Stable optical oxygen sensing materials based on click-coupling of fluorinated platinum(II) and palladium(II) porphyrins-A convenient way to eliminate dye migration and leaching.

Nucleophilic substitution of the labile para-fluorine atoms of 2,3,4,5,6-pentafluorophenyl groups enables a click-based covalent linkage of an oxygen indicator (platinum(II) or palladium(II) 5,10,15,20-meso-tetrakis-(2,3,4,5,6-pentafluorophenyl)-porphyrin) to the sensor matrix. Copolymers of styrene and pentafluorostyrene are chosen as polymeric materials. Depending on the reaction conditions either soluble sensor materials or cross-linked microparticles are obtained. Additionally, we prepared Ormosil-based sensors with linked indicator, which showed very high sensitivity toward oxygen. The effect of covalent coupling on sensor characteristics, stability and photophysical properties is studied. It is demonstrated that leaching and migration of the dye are eliminated in the new materials but excellent photophysical properties of the indicators are preserved.

Introduction

Optical oxygen sensors are among the most important sensor systems and are widely applied in industry and academia alike [1,2]. For example, oxygen sensors are widely used in biology [6,7], food packaging [1], as pressure sensitive paints [3-5] to mention only a few areas. Numerous optical oxygen sensors were reported during the last decades and the materials employed are now fairly established [8,9]. Most state-of-the art sensing materials rely on oxygen indicators physically entrapped in a polymer matrix or in an organically modified silica (Ormosil) [10]. For this reason they possess several major shortcomings, which originate from the fact that dye molecules can freely migrate within the matrix. First, aggregation of the dye can occur with time. Second, calibration plots can be affected due to redistribution of the dye between different environments within the matrix which are known to have different polarity and oxygen permeability [11]. Third, diffusion of the dye molecules into adjusting materials such as polymeric support is thermodynamically favored and readily occurs with time. Evidently, this affects storage stability of the sensor. Since migration processes are accelerated at higher temperatures even short treatment (e.g. during sterilization) may affect the calibration. Moreover, many oxygen sensing materials rely on the use of nano- and microparticles [12,13] dispersed in a polymer matrix [14] where the migration processes can be even more critical due to large contact surface. Finally, simple entrapment of the dye does not completely prevent it from leaching particularly if organic solvents, lipophilic samples or surfactants are present [15].

Obviously, the above-mentioned shortcomings can be overcome by covalently linking the indicator dye to the polymer. This approach is rather common for other indicators (e.g. pH indicators) but is not state-of-the art in case of oxygen sensors. Possible explanations are increasing production costs and significant synthetic effort required for both the preparation of functionalized indicators and the coupling step. A possible route toward such sensing materials could include click chemistry [16,17] which enables functionalization of polymers [18-20] and dyes alike [21] using very mild and straightforward reaction conditions. Within the arsenal of common click reactions [22] the nucleophilic substitution of the labile para-fluorine atoms of a 2,3,4,5,6-pentafluorophenyl groups seems to be a quite promising method [23,24]. Fortunately, highly photostable oxygen indicators such as Pt(II) and Pd(II) complexes of meso-tetra(pentafluorophenyl)porphyrin (TFPP) [25-28] and tetra(pentafluorophenyl)porphyrin lactones [29,30] possess such functionality. Additional positive features of these dyes are commercial availability, excellent photostability and good quantum yields. Although other porphyin based indicators with near infrared (NIR) emission were recently employed [31,32], PtTFPP and PdTFPP are still widely used. Also, nucleophilic substitution reactions have been widely used for derivatization/covalent coupling of TFPP and its complexes [33,34].

Virtually all oxygen sensors make use of polymers or Ormosils which act as a support or solvent for the dye and as a permeation-selective membrane. Important parameters for a good sensor matrix are adequate gas permeability, excellent chemical and photochemical stability, good processability of the sensing materials and commercial availability [9]. Within the set of different polymers, polystyrene (PS) is the most frequently used matrix. For Ormosil based sensors a variety of different materials with varying composition were reported [35-37].

In this contribution we will present a click chemistry based approach to covalently link the indicator to two different sensor matrixes and investigate the influence of the covalent attachment on sensor characteristics, photophysical properties and sensor stability.

Experimental

Reagents

Styrene (St) and tetraethoxysilane (TEOS) were purchased from Fluka (www.sigmaaldrich.com) and pentafluorostyrene (PFS), phenyltrimethoxysilane (PTMS) and 3-mercaptopropyltrimethoxysilane (MPS) from ABCR (www.abcr.de). For St and PFS the stabilizer was removed by filtration over Al2O3. The radical initiator azo-bis-(isobutyronitril) (AIBN) and 1,3-propanedithiol were obtained from ACROS (acros.com), triethylamine (NEt3) from Sigma–Aldrich, platinum(II)- and palladium(II)-meso-tetra(2,3,4,5,6-pentafluorophenyl)porphyrin (PtTFPP and PdTFPP, respectively) and platinum(II) octaethylporphyrin (PtOEP) from Frontier Scientific (www.frontiersci.com), polystyrene (MW 250,000) from Fisher Scientific (www.fishersci.com) and poly(ethylene glycol terephthalate) support (Mylar®) from Goodfellow (www.goodfellow.com). Anhydrous DMF was obtained from Sigma–Aldrich, all other solvents were purchased from Roth (www.carl-roth.de) and used without further purification. Synthetic air and nitrogen were of 99.999% purity and were purchased from Linde gas (www.linde-gas.at).

Characterization and measurement setup

19F NMR spectra were recorded on Varian Mercury 300 spectrometer (282.46 MHz) in deuterated chloroform and referenced against Me4Si for 1H and 13C. Weight and number average molecular weights (Mw and Mn), as well as the polydispersity index PDI = Mw/Mn, were determined by size exclusion chromatography (SEC) with the following set-up: Merck Hitachi L6000 pump, separation columns from Polymer Standards Service (8 mm × 300 mm, STV 5 μm grade size; 106, 104 and 103 Ǻ pore size), refractive index detector (model Optilab DSP Interferometric Refractometer) from Wyatt Technology. Polystyrene standards from Polymer Standard Service were used for calibration. All SEC runs were performed with THF as the eluent. Pt content in the sensor material was determined by ICP-OES after microwave assisted digestion. SEM images were recorded on a Zeiss Ultra 55 scanning electron microscope equipped with a field-emission gun (FEG). The particles were fixed on a double-stick carbon tape on a conventional specimen holder and sputter-coated by using a Pt/Pd (50:50) target.

Absorption spectra were measured at a Cary 50 UV–VIS spectrophotometer (www.lzs-concept.com). Emission spectra were acquired on a Hitachi F-7000 fluorescence spectrometer (www.inula.at) equipped with a red-sensitive photo-multiplier R 9876 from Hamamatsu (www.hamamatsu.com). The emission spectra were corrected for the sensitivity of the PMT which was calibrated using a halogen lamp. Relative luminescence quantum yields were determined using a solution of PtOEP in toluene as a standard [38] (quantum yield = 41.5%). The solutions of the dyes were thoroughly deoxygenated by bubbling argon through.

Luminescence phase shifts were measured with a two-phase lock-in-amplifier (SR830, Stanford Research Inc., www.thinksrs.com). Excitation was performed with the light of a 405 nm LED which was sinusoidally modulated at a frequency of 5 kHz for PtTFPP or 500 Hz for PdTFPP. A bifurcated fiber bundle was used to guide the excitation-light to the sensor film and to guide back the luminescence. A BG12 excitation glass filter and an OG630 emission filter (both from Schott, www.schott.com) were used. The luminescence was detected with a photo-multiplier tube (H5701-02, Hamamatsu, www.sales.hamamatsu.com). Temperature was controlled by a cryostat ThermoHaake DC50. Gas calibration mixtures were obtained using a gas mixing device (MKS, www.mksinst.com). For low oxygen concentrations measurements were performed using the same two-phase lock-in-amplifier and a home-made calibration chamber as described earlier [34].

Polymerization of PS-CO-PFS

Within this work two copolymers of St and PFS were polymerized in a solvent-free manner, one with a high PFS content (PS-PFS-1) and one with a low PFS content (PS-PFS-2). In case of PS-PFS-1 2 ml of St and 1 ml of PFS were used, while for PS-PFS-2 5.5 ml of St and 35 μl of PFS. Both mixtures were deoxygenated for 30 min, then 1 mol% of AIBN was added. The reactions were carried out at 85 °C in a N2 atmosphere for 20 h. In both cases the solid polymerization product was dissolved in CH2Cl2 and precipitated with MeOH, this procedure was repeated three times. The fluoride content in the polymers was determined with a fluoride-sensitive electrode after microwave assisted combustion [39].

Polymer-coupling

As demonstrated in Scheme 1, two routes toward sensor polymers were chosen. In the (route a) PS-PFS-1 and Pt- or PdTFPP were cross-linked using a less than equimolar amount of propanedithiol. In a typical procedure 105 mg of PS-co-PFS-1 were dissolved in 1 ml DMF, then 1–5 mg of PtTFPP or PdTFPP were added. Subsequently 14 μl of propanedithiol and 100 μl of TEA were added. The mixture was stirred at 75 °C for 10–20 min, until it turned into a “jelly”. To ensure complete reaction the “jelly” was left stirring at 75 °C for another 2 h. Finally the “jelly” was washed three times with EtOH and acetone to remove DMF, TEA, unreacted propanedithiol and indicator. After drying the polymer was crushed in a mortar yielding cross-linked sensor particles. In the (route b) PS-PFS-2 was used and propanedithiol was added in excess to prevent cross-linking. In a typical procedure 100 mg of PS-PFS-2 were dissolved in 2 ml of DMF and reacted with 30 μl of propanedithiol (60 equiv.) and 50 μl of TEA (35 equiv.) at 75 °C for 5–6 h. After precipitation and subsequent washing with MeOH the thiol-saturated polymer (TSP) was dried at 60 °C. Subsequently PtTFPP or PdTFPP were introduced: 10 mg of the dye were dissolved in 5 ml of DMF, 50 μl of TEA were added and the mixture was heated to 75 °C, then 100 mg of TSP, dissolved in 3 ml of DMF, were added dropwise. After 5 h the soluble sensor polymers were obtained by precipitation and washing with MeOH.

Scheme 1

Click-based synthesis of oxygen-sensitive materials by grafting indicators onto poly(styrene-co-pentaflurostyrene) or by introducing functionalized indicator to Ormosil-based materials. For the polymer based materials two routes can be distinguished: (route a) leads to cross-linked sensor polymers; (route b) yields soluble sensor polymers with covalently coupled indicator.

Ormosil-coupling

15 mg of PtTFPP or PdTFPP were dissolved in 100 μl of anhydrous DMF, 20 μl of TEA were added as well as 3.2 M equiv. of MPS. The reaction was carried out for 1 h in a closed vial at 75 °C. Reaction progress was monitored using TLC (silica gel): in contrast to the unreacted dye which is hydrophobic, the modified dye binds to the TLC plate. Since the functional dye is prone to hydrolysis it was used without further isolation. To the above mixture 2 ml of EtOH, 1 ml of TEOS, 1 ml of PTMS and 630 μl of water were added. After 15 min at 65 °C the mixture turned into a porous solid and was left at this temperature for another 2 h. Finally the mixture was dried at 200 °C in a drying chamber (3 h). After washing with acetone and CH2Cl2 the dry product was crushed in a mortar to yield Ormosil sensor microparticles (OSP).

Sensor preparation

The “cocktails” for coating the reference sensors were prepared by dissolving 1–5 mg PtTFPP or PdTFPP and 100 mg of polystyrene in 1000 mg of CHCl3. Sensor films containing cross-linked sensor particles were prepared as follows. 20 mg of the CLSPs or the OSP were dispersed in 50 mg of silicon E4 (www.wacker.com) and 150 mg of hexane. In case of the soluble sensor polymer 100 mg of the material were dissolved in 1000 mg of CHCl3. All cocktails were knife-coated on either a glass slide or a PET-foil to give, after solvent evaporation, phosphorescent sensor films.

Results and discussion

Choice of materials

PtTFPP and PdTFPP are doubtless among the most commonly used indicators in optical oxygen sensors. Although other indicators (e.g. NIR emitting benzoporphyrins [31,40,41]) were also found promising, the TFPP based indicators remain very attractive and are commonly used due to commercial availability, excellent photostability and red emission that can easily be visualized. Synthetic modification of such important indicators seems therefore worthwhile. Due to the presence of pentafluorophenyl groups such a modification is possible via a click chemistry pathway. The nucleophilic substitution of the labile para-fluorine atoms of a pentafluorophenyl group with a thiol is an essential step [22], as it yields groups which are suitable for covalent coupling. This approach was used to link the indicators to polymer matrixes as well as to an Ormosil.

Polymer synthesis and click modification

High molecular weight styrene–pentaflurostyrene-copolymers were synthesized via radical polymerization of the monomers similarly to the literature procedure [42]. This method was chosen as it yields polymers with high molecular weight which generally improves the mechanical stability of the sensor film. Other polymerization methods like nitroxide-mediated radical polymerization (NMP) can of course be used and may have better controllability [43], however high molecular weight polymers are not always easily accessible [44,45]. The properties of the obtained polymers are summarized in Table 1.

Table 1

Properties of poly(styrene-co-pentaflurostyrene).

Polymer	PFS (mol%)	Mw (g/mol)	Mw/Mn
PS-PFS-1	23.5	418,000	2.3
PS-PFS-2	0.5	494,000	2.5

In order to graft the indicator dye to the polymer via click chemistry, a dithiol was chosen for the nucleophilic substitution of the labile para-fluorine atoms of the pentafluorophenyl residues. The reaction progress for the soluble sensor polymer with coupled indicator was monitored via 19F NMR. At first, PS-PFS-2 was reacted with a large excess (60 times) of propanedithiol. Excess of the dithiol was used to minimize potential cross-linking and to ensure only one binding between the thiol residues and the polymer. After 6 h nearly every para-fluorine was substituted by a thiol group as seen in the 19F NMR (Fig. 1b). In the next step an indicator dye, either PtTFPP or PdTFPP, was grafted onto the polymer. Also in this step the indicator was used slightly in excess (about 2 equiv.) again to prevent cross-linking.

Fig. 1

19F NMR spectra of (a) PS-PFS-2, (b) PS-PFS-2 after saturating with propanedithiol, (c) the free indicator dye PtTFPP and (d) PtTFPP grafted onto PS-PFS-2; all measurements were performed in CDCl3.

In the final product six different peaks were observed in the 19F NMR and are clearly attributed. The para-flourine peak corresponding to the dye decreased by roughly 25% confirming that the indicator dye is coupled to the polymer, as intended, only once. The platinum content in the final sensing material was determined to be 7.6 mg/g. This indicates that about 82% of the previously introduced pentafluorophenyl groups in the polymer reacted with a PtTFPP via the dithiol linkage. The molecular weight increased to 644,000 g/mol.

In order to generate a cross-linked sensing material with covalently linked indicator PS-PFS-1 was reacted with 0.5 equiv. of propanedithiol in the presence of the indicator. Within 10–20 min a jelly like product was formed, indicating the spontaneous cross-linking of polymer chains. The product was washed with different organic solvents including CH2Cl2 and CHCl3 without being dissolved, indicating that the cross-linking was successful.

Photophysical properties

Covalent bonds between the dye and the matrix are expected to increase sensor stability without negatively affecting the sensor's characteristics especially brightness and sensitivity. As can be seen in Table 2, coupling of the indicator did not affect the position of the absorption and emission bands. Similarly, the phosphorescence quantum yields (Q.Y.) and decay times were only marginally affected.

Table 2

Comparison of the photophysical properties of covalently linked PtTFPP (SPP1) and the non-coupled indicator.

Code	Solvent	λmax (abs/nm)	λmax (em/nm)	Q.Y.	τ (μs)
PtTFPP	Toluene	396, 509, 541	652	0.19	33
PtTFPP coupled to PS-PFS-2	Toluene	397, 509, 542	653	0.16	27
PtTFPP	CHCl3	392, 508, 540	653	0.33	69
PtTFPP coupled to PS-PFS-2	CHCl3	392, 509, 541	654	0.29	58

Sensor characteristics

Cross-linked sensor polymers were crushed into microparticles and dispersed in silicone rubber. The soluble polymers with the coupled indicators were dissolved in CHCl3 to give homogeneous sensor films after solvent evaporation. Furthermore, the indicator was also simply dissolved (physically entrapped) in the polymer. PS based sensors were chosen as a reference.

Fig. 2 shows the calibration plots obtained for the new materials and Table 3 summarizes the respective sensor properties. Non linear Stern–Volmer calibration plots are quite common for optical oxygen sensors [11]. Such a calibration behavior can be adequately described by the two-site model (Eq. (1)) [10,11], where KSV1 and KSV2 are the two different Stern–Volmer constants, and F is the distribution coefficient. Although this model is not physically meaningful, for the decay time measurements fitting with high correlation coefficients (r2 > 0.999) is possible.Sensors S1 and S2 differ only by the amount of loaded indicator, their calibration curves are very similar, even though the decay time of S2 is significantly reduced. Usually a higher dye loading increases brightness, but often aggregation or insolubility are limiting factors. For S2 the decreased decay time (and decreased I0/I − 1) can indicate that PS and PFS tend to form block-like structures, if the PFS content is high due to the difference in polymerization rate of the two monomers. Within the blocks the indicator dyes are closer to each other leading to a decreased decay time. When PtTFPP is only dissolved in PS-PFS-1 (S3) the resulting sensor is more sensitive than when the indicator is coupled to the same polymer (S1). This might be due to a decrease in the free volume upon cross-linking which affects gas permeability.

Fig. 2

Stern–Volmer plots for PtTFPP-based (a) and PdTFPP-based (b) oxygen-sensing materials.

Table 3

Sensor characteristics of new sensor materials compared to the reference sensors.

Code	Matrix	Type	Dye (amount in wt%)	τ0 (μs)	τ0/τ − 1b	I0/I − 1b
S1a	PS-PFS-1	Coupled + cross-linked	PtTFPP (1)	59.3 ± 0.2	2.74	3.57
S2a	PS-PFS-1	Coupled + cross-linked	PtTFPP (5)	49.9 ± 0.1	2.66	3.20
S3	PS-PFS-1	Dissolved	PtTFPP (1)	61.5 ± 0.1	3.02	4.27
S4	PS-PFS-2	Coupled	PtTFPP (5)	56.4 ± 0.1	2.04	3.06
S5	PS-PFS-2	Coupled	PdTFPP (5)	951 ± 7	4.54c	6.74c
Ref 1	PS	Dissolved	PtTFPP (1)	58.8 ± 0.1	2.02	2.88
Ref 2	PS	Dissolved	PtTFPP (5)	49.8 ± 0.1	1.78	2.18
Ref 3	PS	Dissolved	PdTFPP (1)	964 ± 4	3.97c	5.97c

Cross-linked sensor particles of PtTFPP on PS-PFS-1 dispersed in silicone rubber.

At 100% synthetic air.

Calibration from 0 to 42 hPa (τ0/τ − 1 at 42 hPa).

S1 and S2 are insoluble in organic solvents, but the beads can easily be dispersed in common matrixes (such as hydrogels or silicones) also along with other components (e.g. temperature-sensitive beads, titanium dioxide particles or other additives). On the other hand, if the indicator is only coupled to the polymer without crosslinking the chains the resulting material is still soluble (S4 and S5) and can easily be applied on polymeric or inorganic substrates. Compared to PtTFPP the palladium analogue (PdTFPP) has a roughly 15 times longer decay time, which results in higher sensitivity of the sensors using this indicator.

All optical sensors including oxygen sensors show cross-sensitivity to temperature and this is of course valid for the materials presented here. The cross-sensitivity results from two components, namely the thermal quenching of the indicator (which lowers τ0 with increasing temperature) and the change in oxygen diffusion and solubility within the polymer (which normally increases the sensitivity). In this respect the investigated materials show behavior similar to the reference sensors and other reported oxygen sensors [31]. For example in case of S4 the decay time decreased by 0.3% per K and the sensitivity increased by about 0.6% per K.

Sensor stability

Immobilizing the indicator in the sensor matrix is expected to have positive effects on sensor stability. As known, dye migration and aggregation are factors that limit long term stability, and can be caused by high indicator concentrations which can alter the characteristics of the sensor [46]. Notably, dye molecules can also migrate into the support [47] especially if polymeric fibers or polymeric sensor supports are used. The migration of the dye into other polymers is accelerated by heat or by the presence of organic solvents that swell either the matrix or the polymeric support. Evidently, this effect is suppressed by covalently coupling the dye to the sensor matrix.

Poly(methyl methacrylate) (PMMA) plastic optical fibers are commonly used for optical signal transduction and represent a cheap alternative to glass fibers. PMMA fibers were coated with PtTFPP coupled to PS-PFS-2 as well as with PtTFPP in PS (both materials dissolved in chloroform). The coated fibers were dried at 60 °C for 6 h and then compared to a corresponding sensor spot (PtTFPP/PS on PET support). The results are presented in Fig. 3. As it can be observed, sensitivity decreases dramatically (by about 2-fold) if the indicator is not linked to the polymer. This indicates that the migration of the indicator into PMMA (which has lower gas permeability than PS) is very efficient. Similar effects were observed by Badocco et al. who characterized the sensors prepared by coating of PMMA fibers with a Pt(II) porphyrin in polysulfon [48]. In fact, the authors documented decrease in the sensitivity over time and observed very long dynamic response times for the sensors. It is evident (Fig. 3) that if the dye is coupled to the polymer, the negative effects are dramatically reduced. In fact, the sensitivity decreases only by about 4%. Thus, minimized risk of dye migration enables simple coating procedures for polymeric fibers and increases long-term stability of optical sensors.

Fig. 3

Stern–Volmer plots for the oxygen-sensing materials coated on a poly (ethylene terephthalate) support (red squares) and on a PMMA fiber (blue circles). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

A further stability problem of optical oxygen sensors arises from dye leaching. This issue can be solved by covalently coupling the dye to the matrix. For example, cross-linked sensor particles can be easily incorporated into a gas permeable membrane (e.g. silicone or hydrogel). Even when exposed to organic solvents, which are known to swell the particles, leaching of the dye was not observed. Not surprisingly, in case of the indicator coupled to the polymer (S4 and S5) the exposure to pure organic solvents such as chloroform leads to dissolving of the sensor, as the polymer itself is soluble. Exposure of the film to solvents that do not the dissolve but swell the polymer (e.g. acetone) is possible. This is of particular interest as sensor films can sometimes, for example during cleaning, be exposed to such solvents or mixtures of solvents with water. As shown in Fig. 4, dye leaching occurred rapidly when sensors with entrapped indicator are dipped in or are exposed to such solvents or solvent mixtures. When the dye is linked to the polymer it will not be released from the sensor layer even if stored in the solvent for several days.

Fig. 4

Photographic images of the sensor films after dipping into different organic solvents. If PtTFPP is coupled to PS-PFS-2 the sensor films remain colored after several days exposure (top). When the indicator is only physically entrapped in PS most organic solvents remove the indicator from the sensor which in a few seconds (bottom).

Short time exposure of sensor films to organic solvents like acetone can change the sensor signal and which is even more critical can lead to extraction of the dye into the environment. When exposing S4 to acetone or THF/EtOH mixtures τ0 does not change but the sensitivity of the sensor is marginally increased (by 3–4%). On the other hand, if a sensor with the physically entrapped indicator (Ref 1) is exposed to those solvents the sensor signal drops immediately and the indicator is fully extracted.

The previously discussed stability problems arise from the mobility of the indicator in the matrix itself or from the exposure of the sensor to certain solvents. An additional factor, which plays an important role in determining the stability, is photostability. In general, photostability is mostly determined by the indicator rather than by the sensor matrix. To check the effect of covalent linkage of the indicator to the matrix a reference sensor (Ref 1) and S4 were both continuously illuminated. After 5.5 h the intensity of S4 dropped by 3.5% while the reference sensor lost 1.4% of its initial intensity (Fig. 5). Despite this decrease the photostability remains very good. It should also be considered that normally only short light pulses (and not continuous irradiation) are needed for measuring. Additionally, in case of the decay time measurements the effect of photobleaching is negligible.

Fig. 5

Photostability of S4 (top) and Ref 1 (bottom). After covalent coupling of the indicator the photostability slightly decreases.

As was mentioned above, organically modified silicas (Ormosils) represent other highly important sensor matrixes. These materials can be synthesized from a variety of alkoxysilanes by condensation. Depending on the reaction conditions and the precursors used quite different sensor materials can be obtained including, for example, Ormosil films [37], soluble ormosils [10], xerogels [49] or aerogels [50]. Ormosil-based materials are rather versatile sensor matrixes, and as such they are an interesting addition to polymers.

Mercaptopropyltrimethoxysilane was used to modify the indicator to enable covalent coupling to the Ormosil. Following the click reaction scheme (Scheme 1) PtTFPP or PdTFPP were reacted in anhydrous DMF in the presence of TEA. The reaction process was monitored via TLC. Since the silane-modified dye is prone to hydrolysis, it is important to use the functionalized dye immediately.

For the proof of principle the functionalized dye was reacted with tetraethoxysilane (TEOS) and phenyltrimethoxysilane (PTMS). The resulting Ormosil was dried, crushed and afterwards exposed to organic solvents such as acetone or CH2Cl2. The Ormosil, either containing PtTFPP or PdTFPP, did not show any leaching of the dye. A similar Ormosil containing physically entrapped PtTFPP was synthesized for comparison. When this material was exposed to the same organic solvents instantaneous dye leaching was observed and the particles turned essentially colorless within a few seconds.

The Ormosil particles with the linked indicator were dispersed in a silicone matrix. Surprisingly, the prepared materials were found to be highly sensitive to oxygen which makes them promising for trace oxygen sensing (Fig. 6). Again the calibration plots are not linear, so the two-site-model was used to fit the calibration plots. The KSV values given in Fig. 6 represent the values obtained for the better quenchable fraction (KSV2 is more then 10 times smaller). The bimolecular quenching constant (k = KSV/τ0) was calculated to be 120 and 116 Pa−1 s−1 for PdTFPP and PtTFPP, respectively. The Stern–Volmer constants (KSV) are comparable to the values obtained for the recently published trace oxygen sensors based on the porphyrin dyes covalently attached to the surface of the silica particles (≈0.4 hPa−1 for PtTFPP and ≈6.8 hPa−1 for PdTFPP [34]). On the other hand, k of the published sensor is roughly only a half of the value reported here. The bimolecular quenching constants k are also significantly smaller for other sensitive ormosils (k = 30–40 Pa−1 s−1 using PtOEP [35]) or xerogels (k = 54 Pa−1 s−1 using [Ru(dpp)3]2+ [49]).

Fig. 6

Sensing properties and structure of the Ormosil-based material: (a) Stern–Volmer plots; (b) low magnification SEM picture of an Ormosil particle showing the macro porosity; (c) magnification of the square shown in (b) indicating the nano porosity of the material.

The high sensitivity toward oxygen showed by the material presented here can be correlated to the highly porous structure, which was verified via SEM measurements. The obtained Ormosil particles display both a macro porous structure (Fig. 6b) and an additionally nano porous substructure (Fig. 6c). Therefore, oxygen can easily diffuse toward the linked indicator dye.

The chosen example shows that Ormosils are highly interesting materials as the structure and composition can have a strong influence on the sensor performance. As mentioned above a variety of different Ormosil-based materials exist and are used, therefore the concept presented here may also be applied for those materials.

Conclusion

In this article we presented a convenient and quite versatile way to covalently immobilize two frequently used indicators (PtTFPP and PdTFPP) to polymer and Ormosil-based sensor materials. The presented way is based on a single step click reaction involving the nucleophilic substitution of the labile para-fluorine atom of the indicator with a thiol. The impact of covalent coupling on sensitivity, photophysical properties and sensor stability was studied. It was found that covalent coupling increases the sensor stability in terms of dye leaching and migration, while conserving the good photophysical properties. The concept was proved to be useful for two important sensor matrixes; but it can possibly be extended to other materials as well. Notably, the same strategy can be adapted for other luminescent dyes bearing pentafluorophenyl group, as for example NIR emitting phosphorescent porphyrin-lactones [29,30].