Photochemical & Photobiological Sciences View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
PAPER
Cite this: Photochem. Photobiol. Sci., 2014, 13, 397
View Journal | View Issue
Are silicone-supported [C60]-fullerenes an alternative to Ru(II) polypyridyls for photodynamic solar water disinfection?† Francisco Manjón, Montserrat Santana-Magaña,‡ David García-Fresnadillo* and Guillermo Orellana* Different photosensitizing materials manufactured by immobilizing (0.5–3.0 g m−2) tris(4,7-diphenyl1,10-phenanthroline)ruthenium(II) (RDP2+), [C60]-fullerene, or 1-(4-methyl)-piperazinylfullerene (MPF) on porous neutral ( pSil) or surface-modified anionic ( pSil−) poly(dimethylsiloxane) are compared on the grounds of their singlet molecular oxygen (1O2) production and photodynamic solar water disinfection capability. The C60-based sensitizers display a broad weak absorption in the visible and strong absorption in the UV, while absorption of light by RDP2+ supported on pSil is strong in both the UV and blue regions. The 1O2 emission lifetimes (τΔ) determined for RDP2+ and MPF on porous silicone materials under air are similar (40–50 μs) and correspond to the decay of 1O2 generated by sensitizers dissolved in the polymer support. In contrast, τΔ measured for C60 in pSil is similar to that observed for MPF or RDP2+ when immobilized at low loading on pSil, but dramatically increases up to 5 ms if C60 aggregates are formed in the porous material as evidenced by microscopy evaluation. The photosensitizing properties of the dyes, together with their electrical charge and the overall charge of the porous silicone-based materials, lead to highly different sunlight-driven bacteria inactivation efficiencies, as tested with waterborne E. faecalis.
Received 13th October 2013, Accepted 5th December 2013
RDP/pSil provides efficient disinfection by photosensitization unlike MPF/pSil, which leads to reduced bacteria inactivation rates due to poorer 1O2 production. C60/pSil and MPF/pSil− materials, despite their 1
DOI: 10.1039/c3pp50361e
O2 photogeneration, show unsuccessful waterborne bacteria inactivation due to the negative surface charge of fullerene aggregates in contact with water, and to the net negative charge of the pSil−,
www.rsc.org/pps
respectively.
1.
Introduction
The well-known oxidizing properties of singlet molecular oxygen (abbreviated as 1O2 or O2(1Δg)) have led to the use of this reactive species in biomedical applications such as the photodynamic therapy (PDT) of cancer, photodynamic antimicrobial chemotherapy (PACT or APDT), photodynamic dermatological treatments and so on, or for water disinfection, as 1O2 is able to damage many cellular components by oxidation.1–5 Photosensitization is an effective mechanism to generate 1O2 by which, after absorption of light from a suitable
Optical Chemosensors & Applied Photochemistry Group (GSOLFA), Department of Organic Chemistry, Faculty of Chemistry, Complutense University of Madrid, E-28040 Madrid, Spain. E-mail: [email protected], [email protected]; Fax: +34-913944103; Tel: +34-913944220 † This paper is dedicated to the memory of Nick J. Turro. ‡ Current address: Laboratory of Industrial Microbiology, Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, Guadalajara, Jalisco, México.
This journal is © The Royal Society of Chemistry and Owner Societies 2014
dye ( photosensitizer), the excited state of the latter is quenched by the spontaneously dissolved molecular oxygen (O2) to yield 1O2 by a collisional energy transfer (also called “type II” photosensitization).6 The photodynamic action can occur in either homogeneous or heterogeneous media. Homogeneous photocatalysis is more efficient but the sensitizer must be removed afterwards from the solution.7 Photosensitizing dyes such as rose Bengal, methylene blue or (metallo)porphyrins display high 1O2 production quantum yields in water (ΦΔ > 0.5), but their limited stability calls only for specific applications.8,9 Unlike those dyes, Ru(II) coordination complexes with polyazaaromatic chelating ligands are efficient photosensitizers with high stability. For instance, [tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II)], RDP2+, and [tris(1,10-phenanthrolinyl-4,7-bis(benzenesulfonate))ruthenate(II)], RSD4−, display ΦΔ of 0.43 when dissolved in water and near unity in methanol.10 Moreover, this family of dyes can be readily attached to different polymers for optical chemical sensing,11–14 or to carry out heterogeneous sensitization with sunlight,15–19 with the aim of
Photochem. Photobiol. Sci., 2014, 13, 397–406 | 397
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Paper
providing point-of-use water disinfection in less favoured regions and countries.20 On the other side, stable molecular carbon species such as [C60]-fullerene show very interesting properties for the photogeneration of 1O2 because of their high quantum yield of intersystem crossing, long triplet state lifetimes and ΦΔ close to unity in solvents such as toluene, carbon disulfide or DMF/water, and in reverse micelles.21–27 Other advantages of fullerene are its photostability, a strong absorption in the UV-A region and broad capture of light in the visible due to its extended π-conjugated system of molecular orbitals. As a potential photodisinfectant, C60 is limited by its exceedingly low solubility in water,28 but it is possible to increase the level of C60 in water as stable aqueous suspensions of fullerene nanoparticles (“nC60”) using different methods such as stirring powdered C60 in water for a long time, or by its encapsulation within the core of micelles, among others.29–31 On the other hand, pristine fullerene can be chemically modified to produce derivatives with similar photochemical features to those of C60, but with increased solubility in polar media. Moreover, such functionalization may facilitate its immobilization on suitable solid supports to carry out heterogeneous photodynamic processes. For instance, C60 derivatives supported on silica have been used to oxidize organic chemicals or for photodynamic inactivation of microorganisms;32–35 additional applications include photochemical transformations in microreactor systems employing amine-coated beads or other solid supports.36,37 However, a limitation of fullerene derivatives is often the decrease in the efficiency of 1O2 photogeneration due to the partial loss of the extended π conjugation system of fullerene that leads to a decrease in the absorption of light efficiency and intersystem crossing yield.23,24,38–40 Moreover, the type II photosensitization of 1O2 in polar solvents competes with the photogeneration of other reactive oxygen species (ROS) such as the hydroxyl radical (HO•) and the superoxide anion (O2•−). These (“type I”) ROS are strongly oxidizing and therefore less selective than 1O2, and its formation is favoured in polar solvents, especially when reductants present in biological systems such as the cellular NADH are found nearby.39–41 Our aim has been to develop novel facile methods to combine the excellent photosensitizing features of [C60]-fullerene with the optimal characteristics of macroporous silicone (dimethylsiloxane) for developing field-deployable solar water disinfection systems: high O2 permeability, 1O2 friendliness, durability, favourable rheology and chemical inertia. Due to the latter, covalent attachment to silicone is considerably difficult. This paper describes simple procedures to immobilize [C60]-fullerene and a fullerene derivative onto a suitable polymer support to carry out photodynamic water disinfection. Efficient generation of 1O2 by these photosensitizing materials is demonstrated. Based on the photophysical and photochemical properties of the photosensitizer, the efficiency of waterborne bacteria inactivation of the aforementioned materials by photosensitized 1O2 production is compared
398 | Photochem. Photobiol. Sci., 2014, 13, 397–406
Photochemical & Photobiological Sciences
to the ability of known Ru(II) polypyridyl-based photosensitizing materials.15–19
2. Experimental 2.1.
Preparation of sensitizing materials
Tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II) dichloride, abbreviated as RDP2+, was selected as the reference photosensitizer (Fig. 1A) and prepared according to an established procedure.10 1-(4-Methyl)-piperazinylfullerene, abbreviated as MPF (Fig. 1B), was synthesized from [C60]-fullerene (>99%, Aldrich, WI, USA) as described by Goh et al.42 RDP2+, MPF and commercial C60 were immobilized on CulturSil porous silicone, hereinafter called pSil (Cellon, Luxembourg).16 This biocompatible material consists of 150–300 μm interconnected pores with an estimated porosity of 40% by volume, with a density of 1.065 g cm−3 and a specific area of 5 × 104 m2 m−3 according to the manufacturer’s information. RDP2+ was adsorbed by hydrophobic interactions onto pSil up to the sensitizer saturation, as described previously.16 The photosensitizing material prepared in this way is designated as RDP/pSil. Pristine fullerene was trapped by the swell-encapsulation-shrink method into 35 × 8 mm (wet) pieces of 1.5 mm thick pSil that were previously soaked for 30 min in dichloromethane (99.9%, Panreac, Spain) and then were immersed for 30 min in stirred solutions of C60 in toluene (99.9%, Panreac, Spain; 20 mL each of various concentrations up to ca. 3 × 10−3 M, close to its solubility limit).28 In this way, the denominated C60/pSil photosensitizing materials were obtained with different C60 loadings. MPF was immobilized onto pSil by hydrophobic interactions to yield the MPF/ pSil material, or it was attached by electrostatic interactions to an anionic surface-modified pSil film prepared by Bionic Surfaces (Wuerzburg, Germany) according to a proprietary wetchemical procedure (MPF/pSil−).43 MPF immobilizations in porous films were achieved by dipping during 1 day 35 × 8 mm pieces of pSil or pSil− in 10 mL of stirred solutions of MPF in chloroform (5 × 10−5 and 1 × 10−6 M, respectively). C60 was also incorporated into polydimethylsiloxane silicone oil (DMS-S15, 45–85 cSt, 0.9–1.2% silanol-terminated,
Fig. 1 Chemical structure of the selected photosensitizers RDP2+ (A) and MPF (B).
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Photochemical & Photobiological Sciences
ABCR, Germany) and into conformal silicone coating (1-2577, Dow Corning, MI), by stirring 2.0 mg of powdered C60 in 2.0 g of silicone oil or conformal coating contained in standard fluorescence PMMA cells (10 × 10 mm optical pathlength, Kartell, Italy). The photosensitizing materials prepared in this way are designated as C60/PDMS and C60/sil, respectively. Alternatively, although the commercial conformal coating already contains 15–40% toluene by weight, as per the manufacturer’s information, C60 was also added after dissolving it previously in toluene (1.0 mL of a 3 × 10−3 M C60 solution mixed with 2.0 g of conformal coating) to yield the material called C60/sil–tol. Both C60/sil and C60/sil–tol were subjected to vacuum (0.1 Torr) overnight to completely evaporate toluene before performing the 1O2 production. As a control photosensitizing system, TPP (5,10,15,20-tetraphenyl-21H,23H-porphine, 97%, Aldrich) in silicone oil was used (2.0 mg of TPP added to 2.0 g of stirred conformal silicone coating). 2.2. Photophysical and photochemical characterization of the photosensitizing materials The absorption spectra of the sensitizer solutions were recorded with a Varian Cary 3Bio UV-Vis spectrophotometer (CA, USA). Transient absorbance measurements of O2-free sensitizer solutions were carried out with an Edinburgh Instruments (UK) LP-900 laser kinetic spectrometer system upon excitation at 532 nm (20 mJ per pulse) with a Nd:YAG laser head (Minilite II, Continuum, CA). Absorbance changes were monitored at 785 nm with a Hamamatsu R-955 PMT using as a probe beam (90° arrangement) the white light from a 450 W Xe lamp (Edinburgh Instruments) operated at 0.5 Hz. The 1O2 lifetime of the moist photosensitizing materials was measured using an EI LP-900 laser kinetic spectrometer equipped with a Hamamatsu H10330-45 NIR PMT for the time-resolved detection at 1270 nm of the 1O2 luminescence. The emission spectrum of the 1O2 photogenerated by different materials upon laser excitation was obtained in a front-face illumination arrangement after averaging 25 individual emission decays every 10 nm between 1190 and 1360 nm, using the time-resolved emission spectroscopy (TRES) acquisition mode included in the EI software package. Micrographs were obtained using a Horiba (NJ, USA) DynaMic instrument operating in either fluorescence intensity or lifetime imaging (FLIM) mode, equipped with an epifluorescence confocal microscope (Olympus BX51, NY, USA) fitted with a 1600 × 1200 pixels CCD camera (uEye UI-1450-C, IDS, Germany) and a Peltier-cooled Horiba TBX-04 detection module operated in single-photon timing mode. 2.3.
Photodisinfection assays
Microorganism inactivation tests were carried out using 0.8 × 3.5 cm pieces of different photosensitizing materials in contact with the aqueous media. The experiments were performed in outdoor runs using fluorescence plastic cells (4 mL water volume) or in a solar-simulated lab reactor (10 mL water volume, 0.25 mL min−1 flow-rate) based on a collector-free tubular photoreactor placed 10 cm away from a 150-W Xe lamp (Oriel, Stratford, CT, USA).15 The microorganism used to
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Paper
evaluate the photodynamic water disinfection of the novel materials was Enterococcus faecalis (CECT 5143), a representative strain of gram-positive bacteria. The initial microorganism concentration was 2 × 103 CFU mL−1 in purified water (Millipore Direct-Q, MA, USA). The culturability assessment, procedures and evaluation of results were similar to those described in a previous study, unless stated otherwise.16
3. Results and discussion 3.1. Characterization of the photosensitizing dyes and materials Electronic absorption features. The typical absorption spectrum of C60 solutions displays strong bands in the UV region and a broad band in the visible region. For instance, C60 dissolved in toluene shows absorption maxima at 335 (ε = 54 180 M−1 cm−1) and 534 nm (ε = 965 M−1 cm−1) (±10%). Likewise, MPF dissolved in chloroform presents a strong absorption in the UV region with a characteristic shoulder at 284 nm (ε = 60 185 M−1 cm−1, ±15%). The orange RDP2+ sensitizer shows a strong absorption in the UV, with a maximum in the visible at 461 nm (ε = 32 400 M−1 cm−1, ±10%) in aqueous solution. Photosensitizer loading. As described in a previous study,16 the load of RDP/pSil material, after its conditioning by autoclave treatment in order to avoid the presence of the dye in the aqueous phase, was (2.0 ± 0.2) g m−2, corresponding to the maximum sensitizer adsorption onto silicone from water. Similarly, payloads of C60 in pSil were estimated from spectrophotometric analysis of the supernatants at 335 nm. Unlike the Ru dye, prolonged contact between C60 dissolved in toluene and pSil leads to marginal adsorption of fullerene onto the silicone material. However, the fullerene may be entrapped into the poly(dimethylsiloxane) network by prior swelling of the porous silicone in dichloromethane followed by a rapid shrinkage upon dipping it in toluene containing the dissolved C60. Because of this fast volume change upon exchanging the organic solvents, the fullerene loading process into pSil is virtually instantaneous and approximately proportional to the C60 concentration in the toluene solution, providing loads up to 3 g m−2 (±10%). The corresponding sensitizer contents of the MPF/pSil and MPF/pSil− materials, estimated from spectrophotometric analysis of the supernatants at 284 nm, was 0.5 and 1.5 g m−2 (±15%), respectively. The extremely low solubility of MPF and C60 in water prevents any leaching into the aqueous phase after immobilization. Singlet oxygen photogeneration. Singlet molecular oxygen is generated by photosensitization through electronic energy transfer from the long-lived triplet excited state of sensitizers to O2. The production of 1O2 has been evidenced by its characteristic luminescence centred at 1270 nm, a value slightly dependent on the solvent or material nature. The emission spectrum of 1O2 generated by C60/pSil, MPF/pSil, MPF/pSil− and RDP/pSil moist materials measured under air by timeresolved emission spectroscopy (TRES) in the 1190–1360 nm range is depicted in Fig. 2. It must be noticed that at short
Photochem. Photobiol. Sci., 2014, 13, 397–406 | 399
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Paper
Photochemical & Photobiological Sciences
Fig. 2 Time-resolved emission spectroscopy (TRES) of 1O2 photogenerated under air by four photosensitizing materials: C60/pSil (A) in a time window of 10 ms, MPF/pSil− (B) and RDP/pSil (D) in a time window of 200 μs, and MPF/pSil (C) in a time window of 100 μs.
times after the laser pulse, the NIR emission of the 1O2 photogenerated by RDP/pSil is superimposed with the tail of the luminescence band of the Ru(II) complex, in contrast to the non-luminescent [C60]-based materials. Fig. 3 displays a comparison of the emission decays at 1270 nm of singlet oxygen photogenerated by C60 and by RDP2+ immobilized on porous silicone.
Fig. 3 Emission kinetics at 1270 nm of 1O2 photogenerated under air with C60/pSil (gray) and RDP/pSil (black and insert detail) as sensitizing materials.
400 | Photochem. Photobiol. Sci., 2014, 13, 397–406
The singlet oxygen lifetimes (τΔ) measured under air for the photosensitizing materials based on pSil are 40 μs for RDP/ pSil, 43 μs for MPF/pSil− and 47 μs for MPF/pSil (±10%). These similar values of τΔ suggest that 1O2 is generated by the different photosensitizers dwelling in an analogous microenvironment into the silicone materials, as the 1O2 lifetime is strongly dependent on the surrounding media where it is generated.44 However, in the case of C60/pSil, τΔ dramatically increases up to ca. 5 ms, with a significant variation with the fullerene load. The entrapment process of C60 into the porous silicone could lead to a different distribution of the sensitizer, so that the microenvironment around the 1O2 photogeneration site would be different from those of the other photosensitizers even within the same material. In order to account for this fact, the 1O2 emission decay profiles were satisfactorily fitted to a sum of 3-exponential functions and the lifetime components (τΔi) and their weights (Bi) may be analysed with respect to the C60 loading. The least squares fits of the 1O2 emission decays are shown in Table 1. In all cases, it is possible to distinguish a slow-decaying contribution τΔ3 in the ms range and a short-lived component τΔ1 in the order of several tens of μs that is similar to the 1O2 emission lifetimes measured within the other pSil-based materials described above. The possibility of the observed lifetimes being governed by the C60 triplet state decay must be ruled out: a transient absorption experiment at 720 nm (data not shown) demonstrates that the C60 triplet state lifetime in
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Photochemical & Photobiological Sciences
Paper
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Table 1 Three-exponential fits of the emission kinetics of 1O2 photogenerated by C60/pSil materials with different fullerene loads. The uncertainty of the lifetime components (τΔi) and their relative pre-exponential factors (Bi) is ±10%
C60 load/g m−2
τΔ1/ms
%B1
τΔ2/ms
%B2
τΔ3/ms
%B3
1.0 1.5 3.0
0.025 0.033 0.084
60 41 9
0.154 0.264 1.31
29 29 9
2.25 3.01 4.69
11 30 82
pSil/air is under 10 ns under the experimental conditions of Fig. 3. The 1O2 lifetime values of the three components increase with the C60 load while the relative weight of the long lifetime component, B3, notably increases by a larger amount. Moreover, if the bright field microscopy images of the C60/pSil films are analyzed (Fig. 4), C60 particles measuring ca. 1–2 μm are found in all the materials containing this photosensitizer (and most probably also smaller ones below the optical resolution). For ca. 1 g m−2 of C60 load, most of the particles are within that size range but, for 1.5 g m−2, lots of 5–10 μm particles are found and, for the highest loading of C60 in pSil (3 g m−2), fullerene aggregates as large as 20 μm are evident. Taking into account the microscopy results, it is possible to associate the lifetime components of the tri-exponential fits of the emission decay profiles with 1O2 photogenerated by C60 into pSil in different microenvironments. The fast-decaying component τΔ1 might correspond to τΔ generated by molecular C60 genuinely dissolved onto pSil in a similar way to that for
MPF or RDP2+, for which it is hard to find any aggregate (Fig. 5) because their main immobilization mechanism onto the polymer material is adsorption rather than aggregation and inclusion.19 The intermediate component τΔ2 is assigned to 1O2 produced by fullerene sub-micrometer clusters embedded into the pSil, while the ms lifetime component τΔ3 would be compatible with the 1O2 generated by C60 aggregates in the μm scale. The various fullerene loads lead to different distributions of C60 in the above mentioned forms, changing also the pre-exponential factors Bi. Nevertheless, we do not intend to assign the different lifetime values to three discrete 1 O2-generating species within the silicone matrix but they would rather correspond to the centre values of a continuum of lifetime distributions. In order to support those assumptions, the generated 1O2 decay was measured for C60-fullerene in different environments, namely, C60 powder in contact with air placed between two microscopy slides (C60/air), C60 embedded into polydimethylsiloxane fluid (C60/PDMS), and C60 suspended in a silicone solution for conformal coatings. The conformal coating solution contains toluene, but C60 was added either powdered or previously dissolved in additional toluene; in any case, toluene was removed under high vacuum from both mixtures before carrying out the τΔ measurements (these preparations are abbreviated as C60/sil and C60/sil–tol, respectively). The parameters of the multiexponential fits corresponding to the 1O2 emission generated by these C60-containing photosensitizing materials are collected in Table 2.
Fig. 4 Bright field microscopy images of C60/pSil films with different fullerene loads: (A) 1 g m−2, (B) 1.5 g m−2 and (C) 3 g m−2; the particle size analysis has been made with ImageJ software (v. 1.45 s).
Fig. 5 Bright field micrographs (40 μm observation window) of pSil (A), RDP/pSil (B, with detail of a ca. 1 μm RDP2+ crystal), and MPF/pSil (C, with detail of a
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
PAPER
Cite this: Photochem. Photobiol. Sci., 2014, 13, 397
View Journal | View Issue
Are silicone-supported [C60]-fullerenes an alternative to Ru(II) polypyridyls for photodynamic solar water disinfection?† Francisco Manjón, Montserrat Santana-Magaña,‡ David García-Fresnadillo* and Guillermo Orellana* Different photosensitizing materials manufactured by immobilizing (0.5–3.0 g m−2) tris(4,7-diphenyl1,10-phenanthroline)ruthenium(II) (RDP2+), [C60]-fullerene, or 1-(4-methyl)-piperazinylfullerene (MPF) on porous neutral ( pSil) or surface-modified anionic ( pSil−) poly(dimethylsiloxane) are compared on the grounds of their singlet molecular oxygen (1O2) production and photodynamic solar water disinfection capability. The C60-based sensitizers display a broad weak absorption in the visible and strong absorption in the UV, while absorption of light by RDP2+ supported on pSil is strong in both the UV and blue regions. The 1O2 emission lifetimes (τΔ) determined for RDP2+ and MPF on porous silicone materials under air are similar (40–50 μs) and correspond to the decay of 1O2 generated by sensitizers dissolved in the polymer support. In contrast, τΔ measured for C60 in pSil is similar to that observed for MPF or RDP2+ when immobilized at low loading on pSil, but dramatically increases up to 5 ms if C60 aggregates are formed in the porous material as evidenced by microscopy evaluation. The photosensitizing properties of the dyes, together with their electrical charge and the overall charge of the porous silicone-based materials, lead to highly different sunlight-driven bacteria inactivation efficiencies, as tested with waterborne E. faecalis.
Received 13th October 2013, Accepted 5th December 2013
RDP/pSil provides efficient disinfection by photosensitization unlike MPF/pSil, which leads to reduced bacteria inactivation rates due to poorer 1O2 production. C60/pSil and MPF/pSil− materials, despite their 1
DOI: 10.1039/c3pp50361e
O2 photogeneration, show unsuccessful waterborne bacteria inactivation due to the negative surface charge of fullerene aggregates in contact with water, and to the net negative charge of the pSil−,
www.rsc.org/pps
respectively.
1.
Introduction
The well-known oxidizing properties of singlet molecular oxygen (abbreviated as 1O2 or O2(1Δg)) have led to the use of this reactive species in biomedical applications such as the photodynamic therapy (PDT) of cancer, photodynamic antimicrobial chemotherapy (PACT or APDT), photodynamic dermatological treatments and so on, or for water disinfection, as 1O2 is able to damage many cellular components by oxidation.1–5 Photosensitization is an effective mechanism to generate 1O2 by which, after absorption of light from a suitable
Optical Chemosensors & Applied Photochemistry Group (GSOLFA), Department of Organic Chemistry, Faculty of Chemistry, Complutense University of Madrid, E-28040 Madrid, Spain. E-mail: [email protected], [email protected]; Fax: +34-913944103; Tel: +34-913944220 † This paper is dedicated to the memory of Nick J. Turro. ‡ Current address: Laboratory of Industrial Microbiology, Centro Universitario de Ciencias Exactas e Ingenierías, Universidad de Guadalajara, Guadalajara, Jalisco, México.
This journal is © The Royal Society of Chemistry and Owner Societies 2014
dye ( photosensitizer), the excited state of the latter is quenched by the spontaneously dissolved molecular oxygen (O2) to yield 1O2 by a collisional energy transfer (also called “type II” photosensitization).6 The photodynamic action can occur in either homogeneous or heterogeneous media. Homogeneous photocatalysis is more efficient but the sensitizer must be removed afterwards from the solution.7 Photosensitizing dyes such as rose Bengal, methylene blue or (metallo)porphyrins display high 1O2 production quantum yields in water (ΦΔ > 0.5), but their limited stability calls only for specific applications.8,9 Unlike those dyes, Ru(II) coordination complexes with polyazaaromatic chelating ligands are efficient photosensitizers with high stability. For instance, [tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II)], RDP2+, and [tris(1,10-phenanthrolinyl-4,7-bis(benzenesulfonate))ruthenate(II)], RSD4−, display ΦΔ of 0.43 when dissolved in water and near unity in methanol.10 Moreover, this family of dyes can be readily attached to different polymers for optical chemical sensing,11–14 or to carry out heterogeneous sensitization with sunlight,15–19 with the aim of
Photochem. Photobiol. Sci., 2014, 13, 397–406 | 397
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Paper
providing point-of-use water disinfection in less favoured regions and countries.20 On the other side, stable molecular carbon species such as [C60]-fullerene show very interesting properties for the photogeneration of 1O2 because of their high quantum yield of intersystem crossing, long triplet state lifetimes and ΦΔ close to unity in solvents such as toluene, carbon disulfide or DMF/water, and in reverse micelles.21–27 Other advantages of fullerene are its photostability, a strong absorption in the UV-A region and broad capture of light in the visible due to its extended π-conjugated system of molecular orbitals. As a potential photodisinfectant, C60 is limited by its exceedingly low solubility in water,28 but it is possible to increase the level of C60 in water as stable aqueous suspensions of fullerene nanoparticles (“nC60”) using different methods such as stirring powdered C60 in water for a long time, or by its encapsulation within the core of micelles, among others.29–31 On the other hand, pristine fullerene can be chemically modified to produce derivatives with similar photochemical features to those of C60, but with increased solubility in polar media. Moreover, such functionalization may facilitate its immobilization on suitable solid supports to carry out heterogeneous photodynamic processes. For instance, C60 derivatives supported on silica have been used to oxidize organic chemicals or for photodynamic inactivation of microorganisms;32–35 additional applications include photochemical transformations in microreactor systems employing amine-coated beads or other solid supports.36,37 However, a limitation of fullerene derivatives is often the decrease in the efficiency of 1O2 photogeneration due to the partial loss of the extended π conjugation system of fullerene that leads to a decrease in the absorption of light efficiency and intersystem crossing yield.23,24,38–40 Moreover, the type II photosensitization of 1O2 in polar solvents competes with the photogeneration of other reactive oxygen species (ROS) such as the hydroxyl radical (HO•) and the superoxide anion (O2•−). These (“type I”) ROS are strongly oxidizing and therefore less selective than 1O2, and its formation is favoured in polar solvents, especially when reductants present in biological systems such as the cellular NADH are found nearby.39–41 Our aim has been to develop novel facile methods to combine the excellent photosensitizing features of [C60]-fullerene with the optimal characteristics of macroporous silicone (dimethylsiloxane) for developing field-deployable solar water disinfection systems: high O2 permeability, 1O2 friendliness, durability, favourable rheology and chemical inertia. Due to the latter, covalent attachment to silicone is considerably difficult. This paper describes simple procedures to immobilize [C60]-fullerene and a fullerene derivative onto a suitable polymer support to carry out photodynamic water disinfection. Efficient generation of 1O2 by these photosensitizing materials is demonstrated. Based on the photophysical and photochemical properties of the photosensitizer, the efficiency of waterborne bacteria inactivation of the aforementioned materials by photosensitized 1O2 production is compared
398 | Photochem. Photobiol. Sci., 2014, 13, 397–406
Photochemical & Photobiological Sciences
to the ability of known Ru(II) polypyridyl-based photosensitizing materials.15–19
2. Experimental 2.1.
Preparation of sensitizing materials
Tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II) dichloride, abbreviated as RDP2+, was selected as the reference photosensitizer (Fig. 1A) and prepared according to an established procedure.10 1-(4-Methyl)-piperazinylfullerene, abbreviated as MPF (Fig. 1B), was synthesized from [C60]-fullerene (>99%, Aldrich, WI, USA) as described by Goh et al.42 RDP2+, MPF and commercial C60 were immobilized on CulturSil porous silicone, hereinafter called pSil (Cellon, Luxembourg).16 This biocompatible material consists of 150–300 μm interconnected pores with an estimated porosity of 40% by volume, with a density of 1.065 g cm−3 and a specific area of 5 × 104 m2 m−3 according to the manufacturer’s information. RDP2+ was adsorbed by hydrophobic interactions onto pSil up to the sensitizer saturation, as described previously.16 The photosensitizing material prepared in this way is designated as RDP/pSil. Pristine fullerene was trapped by the swell-encapsulation-shrink method into 35 × 8 mm (wet) pieces of 1.5 mm thick pSil that were previously soaked for 30 min in dichloromethane (99.9%, Panreac, Spain) and then were immersed for 30 min in stirred solutions of C60 in toluene (99.9%, Panreac, Spain; 20 mL each of various concentrations up to ca. 3 × 10−3 M, close to its solubility limit).28 In this way, the denominated C60/pSil photosensitizing materials were obtained with different C60 loadings. MPF was immobilized onto pSil by hydrophobic interactions to yield the MPF/ pSil material, or it was attached by electrostatic interactions to an anionic surface-modified pSil film prepared by Bionic Surfaces (Wuerzburg, Germany) according to a proprietary wetchemical procedure (MPF/pSil−).43 MPF immobilizations in porous films were achieved by dipping during 1 day 35 × 8 mm pieces of pSil or pSil− in 10 mL of stirred solutions of MPF in chloroform (5 × 10−5 and 1 × 10−6 M, respectively). C60 was also incorporated into polydimethylsiloxane silicone oil (DMS-S15, 45–85 cSt, 0.9–1.2% silanol-terminated,
Fig. 1 Chemical structure of the selected photosensitizers RDP2+ (A) and MPF (B).
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Photochemical & Photobiological Sciences
ABCR, Germany) and into conformal silicone coating (1-2577, Dow Corning, MI), by stirring 2.0 mg of powdered C60 in 2.0 g of silicone oil or conformal coating contained in standard fluorescence PMMA cells (10 × 10 mm optical pathlength, Kartell, Italy). The photosensitizing materials prepared in this way are designated as C60/PDMS and C60/sil, respectively. Alternatively, although the commercial conformal coating already contains 15–40% toluene by weight, as per the manufacturer’s information, C60 was also added after dissolving it previously in toluene (1.0 mL of a 3 × 10−3 M C60 solution mixed with 2.0 g of conformal coating) to yield the material called C60/sil–tol. Both C60/sil and C60/sil–tol were subjected to vacuum (0.1 Torr) overnight to completely evaporate toluene before performing the 1O2 production. As a control photosensitizing system, TPP (5,10,15,20-tetraphenyl-21H,23H-porphine, 97%, Aldrich) in silicone oil was used (2.0 mg of TPP added to 2.0 g of stirred conformal silicone coating). 2.2. Photophysical and photochemical characterization of the photosensitizing materials The absorption spectra of the sensitizer solutions were recorded with a Varian Cary 3Bio UV-Vis spectrophotometer (CA, USA). Transient absorbance measurements of O2-free sensitizer solutions were carried out with an Edinburgh Instruments (UK) LP-900 laser kinetic spectrometer system upon excitation at 532 nm (20 mJ per pulse) with a Nd:YAG laser head (Minilite II, Continuum, CA). Absorbance changes were monitored at 785 nm with a Hamamatsu R-955 PMT using as a probe beam (90° arrangement) the white light from a 450 W Xe lamp (Edinburgh Instruments) operated at 0.5 Hz. The 1O2 lifetime of the moist photosensitizing materials was measured using an EI LP-900 laser kinetic spectrometer equipped with a Hamamatsu H10330-45 NIR PMT for the time-resolved detection at 1270 nm of the 1O2 luminescence. The emission spectrum of the 1O2 photogenerated by different materials upon laser excitation was obtained in a front-face illumination arrangement after averaging 25 individual emission decays every 10 nm between 1190 and 1360 nm, using the time-resolved emission spectroscopy (TRES) acquisition mode included in the EI software package. Micrographs were obtained using a Horiba (NJ, USA) DynaMic instrument operating in either fluorescence intensity or lifetime imaging (FLIM) mode, equipped with an epifluorescence confocal microscope (Olympus BX51, NY, USA) fitted with a 1600 × 1200 pixels CCD camera (uEye UI-1450-C, IDS, Germany) and a Peltier-cooled Horiba TBX-04 detection module operated in single-photon timing mode. 2.3.
Photodisinfection assays
Microorganism inactivation tests were carried out using 0.8 × 3.5 cm pieces of different photosensitizing materials in contact with the aqueous media. The experiments were performed in outdoor runs using fluorescence plastic cells (4 mL water volume) or in a solar-simulated lab reactor (10 mL water volume, 0.25 mL min−1 flow-rate) based on a collector-free tubular photoreactor placed 10 cm away from a 150-W Xe lamp (Oriel, Stratford, CT, USA).15 The microorganism used to
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Paper
evaluate the photodynamic water disinfection of the novel materials was Enterococcus faecalis (CECT 5143), a representative strain of gram-positive bacteria. The initial microorganism concentration was 2 × 103 CFU mL−1 in purified water (Millipore Direct-Q, MA, USA). The culturability assessment, procedures and evaluation of results were similar to those described in a previous study, unless stated otherwise.16
3. Results and discussion 3.1. Characterization of the photosensitizing dyes and materials Electronic absorption features. The typical absorption spectrum of C60 solutions displays strong bands in the UV region and a broad band in the visible region. For instance, C60 dissolved in toluene shows absorption maxima at 335 (ε = 54 180 M−1 cm−1) and 534 nm (ε = 965 M−1 cm−1) (±10%). Likewise, MPF dissolved in chloroform presents a strong absorption in the UV region with a characteristic shoulder at 284 nm (ε = 60 185 M−1 cm−1, ±15%). The orange RDP2+ sensitizer shows a strong absorption in the UV, with a maximum in the visible at 461 nm (ε = 32 400 M−1 cm−1, ±10%) in aqueous solution. Photosensitizer loading. As described in a previous study,16 the load of RDP/pSil material, after its conditioning by autoclave treatment in order to avoid the presence of the dye in the aqueous phase, was (2.0 ± 0.2) g m−2, corresponding to the maximum sensitizer adsorption onto silicone from water. Similarly, payloads of C60 in pSil were estimated from spectrophotometric analysis of the supernatants at 335 nm. Unlike the Ru dye, prolonged contact between C60 dissolved in toluene and pSil leads to marginal adsorption of fullerene onto the silicone material. However, the fullerene may be entrapped into the poly(dimethylsiloxane) network by prior swelling of the porous silicone in dichloromethane followed by a rapid shrinkage upon dipping it in toluene containing the dissolved C60. Because of this fast volume change upon exchanging the organic solvents, the fullerene loading process into pSil is virtually instantaneous and approximately proportional to the C60 concentration in the toluene solution, providing loads up to 3 g m−2 (±10%). The corresponding sensitizer contents of the MPF/pSil and MPF/pSil− materials, estimated from spectrophotometric analysis of the supernatants at 284 nm, was 0.5 and 1.5 g m−2 (±15%), respectively. The extremely low solubility of MPF and C60 in water prevents any leaching into the aqueous phase after immobilization. Singlet oxygen photogeneration. Singlet molecular oxygen is generated by photosensitization through electronic energy transfer from the long-lived triplet excited state of sensitizers to O2. The production of 1O2 has been evidenced by its characteristic luminescence centred at 1270 nm, a value slightly dependent on the solvent or material nature. The emission spectrum of 1O2 generated by C60/pSil, MPF/pSil, MPF/pSil− and RDP/pSil moist materials measured under air by timeresolved emission spectroscopy (TRES) in the 1190–1360 nm range is depicted in Fig. 2. It must be noticed that at short
Photochem. Photobiol. Sci., 2014, 13, 397–406 | 399
View Article Online
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Paper
Photochemical & Photobiological Sciences
Fig. 2 Time-resolved emission spectroscopy (TRES) of 1O2 photogenerated under air by four photosensitizing materials: C60/pSil (A) in a time window of 10 ms, MPF/pSil− (B) and RDP/pSil (D) in a time window of 200 μs, and MPF/pSil (C) in a time window of 100 μs.
times after the laser pulse, the NIR emission of the 1O2 photogenerated by RDP/pSil is superimposed with the tail of the luminescence band of the Ru(II) complex, in contrast to the non-luminescent [C60]-based materials. Fig. 3 displays a comparison of the emission decays at 1270 nm of singlet oxygen photogenerated by C60 and by RDP2+ immobilized on porous silicone.
Fig. 3 Emission kinetics at 1270 nm of 1O2 photogenerated under air with C60/pSil (gray) and RDP/pSil (black and insert detail) as sensitizing materials.
400 | Photochem. Photobiol. Sci., 2014, 13, 397–406
The singlet oxygen lifetimes (τΔ) measured under air for the photosensitizing materials based on pSil are 40 μs for RDP/ pSil, 43 μs for MPF/pSil− and 47 μs for MPF/pSil (±10%). These similar values of τΔ suggest that 1O2 is generated by the different photosensitizers dwelling in an analogous microenvironment into the silicone materials, as the 1O2 lifetime is strongly dependent on the surrounding media where it is generated.44 However, in the case of C60/pSil, τΔ dramatically increases up to ca. 5 ms, with a significant variation with the fullerene load. The entrapment process of C60 into the porous silicone could lead to a different distribution of the sensitizer, so that the microenvironment around the 1O2 photogeneration site would be different from those of the other photosensitizers even within the same material. In order to account for this fact, the 1O2 emission decay profiles were satisfactorily fitted to a sum of 3-exponential functions and the lifetime components (τΔi) and their weights (Bi) may be analysed with respect to the C60 loading. The least squares fits of the 1O2 emission decays are shown in Table 1. In all cases, it is possible to distinguish a slow-decaying contribution τΔ3 in the ms range and a short-lived component τΔ1 in the order of several tens of μs that is similar to the 1O2 emission lifetimes measured within the other pSil-based materials described above. The possibility of the observed lifetimes being governed by the C60 triplet state decay must be ruled out: a transient absorption experiment at 720 nm (data not shown) demonstrates that the C60 triplet state lifetime in
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Photochemical & Photobiological Sciences
Paper
Published on 05 December 2013. Downloaded by University of Windsor on 29/10/2014 20:12:25.
Table 1 Three-exponential fits of the emission kinetics of 1O2 photogenerated by C60/pSil materials with different fullerene loads. The uncertainty of the lifetime components (τΔi) and their relative pre-exponential factors (Bi) is ±10%
C60 load/g m−2
τΔ1/ms
%B1
τΔ2/ms
%B2
τΔ3/ms
%B3
1.0 1.5 3.0
0.025 0.033 0.084
60 41 9
0.154 0.264 1.31
29 29 9
2.25 3.01 4.69
11 30 82
pSil/air is under 10 ns under the experimental conditions of Fig. 3. The 1O2 lifetime values of the three components increase with the C60 load while the relative weight of the long lifetime component, B3, notably increases by a larger amount. Moreover, if the bright field microscopy images of the C60/pSil films are analyzed (Fig. 4), C60 particles measuring ca. 1–2 μm are found in all the materials containing this photosensitizer (and most probably also smaller ones below the optical resolution). For ca. 1 g m−2 of C60 load, most of the particles are within that size range but, for 1.5 g m−2, lots of 5–10 μm particles are found and, for the highest loading of C60 in pSil (3 g m−2), fullerene aggregates as large as 20 μm are evident. Taking into account the microscopy results, it is possible to associate the lifetime components of the tri-exponential fits of the emission decay profiles with 1O2 photogenerated by C60 into pSil in different microenvironments. The fast-decaying component τΔ1 might correspond to τΔ generated by molecular C60 genuinely dissolved onto pSil in a similar way to that for
MPF or RDP2+, for which it is hard to find any aggregate (Fig. 5) because their main immobilization mechanism onto the polymer material is adsorption rather than aggregation and inclusion.19 The intermediate component τΔ2 is assigned to 1O2 produced by fullerene sub-micrometer clusters embedded into the pSil, while the ms lifetime component τΔ3 would be compatible with the 1O2 generated by C60 aggregates in the μm scale. The various fullerene loads lead to different distributions of C60 in the above mentioned forms, changing also the pre-exponential factors Bi. Nevertheless, we do not intend to assign the different lifetime values to three discrete 1 O2-generating species within the silicone matrix but they would rather correspond to the centre values of a continuum of lifetime distributions. In order to support those assumptions, the generated 1O2 decay was measured for C60-fullerene in different environments, namely, C60 powder in contact with air placed between two microscopy slides (C60/air), C60 embedded into polydimethylsiloxane fluid (C60/PDMS), and C60 suspended in a silicone solution for conformal coatings. The conformal coating solution contains toluene, but C60 was added either powdered or previously dissolved in additional toluene; in any case, toluene was removed under high vacuum from both mixtures before carrying out the τΔ measurements (these preparations are abbreviated as C60/sil and C60/sil–tol, respectively). The parameters of the multiexponential fits corresponding to the 1O2 emission generated by these C60-containing photosensitizing materials are collected in Table 2.
Fig. 4 Bright field microscopy images of C60/pSil films with different fullerene loads: (A) 1 g m−2, (B) 1.5 g m−2 and (C) 3 g m−2; the particle size analysis has been made with ImageJ software (v. 1.45 s).
Fig. 5 Bright field micrographs (40 μm observation window) of pSil (A), RDP/pSil (B, with detail of a ca. 1 μm RDP2+ crystal), and MPF/pSil (C, with detail of a
