- Email: [email protected]
Incorporation of low molecular weight biocides into polystyrene–divinyl benzene beads with controlled release characteristics
Journal of Controlled Release 102 (2005) 223 – 233 www.elsevier.com/locate/jconrel
Incorporation of low molecular weight biocides into polystyrene–divinyl benzene beads with controlled release characteristics S.M. Iconomopouloua, A.K. Andreopouloua,b, A. Sotoa, J.K. Kallitsisa,b, G.A. Voyiatzisa,* a
Foundation for Research and Technology-Hellas (FORTH) Institute of Chemical Engineering and High Temperature Chemical Processes (ICE/HT), P.O. Box 1414, GR-265 04 Rio-Patras, Greece b Department of Chemistry, University of Patras, GR-265 04 Rio-Patras, Greece Received 29 July 2004; accepted 7 October 2004 Available online 10 November 2004
Abstract Triclosan and phosphonium salt biocides have been separately incorporated into polystyrene–divinylbenzene (PS–DVB) beads by suspension polymerization. Ultraviolet (UV) absorption measurements have been used to monitor the release of these low molecular weight biocides out of the PS–DVB beads immersed in water–ethanol mixtures and in physiological saline. The release of the biocide agents is strongly dependent on either the DVB or/and the antimicrobial composition ratio in the beads. An increase of biocide incorporation in the PS/DVB beads was accompanied by a corresponding enhancement of its concentration in liquid mixtures. On the contrary, higher cross-linking densities hindered the biocide migration out of the beads by diminishing its release rate into either the aqueous ethanol solutions or the natural serum. Moreover, Fourier transform Raman (FT-Raman) spectra and Attenuated Total Reflectance Infrared (ATR-FTIR) measurements of the PS–DVB–Triclosan and PS–DVB–phosphonium salt beads, before and after their immersion in water–ethanol solutions, gave a similar qualitative evidence of the biocide release. D 2004 Elsevier B.V. All rights reserved. Keywords: PS–DVB beads; Low molecular weight biocide incorporation; Release of biocide; UV absorption; Vibrational spectroscopy
1. Introduction
* Corresponding author. Tel.: +30 2610 965253; fax: +30 2610 965223. E-mail address: [email protected] (G.A. Voyiatzis). 0168-3659/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2004.10.006
Polymeric materials having antimicrobial or antifouling activity have been developed during the last years since the interest about health and hygiene habits is growing. Depending on the way that the incorporation of the active species into the polymeric
224
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
matrix takes place, these materials have been classified into two major categories. First, the immobilized type, where the active antimicrobial groups are covalently attached onto the polymeric substrate and the final antimicrobial efficiency is based on the contact between the biocide and the microorganisms. Typical representatives of this category are polymers bearing quaternary ammonium salts [1–4], phosphonium salts [5–7] or pyridinium type groups [8,9]. The second category is pertaining to a biocide-controlled release type after the incorporation of the antimicrobial agent into a host polymeric matrix. The latter contains the active species either as counter ion of polymeric bounded charges or as dispersion of low molecular weight active compounds [10]. Triclosan, 2,4,4V-trichloro-2V-hydroxydiphenyl ether, is known to be a broad spectrum antibacterial and antifungal agent [11,12]. Thereby, it is widely used in formulations such as soaps, toothpastes, deodorants, cosmetics, toys, carpets and kitchenware. Moreover, it is used as an additive in textiles and polymers providing to these materials its antimicrobial properties. In vivo inhibition studies have shown [13] that this low molecular weight biocide is active even at very low concentrations. Triclosan acts mainly in a specific target, the enoyl reductase enzyme, which is involved in the synthesis of fatty acids [14]. Phosphonium as well as quaternary ammonium salts are cationic biocides that target the bacterial membranes, as well. Their antimicrobial activity is based on their ability to disrupt and disintegrate the negatively charged cell membranes. In some cases, phosphonium salts show [15] higher activity than the quaternary ammonium salts; this characteristic, combined with their inherent higher thermal stability, makes them potential candidates to be used in polymer formulations. Insoluble cross-linked polymers bearing active antimicrobial groups have been widely reported in the literature including quaternary ammonium type resins [16,17]. However, no reference has been made for the incorporation of a low molecular weight biocide inside a polymer shell like the one formed by cross-linked beads. This method may improve the thermal stability of the end-used material, reduce the volatility of the biocide agent and, moreover, control the leaching-out rate of the antimicrobial entity.
The target of this work is to incorporate low molecular weight biocides, like Triclosan and phosphonium salts, into PS/DVB beads and investigate their release rate in either water–ethanol solutions or physiological saline.
2. Materials and methods 2.1. Materials Triclosan (Irgarguard B 1000) was kindly supplied from Ciba Specialty Chemicals. Tri-n-butyl phosphine, dodecylbromine and 5-sulphoisophthalic acid sodium salt were purchased from Aldrich. All other chemicals and reagents were purchased from Aldrich or Merck and used without purification unless otherwise noticed. Benzoyl peroxide (BPO) was recrystallized from CHCl3/MeOH. Styrene was distilled from CaH2 under reduced pressure, prior to use. Suspension polymerizations were carried out under an argon atmosphere. 2.2. Synthesis of phosphonium salts The phosphonium derivative of tri-n-butyl phosphine was synthesized by reaction with alkyhalides bearing alkyl chains with 12 carbon atoms [18]. The desired 5-sulfoisophthalic acid dimethyl ester tributylalkylphosphonium salt was synthesized by the reaction of the abovementioned phosphonium salt with dimethyl 5-sulfoisophthalate sodium salt [19]. 2.3. Synthesis of PS–DVB–biocide beads In an attempt to create polymeric matrixes with the ability of controlled releasing, antimicrobial substances have been incorporated into cross-linked polystyrene beads using suspension polymerization conditions. More specifically, PS beads of various degrees of cross-linking were synthesized carrying either Triclosan or phosphonium salt amounts as the antimicrobial entity (Table 1). Since Triclosan as well as phosphonium salts can be easily dissolved in styrene, the suspension polymerization technique provides a unique opportunity to encapsulate these active substances into a polymeric more or less bstorage cellQ. In order to examine possible controlled release of both
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
225
Table 1 The PS/DVB/biocide beads formulations synthesized
2.4. Instrumentation
Samples
The electronic absorption spectra were measured with a double beam Lambda 900 UV/Vis/NIR Spectrophotometer of Perkin Elmer. Reference beam was always delivered through an optical cell fill with the corresponding biocide-free ethanol or sodium chloride 0.9% aqueous solution. The spectra were recorded from 500 to 190 nm with a slit width of 0.5 nm. Fourier transform Raman (FT-Raman) measurements were obtained using a Bruker (D) FRA-106/S component attached to an EQUINOX 55 spectrometer. A R510 diode pumped Nd:YAG laser at 1064 nm (with a maximum output power of 500 mW) was used for Raman excitation in a 1808 scattering sample illumination module. An optical filtering reduced the Rayleigh elastic scattering and, in combination with a CaF2 beamsplitter and a high sensitivity liquid N2cooled Ge-detector, allowed the Raman intensities to be recorded from 50 to 3300 cm1 in Stokes-shifted Raman region, all in one spectrum. Attenuated Total Reflectance mid-infrared spectra were measured utilizing the single reflection Diamond (Miracle) crystal of the Horizontal ATR accessory of Pike Technologies (USA) attached to the sample compartment of a Bruker (D) Equinox 55 FTIR spectrometer bearing a DTGS detector.
1 2 3 4 5 6 7 8 9
PS–DVB–biocide beads %DVB
%Triclosan
2 2 2 5 2 5 2 5 10
0 10 5 10
%Phosphonium salt 0
10 10 20 20 20
Triclosan and phosphonium salt as a function of the degree of cross-linking, PS beads with different crosslinking densities (2–10% DVB) were synthesized. Moreover, and since the biocide content is expected to influence the porous’ nature of these beads, various amounts (5–20%) of either Triclosan or phosphonium salt were combined along with the different degrees of cross-linking. PS–DVB–Triclosan and PS–DVB– phosphonium salt beads were synthesized by suspension polymerization according to the following procedure that is given as an example of a typical polymerization for the preparation of the PS–DVB 2%–Tr 10% type beads. Gelatin (0.48 g) dissolved in hot, distilled and degassed water (100 mL), NaOH 10N (0.65 mL), H3BO3 (0.40 g) and Bentonite (0.70 g) were added into a three-necked round flask, equipped with a mechanical stirrer, a thermometer, a reflux condenser and an argon inlet. The mixture was stirred at room temperature for 30 min. Afterwards, 2 g of BPO were suspended in another mixture of 28 mL styrene, 0.65-mL divinylbenzene and 2.9 g Triclosan and it was added to the initial mixture. The suspension thus formed was vigorously stirred at 80 8C for 8 to 12 h. This procedure, in the case of Triclosan, was prolonged to 1 or 2 days to avoid the presence of excess styrene in the beads. After cooling to room temperature, the resulting PS–DVB–biocide beads were filtrated, washed carefully with water and dried under reduced pressure at ambient temperature to constant weight. In the case of PS–DVB–phosphonium salt beads for the removal of the excess of styrene, due to the low vapour pressure of the phosphonium salt, the last step of drying was done under reduced pressure at 50 8C to constant weight.
2.5. Release studies In order to probe the release of the biocides incorporated into PS–DVB beads by UV–visible absorption spectroscopy, PS–DVB–biocide beads have been immersed into different water–ethanol mixtures (10%, 50% and 95%) and into physiological saline in such amount that in the case of total release of the biocide its final concentration in the different solutions to be 0.02 g/100 mL. Nevertheless, a note is made of the fact that the maximum dissolution of the biocides in the water–ethanol solution 10% and in the physiological saline was found (a) for Triclosan equal to 0.0012 and 0.0005 g/100 mL, respectively, and (b) for the phosphonium salt equal to ~0.0085 g/100 mL, in both cases. Anyway, in the present study the degree of dissolution anticipated for the biocidal solutions was a compromise between the above imposed saturation concentrations and the detection limits of the spectroscopic techniques utilized. More precisely,
226
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
a quantity of around 0.1 to 0.4 g of beads, depending on the composition, was placed in a volumetric flask and filled with aqueous ethanol solution or with physiological saline to a total volume of 100 mL. Whenever UV absorption measurements have been scheduled, 2 mL of the solution was transferred to the sample compartment of the spectrophotometer and after the UV absorption measurement it was decanted back to the volumetric flask. For the vibrational spectroscopic measurements, similar but separate experiments were carried out in order to avoid disturbance of the UV-absorption measurements and the release study in progress. More exactly, a quantity of the solid was removed and dried in a vacuum oven at room temperature prior to the acquisition of the vibrational, FT-Raman or ATR-FTIR, spectra.
Fig. 1. FT-Raman spectra of Triclosan, Tr (top), phosphonium salt, Psalt (bottom), PS–DVB 2%, PS–DVB 2%–Tr 10% and PS–DVB 2%–Psalt 10%. Characteristic Raman bands of the biocides are marked with arrows.
3. Results and discussion 3.1. Characterization of biocides Triclosan is a chlorinated aromatic ether, while the phosphonium salt used in present work is a sulfoisophthalate–phosphonium salt. The molecular structures of both antimicrobial agents are depicted in Scheme 1. The presence of biocides in the beads, before and after their immersion in either water– ethanol or sodium chloride 0.9% solutions, might be monitored spectroscopically. To this regard, the Raman spectra of the Triclosan (Tr) and the phos-
Scheme 1. Molecular structures of Triclosan (1) and cationic phosphonium salt (2).
phonium salt (Psalt) are respectively depicted on the top and on the bottom of Fig. 1, in the corresponding Raman Stokes spectral range. For the Triclosan, two main spectral features are distinguished: the low frequency peak at 710 cm1 that was tentatively attributed to the skeletal (C–C) stretching vibration of trisubstituted benzene rings and the high frequency band at 3073 cm1 accurately assigned to the C–H stretching of the phenyl rings. The assignment of the low-frequency peak to benzene-chlorine (Ar-Cl) vibration mode or even to (Ar-O-Ar) ether grouping is less probable but it cannot be securely ruled out. Also in Fig. 1, the Raman spectra of PS–DVB 2%– Triclosan 10%, PS–DVB 2% and PS–DVB 2%–Psalt 10% beads are shown. It is obvious that the appropriate Raman band for the identification of the presence of Triclosan in the beads is that located at 710 cm1, since the high frequency scattering peak at ~3070 cm1 is overlapped with the relevant spectral contribution of the phenyl rings of both PS and DVB components of the beads. In the case of the phosphonium salt, the most identifiable spectral feature is that of the carbonyl stretching at ~1726 cm1. In the physiological saline, the occurrence of this vibrational mode is due to the presence of the 5sulfoisophthalate anion of the sulfoisophthalate–phosphonium salt (see Scheme 1), which gets soluble after the immersion of the cationic part of the phosphonium salt, thus indirectly giving evidence for the release of
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 2. ATR-FTIR spectra of Triclosan, Tr (top), phosphonium salt, Psalt (bottom), PS–DVB 2%, PS–DVB 2%–Tr 10% and PS–DVB 2%–Psalt 10%. Characteristic IR bands of the biocides are marked with arrows.
the latter. In the case of water–ethanol solutions, both parts are soluble simultaneously. In the same context, the ATR-FTIR spectra of both biocides and of the PS–DVB 2% and PS–DVB 2%– biocide 10% beads are depicted in Fig. 2 in the blow frequency rangeQ. Comparing the vibrational absorption spectra of each biocide with the corresponding biocide containing beads, it is evident that both antimicrobial agents exhibit specific infrared vibrational features that might be used to monitor their release out of the polystyrene beads. More precisely, Triclosan might be identified in the beads through the
227
infrared peaks at 860 and 1476 cm1 attributed to the out of plane rocking vibration of hydrogen atoms attached to aromatic ring and to the CMC stretching vibration of benzene rings, respectively. In the case of the phosphonium salt, it might be detected via the vibrational peaks located at 625, 1040 and 1726 cm1 and attributed in that order to C–P, -SO3 and CMO stretching. All above absorption bands are unique spectral features of the biocides with no overlapping vibrational absorption counterparts of the beads. In this context, they are used to determine the biocide remaining after the immersion of either the Triclosan or the phosphonium salt containing beads into water– ethanol solutions or physiological saline. Besides that, it must be noticed that the penetration depth of both vibrational spectroscopic techniques is in the order of micrometer, accounted most probably for bulk than surface sensitive measurements. Preliminary UV absorption spectra have indicated that after the immersion of the corresponding beads in the water–ethanol liquid mixtures, the concentration of both biocides was enhanced in the richer in ethanol solutions. This quick release is in good agreement with the better solubility of both biocides in ethanol compared to that in water. Ethanol most probably influences both the migration of the biocides from the bulk to the bead surface and its release into the water– ethanol solution. In Fig. 3a, the UV absorption spectra of Triclosan that has been released/dissolved in aqueous ethanol solution 95%, after impregnation for 1 week of beads that derived from the suspension
Fig. 3. UV–Visible absorption spectra of (a) ethanol solutions 95% in which three different beads–Triclosan compositions have been immersed for 1 week and (b) ethanol solutions 50% in which three different beads–phosphonium salt compositions have been immersed for 12 h.
228
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 4. FT-Raman spectra of three different PS–DVB–Psalt beads compositions before (1) and after (2) their immersion in ethanol solutions 50% for 12 h. Circle marks the characteristic band of the biocide phosphonium salt at 1726 cm1.
with the same either cross-linking agent or biocide composition has been vibrationally monitored after their immersion in water ethanol solution 50% for 12 h. The corresponding FT-Raman and ATR-FTIR spectra before and after their immersion in the ethanol solution are depicted in Figs. 4 and 5, respectively. In both cases, the relative intensities of the abovementioned spectral features of the phosphonium salt, at 1726 cm1 for Raman (Fig. 4) and at 625, 1040 and 1726 cm1 for infrared (Fig. 5), reveal the increased release rate of the biocide in the water–ethanol solution 50%, with higher biocide or with lower cross-linking agent compositions in the beads, in close comparison with the relevant UV absorption measurements shown in Fig. 3b. Similar vibrational results have been obtained from relevant Triclosan incorporated into PS–DVB beads. 3.2. Release studies
polymerization of various styrene–DVB X%–Triclosan Y% (X=2 and Y=10; X=5 and Y=10; X=2 and Y=5) compositions, are shown. In case all the incorporated quantity of the biostatic agent would be released/dissolved, in each solution a total Triclosan content of 0.02 g/100 mL would have been resulted. In Fig. 3b, the UV absorption spectra of water–ethanol solutions 50%, in which PS–DVB 5%–Psalt 20%, PS–DVB 2%–Psalt 10% or PS–DVB 5%–Psalt 10% beads were immersed for 12 h, are also depicted. In Fig. 3a the broad absorption band centered at ~280 nm is attributed to the Triclosan [20], while in Fig. 3b the absorption zone, with the more intense absorption peaks at 284 and 293 nm, is attributed to the phosphonium salt. These assignments were further confirmed by UV–Vis absorption spectra of pure Triclosan and phosphonium salt dissolved either in ethanol solutions or in natural serum. In both cases, it is clear that for an equal time of immersion in the same water ethanol solution, having the same crosslinking density, there is higher release of the biocides as the biocide load in the beads increases. Inversely, keeping fixed the percentage of the biocide-incorporated beads, a lower release of the biocides was observed with increasing cross-linking density. The same picture holds for both biocides in all water ethanol and physiological saline solutions. Further spectroscopic examination of the three samples of PS–DVB X%–phosphonium salt Y% beads
More systematic release studies have been performed for all biocide incorporated beads in the three aqueous ethanol solutions and in natural serum for a time of period extended up to 7–8 months. In Fig. 6 the UV absorption spectra of three water ethanol solutions 95% are continuously compared after the immersion for a period of time from 1 day to 7–8
Fig. 5. ATR-FTIR spectra of three different PS–DVB–Psalt beads compositions before (2) and after (1) their immersion in ethanol 50% solutions for 12 h. Dot contours mark the characteristics bands of the biocide phosphonium salt at 625 and 1726 cm1.
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
229
Fig. 6. UV–Visible absorption spectra of ethanol solutions 95% in which three different beads–Triclosan compositions, PS–DVB 2%–Triclosan 5% (left), PS–DVB 5%–Triclosan 10% (middle) and PS–DVB 2%–Triclosan 10% (right), have been immersed. In dot lines (on the right) are given the absorption zones that exceed the linearity of the UV detector.
weeks of Triclosan-incorporated beads with different either biocide concentration or cross-linking agent density. Considering that the absorption of the Triclosan at ~280 nm represents the relative concentration of the biocide in the aqueous–ethanol solutions 95%, it is very obvious that, in any cycle of time, there is increased release of Triclosan when either the biocide content in the beads was higher, PS–DVB 2%–Tr 10% vs. PS–DVB 2%–Tr 5%, or the DVB density was lower, PS–DVB 2%–Tr 10% vs. PS– DVB 5%–Tr 10%, keeping stable the antagonist parameter. A note is made of the fact that, in the case of PS–DVB 2%–Tr 10% beads immersed for 3 and 7 weeks in EtOH 95%, the absorption zones are given in dot lines since the absorbance exceeds the linearity of the UV detector. For the same reason, the above measurements have not been taken into account in the subsequent quantitative measurements. For the phosphonium salt-incorporated PS–DVB beads, a pretty similar picture to the PS–DVB–Tr beads holds as well. The corresponding absorption spectra of water–ethanol solutions 10% after the immersion of the related phosphonium salt incorpo-
rated beads are shown in Fig. 7. Once more, it is clearly shown that for the same time framework, increased release of the phosphonium salt is observed at lower cross-linking agent density, PS–DVB 2%– Psalt 10% vs. PS–DVB 5%–Psalt 10%, as well as at higher biocide content, although keeping the same cross-linking over the biocidal agent ratio, PS–DVB 10%–Psalt 20% vs. PS–DVB 5%–Psalt 10%. In both cases, it seems that the biocide effectiveness depends on its ability to diffuse through the cross-linked matrix in the vicinity of the polymeric beads surface. For that reason, the concentration of the biocide is critical because the molecular diffusion is controlled by it as well. Furthermore, the content of DVB seems to play a significant role in the liberation of Triclosan from the beads since it affects the permeability of the plastic formulation. DVB contains two double bonds that might enter in the polymerization process of two different developing macromolecular chains, acting as a cross-linking agent in the PS network. As the mole fraction of DVB in the beads increases, a denser network is developed and in this sense the migration of biocides is restrained. So,
230
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 7. UV–Visible absorption spectra of ethanol solutions 10% in which three different beads–phosphonium salt compositions, PS–DVB 2%– Psalt 10% (left), PS–DVB 5%–Psalt 10% (middle) and PS–DVB 10%–Triclosan 20% (right), have been immersed.
the morphology of the polymer seems to have a significant effect on the liberation of the biocides. A quantitative study was also attempted in order to follow the release of the biocides with time. The diffusion of Triclosan out of the polymer beads is monitored through the detection of the absorbance at 282 nm for different beads samples immersed in different water–ethanol solutions and in physiological saline. Evaluating the optical densities at 282 nm, a quantitative measurement of the released Triclosan has been carried out. The concentration of Triclosan in the solution was estimated according to the Beer’s law: Ak ¼ ek bc
ð1Þ
where Ak is the absorbance at a wavelength k, ek the molar extinction coefficient (cm l mol1), b the optical path (cm) of the cell and c the concentration (mol L1). Since Eq. (1) is linear for each wavelength, the concentration is proportional to the absorption. In principle, we created a calibration curve by recording the absorption spectra for several known different concentrations of Triclosan. The evolution of the Triclosan’s release with time after the impregnation of the different PS/DVB/Triclosan composition beads in
ethanol solutions 50% on the left and 95% on the right is correspondingly depicted in Fig. 8 where the concentration values of Triclosan are plotted against time. As it was expected, using the ethanol solution (EtOH) 50%, the released amount of Triclosan was significantly lower, by one order of magnitude, than from the EtOH 95% immersed beads. The effect of the cross-linking density on the Triclosan release in EtOH solution 50% is also demonstrated in Fig. 8 for PS– 2% DVB and PS–5% PVB containing 10% Triclosan. In both cases, there is a gradual increase in the released amount for a period of about 3–4 months and a plateau is reached. The plateau amount was found at 2.5% and 5% of the total biocide encapsulated quantity for the two formulations, PS–DVB 5%–Tr 10% and PS–DVB 2%–Tr 10%, respectively, when the maximum Triclosan amount that could be released is 0.02 g/100 mL. Coming to the EtOH solution 95%, the relevant maximum released amount of Triclosan is significantly higher, but, nevertheless, even after 7 months of immersion it represents only 50% of the total amount of the incorporated biocide concentration. A more interesting feature concerning the beads immersed in EtOH 95% is the high differentiation of
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
231
Fig. 8. Quantitative measurements of Triclosan released from PS–DVB–Triclosan beads immersed in ethanol solutions 50% (left) and 95% (right) vs. time of impregnation. Lines are drawn to guide the eye.
the Triclosan-releasing ability of the different formulations studied. The beads with lower cross-linking density and higher Triclosan amount, PS–DVB 2%– Tr 10%, show a high release rate of the order of 65% of the total amount in 1 month of immersion. As the cross-linking density was increased, the release rate was decreased to 50% of the total amount in 7 months of immersion. A further decrease in the release rate of Triclosan has taken place by decreasing the biocide content in the PS–DVB 2%–Tr 5% type of beads. In this case, a very slow but continuously increased release was observed in the time frame of 7 months with the relevant maximum released quantity being 5– 10% of the total amount. As it is clearly shown in this picture the Triclosan release rate can be effectively
tuned by using the proper combination of the crosslinking density and the active material loading. Furthermore, it is remarkable that the total amount of Triclosan (C max=0.020 g/100 mL) has not migrated into the solution, but a certain part of it remains in the bulk after 7 months of immersion in ethanol solutions, even in EtOH 95%. These findings suggest that two contingencies may occur: either an amount of Triclosan remains in the bulk because it is captured in the polymer network and so cannot be in contact with ethanol, or the period of time employed is not enough for the total migration/release of Triclosan to the solution. A rather similar release study was carried out for the phosphonium salt containing beads. The results
Fig. 9. Quantitative measurements of phosphonium salt released from PS–DVB–Psalt beads immersed in saline (left) and ethanol solution 10% (right) vs. time of impregnation. Lines are drawn to guide the eye.
232
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
indicating the evolution of the release of the phosphonium salt with time are shown in Fig. 9. More precisely, the concentration values of the phosphonium salt, determined via relevant calibration curves, in terms of optical densities at 293 nm, for three PS– DVB–Psalt type beads are plotted against time after being immersed in the natural serum (on the left) or in water–ethanol solution 10% (on the right). Once more, it is clear that the release of the low molecular weight biocide is favored when the amount of the crosslinking agent was decreased, i.e., going from the PS– DVB 5%–Psalt 10% to the PS–DVB 2%–Psalt 10% type of beads. A note is made here again of the fact that the third PS–DVB 10%–Psalt 20% type of beads exhibits higher release rate than the PS–DVB 5%– Psalt 10% formulation in both EtOH 10% and physiological saline solutions, although they contain the same ratio of the cross-linking to the biocide agent content. This is tentatively attributed to the higher dependence of the release on the biocide concentration itself in the polymeric matrix. Such a release at higher biocide loading might be due to an eventual increased biocide concentration gradient inside the polymeric cross-linked beads, making it diffuse more quickly. In a similar context, since the biocide behaves as an inert material, at high biocide concentration levels, it might ensure released channels that are getting less and less dependent on the cross-linking density as well. This reveals the importance of both the interplay between the two antagonistic to the release rate parameters and the overall concentration range employed. We may thus infer for both cases that for the controlled release, of great importance are the concentration of the biocide, the percentage of the crosslinks determined by the amount of the DVB added at the stage of synthesis, the percentage of their total concentration in the PS–DVB–biocide beads and the liquid mixture the beads are impregnated in.
4. Conclusions A method of incorporation of low molecular weight biostatic agents in polymer matrices that have been subject to some degree of cross-linking has been described. PS–DVB–biocide beads were synthesized by the simultaneous polymerization of a suspension of the monomer (styrene), the initiator, the cross-linking
agent (DVB) and the biostatic compound (Triclosan or cationic phosphonium salt) to be incorporated. The relative concentration of the biostatic agent and the cross-linking agent provides the capability to control the release/diffusion of the biostatic agent in liquids, such as natural serum and water–ethanol mixtures in contact to the polymeric beads. Ultraviolet–visible absorption spectrometry was used for controlling the rate of release/diffusion to the particular liquid phases of either the Triclosan or the phosphonium salt, which have been separately incorporated in PS–DVB polymeric beads matrix. Both quantitative (via UV absorption) and qualitative (via vibrational spectra) measurements indicate that increased biocide incorporation in the PS/DVB beads was accompanied by a corresponding enhancement of its concentration in either the water ethanol solutions or the physiological saline, while higher cross-linking densities hindered the biocide migration out of the beads by lowering its release rate into the liquid mixtures.
Acknowledgements The financial support for the present work by the Growth CEU SPAN (contract no. G5RD-CT-200100568) and the PAVET-2000 GSRT 1463 projects as well as the constructive cooperation with the ARGO S.A. company are greatly acknowledged. This work was also performed in the framework of the Interdepartmental Operation Program for Education and Initial Vocational Training on Polymer Science and Technology- 32a, 33H6 of the University of Patras, administered through the Minister of Education and Religious Affairs of Greece.
References [1] C.P. Gerba, G.E. Janauer, M. Costello, Removal of poliovirus and rotavirus from tapwater by a quaternary ammonium resin, Water Res. 18 (1) (1984) 17 – 19. [2] J. Hazzizalaskar, N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 1. Synthesis of polybutadiene with pendent quaternary ammonium groups, J. Appl. Polym. Sci. 50 (4) (1993) 651 – 662; N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 2. Biological evaluation of polyurethane coatings with pendant quaternary ammonium-salts, J. Appl. Polym. Sci. 50 (4) (1993) 663 – 670;
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 3. Aging of biological polyurethane coatings in water, J. Appl. Polym. Sci. 50 (4) (1993) 671 – 678. A.J. Isquith, E.A. Abbott, P.A. Walters, Surface-bonded antimicrobial activity of an organosilicon quaternary ammonium chloride, Appl. Microbiol. 24 (6) (1972) 859 – 863; P.A. Walters, E.A. Abbott, A.J. Isquith, Algicidal activity of a surface-bonded organosilicon ammonium chloride, Appl. Microbiol. 25 (2) (1973) 253 – 256. C. Hugues, C. Bessy, P. Bartolomeo, A. Margaillan, Complexation of an acrylic resin by tertiary amines: synthesis and characterization of new binders for antifouling paints, Eur. Polym. J. 39 (2) (2003) 319 – 326. A. Kanazawa, T. Ikeda, Multifuctional tetracoordinate phosphorus species with high self-organizing ability, Coord. Chem. Rev. 198 (2000) 117 – 131. G. Cerichelli, C. La Mesa, L. Luchetti, G. Mancini, Role of counterions in the catalytic activity and phase equilibria of phosphonium salts in water, Langmuir 16 (1) (2000) 166 – 171. E.R. Kenawy, F.I. Abdel-Hay, A. El-Raheem, R. El-Shanshoury, M.H. El-Newehy, Biologically active polymers: synthesis and antimicrobial activity of modified glycidyl methacrylate polymers having a quaternary ammonium and phosphonium groups, J. Control. Release 50 (1–3) (1998) 145 – 152. G.J. Li, J.R. Shen, Y.L. Zhu, A study of pyridinium-type functional polymers. III. Preparation and characterization of insoluble pyridinium-type polymers, J. Appl. Polym. Sci. 78 (3) (2000) 668 – 675. G.J. Li, J.R. Shen, Preparation and antibacterial activity of pyridinium-carrying polypropylene particles, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 42 (2) (2001) 360 – 361. T. Tarvainen, T. Karjalainen, M. Malin, S. Pohjolainen, J. Tuominen, J. Seppala, K. Jarvinen, Degradation of and drug release from a novel 2,2-bis(2-oxazoline) linked poly(lactic acid) polymer, J. Control. Release 81 (3) (2002) 251 – 261; T. Tarvainen, T. Karjalainen, M. Malin, K. Perakorpi, S. Pohjolainen, J. Tuominen, J. Seppala, K. Jarvinen, Drug release profiles from and degradation of a novel biodegradable polymer, 2,2-bis(2-oxazoline) linked poly(epsilon-caprolactone), Eur. J. Pharm. Sci. 16 (4–5) (2002) 323 – 331. T.E. Furia, A.G. Schenkel, New, broad spectrum bacteriostat, Soap Chem. Special 44 (1968) 47–50, 116–122.
233
[12] H.N. Bhargava, P.A. Leonard, Triclosan: applications and safety, Am. J. Infect. Control 24 (1996) 209 – 218. [13] R.D. Jones, H.B. Jampani, J.L. Newman, A.S. Lee, Triclosan: a review of effectiveness and safety in health care settings, Am. J. Infect. Control 28 (2) (2000) 184 – 196. [14] L.M. McMurry, M. Oethinger, S.B. Levy, Triclosan targets lipid synthesis, Nature 394 (1998) 531 – 532; C.W. Levy, A. Roujeinikova, S. Sedelnikova, P.J. Baker, A.R. Stuitjet, A.R. Slabast, D.W. Rice, J.B. Rafferty, Molecular basis of triclosan activity, Nature 398 (1999) 383 – 384. [15] A. Kanazawa, T. Ikeda, T. Endo, Polymeric phosphonium salts as a novel class of cationic biocides. II. Effects of counter anion and molecular weight on antibacterial activity of polymeric phosphonium salts, J. Polym. Sci., Polym. Chem. 31 (6) (1993) 1441 – 1447. [16] D.W. Emerson, Slow release of active chlorine and bromine from styrene divinylbenzene copolymers bearing N,Ndichlorosulfonamide, N-chloro-N-alkylsulfonamide, and Nbromo-N-alkylsulfonamide functional-groups-polymer-supported reagents. 6, Ind. Eng. Chem. Res. 30 (11) (1991) 2426 – 2430. [17] T. Tashiro, Removal of Escherichia coli from water by systems based on insoluble polystyrene-poly(ethylene glycol)s, polystyrene-polyethylenimides, and polystyrene-polyethylenepolyamines quaternized, J. Appl. Polym. Sci. 43 (7) (1991) 1369 – 1377. [18] M. Grayson, P.T. Keough, Phosphonium compounds. I. Elimination reactions of 2-cyanoethyl phosphonium salts. Asymmetric phosphines and alkylenediphosphines, J. Am. Chem. Soc. 81 (18) (1959) 4803 – 4807; M. Grayson, P.T. Keough, Phosphonium compounds: II. Decomposition of phosphonium alkoxides to hydrocarbon, ether and phosphine oxide, J. Am. Chem. Soc. 82 (15) (1960) 3919 – 3924. [19] A. Kanazawa, T. Ikeda, T. Endo, Polymeric phosphonium salts as a novel class of cationic biocides: 5. Synthesis and antibacterial activity of polyesters releasing phosphonium biocides, J. Polym. Sci., Polym. Chem. 31 (12) (1993) 3003 – 3011. [20] B.D. Kalyon, U. Olgun, Antibacterial efficacy of triclosanincorporated polymers, Am. J. Infect. Control 29 (2) (2001) 124 – 125.
Incorporation of low molecular weight biocides into polystyrene–divinyl benzene beads with controlled release characteristics S.M. Iconomopouloua, A.K. Andreopouloua,b, A. Sotoa, J.K. Kallitsisa,b, G.A. Voyiatzisa,* a
Foundation for Research and Technology-Hellas (FORTH) Institute of Chemical Engineering and High Temperature Chemical Processes (ICE/HT), P.O. Box 1414, GR-265 04 Rio-Patras, Greece b Department of Chemistry, University of Patras, GR-265 04 Rio-Patras, Greece Received 29 July 2004; accepted 7 October 2004 Available online 10 November 2004
Abstract Triclosan and phosphonium salt biocides have been separately incorporated into polystyrene–divinylbenzene (PS–DVB) beads by suspension polymerization. Ultraviolet (UV) absorption measurements have been used to monitor the release of these low molecular weight biocides out of the PS–DVB beads immersed in water–ethanol mixtures and in physiological saline. The release of the biocide agents is strongly dependent on either the DVB or/and the antimicrobial composition ratio in the beads. An increase of biocide incorporation in the PS/DVB beads was accompanied by a corresponding enhancement of its concentration in liquid mixtures. On the contrary, higher cross-linking densities hindered the biocide migration out of the beads by diminishing its release rate into either the aqueous ethanol solutions or the natural serum. Moreover, Fourier transform Raman (FT-Raman) spectra and Attenuated Total Reflectance Infrared (ATR-FTIR) measurements of the PS–DVB–Triclosan and PS–DVB–phosphonium salt beads, before and after their immersion in water–ethanol solutions, gave a similar qualitative evidence of the biocide release. D 2004 Elsevier B.V. All rights reserved. Keywords: PS–DVB beads; Low molecular weight biocide incorporation; Release of biocide; UV absorption; Vibrational spectroscopy
1. Introduction
* Corresponding author. Tel.: +30 2610 965253; fax: +30 2610 965223. E-mail address: [email protected] (G.A. Voyiatzis). 0168-3659/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2004.10.006
Polymeric materials having antimicrobial or antifouling activity have been developed during the last years since the interest about health and hygiene habits is growing. Depending on the way that the incorporation of the active species into the polymeric
224
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
matrix takes place, these materials have been classified into two major categories. First, the immobilized type, where the active antimicrobial groups are covalently attached onto the polymeric substrate and the final antimicrobial efficiency is based on the contact between the biocide and the microorganisms. Typical representatives of this category are polymers bearing quaternary ammonium salts [1–4], phosphonium salts [5–7] or pyridinium type groups [8,9]. The second category is pertaining to a biocide-controlled release type after the incorporation of the antimicrobial agent into a host polymeric matrix. The latter contains the active species either as counter ion of polymeric bounded charges or as dispersion of low molecular weight active compounds [10]. Triclosan, 2,4,4V-trichloro-2V-hydroxydiphenyl ether, is known to be a broad spectrum antibacterial and antifungal agent [11,12]. Thereby, it is widely used in formulations such as soaps, toothpastes, deodorants, cosmetics, toys, carpets and kitchenware. Moreover, it is used as an additive in textiles and polymers providing to these materials its antimicrobial properties. In vivo inhibition studies have shown [13] that this low molecular weight biocide is active even at very low concentrations. Triclosan acts mainly in a specific target, the enoyl reductase enzyme, which is involved in the synthesis of fatty acids [14]. Phosphonium as well as quaternary ammonium salts are cationic biocides that target the bacterial membranes, as well. Their antimicrobial activity is based on their ability to disrupt and disintegrate the negatively charged cell membranes. In some cases, phosphonium salts show [15] higher activity than the quaternary ammonium salts; this characteristic, combined with their inherent higher thermal stability, makes them potential candidates to be used in polymer formulations. Insoluble cross-linked polymers bearing active antimicrobial groups have been widely reported in the literature including quaternary ammonium type resins [16,17]. However, no reference has been made for the incorporation of a low molecular weight biocide inside a polymer shell like the one formed by cross-linked beads. This method may improve the thermal stability of the end-used material, reduce the volatility of the biocide agent and, moreover, control the leaching-out rate of the antimicrobial entity.
The target of this work is to incorporate low molecular weight biocides, like Triclosan and phosphonium salts, into PS/DVB beads and investigate their release rate in either water–ethanol solutions or physiological saline.
2. Materials and methods 2.1. Materials Triclosan (Irgarguard B 1000) was kindly supplied from Ciba Specialty Chemicals. Tri-n-butyl phosphine, dodecylbromine and 5-sulphoisophthalic acid sodium salt were purchased from Aldrich. All other chemicals and reagents were purchased from Aldrich or Merck and used without purification unless otherwise noticed. Benzoyl peroxide (BPO) was recrystallized from CHCl3/MeOH. Styrene was distilled from CaH2 under reduced pressure, prior to use. Suspension polymerizations were carried out under an argon atmosphere. 2.2. Synthesis of phosphonium salts The phosphonium derivative of tri-n-butyl phosphine was synthesized by reaction with alkyhalides bearing alkyl chains with 12 carbon atoms [18]. The desired 5-sulfoisophthalic acid dimethyl ester tributylalkylphosphonium salt was synthesized by the reaction of the abovementioned phosphonium salt with dimethyl 5-sulfoisophthalate sodium salt [19]. 2.3. Synthesis of PS–DVB–biocide beads In an attempt to create polymeric matrixes with the ability of controlled releasing, antimicrobial substances have been incorporated into cross-linked polystyrene beads using suspension polymerization conditions. More specifically, PS beads of various degrees of cross-linking were synthesized carrying either Triclosan or phosphonium salt amounts as the antimicrobial entity (Table 1). Since Triclosan as well as phosphonium salts can be easily dissolved in styrene, the suspension polymerization technique provides a unique opportunity to encapsulate these active substances into a polymeric more or less bstorage cellQ. In order to examine possible controlled release of both
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
225
Table 1 The PS/DVB/biocide beads formulations synthesized
2.4. Instrumentation
Samples
The electronic absorption spectra were measured with a double beam Lambda 900 UV/Vis/NIR Spectrophotometer of Perkin Elmer. Reference beam was always delivered through an optical cell fill with the corresponding biocide-free ethanol or sodium chloride 0.9% aqueous solution. The spectra were recorded from 500 to 190 nm with a slit width of 0.5 nm. Fourier transform Raman (FT-Raman) measurements were obtained using a Bruker (D) FRA-106/S component attached to an EQUINOX 55 spectrometer. A R510 diode pumped Nd:YAG laser at 1064 nm (with a maximum output power of 500 mW) was used for Raman excitation in a 1808 scattering sample illumination module. An optical filtering reduced the Rayleigh elastic scattering and, in combination with a CaF2 beamsplitter and a high sensitivity liquid N2cooled Ge-detector, allowed the Raman intensities to be recorded from 50 to 3300 cm1 in Stokes-shifted Raman region, all in one spectrum. Attenuated Total Reflectance mid-infrared spectra were measured utilizing the single reflection Diamond (Miracle) crystal of the Horizontal ATR accessory of Pike Technologies (USA) attached to the sample compartment of a Bruker (D) Equinox 55 FTIR spectrometer bearing a DTGS detector.
1 2 3 4 5 6 7 8 9
PS–DVB–biocide beads %DVB
%Triclosan
2 2 2 5 2 5 2 5 10
0 10 5 10
%Phosphonium salt 0
10 10 20 20 20
Triclosan and phosphonium salt as a function of the degree of cross-linking, PS beads with different crosslinking densities (2–10% DVB) were synthesized. Moreover, and since the biocide content is expected to influence the porous’ nature of these beads, various amounts (5–20%) of either Triclosan or phosphonium salt were combined along with the different degrees of cross-linking. PS–DVB–Triclosan and PS–DVB– phosphonium salt beads were synthesized by suspension polymerization according to the following procedure that is given as an example of a typical polymerization for the preparation of the PS–DVB 2%–Tr 10% type beads. Gelatin (0.48 g) dissolved in hot, distilled and degassed water (100 mL), NaOH 10N (0.65 mL), H3BO3 (0.40 g) and Bentonite (0.70 g) were added into a three-necked round flask, equipped with a mechanical stirrer, a thermometer, a reflux condenser and an argon inlet. The mixture was stirred at room temperature for 30 min. Afterwards, 2 g of BPO were suspended in another mixture of 28 mL styrene, 0.65-mL divinylbenzene and 2.9 g Triclosan and it was added to the initial mixture. The suspension thus formed was vigorously stirred at 80 8C for 8 to 12 h. This procedure, in the case of Triclosan, was prolonged to 1 or 2 days to avoid the presence of excess styrene in the beads. After cooling to room temperature, the resulting PS–DVB–biocide beads were filtrated, washed carefully with water and dried under reduced pressure at ambient temperature to constant weight. In the case of PS–DVB–phosphonium salt beads for the removal of the excess of styrene, due to the low vapour pressure of the phosphonium salt, the last step of drying was done under reduced pressure at 50 8C to constant weight.
2.5. Release studies In order to probe the release of the biocides incorporated into PS–DVB beads by UV–visible absorption spectroscopy, PS–DVB–biocide beads have been immersed into different water–ethanol mixtures (10%, 50% and 95%) and into physiological saline in such amount that in the case of total release of the biocide its final concentration in the different solutions to be 0.02 g/100 mL. Nevertheless, a note is made of the fact that the maximum dissolution of the biocides in the water–ethanol solution 10% and in the physiological saline was found (a) for Triclosan equal to 0.0012 and 0.0005 g/100 mL, respectively, and (b) for the phosphonium salt equal to ~0.0085 g/100 mL, in both cases. Anyway, in the present study the degree of dissolution anticipated for the biocidal solutions was a compromise between the above imposed saturation concentrations and the detection limits of the spectroscopic techniques utilized. More precisely,
226
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
a quantity of around 0.1 to 0.4 g of beads, depending on the composition, was placed in a volumetric flask and filled with aqueous ethanol solution or with physiological saline to a total volume of 100 mL. Whenever UV absorption measurements have been scheduled, 2 mL of the solution was transferred to the sample compartment of the spectrophotometer and after the UV absorption measurement it was decanted back to the volumetric flask. For the vibrational spectroscopic measurements, similar but separate experiments were carried out in order to avoid disturbance of the UV-absorption measurements and the release study in progress. More exactly, a quantity of the solid was removed and dried in a vacuum oven at room temperature prior to the acquisition of the vibrational, FT-Raman or ATR-FTIR, spectra.
Fig. 1. FT-Raman spectra of Triclosan, Tr (top), phosphonium salt, Psalt (bottom), PS–DVB 2%, PS–DVB 2%–Tr 10% and PS–DVB 2%–Psalt 10%. Characteristic Raman bands of the biocides are marked with arrows.
3. Results and discussion 3.1. Characterization of biocides Triclosan is a chlorinated aromatic ether, while the phosphonium salt used in present work is a sulfoisophthalate–phosphonium salt. The molecular structures of both antimicrobial agents are depicted in Scheme 1. The presence of biocides in the beads, before and after their immersion in either water– ethanol or sodium chloride 0.9% solutions, might be monitored spectroscopically. To this regard, the Raman spectra of the Triclosan (Tr) and the phos-
Scheme 1. Molecular structures of Triclosan (1) and cationic phosphonium salt (2).
phonium salt (Psalt) are respectively depicted on the top and on the bottom of Fig. 1, in the corresponding Raman Stokes spectral range. For the Triclosan, two main spectral features are distinguished: the low frequency peak at 710 cm1 that was tentatively attributed to the skeletal (C–C) stretching vibration of trisubstituted benzene rings and the high frequency band at 3073 cm1 accurately assigned to the C–H stretching of the phenyl rings. The assignment of the low-frequency peak to benzene-chlorine (Ar-Cl) vibration mode or even to (Ar-O-Ar) ether grouping is less probable but it cannot be securely ruled out. Also in Fig. 1, the Raman spectra of PS–DVB 2%– Triclosan 10%, PS–DVB 2% and PS–DVB 2%–Psalt 10% beads are shown. It is obvious that the appropriate Raman band for the identification of the presence of Triclosan in the beads is that located at 710 cm1, since the high frequency scattering peak at ~3070 cm1 is overlapped with the relevant spectral contribution of the phenyl rings of both PS and DVB components of the beads. In the case of the phosphonium salt, the most identifiable spectral feature is that of the carbonyl stretching at ~1726 cm1. In the physiological saline, the occurrence of this vibrational mode is due to the presence of the 5sulfoisophthalate anion of the sulfoisophthalate–phosphonium salt (see Scheme 1), which gets soluble after the immersion of the cationic part of the phosphonium salt, thus indirectly giving evidence for the release of
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 2. ATR-FTIR spectra of Triclosan, Tr (top), phosphonium salt, Psalt (bottom), PS–DVB 2%, PS–DVB 2%–Tr 10% and PS–DVB 2%–Psalt 10%. Characteristic IR bands of the biocides are marked with arrows.
the latter. In the case of water–ethanol solutions, both parts are soluble simultaneously. In the same context, the ATR-FTIR spectra of both biocides and of the PS–DVB 2% and PS–DVB 2%– biocide 10% beads are depicted in Fig. 2 in the blow frequency rangeQ. Comparing the vibrational absorption spectra of each biocide with the corresponding biocide containing beads, it is evident that both antimicrobial agents exhibit specific infrared vibrational features that might be used to monitor their release out of the polystyrene beads. More precisely, Triclosan might be identified in the beads through the
227
infrared peaks at 860 and 1476 cm1 attributed to the out of plane rocking vibration of hydrogen atoms attached to aromatic ring and to the CMC stretching vibration of benzene rings, respectively. In the case of the phosphonium salt, it might be detected via the vibrational peaks located at 625, 1040 and 1726 cm1 and attributed in that order to C–P, -SO3 and CMO stretching. All above absorption bands are unique spectral features of the biocides with no overlapping vibrational absorption counterparts of the beads. In this context, they are used to determine the biocide remaining after the immersion of either the Triclosan or the phosphonium salt containing beads into water– ethanol solutions or physiological saline. Besides that, it must be noticed that the penetration depth of both vibrational spectroscopic techniques is in the order of micrometer, accounted most probably for bulk than surface sensitive measurements. Preliminary UV absorption spectra have indicated that after the immersion of the corresponding beads in the water–ethanol liquid mixtures, the concentration of both biocides was enhanced in the richer in ethanol solutions. This quick release is in good agreement with the better solubility of both biocides in ethanol compared to that in water. Ethanol most probably influences both the migration of the biocides from the bulk to the bead surface and its release into the water– ethanol solution. In Fig. 3a, the UV absorption spectra of Triclosan that has been released/dissolved in aqueous ethanol solution 95%, after impregnation for 1 week of beads that derived from the suspension
Fig. 3. UV–Visible absorption spectra of (a) ethanol solutions 95% in which three different beads–Triclosan compositions have been immersed for 1 week and (b) ethanol solutions 50% in which three different beads–phosphonium salt compositions have been immersed for 12 h.
228
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 4. FT-Raman spectra of three different PS–DVB–Psalt beads compositions before (1) and after (2) their immersion in ethanol solutions 50% for 12 h. Circle marks the characteristic band of the biocide phosphonium salt at 1726 cm1.
with the same either cross-linking agent or biocide composition has been vibrationally monitored after their immersion in water ethanol solution 50% for 12 h. The corresponding FT-Raman and ATR-FTIR spectra before and after their immersion in the ethanol solution are depicted in Figs. 4 and 5, respectively. In both cases, the relative intensities of the abovementioned spectral features of the phosphonium salt, at 1726 cm1 for Raman (Fig. 4) and at 625, 1040 and 1726 cm1 for infrared (Fig. 5), reveal the increased release rate of the biocide in the water–ethanol solution 50%, with higher biocide or with lower cross-linking agent compositions in the beads, in close comparison with the relevant UV absorption measurements shown in Fig. 3b. Similar vibrational results have been obtained from relevant Triclosan incorporated into PS–DVB beads. 3.2. Release studies
polymerization of various styrene–DVB X%–Triclosan Y% (X=2 and Y=10; X=5 and Y=10; X=2 and Y=5) compositions, are shown. In case all the incorporated quantity of the biostatic agent would be released/dissolved, in each solution a total Triclosan content of 0.02 g/100 mL would have been resulted. In Fig. 3b, the UV absorption spectra of water–ethanol solutions 50%, in which PS–DVB 5%–Psalt 20%, PS–DVB 2%–Psalt 10% or PS–DVB 5%–Psalt 10% beads were immersed for 12 h, are also depicted. In Fig. 3a the broad absorption band centered at ~280 nm is attributed to the Triclosan [20], while in Fig. 3b the absorption zone, with the more intense absorption peaks at 284 and 293 nm, is attributed to the phosphonium salt. These assignments were further confirmed by UV–Vis absorption spectra of pure Triclosan and phosphonium salt dissolved either in ethanol solutions or in natural serum. In both cases, it is clear that for an equal time of immersion in the same water ethanol solution, having the same crosslinking density, there is higher release of the biocides as the biocide load in the beads increases. Inversely, keeping fixed the percentage of the biocide-incorporated beads, a lower release of the biocides was observed with increasing cross-linking density. The same picture holds for both biocides in all water ethanol and physiological saline solutions. Further spectroscopic examination of the three samples of PS–DVB X%–phosphonium salt Y% beads
More systematic release studies have been performed for all biocide incorporated beads in the three aqueous ethanol solutions and in natural serum for a time of period extended up to 7–8 months. In Fig. 6 the UV absorption spectra of three water ethanol solutions 95% are continuously compared after the immersion for a period of time from 1 day to 7–8
Fig. 5. ATR-FTIR spectra of three different PS–DVB–Psalt beads compositions before (2) and after (1) their immersion in ethanol 50% solutions for 12 h. Dot contours mark the characteristics bands of the biocide phosphonium salt at 625 and 1726 cm1.
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
229
Fig. 6. UV–Visible absorption spectra of ethanol solutions 95% in which three different beads–Triclosan compositions, PS–DVB 2%–Triclosan 5% (left), PS–DVB 5%–Triclosan 10% (middle) and PS–DVB 2%–Triclosan 10% (right), have been immersed. In dot lines (on the right) are given the absorption zones that exceed the linearity of the UV detector.
weeks of Triclosan-incorporated beads with different either biocide concentration or cross-linking agent density. Considering that the absorption of the Triclosan at ~280 nm represents the relative concentration of the biocide in the aqueous–ethanol solutions 95%, it is very obvious that, in any cycle of time, there is increased release of Triclosan when either the biocide content in the beads was higher, PS–DVB 2%–Tr 10% vs. PS–DVB 2%–Tr 5%, or the DVB density was lower, PS–DVB 2%–Tr 10% vs. PS– DVB 5%–Tr 10%, keeping stable the antagonist parameter. A note is made of the fact that, in the case of PS–DVB 2%–Tr 10% beads immersed for 3 and 7 weeks in EtOH 95%, the absorption zones are given in dot lines since the absorbance exceeds the linearity of the UV detector. For the same reason, the above measurements have not been taken into account in the subsequent quantitative measurements. For the phosphonium salt-incorporated PS–DVB beads, a pretty similar picture to the PS–DVB–Tr beads holds as well. The corresponding absorption spectra of water–ethanol solutions 10% after the immersion of the related phosphonium salt incorpo-
rated beads are shown in Fig. 7. Once more, it is clearly shown that for the same time framework, increased release of the phosphonium salt is observed at lower cross-linking agent density, PS–DVB 2%– Psalt 10% vs. PS–DVB 5%–Psalt 10%, as well as at higher biocide content, although keeping the same cross-linking over the biocidal agent ratio, PS–DVB 10%–Psalt 20% vs. PS–DVB 5%–Psalt 10%. In both cases, it seems that the biocide effectiveness depends on its ability to diffuse through the cross-linked matrix in the vicinity of the polymeric beads surface. For that reason, the concentration of the biocide is critical because the molecular diffusion is controlled by it as well. Furthermore, the content of DVB seems to play a significant role in the liberation of Triclosan from the beads since it affects the permeability of the plastic formulation. DVB contains two double bonds that might enter in the polymerization process of two different developing macromolecular chains, acting as a cross-linking agent in the PS network. As the mole fraction of DVB in the beads increases, a denser network is developed and in this sense the migration of biocides is restrained. So,
230
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
Fig. 7. UV–Visible absorption spectra of ethanol solutions 10% in which three different beads–phosphonium salt compositions, PS–DVB 2%– Psalt 10% (left), PS–DVB 5%–Psalt 10% (middle) and PS–DVB 10%–Triclosan 20% (right), have been immersed.
the morphology of the polymer seems to have a significant effect on the liberation of the biocides. A quantitative study was also attempted in order to follow the release of the biocides with time. The diffusion of Triclosan out of the polymer beads is monitored through the detection of the absorbance at 282 nm for different beads samples immersed in different water–ethanol solutions and in physiological saline. Evaluating the optical densities at 282 nm, a quantitative measurement of the released Triclosan has been carried out. The concentration of Triclosan in the solution was estimated according to the Beer’s law: Ak ¼ ek bc
ð1Þ
where Ak is the absorbance at a wavelength k, ek the molar extinction coefficient (cm l mol1), b the optical path (cm) of the cell and c the concentration (mol L1). Since Eq. (1) is linear for each wavelength, the concentration is proportional to the absorption. In principle, we created a calibration curve by recording the absorption spectra for several known different concentrations of Triclosan. The evolution of the Triclosan’s release with time after the impregnation of the different PS/DVB/Triclosan composition beads in
ethanol solutions 50% on the left and 95% on the right is correspondingly depicted in Fig. 8 where the concentration values of Triclosan are plotted against time. As it was expected, using the ethanol solution (EtOH) 50%, the released amount of Triclosan was significantly lower, by one order of magnitude, than from the EtOH 95% immersed beads. The effect of the cross-linking density on the Triclosan release in EtOH solution 50% is also demonstrated in Fig. 8 for PS– 2% DVB and PS–5% PVB containing 10% Triclosan. In both cases, there is a gradual increase in the released amount for a period of about 3–4 months and a plateau is reached. The plateau amount was found at 2.5% and 5% of the total biocide encapsulated quantity for the two formulations, PS–DVB 5%–Tr 10% and PS–DVB 2%–Tr 10%, respectively, when the maximum Triclosan amount that could be released is 0.02 g/100 mL. Coming to the EtOH solution 95%, the relevant maximum released amount of Triclosan is significantly higher, but, nevertheless, even after 7 months of immersion it represents only 50% of the total amount of the incorporated biocide concentration. A more interesting feature concerning the beads immersed in EtOH 95% is the high differentiation of
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
231
Fig. 8. Quantitative measurements of Triclosan released from PS–DVB–Triclosan beads immersed in ethanol solutions 50% (left) and 95% (right) vs. time of impregnation. Lines are drawn to guide the eye.
the Triclosan-releasing ability of the different formulations studied. The beads with lower cross-linking density and higher Triclosan amount, PS–DVB 2%– Tr 10%, show a high release rate of the order of 65% of the total amount in 1 month of immersion. As the cross-linking density was increased, the release rate was decreased to 50% of the total amount in 7 months of immersion. A further decrease in the release rate of Triclosan has taken place by decreasing the biocide content in the PS–DVB 2%–Tr 5% type of beads. In this case, a very slow but continuously increased release was observed in the time frame of 7 months with the relevant maximum released quantity being 5– 10% of the total amount. As it is clearly shown in this picture the Triclosan release rate can be effectively
tuned by using the proper combination of the crosslinking density and the active material loading. Furthermore, it is remarkable that the total amount of Triclosan (C max=0.020 g/100 mL) has not migrated into the solution, but a certain part of it remains in the bulk after 7 months of immersion in ethanol solutions, even in EtOH 95%. These findings suggest that two contingencies may occur: either an amount of Triclosan remains in the bulk because it is captured in the polymer network and so cannot be in contact with ethanol, or the period of time employed is not enough for the total migration/release of Triclosan to the solution. A rather similar release study was carried out for the phosphonium salt containing beads. The results
Fig. 9. Quantitative measurements of phosphonium salt released from PS–DVB–Psalt beads immersed in saline (left) and ethanol solution 10% (right) vs. time of impregnation. Lines are drawn to guide the eye.
232
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
indicating the evolution of the release of the phosphonium salt with time are shown in Fig. 9. More precisely, the concentration values of the phosphonium salt, determined via relevant calibration curves, in terms of optical densities at 293 nm, for three PS– DVB–Psalt type beads are plotted against time after being immersed in the natural serum (on the left) or in water–ethanol solution 10% (on the right). Once more, it is clear that the release of the low molecular weight biocide is favored when the amount of the crosslinking agent was decreased, i.e., going from the PS– DVB 5%–Psalt 10% to the PS–DVB 2%–Psalt 10% type of beads. A note is made here again of the fact that the third PS–DVB 10%–Psalt 20% type of beads exhibits higher release rate than the PS–DVB 5%– Psalt 10% formulation in both EtOH 10% and physiological saline solutions, although they contain the same ratio of the cross-linking to the biocide agent content. This is tentatively attributed to the higher dependence of the release on the biocide concentration itself in the polymeric matrix. Such a release at higher biocide loading might be due to an eventual increased biocide concentration gradient inside the polymeric cross-linked beads, making it diffuse more quickly. In a similar context, since the biocide behaves as an inert material, at high biocide concentration levels, it might ensure released channels that are getting less and less dependent on the cross-linking density as well. This reveals the importance of both the interplay between the two antagonistic to the release rate parameters and the overall concentration range employed. We may thus infer for both cases that for the controlled release, of great importance are the concentration of the biocide, the percentage of the crosslinks determined by the amount of the DVB added at the stage of synthesis, the percentage of their total concentration in the PS–DVB–biocide beads and the liquid mixture the beads are impregnated in.
4. Conclusions A method of incorporation of low molecular weight biostatic agents in polymer matrices that have been subject to some degree of cross-linking has been described. PS–DVB–biocide beads were synthesized by the simultaneous polymerization of a suspension of the monomer (styrene), the initiator, the cross-linking
agent (DVB) and the biostatic compound (Triclosan or cationic phosphonium salt) to be incorporated. The relative concentration of the biostatic agent and the cross-linking agent provides the capability to control the release/diffusion of the biostatic agent in liquids, such as natural serum and water–ethanol mixtures in contact to the polymeric beads. Ultraviolet–visible absorption spectrometry was used for controlling the rate of release/diffusion to the particular liquid phases of either the Triclosan or the phosphonium salt, which have been separately incorporated in PS–DVB polymeric beads matrix. Both quantitative (via UV absorption) and qualitative (via vibrational spectra) measurements indicate that increased biocide incorporation in the PS/DVB beads was accompanied by a corresponding enhancement of its concentration in either the water ethanol solutions or the physiological saline, while higher cross-linking densities hindered the biocide migration out of the beads by lowering its release rate into the liquid mixtures.
Acknowledgements The financial support for the present work by the Growth CEU SPAN (contract no. G5RD-CT-200100568) and the PAVET-2000 GSRT 1463 projects as well as the constructive cooperation with the ARGO S.A. company are greatly acknowledged. This work was also performed in the framework of the Interdepartmental Operation Program for Education and Initial Vocational Training on Polymer Science and Technology- 32a, 33H6 of the University of Patras, administered through the Minister of Education and Religious Affairs of Greece.
References [1] C.P. Gerba, G.E. Janauer, M. Costello, Removal of poliovirus and rotavirus from tapwater by a quaternary ammonium resin, Water Res. 18 (1) (1984) 17 – 19. [2] J. Hazzizalaskar, N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 1. Synthesis of polybutadiene with pendent quaternary ammonium groups, J. Appl. Polym. Sci. 50 (4) (1993) 651 – 662; N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 2. Biological evaluation of polyurethane coatings with pendant quaternary ammonium-salts, J. Appl. Polym. Sci. 50 (4) (1993) 663 – 670;
S.M. Iconomopoulou et al. / Journal of Controlled Release 102 (2005) 223–233
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
N. Nurdin, G. Helary, G. Sauvet, Biocidal polymers active by contact. 3. Aging of biological polyurethane coatings in water, J. Appl. Polym. Sci. 50 (4) (1993) 671 – 678. A.J. Isquith, E.A. Abbott, P.A. Walters, Surface-bonded antimicrobial activity of an organosilicon quaternary ammonium chloride, Appl. Microbiol. 24 (6) (1972) 859 – 863; P.A. Walters, E.A. Abbott, A.J. Isquith, Algicidal activity of a surface-bonded organosilicon ammonium chloride, Appl. Microbiol. 25 (2) (1973) 253 – 256. C. Hugues, C. Bessy, P. Bartolomeo, A. Margaillan, Complexation of an acrylic resin by tertiary amines: synthesis and characterization of new binders for antifouling paints, Eur. Polym. J. 39 (2) (2003) 319 – 326. A. Kanazawa, T. Ikeda, Multifuctional tetracoordinate phosphorus species with high self-organizing ability, Coord. Chem. Rev. 198 (2000) 117 – 131. G. Cerichelli, C. La Mesa, L. Luchetti, G. Mancini, Role of counterions in the catalytic activity and phase equilibria of phosphonium salts in water, Langmuir 16 (1) (2000) 166 – 171. E.R. Kenawy, F.I. Abdel-Hay, A. El-Raheem, R. El-Shanshoury, M.H. El-Newehy, Biologically active polymers: synthesis and antimicrobial activity of modified glycidyl methacrylate polymers having a quaternary ammonium and phosphonium groups, J. Control. Release 50 (1–3) (1998) 145 – 152. G.J. Li, J.R. Shen, Y.L. Zhu, A study of pyridinium-type functional polymers. III. Preparation and characterization of insoluble pyridinium-type polymers, J. Appl. Polym. Sci. 78 (3) (2000) 668 – 675. G.J. Li, J.R. Shen, Preparation and antibacterial activity of pyridinium-carrying polypropylene particles, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 42 (2) (2001) 360 – 361. T. Tarvainen, T. Karjalainen, M. Malin, S. Pohjolainen, J. Tuominen, J. Seppala, K. Jarvinen, Degradation of and drug release from a novel 2,2-bis(2-oxazoline) linked poly(lactic acid) polymer, J. Control. Release 81 (3) (2002) 251 – 261; T. Tarvainen, T. Karjalainen, M. Malin, K. Perakorpi, S. Pohjolainen, J. Tuominen, J. Seppala, K. Jarvinen, Drug release profiles from and degradation of a novel biodegradable polymer, 2,2-bis(2-oxazoline) linked poly(epsilon-caprolactone), Eur. J. Pharm. Sci. 16 (4–5) (2002) 323 – 331. T.E. Furia, A.G. Schenkel, New, broad spectrum bacteriostat, Soap Chem. Special 44 (1968) 47–50, 116–122.
233
[12] H.N. Bhargava, P.A. Leonard, Triclosan: applications and safety, Am. J. Infect. Control 24 (1996) 209 – 218. [13] R.D. Jones, H.B. Jampani, J.L. Newman, A.S. Lee, Triclosan: a review of effectiveness and safety in health care settings, Am. J. Infect. Control 28 (2) (2000) 184 – 196. [14] L.M. McMurry, M. Oethinger, S.B. Levy, Triclosan targets lipid synthesis, Nature 394 (1998) 531 – 532; C.W. Levy, A. Roujeinikova, S. Sedelnikova, P.J. Baker, A.R. Stuitjet, A.R. Slabast, D.W. Rice, J.B. Rafferty, Molecular basis of triclosan activity, Nature 398 (1999) 383 – 384. [15] A. Kanazawa, T. Ikeda, T. Endo, Polymeric phosphonium salts as a novel class of cationic biocides. II. Effects of counter anion and molecular weight on antibacterial activity of polymeric phosphonium salts, J. Polym. Sci., Polym. Chem. 31 (6) (1993) 1441 – 1447. [16] D.W. Emerson, Slow release of active chlorine and bromine from styrene divinylbenzene copolymers bearing N,Ndichlorosulfonamide, N-chloro-N-alkylsulfonamide, and Nbromo-N-alkylsulfonamide functional-groups-polymer-supported reagents. 6, Ind. Eng. Chem. Res. 30 (11) (1991) 2426 – 2430. [17] T. Tashiro, Removal of Escherichia coli from water by systems based on insoluble polystyrene-poly(ethylene glycol)s, polystyrene-polyethylenimides, and polystyrene-polyethylenepolyamines quaternized, J. Appl. Polym. Sci. 43 (7) (1991) 1369 – 1377. [18] M. Grayson, P.T. Keough, Phosphonium compounds. I. Elimination reactions of 2-cyanoethyl phosphonium salts. Asymmetric phosphines and alkylenediphosphines, J. Am. Chem. Soc. 81 (18) (1959) 4803 – 4807; M. Grayson, P.T. Keough, Phosphonium compounds: II. Decomposition of phosphonium alkoxides to hydrocarbon, ether and phosphine oxide, J. Am. Chem. Soc. 82 (15) (1960) 3919 – 3924. [19] A. Kanazawa, T. Ikeda, T. Endo, Polymeric phosphonium salts as a novel class of cationic biocides: 5. Synthesis and antibacterial activity of polyesters releasing phosphonium biocides, J. Polym. Sci., Polym. Chem. 31 (12) (1993) 3003 – 3011. [20] B.D. Kalyon, U. Olgun, Antibacterial efficacy of triclosanincorporated polymers, Am. J. Infect. Control 29 (2) (2001) 124 – 125.