer complexation. The interaction of TTC with PSS in the absence of NaNO3 is reflected in UV-Vis spectroscopic studies by a decrease on the intensity of the maximum at 248 nm.


1. Introduction

The physical behavior of mixed solutions of polymers and low molecular-weight molecules (LMWM) is a large research area.1-11 The strong interest in these mixtures is due to their importance in biological processes and their widespread use in different industrial procedures with increasing potential applications. The behavior of these mixtures is determined by the respective interactions between all their components. Both water-soluble polymers and low molecular-weight molecules may form aggregates in solution. When these two kind of different species are mixed together, their mutual interactions may yield in phase separations, changes in surface tension, flocculation or dispersion, etc. At a molecular level, these changes are related to formation of higher-order structures, complex formation, conformational changes on the polymer, etc. The nature of the polymer-LMWM interactions may be electrostatic, hydrogen bond mediated, or hydrophobic.

Ultrafiltration has emerged as a useful technique to detect and quantify these interactions.12-17 This technique is based on the separation of particles whose size is greater than the ultrafiltration membrane pores (as water-soluble polymers, WSP) from smaller molecules (as LMWM). The rate of filtration of the LMWM under the washing method (analogue to a Batch method) is strongly influenced by their interactions with the water-soluble polymer. In this paper, the theory involved in the evaluation and quantification of the interactions between polymers and LMWM in aqueous solutions is introduced, and some experimental results are contrasted concerning the behavior of solutions of the polyelectrolyte poly(sodium 4-styrenesulfonate) (PSS) and the LMWM L-tryptophan (Try), L-phenylalanine (Phe), chlorpheniramine maleate (CPM), and 2,3,5-triphenyl-2H-tetrazolium chloride (TTC). All these LMWM show characteristic bands in UV-Vis spectroscopy that allow easily analyzing their concentration in water. The first two molecules are natural occurring quiral aminoacids. Chlorpheniramine maleate is an antihistaminic drug used for topical ointments, (especially for skin

disorders such as sunburn, urticaria, angioedema, pruritus, and insect bites) as well as in orally administrable formulations. Pharmaceutical investigations focus its inclusion in modified release matrixes.18-21 2,3,5-triphenyl-2H-tetrazolium chloride is used as indicator of reductive enzyme activity, as oxidizing agent of sugars and steroids, and as a reagent for photometric determination of metals. It possesses valuable staining properties in various biological applications, such as counting bacterial colonies, indicating viability of seeds, differential staining of tissues, or bacteriostatic action.22-25

2. Theory for ultrafiltration

2.1. Ultrafiltration in the absence of a water-soluble polymer; inert system: In this section we will analyze which are the expected results for ultrafiltration experiments where no interaction is affecting the elution of LMWM during filtration. Ultrafiltration experiments are performed keeping constant the volume of the solution inside the cell (Vcell), by creating a continuous flux of liquid across the membrane through the cell. A mass balance may relate the amounts of LMWM that traverse the membrane so that

cLMWMfiltrate dVf = -Vcell dcLMWMcell (1)

where cLMWMfiltrate is the instantaneous concentration of the LMWM in the filtrate, cLMWMcell is the concentration of the LMWM in the cell in every instant, and Vf is the volume in the filtrate. Let F be the filtration factor defined as the volume ratio of the filtrate, versus the volume in the cell, Vf / Vcell, so that dividing expression (1) by Vcell we obtain

- dcLMWMcell / dF = cLMWMfiltrate (2)

As no interaction is affecting the elution of the LMWM, cLMWMcell = cLMWMfiltrate, and then substituting and integrating

cLMWMfiltrate = cLMWMfiltrate-init exp(­ F) (3)

cLMWMcell= cLMWMcell-init exp(­ F) (4)

where init is referred to initial values (for F = 0).

2.2. Ultrafiltration in the absence of a WSP; no inert system: It is probable that any of the components of the ultrafiltration system interact with the LMWM, resulting in a difference between the total concentration of LMWM in the cell and in the filtrate. If a blank experiment (in the absence of the water-soluble polymer) is done, it is normally found that

cLMWMfiltrate = cLMWMfiltrate-init exp(­ km · F) (5)

where km is a constant. Applying equation (2) we have

cLMWMcell = cLMWMcell-init [um + vm exp(­km · F)] (6)

where

cLMWMfiltrate-init = cLMWMcell-init vm km (7)

um and vm are constants, and

um + vm = 1 (8)

As km is given by the slope of the plot of ln cLMWMfiltrate versus F in blank experiments and cLMWMcell-init is known, the value of vm is easily calculated. If vm = 1, and therefore um = 0, it is obtained that

km = cLMWMfiltrate / cLMWMcell (9)

This parameter is giving a measure of the difference on the rate of decay of the concentration of the LMWM in the cell as a consequence of its interaction with the system. If this interaction is only attributed to the membrane and provided that km £ 1 the parameter is named the coefficient of membrane retardation. Note that if vm = km = 1, the system behaves as the inert system described in section 2.1.

2.3. Ultrafiltration in the presence of a water-soluble polymer: Upon interaction with the ultrafiltration system a decrease on the rate of decay of LMWM in the cell is observed. It could be expected that the same effect would be produced by the WSP interacting with the LMWM. Experiments show that

cLMWMfiltrate = cLMWMfiltrate-init exp(­ j · F) (10)

and applying equation (2) we also have that

cLMWMcell = cLMWMcell-init [u + v exp(­j · F)] (11)

where

cLMWMfiltrate-init = cLMWMcell-init v j (12)

j, u and v are constant experimental parameters, and

u + v = 1 (13)

The binding processes of the LMWM to the polymer may be due to both long- and short-range interactions. We define as a LMWM bound to the polymer (cLMWMbound) the molecule that is attracted to the polymer with such a strength that will not pass through the ultrafiltration membrane. On the contrary, the LMWM free in the solution (cLMWMfree) are able to pass through the ultrafiltration membrane. Only these free LMWM are susceptible to traverse the membrane, so that, expression (9) should be reformulated to

km = cLMWMfiltrate / cLMWMfree (14)

Then, combining equations (10-12, 14) and provided that cLMWMcell = cLMWMbound + cLMWMfree we obtain that

(15)

(16)

which are the expressions of the LMWM distributions (bound to the polymer or free in solution). The u value is the fraction of cLMWMcell-init that remains in the cell at F = ¥, while v represents the fraction of cLMWMcell-init that traverses the membrane during filtration. The j parameter gives a measure of the ability of the polymer phase to transfer LMWM to the aqueous phase during filtration. When j = km, the amounts of LMWM bound to the polymer do not change during filtration and the free LMWM are filtered out of the ultrafiltration cell. If j < km, the equilibrium in the solution inside the cell is furnishing LMWM to the aqueous phase from the polymer phase while they are being filtered out during filtration. In this case, the amounts of LMWM bound to the polymer decrease during filtration until the value cLMWMcell-init u is achieved. It can be then considered that this limit value corresponds to the LMWM irreversibly bound to the polymer. Here we mean for irreversible bound that the dissociation constant is low enough that it can not to be detected in the ultrafiltration experiment. The cLMWMcell-init v fraction may be reversible bound to the polymer, and an apparent dissociation constant is defined as

(17)

where (cLMWMbound)rev = cLMWMbound - cLMWMcell-init u.

On the contrary, if j > km, apart from the loss of LMWM in the aqueous phase because of filtration, the polymer is capturing free LMWM from the solution. This situation is normally found when the ionic strength is allowed to decrease during filtration. In these cases, the amount of LMWM bound to the polymer increases during filtration, and the dissociation constant becomes a function of F following the expression14

(18)

From equations (15), (16), and (18), it can be deduced that vj £ km, in order not to obtain negative amounts of LMWM bound to the polymer, concentration of free LMWM larger than the total concentration in the cell, or negative dissociation constants, respectively.

Deviations of the ln cLMWMfiltrate from linearity when plotted versus F may indicate saturation of the polymer, conformational changes in the macromolecules, or changes in the predominant species present in the equilibrium.12

2.4. Mean and instantaneous concentrations: In practice, the filtrate is collected in several filtration fractions, and the concentration of LMWM in the fractions is analyzed. When volume equivalent fractions (DV) are collected it is normally found that

<cLMWMfiltrate > = <cLMWMfiltrate>init exp(­ j · F) (19)

where <cLMWMfiltrate> is the concentration of the LMWM in the filtration fractions, a value greater than the concentration of the last drop in the fraction and smaller than the first. Since in the limit of small volume fraction equation (19) becomes equation (10), a mass balance can be calculated so that

(20)

where DF is DV / Vcell.

Making Fa = Fb - DF, the instantaneous concentration of the LMWM in the filtrate is then given by the following expression:

(21)

Combining equations (14), (16) and (21), is clear that

(22)

This is an interesting relation, since the fraction of LMWM that will remain bound to the polymer at F = · can be calculated without achieving F values so high.

3. Experimental Section

3.1. Reagents. Commercially available poly(sodium 4-styrenesulfonate) (PSS) (Aldrich, synthesized from the para-substituted monomer) was fractionated by ultrafiltration over a membrane of a molecular weight cut-off (MWCO) of 100,000 Dalton (Biomax, 63.5 mm diameter), in the presence of 0.15 M NaNO3 as a firs step, and then in the absence of the electrolyte in order to remove the excess of NaNO3. NaNO3 (Merck), NaCl (Merck), Try (Sigma), Phe (Sigma), CPM (Munnich, provided as a racemic mixture) and TTC (Merck) were used to prepare the solutions. The structures of PSS, Try, Phe, CPM, and TTC are shown in Figure 1. The pH was adjusted with NaOH and HNO3.


 
Fig. 1. Molecular structures

3.2. Equipment. The unit used for ultrafiltration studies consisted of a filtration cell (Amicon 8010, 10 mL capacity) with a magnetic stirrer, a polyethersulfone membrane with a MWCO of 10,000, or 100,000 Dalton (Biomax, 25 mm diameter), a reservoir, a selector, and a pressure source (see Figure 2). The pH was controlled on a Quimis Q400M2 pH meter, and on a portable Oyster pH meter. UV-Vis experiments and analyses were performed in a UNICAM UV5 spectrophotometer at 20C and 1 cm of path length.


 
Fig. 2. Ultrafiltration equipment: 1. filtration cell; 2. ultrafiltration membrane; 3. magnetic stirrer; 4. pressure source; 5. selector; 6. reservoir.

3.3. Procedure for ultrafiltration. Polymeric molecular weight fractions over 100,000 Dalton were dissolved in twice distilled water together with NaNO3 and the corresponding LMWM to obtain the concentrations showed in Table 1. The solutions (10 mL) were placed into the ultrafiltration cell. The pH and the concentration of NaNO3 or NaCl of the aqueous solution contained in the reservoir were adjusted to the same value as in the cell solution. In order no macromolecule is filtered, the filtration runs were carried out over a membrane with a molecular weight cut-off of 10,000 Dalton under a total pressure of 3 bar, keeping constant the solution volume in the cell by creating a continuous flux of liquid through the cell solution from the reservoir. Filtration fractions (ranging between 6.0 and 8.0 mL) were collected and LMWM concentrations analyzed by UV-Vis spectroscopy. No macromolecule was found in the filtrates. Blank experiments were performed with the same procedure, in the absence of the water-soluble polymer (Table 1).


Table 1. Values of the experimental variables for ultrafiltration experiments, and linear adjustments for the corresponding results: y = ln <cLMWMfiltrate>; x = F; R2 = linear regression factor.

Experiment NaNO3 a) or
NaCl b)
conc.c) (M)
Initial
LMWM
conc. (M)
PSS conc.
(momom.
units) (M)
pHc) Linear adjustments for
the experimental data
R2

TTC-01 - 0.001 - 7.5 y = -0.966x ­ 6.77 0.996
TTC-02 0.10 a) 0.001 - 7.5 y = -0.819x ­ 6.85 0.996
TTC-03 - 0.001 0.002 7.5 y = -0.202x ­ 9.88 0.946
TTC-04 0.10 a) 0.001 0.002 7.5 y = -0.548x ­ 7.38 0.996
CPM-01 - 0.001 - 7.5 y = -0.952x ­ 6.64 0.999
CPM-02 0.13 b) 0.001 - 7.5 y = -1.00x ­ 6.62 0.999
CPM-03 - 0.001 0.002 7.5 y = -0.231x ­ 9.34 0.947
CPM-04 0.13 b) 0.001 0.002 7.5 y = -0.546x ­ 7.30 0.999
Try-01 - 0.001   7.0 y = -0.915x ­ 6.81 0.995
Phe-01 - 0.01   7.0 y = -0.944x ­ 4.35 0.995

c) Values for both the cell solution and the reservoir solution.

For LMWM analyses, calibration curves were obtained at the conditions given in Table 2.


Table 2. Calibration curves for UV-Vis spectroscopic analyses: y = absorbance; x = [LMWM]; R2 = linear regression factor, at wavelengths 248 nm (TTC), 262 nm (CPM), 280 nm (Try), 258 nm (Phe).

Molecule NaNO3
conc. (M)
Calibration curve R2 pH Concentration range
(M)

TTC - y = 26305 x 1.00 7.5 [2·10-6, 4·10-5]
TTC 0.10 y = 24608 x 1.00 7.5 [2·10-6, 4·10-5]
CPM - y = 5040.3 x 1.00 7.5 [2·10-5, 4·10-4]
CPM 0.13 y = 5200.8 x 1.00 7.5 [2·10-5, 4·10-4]
Try - y = 5507.2 x 1.00 7.0 [1·10-5, 1.5·10-4]
Try 0.10 y = 5418.7 x 1.00 7.0 [1·10-5, 1.5·10-4]
Phe - y = 167.8 x 1.00 7.0 [2.5·10-4, 5·10-3]
Phe 0.10 y = 174.2 x 1.00 7.0 [2.5·10-4, 5·10-3]

4. Results and Discussion

The water-soluble poly(sodium styrenesulfonate) (PSS) has been taken as model of a polyelectrolyte (see Figure 1). PSS undergoes electrostatic interactions with counterions. Due to the hydrocarbon nature of this water-soluble polyelectrolyte, it can also undergo hydrophobic interactions with LMWM. Ultrafiltration of the LMWM in the absence of PSS (blank experiments) shows negligible interactions between the ultrafiltration system and the four species at all the pH and ionic strengths considered, since the values of vm and km are very close to 1 (see Table 3).


Table 3. Ultrafiltration experimental parameters, apparent dissociation constants, and its inverse, the apparent association constant at polymer concentrations of 0.002 M in monomeric units.

Experiment n u j km KLMWMdiss KLMWMass

TTC-01 0.83 0.17 - 0.966 - -
TTC-02 0.96 0.04 - 0.819 - -
TTC-03 0.24 0.76 0.202 - 0.26 3.84
TTC-04 0.92 0.08 0.548 - 2.02 0.50
CPM-01 0.99 0.01 - 0.952 - -
CPM-02 0.94 0.06 - 1.00 - -
CPM-03 0.34 0.66 0.231 - 0.32 3.13
CPM-04 1.04 - 0.546 - 1.2 0.83
Try-01 0.81 0.19 - 0.915 - -
Phe-01 0.96 0.04 - 0.944 - -

As an example, the data for the blank experiments performed at pH 7 for Try and Phe, and 7.5 for CPM and TTC, in the absence of NaNO3 are plotted in Figure 3.


 
Fig. 3. Plot of the ln <cLMWMfiltrate> versus F for experiments performed in the absence of PSS and NaNO3 and linear adjustments; (¨) TTC; () Try; ( ) Phe; (C) CPM.

The interaction of these LMWM with PSS is analyzed under different pH and ionic strength conditions. Negligible interactions (or undetectable by ultrafiltration at these conditions) were found for both Try and Phe. The aminoacids are zwitterionic species showing an overall null charge at the isoelectric point (5.89 and 5.91, respectively) so that long-range electrostatic attractive forces are thus not produced. These two aminoacids show pKa values of pK1= 2.38, pK2= 9.39 for Try, and pK1= 2.58, pK2= 9.24 for Phe. Even at pH 4, the concentration of net positively charged molecules is very low. Thus, the ultrafiltration results in the presence of PSS were in practice equivalent to those of the blank experiments (data not shown). On the contrary, both TTC and CPM show positive charges on their structure, so that long-range electrostatic interactions occur as is shown in Figure 4 where it can be seen that the filtration rates decrease. When no added NaNO3 or NaCl is present in the solution, j values of 0.23 for CPM and 0.20 for TTC are found (experiments TTC-03 and CPM-03). Nevertheless, it can be noticed a deviation of the experimental points from linearity. This may be attributed to the elution of the Na+ counterions of PSS yielding higher attraction forces, and a change on the predominant species in solution from long-range attracted molecules to molecules subjected to short-range electrostatic attraction forces. Assuming the approximation of a linear behavior, it can be noticed in Table 3 that a large percentage of the total initial TTC and CPM remains irreversible bound to PSS (76 and 66 % respectively). Long-range interactions are sensitive to the presence of other ions in solution. The addition of NaNO3 or NaCl to both the initial solution and the reservoir solution produces screening of the long-range electrostatic interactions as observed for experiments TTC-04 and CPM-04. However, it can be seen in Figure 4 that despite the addition of a large excess of Na+ in the solution, the interactions of TTC and CPM are still remarkable, suggesting a strong influence of the hydrophobicity and / or other type of short-range interactions as hydrogen bonding or charge transfer complexation. The magnitude of the dissociation constants is given in Table 3, following equation (17). It can be seen that at these concentration ranges, the apparent dissociation constants obtained in the presence of NaNO3 or NaCl are one order of magnitude higher than those obtained in the absence of the electrolyte.


 
Fig. 4. Plot of the ln <cLMWMfiltrate> versus F for experiments performed in the absence of PSS and in the presence of 0.1 M NaNO3 for TTC (·) and 0.13 M NaCl for CPM (); in the presence of 2·10-3 M of PSS (in monomeric units) and in the absence of NaNO3 or NaCl for TTC (¨) and CPM ( ); and in the presence of 2·10-3 M of PSS (in monomeric units) and in the presence of 0.1 M NaNO3 for TTC () and 0.13 M NaCl for CPM (C).

Although ultrafiltration experiments show an interaction between TTC and CPM with PSS in water, direct observation of the interaction by UV-Vis spectroscopy yields in changes in the spectrum profile for TTC in the absence of NaNO3 (see Figure 5), but these changes are not observed in the presence of 0.1 M of NaNO3 for TTC, neither for CPM in the presence of 0.13 M of NaCl or in the absence of the simple electrolyte. It can be seen that the presence of the polymer produces a decrease on the intensity of the absorbance of TTC of the maximum at 248 nm of 32 % at pH 7.5 and absence of NaNO3. This decrease is still found independently of the TTC concentration ranging between 2·10-6 to 4·10-5 M.


 
Fig. 5. UV-Vis spectra of PSS (···), 4·10-5 M of TTC (a), 4·10-5 M of TTC in the presence of 8·10-5 M of PSS in monomeric units (b), 2.2·10-4 M of CPM (c), 2.2·10-4 M of CPM in the presence of 4.4·10-4 M of PSS in monomeric units (d), and 2.2·10-4 M of CPM in the presence of 4.4·10-4 M of PSS in monomeric units and 0.13 M of NaCl (e).

5. Conclusions

The interaction of the water-soluble polymer poly(sodium 4-styrenesulfonate) with L-tryptophan (Try), L-phenylalanine (Phe), chlorpheniramine maleate (CPM), and 2,3,5-triphenyl-2H-tetrazolium chloride (TTC) in aqueous media was studied by ultrafiltration under a theoretical basis presented. Negligible interactions were found for Try or Phe at pHs ranging between 4 and 10 and NaNO3 concentrations ranging between 0.0 and 0.15 M. On the contrary, long-range electrostatic interactions were found at pH 7.5, and different ionic strength values for TTC and CPM with apparent dissociation constants of 0.26 and 0.32 in the absence of NaNO3 or NaCl, and 2.02 and 1.2 in the presence of 0.10 of NaNO3 and 0.13 M of NaCl respectively. The noticeable interaction in the presence of an excess of NaNO3 may suggest an important role of hydrophobic interactions or charge transfer complexation in the overall observed interaction. The interaction of TTC with PSS in the absence of NaNO3 is reflected in UV-Vis spectroscopic studies by a decrease on the intensity of the maximum at 248 nm.

 

Acknowledgment

The authors thank Fondecyt (Grants No 1030669 and No 1020198), and the Dirección de Investigación of the Universidad Austral de Chile (Grant No S-200126) for financial support.

REFERENCES

1. T. J. Dickerson, N. N. Reed, K. D. Janda, Chem. Rev. 102, 3325- 3344 (2002).         [ Links ]

2. D. E. Bergbreiter, Chem. Rev. 102, 3345-3384 (2002).         [ Links ]

3. A. Svensson, L. Piculell, B. Cabane, Ph. Ilekti, J. Phys. Chem. B 106, 1013-1018 (2002).         [ Links ]

4. D. Dhara, P. R. Chatterji, J. Macromol. Sci.-Rev. Macromol. Chem. Phys. C40, 51-68 (2000).         [ Links ]

5. T. Kaneko, S. Orita, J. P. Gong, Y. Osada., Langmuir 15, 5670- 5675 (1999).         [ Links ]

6. L. Masaro, X. X. Zhu, Langmuir 15, 8356-8360 (1999).         [ Links ]

7. W. B. Heuer, H. S. Lee, O.-K. Kim, Macromol. Symp. 138, 79-84 (1999).         [ Links ]

8. S. Nilsson, K. Thuresson, P. Hansson, B. Lindman, J. Phys. Chem. B 102, 7099-7105 (1998).         [ Links ]

9. A. Harada, Adv. Polym. Sci. 133, 141-191 (1997).         [ Links ]

10. M. Tanaka, Y. Asahi, S. Masuda, J. Macromol. Sci.-Pure Appl. Chem. A32, 339-347 (1995).         [ Links ]

11. H. S. Soedjak, Anal. Chem. 66, 4514-4518 (1994).         [ Links ]

12. I. Moreno-Villoslada, C. Muñoz, B.L. Rivas, J. Polym. Sci. B, Polym. Phys. 40, 2587-2593 (2002).         [ Links ]

13. B. L. Rivas, E. Pereira, I. Moreno-Villoslada, Prog. Polym. Sci. 28, 173-208 (2003).         [ Links ]

14. B. L. Rivas, E. Pereira, I. Moreno-Villoslada, J. Phys. Chem. B 102, 6994-6999 (1998).         [ Links ]

15. B. L. Rivas, E. Pereira, I. Moreno-Villoslada, J. Phys. Chem. B 102, 11024-11028 (1998).         [ Links ]

16. Y. Uludag, H. O. Özbelgge, L. Yilmaz, J Membrane Sci. 129, 93- 99 (1997).         [ Links ]

17. P. Vonk, R. Noordman, D. Schippers, B. Tilstra, H. Wesselingh, J Membrane Sci. 130, 249-263 (1997).         [ Links ]

18. Ç. Tas, Y. Özkan, A. Savaser, T. Baykara, Il Farmaco 58, 605-611 (2003).         [ Links ]

19. D. S. Roy, B. D. Rohera, Eur. J. Pharm. Sci. 16, 193-199 (2002).         [ Links ]

20. C. W. Rowe, W. E. Katstra, R. D. Palazzolo, B. Giritlioglu, P. Teung, M. J. Cima, J Control. Release 66, 11-17 (2000).         [ Links ]

21. L. Siepmann, H. Kranz, N. A. Peppas, R. Bodmeier, Int. J. Pharm. 201, 151-164 (2000).         [ Links ]

22. A. W. Ninehham, Chem. Rev. 55, 355-483 (1955).         [ Links ]

23. B. Jámbor, Nature 176, 603 (1955).         [ Links ]

24. O. W. Maender, G. A. Russell, J. Org. Chem. 31, 442-446 (1966).         [ Links ]

25. M. C. González, E. San Román, J. Phys. Chem., 93, 3536-3540 (1989).         [ Links ]

 

Correspondencia a: e-mail: imorenovilloslada@uach.cl

(Received: November 5, 2003 ­ Accepted : December 16, 2003)