Luminescence temperature sensing using poly(vinyl alcohol)-encapsulated Ru(bpy)₃²⁺ films

Abstract

The ruthenium(II) diimine complexes, such as ruthenium(II) tris(bipyridyl), Ru(bpy)₃²⁺, possess highly luminescent excited states that are not only readily quenched by oxygen but also by an increase in temperature. The former effect can be rendered insignificant by encapsulating the complex in an oxygen impermeable polymer, although encapsulation often leads also to a loss of temperature sensitivity. The luminescence properties of Ru(bpy)₃²⁺ encapsulated in PVA were studied as a function of oxygen concentration and temperature and found to be independent of the former, but still very sensitive towards the latter. The results were fitted to an established Arrhenius-type equation, based on thermal quenching of the emitting state by a slightly higher (ΔE = 3100 cm⁻¹) ³d-d state that deactivates very rapidly (10⁻¹³ s) via a non-radiative process.

---

Introduction

Temperature sensing based on luminescence measurements, lifetime or intensity, represents a well established area of fiber optic sensing.¹ This approach to remote temperature sensing is sometimes preferred to the more common methods, based mainly on electronic, usually resistance, measurements. For example, luminescence-based temperature measurements are more appropriate for use in explosive environments and in remote, often marine, environments that require multi-analyte analysis, of which temperature is usually a key component.^1,2

There are many different materials that can be used for luminescence-based temperature measurements, including: rare earth phosphors, ruby crystals, chromium-doped YAG phosphors and fluorescent organic dyes, such as coumarin and perylene.^1–5 In many cases these materials cannot be readily manufactured as thin films nor miniaturized for use with fibre optics.

A necessary requirement of all these systems is that the luminescent species must be oxygen-insensitive, since any quenching by oxygen would also be expected to be effected by temperature. Indeed, this feature is a major problem in the large area of research that focuses on pressure-sensitive paints (PSPs).^6–8 The requirement of temperature-sensitivity, coupled with oxygen-insensitivity has, understandably, usually ruled out the use of the many ruthenium(II) diimine complexes, such as ruthenium tris(bipyridyl), Ru(bpy)₃²⁺, which have featured so strongly in the development of many types of optical sensor, especially oxygen sensors.^9–11 These highly coloured and luminescent metal complexes are very photostable and readily quenched by oxygen. However, they also have lifetimes that are temperature-sensitive, due to the presence of a high-energy, ³d-d state, which readily decays via a non-radiative route and can be thermally populated at ambient temperatures by the ³MLCT (i.e., metal-to-ligand charge transfer) type emitting state.^11–15

A schematic diagram of the various electronic states and their intramolecular deactivation pathways for a typical Ru(II) diimine complex,¹³ such as Ru(bpy)₃²⁺, are illustrated in Fig. 1. Based on this scheme, in the absence of oxygen, it has been shown^12,13 that the observed variation in luminescence intensity and lifetime can be described by the following Arrhenius-type equation:

τ_obs = [k₀ + k₁ exp(−ΔE/RT)]⁻¹ (1)

where τ_obs is the observed radiative lifetime, k₀ (= k_r + k_nr, see Fig. 1) is the temperature-independent rate constant for the excited state decay of the ³MCLT states and is the sum of the radiative and non-radiative rate constants, k₁ is the decay rate constant of the non-radiative ³d-d state and ΔE is the energy difference between the ³d-d and ³MCLT states.

Fig. 1 Electronic energy level diagram for a ruthenium(II) diimine complex, such as Ru(bpy)₃²⁺. The temperature independent rate constant, k₀, referenced in eqn. (1), is the sum of the rate constants for the non-radiative, k_nr and radiative, k_r, decays.

Given the above features, it is not surprising to note that when Ru(II) diimine complexes are used in optical oxygen sensing,¹⁶ as in PSP work,^6–8 it is essential that either the temperature is fixed or, at least known, so that some compensation, usually by dual lumophore referencing, can be made for the inherent temperature sensitivity of their luminescence lifetime/intensity.^17–21

It has been shown that Ru(II) diimine complexes can be used as temperature, and not oxygen, sensors by incorporating them into beads of a polymer, poly(acrylonitrile) (PAN), which has a very low oxygen permeability (P_M(O₂) = 0.000 15 cm³ cm cm⁻² s⁻¹ Pa⁻¹).^22,23 These beads can then be readily incorporated into a luminescence-based oxygen sensor film and so provide information regarding the ambient temperature²⁴ or act as a reference material via dual-lumophore referencing.²² However, such beads are not trivial to make, since the water soluble Ru(bpy)₃²⁺ complex needs first to be rendered soluble in a solvent able to dissolve PAN, such as DMF, and this usually requires ion-pairing with a lipophilic anion, such as phosphorus hexafluoride.²⁴ As an alternative, this paper describes a simple method of making a generic type of luminescence-based temperature sensor, in which a ruthenium(II) diimine complex, Ru(bpy)₃²⁺, is encapsulated in an oxygen-impermeable polymer film, poly(vinyl alcohol), PVA, both of which are water soluble and readily available commercially.

Experimental

Unless stated otherwise all chemicals, including ruthenium(II) tris(bipyridyl) dichloride hexahydrate, Ru(bpy)₃Cl₂·6H₂O, were purchased from Aldrich Chemicals and used as received. All gases were purchased from BOC. All luminescence spectra were recorded using a PerkinElmer LS50B fluorimeter (λ_(excitation) = 455 nm) and lifetime measurements were made with an IBH Fluorocube time correlated single-photon counting system, using a NanoLED-01 source, which has its excitation peak wavelength at 490 nm. All lifetime and intensity measurements reported here exhibited a similar, low, error in repeatability, i.e., < ± 5%. A typical Ru(bpy)₃²⁺/PVA luminescence-based temperature-sensitive film was prepared as follows: 1.5 mg of [Ru(bpy)₃Cl₂·6H₂O] were dissolved in 1 ml of water and added to 10 g of a 10% (w/v) aqueous solution of PVA (MW: 124 000–186 000) and left stirring overnight to ensure complete dissolution and mixing. To create a typical Ru(bpy)₃²⁺/PVA film, this solution was cast through a 100 µm brass metal template, with a 0.8 × 1.5 cm rectangular hole cut out of its centre, onto a glass microscope slide. The resulting film was allowed to dry in the dark in a desiccator for approximately 4 hours and used without any further treatment; its final thickness was typically 35 µm as measured using an electronic micrometer.

Results and discussion

The key features of the Ru(bpy)₃²⁺/PVA films, namely, oxygen insensitivity and temperature sensitivity, were demonstrated by recording the luminescence spectra of a typical film (λ_excitation = 455 nm) in the absence and presence of oxygen at 21 and 45 °C. The results of this work are illustrated in Fig. 2 and show that the luminescence intensity of the film is significantly reduced at the higher temperature, thereby illustrating the temperature sensitivity of the films. In addition, the results show that when at these two different temperatures the measured intensities are largely independent of the ambient level of oxygen (1 atm or 0 atm), thus highlighting the oxygen insensitivity of the luminescent film at 21 and 45 °C. Similar results were observed at relative humidity levels of 0% and 100%, indicating that the film's response characteristics are largely independent of relative humidity.

Fig. 2 Luminescence spectra of a Ru(bpy)₃²⁺/PVA film (λ_(excitation) = 455 nm) recorded in the absence (thick continuous line) and presence (thin continuous line) of oxygen at 21 °C (highest intensity peak pair) and 45 °C.

The oxygen insensitivity of the Ru(bpy)₃²⁺/PVA film is not too surprising given that PVA is a recognized highly oxygen-impermeable polymer. Indeed, its oxygen permeability (P_M(O₂) = 0.006 65 cm³ cm cm⁻² s⁻¹ Pa⁻¹)²⁵ is 55 000 times smaller than that of silicone rubber (P_M(O₂) = 367 cm³ cm cm⁻² s⁻¹ Pa⁻¹), a material that is often to be found as the encapsulation material for Ru(bpy)₃²⁺-based oxygen sensors.^9,11,16

As a result of the oxygen-insensitivity of the Ru(bpy)₃²⁺/PVA films it was possible to assess in isolation and in more detail its temperature sensitivity, by recording the luminescence spectra of the film in air over a range of different temperatures, the results of which are illustrated in Fig. 3(a). In a parallel study the luminescence decay profiles of the film were also determined as a function of temperature and the results of this work are illustrated in Fig. 3(b). From these results it is clear that both the intensity and lifetime of luminescence of the Ru(bpy)₃²⁺ complex in the Ru(bpy)₃²⁺/PVA films decreases with increasing temperature. Other work shows that there is no hysteresis associated with this temperature sensitivity, i.e., both the luminescence intensities and lifetimes recorded at different temperatures as the temperature of the sensor was raised from 20 °C to 50 °C were also observed at the same temperatures as it was lowered. The response time of a Ru(bpy)₃²⁺/PVA film to either an increase or decrease in temperature appears almost instantaneous and well within the response time of the instrumentation or techniques involved in making the measurements.

Fig. 3 (a) Luminescence spectra of a Ru(bpy)₃²⁺/PVA film, λ_(excitation) = 455 nm, recorded in air at the following temperatures (from top to bottom): 25, 30, 40, 45, 50, 55 and 61 °C. (b) Luminescence decay profiles for the same film recorded at the following temperatures (from top to bottom): 28, 40, 50, 60 and 66 °C.

In contrast to work carried out on the temperature sensitivity of Ru(bpy)₃²⁺, the luminescence decay profiles recorded for the Ru(bpy)₃²⁺/PVA films at different temperatures, see fig. 3(b), were not simple first order, due to the heterogeneous nature of the polymer environment.¹⁶ As a consequence, each decay illustrated in fig. 3(b) was fitted to a double exponential and the pre-exponential weighted mean lifetime, τ_M, calculated (τ_M = α₁τ₁ + α₂τ₂)¹⁶ and reported.

The observed variation in luminescence intensity at λ_max(emission), i.e.I, and average lifetime (τ_M) as a function of temperature, obtained from the data illustrated in Fig. 3, can be readily fitted to eqn. (1), assuming τ_M ∝ I (as has been shown by Carraway et al¹⁶ to be the case) and τ_M = τ_obs in eqn, (1). From the optimized fit parameters, details of which are given in Table 1, plots of the predicted variations in the first order rate constant for luminescence decay, k_M (= 1/τ_M) and the reciprocal of the observed luminescence intensity, I⁻¹, (∝ 1/τ_M) versus temperature were generated for the Ru(bpy)₃²⁺/PVA films and are illustrated in figure 4(a) and (b) respectively, (solid lines), along with the actual measured values for the Ru(bpy)₃²⁺/PVA films (data points). From the plots in Fig. 4 it is clear that the kinetic model fits are in good agreement with the observed trends.

Fig. 4 Plot of the reciprocals of (a) the observed luminescence intensity at λ_max(emission) = 581 nm, λ(excitation) = 455 nm, and (b) the calculated weighted mean lifetime, i.e., k_M since k_M = 1/τ_M, as a function of temperature for a typical Ru(bpy)₃²⁺/PVA film. In both cases the solid lines have been generated using eqn. (1) and the optimized fit data in Table 1.

Table 1 Optimised fit parameters, based on eqn. (1), for the observed temperature-dependent luminescence lifetime and intensity of Ru(bpy)₃²⁺ in aqueous solution and in PVA and CA films

System		Lifetime measurements			Intensity measurements
	ΔE/cm^−1	k 0/10^6 s^−1	k 1/10^11 s^−1	r ^2	I 0 ^−1 × 10^−3	I 1 ^−1 × 10^4	r ^2
Ru(bpy)3^2+ in aqueous solution	3527 ± 5	1.588 ± 0.03	100 ± 1.000	0.999	8.94 ± 0.18	6.26 ± 0.16	0.997
Ru(bpy)3^2+ in aqueous solution^26	3560	1.27	100	—
Ru(bpy)3^2+ in aqueous solution^12	3559	1.29	100	—
Ru(bpy)3^2+ in PVA	3100 ± 10	0.53 ± 0.01	2.48 ± 0.07	0.999	1.60 ± 0.01	0.111 ± 0.001	0.999
Ru(bpy)3^2+ in CA^14	390	0.22	3 × 10^−5	—

---

The variations in I and k_M were also measured as a function of temperature for Ru(bpy)₃²⁺ in nitrogen-purged (i.e., oxygen free) aqueous solution and the results fitted to eqn. (1). The optimized fit parameters for which are also reported in Table 1, and are similar to those reported by Dressick et al,²⁶ and Van Houten and Watts,¹² which are also reported in Table 1. In all three cases involving the Ru(bpy)₃²⁺/H₂O system, the pre-exponential factor is large (ca. 10¹³ s⁻¹) and typical for a rapidly deactivating ³d-d state. In addition, the value for ΔE is significant, typically ca. 3550 cm⁻¹, and as a result of this, and a reasonably large value for k₀, the marked deactivation of the emitting ³MLCT state via thermal population of the ³d-d state occurs fortuitously at ambient temperatures.²⁷ This feature makes Ru(bpy)₃²⁺, in water at least, appear a promising material for use in temperature sensors for work in the range 0–100 °C.

In contrast to Ru(bpy)₃²⁺ in water, encapsulation of the dye in a rigid material, such as cellulose acetate (CA)¹⁴ or zeolite,¹³ usually renders the dye much less temperature-sensitive, as indicated by the values of k₀, k₁ and ΔE reported by Allsopp et al.¹⁴ for a Ru(bpy)₃²⁺/CA film, which are given in Table 1. Plots of model-predicted lifetime versus temperature profiles, based on eqn. (1) and the data in Table 1, for Ru(bpy)₃²⁺ in CA, water and PVA systems are illustrated in Fig. 5. The striking loss in temperature sensitivity exhibited by Ru(bpy)₃²⁺ in cellulose acetate, and other rigid media, is ascribed to the inhibition of the ³d-d decay pathway and the consequent domination of the thermal activation decay route by a pathway involving an additional, low-lying (ΔE′ ≅ 800–900 cm⁻¹) MLCT state.¹³ Inhibition of the ³d-d decay pathway is attributed to the rigidity of the encapsulating medium, which causes an increase in ΔE of sufficient magnitude to eliminate its contribution to the overall decay process. Fortunately, as is indicated by the results in Fig. 4 and the model predicted curves in Fig. 5, this loss of temperature sensitivity as a result of encapsulation does not appear to occur when Ru(bpy)₃²⁺ is encapsulated in PVA, presumably due to the inherent lower rigidity of PVA. Some indication of this feature is the glass transition temperature, which is 358 K for PVA but can be as high as 751 K for CA. As prepared, the PVA film will contain some tightly bound water and it is likely that this and PVA contribute to the deactivation of the ³MLCT excited state by nearly resonant activation of OH vibrational stretching modes (~3500 cm⁻¹ and usually very broad in energy), which are rapidly deactivated by non-radiative V–V and V–T energy transfer.

Fig. 5 Plot of variation in luminescent lifetime of Ru(bpy)₃²⁺ in water (dark continuous line), PVA (light continuous line) and cellulose acetate (broken line) as a function of temperature. Plots generated using eqn. (1) and the optimized fit data in Table 1.

From the results in Table 1 and plots in Fig. 5, it is clear that the lifetime of Ru(bpy)₃²⁺ is significantly higher in a polymeric encapsulating medium, such as PVA or CA, compared with that in water and this is due to the loss of radiationless decay pathways associated with structural changes in the solvation sphere. As a consequence, not only do Ru(bpy)₃²⁺ in PVA films exhibit a similar high temperature sensitivity as the dye in water, their excited state luminescence lifetimes are much longer, and as a result easier to measure.

Conclusion

Ruthenium(II) diimine complexes, such as ruthenium(II) tris(bipyridyl), Ru(bpy)₃²⁺, possess highly luminescent excited states that are not only readily quenched by oxygen but also by an increase in temperature. The former effect is rendered insignificant by encapsulating the complex in an oxygen impermeable polymer, such as PAN or PVA. In order to dissolve these complexes in the former polymer, they need to be rendered solvent, rather than water, soluble; usually by ion-pairing. In contrast, both PVA and most commercially available forms of ruthenium(II) diimine complexes, such as Ru(bpy)₃²⁺, are water soluble and so their production as luminescence-based temperature sensitive films is simple. A study of the luminescence properties of Ru(bpy)₃²⁺ encapsulated in PVA revealed that they were insensitive towards ambient oxygen levels but very temperature sensitive. The results of this work fitted an established Arrhenius type equation, based on thermal quenching of the emitting state by a slightly higher (ΔE = 3100 cm⁻¹) ³d-d state which deactivates very rapidly (10⁻¹¹–10⁻¹³ s) via a non-radiative process. From the results of this work it appears that ruthenium(II) diimine complexes, such as Ru(bpy)₃²⁺, can be readily incorporated into PVA to create a luminescence-based temperature sensitive indicator.

References

1. K. T. V. Grattan and Z. Y. Zhang, in Topics in Fluorescent Spectroscopy, ed. J. R. Lakowicz, Plenum Press, New York, vol. 4, 1994, pp. 355–376 and references therein.
2. K. A. Wickershein and M. H. Sun, J. Microwave Power, 1987, 85–94.
3. Y. L. Hu, Z. Y. Zhang, K. T. V. Grattan, A. W. Palmer and B. T. Meggit, Sens. Actuators, A, 1997, 63, 85–90.
4. T. Liu, B. T. Campbell, S. P. Burns and J. P. Sullivan, Appl. Mech. Rev., 1997, 50, 227–246.
5. S. Bertrand, A. Jalocha, G. Tribillon, M. Bouazaoui and J. Roahet, Opt. Laser Technol., 1996, 28, 363–366.
6. M. E. Cose, J. M. Bedlek-Anslow, J. P. Hubner, B. F. Carroll and K. S. Schanze, Topics in Fluorescent Spectroscopy, eds. C. D. Geddes and J. R. Lakowicz, Springer Press, New York, vol. 9, 2005, pp. 285–305 and references therein.
7. J. H. Bell, E. T. Schairer, L. A. Hand and R. D. Mehta, Annu. Rev. Fluid Mech., 2001, 33, 155–206.
8. J. Kavandi, J. Callis, M. Gouterman, G. Khalil, D. Wright, E. Green, D. Burns and B. McLachlan, Rev. Sci. Instrum., 1990, 61, 3340–3347.
9. A. Mills, Platinum Met. Rev., 1997, 41, 115–127.
10. C. McDonagh, B. C. McCraith and A. K. McEvoy, Anal. Chem., 1998, 70, 45–50.
11. J. N Demas and B. A. DeGraff, Topics in Fluorescent Spectroscopy, ed. J. R. Lakowicz, Plenum Press, 1994, vol. 4, New York, 71–108.
12. J. Van Houten and R. J. Watts, J. Am. Chem. Soc., 1976, 98, 4853–4858.
13. K. Maruszewski, D. P. Strommen and J. R. Kincaid, J. Am. Chem. Soc., 1993, 115, 8345–8350.
14. S. R. Allsopp, A. Cox, T. J. Kemp and W. J. Reid, J. Chem. Soc., Faraday Trans. 1, 1978, 74, 1275–1289.
15. R. S. Lumpkin, E. M. Kober, L. A. Worl, Z. Murtaza and T. J. Meyer, J. Phys. Chem., 1992, 94, 239–243.
16. E. R. Carraway, J. N. Demas and B. A. DeGraff, Anal. Chem., 1991, 63, 332–336.
17. K. S. Schanze, B. F. Carroll, S. Korotkevitch and M. J. Morris, AIAA J., 1997, 35, 306–310.
18. H. F. Ji, Y. Shen, J. P. Hubner, B. F. Carroll, R. H. Schmehl, J. A. Simon and K. S. Schanze, Appl. Spectrosc., 2000, 54, 856–863.
19. J. M. Bedlek-Anslow, J. P. Hubner, B. F. Carroll and K. S. Schanze, Langmuir, 2000, 16, 9137–9141.
20. G. E. Khalil, C. Costin, J. Crafton, G. Jones, S. Grenoble, M. Gouterman, J. B. Callis and L. R. Dalton, Sens. Actuators, B, 2004, 97, 13–21.
21. L. M. Coyle and M. Gouterman, Sens. Actuators, B, 1999, 61, 92–99.
22. C. von Bültzingslöwen, A. K. McEvoy, C. McDonagh, B. D. McCraith, I. Klimant, C. Krause and O. S. Wolfbeis, Analyst, 2002, 127, 1478–1483.
23. http://www.optosense.de (accessed November, 2005).
24. G. Liebsch, I. Klimant and O. S. Wolfbeis, Adv. Mater., 1999, 11, 1296–1299.
25. S. Pauly, Polymer Handbook, eds. J. Brandrup and E. H. Immergut, Wiley-Interscience, New York, 3rd edn., 1989, VI-435–VI-449.
26. W. J. Dressick, J. Kline, J. N. Demas and B. A. DeGraf, J. Am. Chem. Soc., 1986, 108, 7567–7574.
27. J. N. Demas and B. A. DeGraff, Sens. Actuators, B, 1993, 11, 35–41.

---