Anal. Chem. 2009, 81, 378–384
Determining Proton Diffusion in Polymer Films by Lifetimes of Luminescent Complexes Measured in the Frequency Domain Walter J. Bowyer,*,† Wenying Xu,‡ and J. N. Demas*,‡ Department of Chemistry, Hobart and William Smith Colleges, Geneva, New York 14456, and Department of Chemistry, University of Virginia, McCormick Road, Charlottesville, Virginia 22904-4319 Polymer-supported luminescent metal complexes represent an important class of oxygen, pH, and ion sensors. The diffusion properties of the analyte into the sensing film are important for rational sensor and support design and development. We describe a technique using lifetime measurements in the frequency domain for determining the diffusion coefficient of hydrochloric acid through various polymeric pH sensor films. Two types of polymers are doped with [Ru(4,7-diphenyl-1,10-phenanthroline)2(4,4′-dicarboxy-2,2′-bipyridine)]Cl2. We monitor the phase shift of luminescence (from which we calculate the apparent lifetime, τapp) versus time after applying a step increase in the aqueous HCl concentration at the surfaces of the film. We model the decrease in τapp as a function of time using the diffusion coefficient of HCl in the polymer as the only adjustable parameter. The model accurately predicts the lifetime versus time curves, and the resulting diffusion coefficients are highly dependent on the polymer. Relative to bulk water, diffusion of protons within very hydrophilic hydrated D4 polymer (a polyethylene oxide cross-linked siloxane ring polymer) films is hindered ∼4-fold, while within a more hydrophobic sol gel it is hindered by over 1 order of magnitude. The methodology is adaptable for measuring diffusion coefficients of a variety of analytes in different sensor films as long as the bound and unbound forms luminesce and the excited states have different lifetimes. Analytical chemists are devising an ever expanding variety of sensors by doping reporting species into a wide variety of polymer films.1,2 We have long been interested in sensors based on polymer-supported luminescent transition metal complexes. For example, the luminescence of many ruthenium complexes is strongly quenched by molecular oxygen, and the Stern-Volmer equation describes the relationship between the oxygen concentration and the luminescence intensity or lifetime of the complex. * To whom correspondence should be addressed. E-mail:
[email protected];
[email protected]. † Hobart and William Smith Colleges. ‡ University of Virginia. (1) Demas, J. N.; DeGraff, B. A. Coord. Chem. Rev. 2001, 211, 317–351. (2) Sanchez-Barragan, I.; Costa-Fernandez, J. M.; Sanz-Medel, A.; Valledor, M.; Campo, J. C. Trends Anal. Chem. 2006, 25 (10), 958–967.
378
Analytical Chemistry, Vol. 81, No. 1, January 1, 2009
This is the basis for very versatile oxygen sensors that have seen a wide range of applications.3-7 Quantification of pH is another important area. We have described the application of [Ru(4,7-diphenyl-1,10-phenanthroline)2(4,4′-dicarboxy-2,2′-bipyridine)]2+ ([Ru(Ph2phen)2(DCbpy)]2+), in pH sensors.8 Measuring either the intensity of luminescence or the lifetime of the excited species allowed rapid determination of the pH of a solution in which the polymer/complex sensor was immersed. Traditionally, polymers have been selected for sensors on empirical grounds; however, much of our recent effort has been to understand the fundamental interactions between the sensing molecule and the support in order to facilitate a more rational sensor design.3 For example, oxygen sensing is critically dependent on the polymer support and its interactions with the sensor molecule. In pH sensing, polymers can protect the sensor molecule from oxygen quenching. Multiple lifetimes are common in polymer-supported sensors due to heterogeneity. In our efforts to understand the diffusion of analytes into sensors, we have published two methods of determining oxygen diffusion in polymers. One was an intensity-based method using the luminescence intensity of ruthenium-complex-doped sensor films as oxygen diffused into it.9 After equilibrating the film in the absence of oxygen, we applied a step increase in the oxygen pressure at one surface of the film and measured the decrease in luminescence versus time as oxygen diffused into the film. The intensity versus time profiles were fit using a rigorous model with the oxygen diffusion coefficient, D, as the only adjustable parameter. Our model accounted for the time-dependent nonuniform oxygen concentrations in the film, nonlinear Stern-Volmer quenching of the luminescence and nonuniform excitation due to sample optical density of the film. This yielded accurate values of D for a variety of sensor films. (3) Demas, J. N.; DeGraff, B. A.; Coleman, P. Anal. Chem. 1999, 71, 793A– 800A. (4) Lu, X.; Winnik, M. A. Chem. Mater. 2001, 13 (10), 3449–3463. (5) DeGraff, B. A.; Demas, J. N. In Reviews in Fluorescence; Geddes, C., Lakowicz, J. R., Eds.; Springer Science: New York, 2005; Vol. 2, pp 125151. (6) Hartzell, K. A.; Danowski, K. L.; Pantano, P. Appl. Spectrosc. Rev. 2008, 43 (1), 1–21. (7) Takeuchi, Y.; Amao, Y. Springer Ser. Chem. Sens. Biosens. 2005, 3, 303– 322. (8) Clark, Y.; Xu, W.; Demas, J. N.; DeGraff, B. A. Anal. Chem. 2000, 72, 3468–3475. (9) Kneas, K. A.; Demas, J. N.; Nguyen, B.; Lockhart, A.; Xu, W.; DeGraff, B. A. Anal. Chem. 2002, 74, 1111–1118. 10.1021/ac8016554 CCC: $40.75 2009 American Chemical Society Published on Web 11/26/2008
Lifetime measurements offer several advantages over intensity measurements. For example, a lifetime method is not greatly affected by photobleaching, source fluctuations, or long-term drift. On the other hand, calculating the apparent lifetime, τapp, is not as simple as measuring the total intensity. As oxygen diffuses into the polymer, the lifetime of a luminescent complex at any point in the film depends on the oxygen concentration at that point. The measured intensity is the sum of the intensity contributions throughout the film, but apparent lifetimes must be derived from an additional measurement and calculation (e.g., phase shift, multiexponential fit, or rapid lifetime determination of a pulse excited decay curve) since one cannot simply integrate the apparent lifetimes in the same way as intensities. We developed and demonstrated the applicability of a frequency domain phase shift lifetime measurement to determine diffusion coefficients of oxygen in polymer films.10 We report here on a similar phase shift strategy for measuring the diffusion of an acid into polymer pH sensor films. In this case, the acid and base forms of the luminescent complex must have different lifetimes. We begin with a polymer equilibrated in a solution sufficiently basic to have virtually all of the sensor complex in the base form. We then apply a step pH change to make the solution highly acidic and monitor the apparent sensor lifetime, τapp, as acid diffuses into the film and protonates the complex. We provide a detailed analysis of the data fitting and discuss the effect of the polymer on diffusion. As expected, different polymers allowed for significantly different rates of diffusion of acid. EXPERIMENTAL SECTION Polymers. The sol-gel was made by the reaction of Ntriethoxysilylpropyl-o-poly(ethylene oxide) urethane (N-urethane Mw ) 441, from United Chemicals Technologies, Bristol, PA) and n-octyltriethoxysilane (Otesilane, Mw )276.5 from Acros) in the presence of HCl and ethanol. A typical preparation was as follows: 1 g of Otesilane and 0.75 mL of HCl (0.1M) were added to 2 g of the N-urethane plus 1.2 mL of ethanol. This mixture was stirred for 1 h and cast on a Teflon surface to yield a film with the desired thickness. Films on the Teflon were kept at room temperature overnight and at 70 °C for ∼4 h to obtain a clear, colorless sol-gel. The synthesis of the D4 polymer has been described earlier.11,12 The D4 network polymer consists of discrete very hydrophobic siloxane pockets and urea cross-linkages using a 2000 MW diaminopoly(ethylene glycol). It is a clear, colorless elastomer. Sensor Preparation. Circular polymer films of ∼5 cm in diameter and ranging from 0.5 to 1 mm thick were cast. Film thickness was determined for the water-swollen films by sandwiching the film between two premeasured microscope slides. The film was snapped into a plastic holder, and the holder was then placed into a polycarbonate tank containing 90 mL of aqueous 10 mM NaCl to maintain a relatively constant ionic strength, exposing both sides of the film to the solution. Absorption Measurements. Sample optical densities (OD) of the doped films were determined with a Hewlett-Packard 8452A (10) Bowyer, W. J.; Xu, W.; Demas, J. N. Anal. Chem. 2004, 76, 4374–4378. (11) Xu, W.; McDonough, R. C., III.; Langsdorf, B.; Demas, J. N.; DeGraff, B. A. Anal. Chem. 1994, 66, 4133–41. (12) Price, J. M.; Xu, W.; Demas, J. N.; DeGraff, B. A. Anal. Chem. 1998, 70, 265–270.
diode array spectrometer. Baseline-corrected ODs at the 450-nm excitation wavelength were recorded in triplicate and averaged. Lifetime Measurements. Luminescence lifetimes were measured with a phase fluorometer supplied by Airak Inc.10 Excitation was with a blue (∼450 nm) LED modulated at 80 kHz. The emission was detected with a silicon photodiode, and the signal was amplified with a current-to-voltage converter. Complementary excitation-emission filters were used for selecting wavelengths. Phase shifts between the excitation and emission were determined with a Stanford Research Systems SR844 dual-channel lock-in amplifier, and data were acquired with a program locally written in Laboratory View (National Instruments, Austin, TX). The LED and detector were integrated into a single head. The probe of the fluorometer was positioned against the outside of the sample tank at the same level as the film, while the film was ∼1 cm from the inside wall of the tank. Because the working distance to the sample was relatively long, the signal-tonoise ratio was not as good as in our earlier work10 but proved sufficient to determine D with good precision. One datum point was acquired about every 2 s. Phase shifts were corrected for the instrument phase shift by measuring a reference solution of rhodamine B where its
