Can Luminescent Ru(II) Polypyridyl Dyes Measure pH Directly?

Anal. Chem. 2010, 82, 5195–5204

Can Luminescent Ru(II) Polypyridyl Dyes Measure pH Directly? Laura Tormo, Nelia Bustamante, Gonzalo Colmenarejo,*,† and Guillermo Orellana* Optical Chemosensors & Applied Photochemistry Group, Department of Organic Chemistry, Faculty of Chemistry, Complutense University of Madrid, E-28040 Madrid, Spain Two molecularly engineered Ru(II) complexes for direct pH optosensing in environmental or physiological media based on luminescence lifetime measurementssnamely, Na2[Ru(bpds)2(F15ap)]andNa2[Ru(pbbs)2(pyim)](where bpds ) 2,2′-bipyridine-4,4′-disulfonate, F15ap ) 5-perfluorooctanamide-1,10-phenanthroline, pbbs ) 1,10phenanthroline-4,7-(diyl)bis(benzenesulfonate), and pyim ) 2-(2′-pyridyl)imidazole)shave been prepared. The suitability of these two luminophores as generalpurpose pH indicators has been assessed to determine the general features of Ru(II) dyes required for such application. Their photochemical properties were investigated at different pH values in various buffer solutions using absorption spectroscopy, as well as steady-state and time-resolved luminescence. Both dyes display a parallel absorption and emission behavior as a function of pH (2-10), namely, higher luminescence in acidic solutions together with a 8-10 nm bathochromic shift in their (blue) absorption and 6-39 nm bathochromic shift in their (red) luminescence maxima in basic media, respectively. Similar ground-state acidity values (pKa) of 6.5 ( 0.2 for the amide group of the F15ap complex and 6.9 ( 0.2 for the imidazole NH moiety of the pyim complex have been measured. However, dramatic differences in their luminescence lifetimes as a function of pH were found. The HA and A- forms of *[Ru(bpds)2(F15ap)]2- conveniently display lifetimes of 372 and 263 ns, respectively, regardless of the solution acidity and buffer nature. Their relative contributions to the overall decay (0%-100%) are dependent on the solution pH indicating excited-state proton exchange rates well below the decay rates of the acidic and basic forms. However, *[Ru(pbbs)2(pyim)]2- deactivation kinetics show a pHindependent component of 80 ns at high pH and an acidity-sensitive one that varies from 610 ns (at pH 2) to 170 ns (at pH 10). Both components are also dependent on the buffer nature and concentration, also indicating the lack of an acid-base equilibrium in its excited-state but an irreversible proton transfer by the buffer species. Density functional theory calculations * Authors to whom correspondence should be addressed: Tel.: +34-913 944 220. Fax: +34-913 944 103. E-mails: [email protected], [email protected]. † Permanent address: Department of Biochemistry and Molecular Biology I, Faculty of Chemistry, Complutense University of Madrid, E-28040 Madrid, Spain. 10.1021/ac1005266  2010 American Chemical Society Published on Web 05/14/2010

have demonstrated the difficult accessibility of the base to the acidic perfluoroamide proton of the F15ap complex, severely slowing the excited-state proton transfer kinetics of the luminescent dye. Therefore, we conclude that the design of Ru(II) polypyridyl lifetimebased pH indicators not affected by the buffer nature and concentration requires the absence of proton exchange during the radiative deactivation of both the acidic and basic species, which then would remain in their ground-state relative ratio. This feature may be achieved by (a) mild excited-state acidities and (b) structural features that shield the exchangeable proton from the buffer access. Continuous monitoring of pH is required in practically all types of chemical, biomedical, and environmental analysis in aqueous media.1 IUPAC-recommended definitions, procedures, and terminology relative to pH measurements have been issued.2 Although the pH electrode is probably still irreplaceable in most situations, optical sensors (or optodes) for in situ pH monitoring based on colorimetric or fluorometric indicator dyes are an excellent option, because of their simplicity, minute size, and robustness for a wide variety of (bio)applications where those features are a must.3 In particular, optodes based on luminescence lifetime measurements display some decisive advantages over emission intensity-based devices since, in addition to the high sensitivity of the luminescence technique, the effect of lamp and detector fluctuation or drift, and the indicator leaching/bleaching are avoided.4 However, instrumentation for nanosecond emission lifetime determinations is expensive and very often fluorescent (1) McMillan, G. K.; Cameron, R. A. Advanced pH Measurement and Control, 3rd Ed.; ISA: Research Triangle Park, NC, 2004. (2) Buck, R. P.; Rondini, S.; Covington, A. K.; Baucke, F. G. K.; Brett, C. M. A.; Camo ës, M. F.; Milton, M. J. T.; Mussini, T.; Naumann, R.; Pratt, K. W. P.; Spitzer, P.; Wilson, G. S. Pure Appl. Chem. 2002, 74, 2169–2200. (3) (a) Cai, W.-J.; Reimers, C. E. In In Situ Monitoring of Aquatic Systems: Chemical Analysis and Speciation; Buffle, J., Horvai, G., Eds.; Wiley: Chichester, U.K., 2000. (b) Narayanaswamy, R., Wolfbeis, O. S., Eds.; Optical Sensors: Industrial, Environmental and Diagnostic Applications; Springer Series on Chemical Sensors and Biosensors, Vol. 1; Springer: Berlin, Heidelberg, 2004. (c) Wolfbeis, O. S. Anal. Chem. 2006, 78, 3859– 3874 (and references therein). (d) McMahon, G. Analytical Instrumentation: A Guide to Laboratory, Portable and Miniaturized Instruments; Wiley: Chichester, U.K., 2007. (4) (a) Orellana, G. Anal. Bioanal. Chem. 2004, 379, 344–346. (b) Orellana, G. In Optical Chemical Sensors; Baldini, F., Chester, A. N., Homola, J., Martellucci, S., Eds.; NATO Science Series II: Mathematics, Physics and Chemistry, Vol. 224; Springer: Dordrecht, The Netherlands, 2006; (c) Demchenko, A. P. Introduction to Fluorescence Sensing: Springer: Berlin, 2009.

Analytical Chemistry, Vol. 82, No. 12, June 15, 2010

5195

indicator dyes do not display a change in their luminescence decay profile with pH. This fact is usually a consequence of emission arising exclusively from either the basic or the acidic form of their excited state. For instance, the widely used pH-sensitive fluorophore 8-hydroxypyrene-1,3,6-trisulfonate (HPTS, pyranine) is excited at either 405 nm (HPTS) or 470 nm (PTS-) to collect its emission at 520 nm exclusively from the phenoxide form and perform ratiometric intensity measurements; therefore, its fluorescence decay profile remains essentially unchanged with the solution pH.5 The analytical applications of luminescent Ru(II) polypyridyl complexes have attracted considerable attention in the last two decades. These coordination compounds have carried fiber-optic oxygen sensing to commercial applications, phasing out the Clark electrode (because of their intense absorption in the blue region), >150-nm Stokes shift, significant emission quantum yields (up to 0.4), excellent photochemical and thermal stability, and near diffusion-controlled oxygen quenching, together with microsecondscale excited-state lifetimes that allow inexpensive emission lifetime-based instrumentation.6 Similarly, it would be desirable to design pH indicator dyes belonging to the family of Ru(II) polyazaheterocyclic complexes to capitalize on their advantages and optoelectronic instrumentation already developed for O2 monitors. A proper design of the pH indicator/solid support couple must ensure little or zero cross-sensitivity to such ubiquitous gas.6,7 The emission from Ru complexes with hydroxy-, amino-, or carboxy-substituted pyridine, phenanthroline, and terpyridine ligands, or bearing heterocyclic ligands with protonatable/deprotonatable N atoms, is influenced by the solution pH. The acid-base properties of the ground and the excited states of this type of coordination compound have been investigated in detail, both in solution8-23 and with polymer-supported indicator dyes.24-27 However, verification of a pH-dependent emission intensity or lifetime does not guarantee that the luminophore may be used as a (5) (a) Schulman, S. G.; Chen, S.; Bai, F.; Leiner, M. J. P.; Weis, L.; Wolfbeis, O. S. Anal. Chim. Acta 1995, 304, 165–170. (b) Zhujun, Z.; Seitz, W. R. Anal. Chim. Acta 1984, 160, 47–55. (c) Offenbacher, H.; Wolfbeis, O. S.; Fu ¨ rlinger, E. Sens. Actuators 1986, 9, 73–84. (6) (a) Orellana, G.; Garcıá-Fresnadillo, D. In Optical Sensors: Industrial, Environmental and Diagnostic Applications; Narayanaswamy, R.; Wolfbeis, O. S. Springer Series on Chemical Sensors and Biosensors, Vol. 1; Springer: Berlin, Heidelberg, 2004; pp 309-357; (b) DeGraff, B. A.; Demas, J. N. In Reviews in Fluorescence, Vol. 2; Geddes, C., Lakowicz, J. R., Eds.; Springer Science: New York, 2005; pp 125-151. (7) Orellana, G.; Moreno-Bondi, M. C.; Garcıá-Fresnadillo, D.; Marazuela, M. C. In Frontiers in Chemical Sensors: Novel Principles and Techniques; Orellana, G., Moreno-Bondi, M. C., Eds.; Springer Series on Chemical Sensors and Biosensors, Vol. 3; Springer: Berlin, Heidelberg, 2005; pp 189-225. (8) Vos, J. G. Polyhedron 1992, 11, 2285–2299. (9) Hicks, C.; Ye, G.; Levi, C.; Gonzales, M.; Rutenburg, I.; Fan, J.; Helmy, R.; Kassis, A.; Gafney, H. D. Coord. Chem. Rev. 2001, 211, 207–222. (10) Nam, H.; Jeong, M.; Sohn, O. J.; Rhee, J.; Oh, J.; Kim, Y.; Lee, S. Inorg. Chem. Commun. 2007, 10, 195–198. (11) Ellerbrock, J. C.; Mcloughlin, S. M.; Baba, A. I. Inorg. Chem. Commun. 2002, 5, 555–559. (12) Bingwen, J.; Wu, Tau; Tian, C.; Zhang, M.; Shen, T. Bull. Chem. Soc. Jpn. 2000, 73, 1749–1755. (13) Constable, E. C.; Housecroft, C. E.; Thompson, A. C.; Passatini, P.; Silvi, S.; Maestri, M.; Credi, A. Inorg. Chim. Acta 2007, 360, 1102–1110. (14) Giordano, P. J.; Bock, C. R.; Wrighton, M. S. J. Am. Chem. Soc. 1978, 100, 6960–6965. (15) Higgins, B.; DeGraff, B. A.; Demas, J. N. Inorg. Chem. 2005, 44, 6662– 6669. (16) Su, C.-H.; Chen, H.-Y.; Tsai, K. Y.-D.; Chang, I.-J. J. Phys. Chem B 2007, 111, 6857–6860.

5196


pH indicator dye. Actually, there have been a few accounts of a buffer-dependent luminescence intensity and lifetime variation of Ru(II) polypyridyl complexes,8,28 but most authors seem to ignore these effects, although they have been thoroughly characterized from the kinetic point of view, both at the ensemble and singlemolecule levels.29 Three ways of designing a pH sensor based on Ru(II) complexes may be envisaged:30 (a) To introduce chelating ligands bearing pH-sensitive groups such as 4,7-dihydroxy-1,10-phenanthroline,14,24,27 4-carboxy-2,2′bipyridine,16–18,25 4-(4-pyridyl-)-2,2′-bipyridine,21 or even CN;31 alternatively, five- and six-membered heterocyclic ligands with intrinsic acid-base properties (e.g., imidazoles,12,20,23 triazoles,26 or pyrazines9) may be used around the metal core. (b) To build a composite molecule having the luminescent Ru(II) coordination complex covalently attached to a moiety (typically a protonatable arylamino fragment) that quenches the metal complex emission by electron transfer, depending on the solution pH.32 (c) To perform an indirect quenching based on efficient Fo¨rster resonance energy transfer (FRET) between the luminescent Ru(II) complex and a conventional colorimetric pH indicator dye, the absorption of which overlaps significantly with the emission of the former.33 (17) Nazeeruddin, K. Md.; Zakeeruddin, S. M.; Humphry-Baker, R.; Jirousek, M.; Liska, P.; Vlachopoulos, N.; Shklover, V.; Fischer, C.-H.; Gra¨tzel, M. Inorg. Chem. 1999, 38, 6298–630. (18) (a) Xie, P.-H.; Hou, Y.-J.; Zhang, B.-W.; Cao, Y.; Wu, F.; Tian, W.-J.; Shen, J.-C. J. Chem. Soc., Dalton Trans. 1999, 4217–4221. (b) Nazeeruddin, K. Md.; Kalyanasundaram, K. Inorg. Chem. 1989, 28, 4251–4259. (19) Kovaćs, M. Inorg. Chim. Acta 2007, 360, 345–352. (20) Lu, Y.-Y.; Gao, L.-H.; Han, M.-J.; Wang, K.-Z. Eur. J. Chem. 2006, 430–436. (21) Thompson, A. M. W. C.; Smailes, M. C. C.; Jeffery, J. C.; Ward, M. D. J. Chem. Soc., Dalton Trans. 1997, 737–743. (22) Murtaza, Z.; Chang, Q.; Rao, G.; Lin, H.; Lakowicz, J. R. Anal. Biochem. 1997, 247, 216–222. (23) (a) Jing, B.; Song, A.; Zhang, M.; Shen, T. Chem. Lett. 1999, 789–790. (b) Wang, K.-Z.; Gao, L. H.; Bai, G.-Y.; Jin, L.-P. Inorg. Chem. Commun. 2002, 5, 841–843. (c) Liu, F.; Wang, K.; Bai, G.; Zhang, Y.; Gao, L. Inorg. Chem. 2004, 43, 1799–1806. (d) Bai, G.-Y.; Wang, K.-Z.; Duan, Z.-M.; Gao, L.-H. J. Inorg. Biochem. 2004, 98, 1017–1022. (24) Price, J. M.; Xu, W.; Demas, J. N.; DeGraff, B. A. Anal. Chem. 1998, 70, 265–270. (25) Clarke, Y.; Xu, W.; Demas, J. N.; DeGraff, B. A. Anal. Chem. 2000, 72, 3468–3475. (26) Malins, C.; Glever, H. G.; Keyes, T. E.; Vos, J. G.; Dressick, W. J.; MacCraith, B. D. Sens. Actuators B 2000, 67, 89–95. (27) Chan, C.-M.; Fung, C.-S.; Wong, K.-Y.; Lo, W. Analyst 1998, 123, 1843– 1847. (28) (a) Orellana, G.; Moreno-Bondi, M. C.; Segovia, E.; Marazuela, M. D. Anal. Chem. 1992, 64, 2210–2215. (b) Marazuela, M. D.; Moreno-Bondi, M. C.; Orellana, G. Appl. Spectrosc. 1998, 52, 1314–1320. (29) (a) Boens, N.; Basaric, N.; Novikov, E.; Crovetto, L.; Orte, A.; Talavera, E. M.; Alvarez-Pez, J. M. J. Phys. Chem. A 2004, 108, 8180–8189. (b) Boens, N.; Qin, W.; Basaric, N.; Orte, A.; Talavera, E. M.; Alvarez-Pez, J. M. J. Phys. Chem. A 2006, 110, 9334–9343. (c) Paredes, J. M.; Crovetto, L.; Rios, R.; Orte, A.; Alvarez-Pez, J. M.; Talavera, E. M. Phys. Chem. Chem. Phys. 2009, 11, 5400–5407. (30) Demas, J. N.; DeGraff, B. A. J. Chem. Educ. 1997, 74, 690–695. (31) (a) Peterson, S. H.; Demas, J. N. J. Am. Chem. Soc. 1976, 98, 7880–7881. (b) Peterson, S. H.; Demas, J. N. J. Am. Chem. Soc. 1979, 101, 6571– 6577. (32) Grigg, R.; Norbert, W. D. J. A. J. Chem. Soc., Chem. Commun. 1992, 1300– 1302. (33) (a) Lakowicz, J. R.; Szmacinski, H.; Karakelle, M. Anal. Chim. Acta 1993, 272, 179–186. (b) Kosh, U.; Klimant, I.; Werner, T.; Wolfbeis, O. S. Anal. Chem. 1998, 70, 3892–3897. (c) Kosh, U.; Klimant, I.; Wolfbeis, O. S. Fresenius J. Anal. Chem. 1999, 364, 48–53.

Strategy (b) is limited by the nature of an adequate quencher moiety (aminoaryl), so that it is impossible to prepare Ru(II) indicator dyes for most of the pH scale. Strategy (c) requires high concentrations of the pH-sensitive acceptor dye, unless it is covalently tethered to the luminescent Ru(II) donor, and it mandates a proper absorption spectrum of the former to be efficient enough, thus limiting the number of usable colorimetric indicators. The most versatile method is strategy (a), because of the large number of functional groups and heterocyclic moieties available, so that Ru(II) pH indicator dyes may be designed for the entire pH range (and even outside it!).28 Development of usable pH sensors based on strategy (a) requires the design of appropriate combinations of the protonexchanging and ancillary ligands. The electron density of the longlived 3MLCT emissive state of Ru(II) complexes can be located preferentially in the protonatable/deprotonatable ligand, in any other ligand or in both, resulting in different acid-base properties of their excited state. Therefore, it is no surprise that there are very few reports so far on the use of Ru(II) complexes for direct pH optosensing (although many studies on the effect of the solution acidity on the emission of ruthenium polypyridyls have been published).15,22,25 There are two limiting situations that yield a useful lifetimebased pH indicator: fast proton exchange and slow proton exchange in the excited state (both of them with respect to the excited acid/base species decay rates). In these two cases, the luminescence decays as a function of pH are not complicated by the re-equilibration of the acidic and basic forms upon excitation because, in the first situation, the excited-state equilibration is achieved immediately before emission takes place, whereas, in the second one, it does not take place at all. In the fast exchange regime, a single emission lifetime (τ) that varies with pH is observed, and from this measurement the solution pH can be determined (eq 1):34

[(

pH ) pK*a + log

)(

1 1 1 1 - / τHA τ τ τA

)]

(1)

where τHA and τA are the lifetimes of the pure acidic and basic forms, respectively, obtained at extreme pH values. In the slow exchange regime, two lifetimes that do not change with pH are observed; only the pre-exponential factors (R) of the decay kinetics of the acidic (HA*) and basic (A-*) forms of the excited-state change according to the ground-state pKa value. Therefore, the pH can be determined in this case from eq 2:

Scheme 1. Chemical Structure of the bpds, F15ap, pbbs, and pyim Chelating Ligandsa

a

The exchangeable proton atom is marked in red boldface.

confer buffer dependency to the indicator dye, a handicapping feature. To develop luminescent Ru(II) polypyridyls that can actually be employed for direct pH optosensing (a mechanism that, so far, seems to be difficult to achieve or elusive to evidence in luminescent Ru(II) polypyridyls) and to shed some light on the features that determine existence of an excited-state equilibrium, we have prepared two novel luminescent Ru(II) complexes (see Scheme 1)snamely, Na2[Ru(bpds)2(F15ap)] (1) and Na2[Ru(pbbs)2(pyim)] (2), where bpds ) (2,2′-bipyridine)-4,4′disulfonate, F15ap ) N-(1,10-phenanthrolin-5-yl)perfluorooctadecanamide, pbbs ) (1,10-phenanthroline-4,7-diyl)bis(benzenesulfonate), and pyim ) 2-(2′-pyridyl)imidazole. These metal complexes contain two ancillary polyazaheterocyclic ligands (bpds and pbbs) that provide a polar anionic periphery around the cationic metal center and sufficient solubility in water. Because attachment of the indicator dye to an insoluble polymer is unavoidable for fiber-optic chemical sensing, those ancillary ligands also will serve as covalent anchors of the pH indicator dye to amino-derivatized solid supports (glass, sol-gel materials, organic polymers, etc.).36 The third chelating ligand (F15ap or pyim) bears the pH-sensitive moiety (perfluoroamide or imidazole, respectively; see Scheme 1). For the sake of a meaningful comparison, both complexes display the same overall electrostatic charge (2-). With the aid of the observed changes in the absorption spectra, luminescence intensity, and lifetime of the metal complexes using various buffer solutions, we are now able to assess the suitability of Ru(II) indicator dyes for pH measurements.

In addition, if the luminescent indicator dye is to be of general usage (i.e., a “true” universal indicator), it must be devoid of quenching by irreversible excited-state proton transfer from/to the buffer species at competitive rates with those of the excited acid/base species. Such processes would complicate the interpretation of results (particularly with the usual luminescence phase-shift instrumentation for field measurements)35 and would

EXPERIMENTAL SECTION Chemicals. All the organic solvents used (analytical or HPLC grade) were supplied by Merck (Darmstadt, Germany), SDS (Peypin, France), or Panreac (Barcelona, Spain). The reagents for synthesis were supplied by Lancaster Synthesis, Ltd. (Morecambe, England), Sigma-Aldrich Quimica (Madrid, Spain), or Panreac. Sephadex LH20 adsorbent was from GE Biosciences (Upsala, Sweden). Water was from a Millipore Direct-Q purification system. Argon of +99.995% purity was from cylinders supplied by Praxair (Madrid, Spain). The luminescent indicator dyes require preparation of cis-dichlorobis(chelate) complexes followed

(34) Ireland, J. F.; Wyatt, P. A. H. Adv. Phys. Org. Chem. 1976, 12, 131–221. (35) Bowyer, W. J.; Xu, W.; Demas, J. N. Anal. Chem. 2009, 81, 378–384.

(36) Xavier, M. P.; Garcıá-Fresnadillo, D.; Moreno-Bondi, M. C.; Orellana, G. Anal. Chem. 1998, 70, 5184–5189.

( )

R*ApH ) pKa + log R*HA

(2)


5197

by reaction with excess of the third ligand (pyim or F15ap). The synthesis of Na4[Ru(bpds)2Cl2] and Na4[Ru(pbbs)2Cl2] has been reported.37 The preparation of F15ap, Na2[Ru(bpds)2(F15ap)], and Na2[Ru(pbbs)2(pyim)] is described in the Supporting Information. Steady-State Spectroscopic Measurements. Electronic spectroscopy was performed at 298 ± 1 K in buffer solutions contained in quartz cells with a path length of 1.0 cm (Hellma Hispania, Badalona, Spain). UV-vis absorption spectra were recorded with a Varian Cary 3Bio spectrophotometer. Luminescence spectra, uncorrected for the instrumental response, were measured with either a Horiba-Jobin-Yvon SPEX FluoroMax-2 or a Perkin-Elmer LS50B spectrofluorometer. Luminescence quantum yields (φem) in water were obtained using [Ru(bpy)3]2+ as a reference luminophore (φem ) 0.042 ± 0.002 in water at 298 K under an argon atmosphere).38 Time-Resolved Luminescence Measurements. Emission lifetimes (τ) were determined by single photon timing (SPT), using an Edinburgh Instruments Model FL-900 spectrometer fitted with an Horiba Model NanoLed-07N violet laser diode (405-nm, 900-ps fwmh pulses). Details on the instrument and procedures for analyzing the data have been described.39 Time-resolved luminescence spectra (TRLS) were obtained with a laser flash photolysis apparatus (Edinburgh Instruments, Model FS-900). The excitation source was a tripled Nd:YAG laser (355-nm, 5-ns per pulse, Continuum Minilite II). Deoxygenated solutions were prepared by purging them with solvent-saturated argon for at least 20 min prior to the measurement. Titrations and Experiments at Variable pH. Solution pH values (±0.02 pH units) were determined with a Crison GLP21 pH meter. Constant ionic strength buffers used include 10-, 50-, and 100-mM phosphate, 10-mM bis-tris propane (1,3-bis[tris(hydroxymethyl)methylamino]propane)orimidazolefromSigma-Aldrich, Panreac, or Fluka (Buchs, Switzerland). Computational Details. Quantum chemical calculations were performed with the Gaussian 03 program.40 The approximation used was density functional theory (DFT) with a hybrid exchangecorrelation B3LYP functional, together with the LANL2DZ basis set. The Ru inner core electrons were represented by a scalar relativistic electron core potential (ECP). Geometry minimizations were performed in vacuo with the default options in Gaussian 03. The lowest triplet state was chosen for minimization, as it is (37) Garcıá-Fresnadillo, D.; Orellana, G. Helv. Chim. Acta 2001, 84, 2708–2730. (38) (a) Van Houten, J.; Watts, R. J. J. Am. Chem. Soc. 1976, 98, 4853–4858. (b) Caspar, J. V.; Meyer, T. J. J. Am. Chem. Soc. 1983, 105, 5583–5590. (39) Castro, A. M.; Delgado, J.; Orellana, G. J. Mater. Chem. 2005, 15, 2952– 2958. (40) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P. ; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A. ; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; and Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.

5198


Figure 1. Normalized absorption (Abs) and uncorrected luminescence spectra (Em) of Ru(II) complexes (A) 1 and (B) 2 in 10 mM phosphate buffer solution at pH 2 (solid lines) and pH 11 (dotted lines).

representative for the emissive state of the ruthenium complexes. For the orbital energy calculations, single-point calculations were conducted in the presence of a polarizable continuum model (PCM) with appropriate parameters to model the effect of water in the system. The molecular surface calculations were performed with the Connolly algorithm,41 as implemented in the Materials Studio software.42 Initial drawing of 2D structures, as well as conversion to 3D structures as starting points for the Gaussian 03 geometry optimizations, were performed with ChemAxon’s Marvin 5.0.7 (http://www.chemaxon.com). RESULTS AND DISCUSSION Photophysical Properties of the Indicator Dyes. The electronic absorption spectra of both complexes in aqueous solution exhibit the typical bands of the Ru(II) polypyridyl family:6,43 Bands in the near-UV region (297 nm for 1 and 277 nm for 2) correspond to π-π* ligand-centered transitions while a broad band in the visible region (465 nm for 1, and 449 and 473 nm for 2) is assigned to a d-π* transition to the 1MLCT (metal-to-ligand charge transfer) excited state. Their bright orange-red emission is also characteristic of Ru(II) polypyridyls. Figure 1 depicts the absorption and luminescence spectra of the metal complexes in acidic and basic media. In both cases, there (41) Connolly, M. L. Science 1983, 221, 709–713. (42) Materials Studio, V. 2.2; Accelrys, Inc.: San Diego, 2002. (43) (a) Juris, A.; Balzani, V.; Barigelleti, F.; Campagna, S.; Bleser, P.; Von Zelewski, A. Coord. Chem. Rev. 1988, 84, 85–277. (b) Campagna, S.; Puntoriero, F.; Nastasi, F.; Bergamini, G.; Balzani, V. Top. Curr. Chem. 2007, 280, 117–214.

Table 1. Photophysical Parameters of [Ru(bpds)2(F15ap)]2- (1) and [Ru(pbbs)2(pyim)]2- (2) in 10 mM Phosphate Buffer Solutions at 298 K τ (ns) λmaxabs dye 1 2

em

In Air Atmosphere

λmax

In Argon-Purged Solution

at pH 2.0

at pH 10.0

at pH 2.0

at pH 10.0

at pH 2.0

at pH 10.0

at pH 2.0

at pH 10.0

463 451

473 459

641 644

647 673

372 600

263 166 (23), 75 (77)b

445 1425

309 209 (22), 95 (78)b

a The uncertainty values are ±1 nm for the spectral maxima; 1% and 3% for the emission lifetimes extracted from the fit to single- or doubleexponential functions, respectively. b The values given in parentheses represent the relative contribution (%) of the pre-exponential factors: IL(t) ) R1 exp(-t/τ1) + R2 exp(-t/τ2); %i ) Ri/ΣRi.

is a bathochromic shift in the absorption and emission bands when the pH increases. Furthermore, the complex 1 shows an n-π* absorption band at 344 nm attributed to the perfluoroamide (F15ap) ligand. This band undergoes a hyperchromic shift at high pH, i.e., after deprotonation. A similar n-π* absorption band has also been found in heteroleptic ruthenium complexes with nonfluorinated 5-acetamide and 5-tetradecanamide phenanthroline ligands.44 A large Stokes shift is observed due to the vibrational relaxation and subsequent emission from a 3MLCT excited state. The sulfonate groups of the bpds and pbbs chelating ligands determine an important stabilization of the LUMO (π*) orbital due to mesomeric effects, particularly in the former one. Luminescence quantum yields (Φem) of 2.6 (± 0.1) × 10-2 and 1.9 (± 0.1) × 10-2 in argon-saturated 10 mM phosphate buffer at pH 2.5 and pH 9.5, respectively, have been measured at 298 K for 1. However, a Φem value of only 0.43 (± 0.03) × 10-2 was determined for 2 in water, compared to 2.2 (± 0.1) × 10-2 for 1 under the same conditions.45 The lower value for [Ru(pbbs)2(pyim)]2- may be attributed to the imidazole ligand, because of a rapid nonradiative deactivation of the excited state in which the excitation energy may be vibrationally dissipated to the solvent through the N-H bond of the ligand.46 The photophysical studies of Ru(II) complexes containing imidazole moieties have shown that the wider “bite” angle of the fivemembered chelating ligand reduces its ligand field strength, increasing the nonradiative deactivation pathway by thermal activation to the silent 3MC state.12,20,23,46 These complexes are usually weakly emissive or nonemissive at all, so that a electronwithdrawing diphenylphenanthroline derivative has been introduced in 2 as an ancillary ligand to prolong the lifetime of the 3 MLCT excited state.43 Selected photophysical parameters of the novel pH-sensitive dyes in 10 mM phosphate buffer solution are compiled in Table 1. Higher luminescence intensity and lifetimes are observed for acidic solutions, while there is a bathochromic shift in the MLCT absorption and luminescence bands in basic media. As the energy gap rule predicts, the closer the emitting state to the ground state, (44) Garcıá-Fresnadillo, D.; Orellana, G. Helv. Chim. Acta 2001, 84, 2708–2730. (45) No attempt was made to determine the emission quantum yields of 2 in buffer media, because of the strong dependence of this parameter and the luminescence lifetime on the buffer nature and concentration, in addition to the solution acidity (see below). (46) See, for example: (a) Haga, M. A. Inorg. Chim. Acta 1983, 77, 39–41. (b) Haga, M. A. Inorg. Chim. Acta 1983, 75, 29–35. (c) Orellana, G.; Quiroga, M. L.; De Dios, C. Trends. Inorg. Chem. 1993, 3, 109–129. (d) Liu, F. R.; Wang, K. Z.; Bai, G. Y.; Zhang, Y. A.; Gao, L. H. Inorg. Chem. 2004, 43, 1799–1806.

Figure 2. Absorption spectra of (A) 1.1 × 10-5 M dye 1 and (B) 1.3 × 10-5 M dye 2 in 10 mM phosphate buffer solution at different pH values. The arrows indicate the direction of the spectral changes that occur from pH 2 to pH 11.

the more efficient will be the radiationless deactivation. The effect of O2 quenching is significant in the case of longer luminescence lifetimes. Table 1 shows that the 610 ns emission lifetime of the acidic form of 2 increases to 1400 ns in argon-saturated aqueous solution, corresponding to a radiative rate constant of ca. 3.5 × 109 s-1, which is a typical value for luminescent Ru(II) complexes.47 Excited states of the latter that live long enough to display a sufficient analytical emission signal but display shorter lifetime to minimize interference from O2 are preferred for pH-sensitive indicator dyes. Ground-State pKa Evaluation. Figure 2 shows the changes in the absorption spectra of the complexes at pH 2-11, using phosphate buffer solutions. The presence of isosbestic points at 478 nm for 1 and at 445 nm for 2 is a signature of the equilibrium (47) Garcıá-Fresnadillo, D.; Georgiadou, Y.; Orellana, G.; Braun, A. M.; Oliveros, E. Helv. Chim. Acta 1996, 79, 1222–1238.


5199

compared to acetanilide (pKa ) 17.6).49 The even stronger acidity enhancement observed for 1 is due to the additional electron-withdrawing effect of the F atoms and the ruthenium(II) core. For complex 2, the measured pKa is much lower than that of the free pyim ligand (pKa ) 13.4),50 because of the strong ligand-to-metal σ donation and stabilization of the imidazolate anion. The reported pKa value of the [Ru(bpy)2(pyim)]2+ complex is 7.9 in methanol-water (5% v/v).46 The lower acidity of the latter, compared to the bispbbs analogue, is due to the stronger electron-withdrawing character of the disulfonated bpy ligand. All the changes observed in the absorption spectra are fully reversible if the solution acidity or basicity is neutralized. Excited-State pKa (pKa*) Evaluation. A photochemical characterization of novel pH indicator dyes requires evaluation of the acidity constant of their luminescent excited state (pKa*). Fo ¨rster employed a thermodynamic cycle to describe the relationship between pKa and pKa*.51 According to his simple model, the free energy for a proton dissociation in the excited state can be approximated by eq 4: ∆G* ≈ ∆G + [(hvA-) - hvHA]NA Figure 3. (A) Normalized absorbance values of Ru(II) dyes (4) 1 and (O) 2 at 344 and 460 nm, respectively, as a function of pH in phosphate buffer. (B) Normalized luminescence intensity of (4) 1 and (O) 2 at their emission maximum (Figure 1), as a function of pH in 10 mM phosphate buffer. Excitation was performed at the isosbestic points in the absorption blue region (Figure 2). Lines are the best fits to sigmoidal functions.

between the respective (ground state) protonated and deprotonated species for both Ru(II) complexes. These equilibriums were characterized by analyzing the absorption spectral changes. The titration curves obtained by plotting the absorbance values at selected wavelengths, as a function of pH, are shown in Figure 3A. The inflection points of the sigmoidal curves corresponding to the pKa values of the indicator dyes can be determined from eq 3:48

[

(AA-) - ApH pKa ) pH + log ApH - AHA

]

(3)

where AHA and AA- represent the absorbance of the protonated and deprotonated forms at the selected wavelength, respectively, and ApH is the absorbance of the solution at each pH value. In this way, pKa values of 6.5 (± 0.2) and 6.9 (± 0.2) for 1 and 2, respectively, are obtained. The pKa of complex 1 indicates that the perfluoroamide group of the F15ap ligand is significantly more acidic than typical amides. The electronwithdrawing effect of the 15 F atoms produces a strong decrease in the amide pKa value. This decrease in pKa, which is due to the effect of halogen atoms on the amide group, has been observed, for instance, in trichloroacetanilide (pKa ) 9.5) (48) Albert, A.; Serjeant, E. P. The Determination of Ionization Constants; Chapman and Hall: New York, 1984.

5200


(4)

where ∆G is the free energy of the reaction in the ground state; νA- and νHA refer to the frequencies of the lowest energy electronic transitions of the basic and acidic forms of the compound, respectively; h is Planck’s constant; and NA is Avogadro’s number. In the Fo ¨rster-cycle approximation, it is assumed that the entropic component of the proton exchange reaction does not change significantly in the excited state compared to that of the ground state, so that the main difference in the ∆G values is enthalpic in nature. In this way, it is possible to obtain the (hypothetical or real, depending on the magnitude of the proton-exchange kinetic constants, compared to the deactivation rate constants) excited-state pKa value (pKa*) by transforming eq 4 into eq 5:

[( ) ( )]

pK*a ) pKa + 2.1 × 104

1 1 λAλHA

(5)

where λA- and λHA are the wavelengths (in nanometers) of the lowest energy (0-0) electronic transitions of the basic and acidic forms of the indicator molecule. These values can be determined by absorption or emission measurements; yet, those from luminescence data are usually more accurate, because of the thermally equilibrated and solvent-equilibrated nature of the emissive state.52 The bathochromic shifts of the absorption and emission bands of both 1 and 2 upon deprotonation (Table 1) suggest pKa* values slightly more acidic than the corresponding ground-state pKa. (49) Homer, R. B.; Johnson, C. D. Acid-Base and Complexing Properties of Amides. In The Chemistry of Amides; Zabicky, J., Ed.; Interscience: London, 1970; Chapter 3, pp 187-243. (50) Boggess, R. K.; Martin, R. B. Inorg. Chem. 1974, 13, 1525–1527. (51) (a) Fo ¨rster, T. Z. Elektrochem. 1950, 54, 42–46. (b) Fo ¨rster, T. Z. Elektrochem. 1950, 54, 531–535. (52) (a) Weller, A. Prog. React. Kinet. 1961, 1, 189–214. (b) Shizuka, H.; Tobita, S. Proton Transfer Reaction in the Excited State. In Molecular and Supramolecular Photochemistry; Ramamurthy, V.,; Schanze, K., Eds.; Organic Chemistry and Photophysics, Vol. 14; Taylor and Francis Group, CRC Press: New York, 2006; Chapter 2, pp 37-74.

By introducing the emission maxima in Table 1 into eq 5, pKa* values of 6.2 (± 0.2) and 5.5 (± 0.2) for 1 and 2, respectively, are obtained. These values are approximate because the emission band maxima have been employed for the calculation instead of the actual 0-0 transition energy values, which would lie at slightly higher energies. The small differences (-0.3 to -1.4 units) between the measured pKa value and the Fo ¨rster pKa* values can be attributed to the fact that the 3MLCT (d-π*) excited state of complexes 1 and 2 is preferentially located in the ancillary sulfonated chelating ligands and not in the pHsensitive ligands, because of the higher energy of the π* orbital of the latter. For comparison, the homoleptic complex [Ru(pzth)3]2+ (here, pzth denotes 2-(2′-pyrazinyl)thiazole)) shows a difference of +6.4 units between its pKa and pKa* values, because the photoexcited electron is mostly located in the π* orbital of the protonatable pyrazine moiety of the pzth ligand and not in the higher-lying π* orbital of the thiazole ring.28 Figure 3B shows the emission titration curve for the novel Ru(II) complexes 1 and 2, as a function of pH. The inflection points obtained thereof are 6.6 (± 0.2) and 6.4 (± 0.1), respectively; note that the emission titration curves only provide values of the apparent acidity constants (pKap*). These values differ from the actual acid-base equilibrium constants for kinetic reasons related to the lifetime of the species involved (see below).34 Luminescence Lifetime Features. The absorption and emission spectral shifts, and the observed changes of the emission intensity, indicate that complexes 1 and 2 are luminescent in both their protonated and deprotonated forms. Those changes are also indicative of different excited state lifetimes for the *HA and *Aspecies. The luminescence decays of each complex were satisfactorily fitted to a sum of two exponentials for a wide range of pH values (see Figure 4), except for the extreme values of the investigated pH range 2-10, where single exponential functions were just required to achieve a good fit in most cases. The emission lifetimes of the corresponding *HA and *A- forms of complexes 1 and 2 measured at the plateaus of the luminescence lifetime titrations (Figure 4) are listed in Table 1. Regardless of the solution pH, the luminescence decay of the F15ap complex (1) can always be fitted to a linear combination of two exponential decay functions with lifetimes of 372 and 263 ns (see Figure 4A), which are exactly those measured for its *HA and *A- forms. This result indicates that the proton exchange rates in the excited state are well below their respective deactivation rates, so that the two excited species decay independently of each other. Using the pKa* value of this complex determined in the previous section, we can obtain an upper estimate for the proton exchange rate constants in the excited state by assuming a diffusion-limited protonation with k2* ) 1010 M-1 s-1 and considering that k1* ) k2* × 10-pKa*, where k1* is the proton dissociation rate constant. In this way, a pKa* value of 6.2 would yield a k1* value of 6.3 × 103 s-1. The corresponding excited-state deactivation rate constants are 2.7 × 106 s-1 (for the 372 ns lifetime of *HA) and 3.8 × 106 s-1 (for the 263 ns lifetime of *A-), i.e., much faster than the proton exchange rate constants at pH >5 ([H3O+] < 10-5), where the ground-state basic form of the indicator dye exists and, therefore, can also be excited.

Figure 4. Luminescence lifetimes of the acidic and basic forms of (A) [Ru(bpds)2(F15ap)]2- at 640 nm and (B) [Ru(pbbs)2(pyim)]2- at 650 nm upon excitation at 405 nm, as a function of pH in 10 mM phosphate buffer under air. The insets show the corresponding amplitudes of the luminescence lifetimes obtained from the biexponential fits of the emission decays at each pH value. The solid lines in panel (B) are intended to guide the reader in the discussion (see text).

Moreover, protonation of the photoexcited basic species of complex 1 might be far from the diffusion-limit control if we take into account the geometry of the energy-minimized structure of this complex (Figure 5A). The amide group is coplanar to the phenanthroline ring and its exchangeable H atom is sandwiched between the peri H-4 atom of the phenanthroline moiety and a F atom of the perfluorooctanoyl chain. In this way, the restricted access to the amide proton (and also to its conjugated basic N atom) would slow protonation from the usual diffusional limit of proton transfer reactions between heteroatoms, with catastrophic consequences for achieving equilibration in the excited state. The presence of multiple lipophilic F atoms in close proximity to the protonation site is expected to distort the water structure around this region, further hampering a diffusion-limited fast protonation. As we will discuss below, such structural features of 1 can explain the insensitivity of this complex to proton transfer from its excited acidic form by the abundant basic species of the different buffers tested at pH ∼7. Under a slow proton exchange regime in the excited state, the luminescence decay of the indicator dye 1 must obey biexponential kinetics (eq 6):

( )

IL(t) ) (RHA*)exp -

( )

t t + (R*A-)exp τHA τA-


(6) 5201

Figure 5. Geometry optimized structures of Ru(II) complexes (A) 1 and (B) 2. The structures are displayed in CPK rendering, together with a molecular surface generated with the Connolly algorithm using a spherical probe with a radius of 3 A. The molecular surfaces are shown in blue and transparent, to allow the visualization of inaccessible atoms located under the molecular surface. Areas where the atomic van der Waals surfaces contact the probe and therefore coincide with the molecular surface appear with the corresponding element-based CPK color. In panel (A), the amide group of 1 turns out to be coplanar to the phenanthroline (phen) moiety, and its proton is sandwiched between the phen peri-H atom and a F atom. Geometry optimizations initiated from various amide-phen torsion angles converged to the displayed structure. Only an alternative local minimum was found where the amide is rotated 140° in the opposite direction; however, this structure is 6.96 kcal/mol higher in energy. Panel (B) shows that the imidazol proton of 2 has a large contact area with the probe and, therefore, is expected to be readily accessible to water and buffer molecules.

where IL(t) is the normalized time-dependent emission decay, τHA and τA- are the luminescence lifetimes of its excited acidic and basic forms, respectively, and RHA* and R*A- are the decay amplitudes of the corresponding excited forms at a particular pH value. The latter amplitudes equal the mole fractions of the corresponding ground-state species before excitation, because of the absence of equilibration in the excited state. From the decay amplitudes in Figure 4 (inset) and eq 2, the indicator pKa can also be calculated and matches the value obtained from eq 3, within the experimental uncertainties. Consequently, after extracting the mole fractions (decay amplitudes) from a single time-resolved luminescence measurement of the indicator 1 in an unknown sample, its pH can be easily determined using eq 2. This self-referencing feature will be valuable for using [Ru(bpds)2(F15ap)]2+ as an intracellular pH probe in FLIM observations. Unlike the indicator dye 1, the pyim complex (2) shows a pHindependent luminescence lifetime of 80 ns (at pH g6) and a pHsensitive emission lifetime that varies from 610 ns (at pH 2) to 170 ns (at pH 10) in an air-equilibrated buffer solution (Figure 4B). The presence of two distinct lifetimes indicates again that an acid-base equilibrium is not achieved in the excited state. Similarly, as discussed above for the luminescent complex 1, the upper limit for the deprotonation rate constant is estimated to be k1 ) 3.2 × 104 s-1 for pKa* ) 5.5. This value and the protonation rate constant at pH >5 ([H3O+] < 10-5), assuming diffusionlimited kinetics (see above), are also much smaller than the rate constants for excited-state deactivation of the acidic and basic species: 1.6 × 106 s-1 (for the 610 ns lifetime) and 1.3 × 107 s-1 (for the 80 ns lifetime), respectively. The emission of complex 2 in an acid medium is dramatically quenched when the solution pH increases beyond pH 5 (Figure 4B). Therefore, in this case, there must be an additional deactivation pathway by proton transfer quenching from the excited Ru 5202


complex to the monohydrogen phosphate buffer species (pKa(H2PO4-) ) 7.2), an increasingly abundant base at pH > 5 where OH- cannot compete. Such quenching is responsible for the decrease of the lifetime of the acidic form of 2 in the pH 5-7 interval. The less-basic (but more concentrated at 2 < pH < 7) H2PO4- does not seem to provide a fast-enough deprotonation to the excited indicator acid form, so that a single exponential lifetime is observed up to pH >5. The involvement of HPO42- in the proton abstraction from the acidic form of the photoexcited Ru complex is demonstrated by the linear Stern-Volmer plot measured from the HPO42- concentrations in the samples, as a function of their pH (KSV ) 936 ± 15 M-1, r2 ) 0.998; see Figure 1S in the Supporting Information). Since KSV can be expressed as kqτ0, the bimolecular rate constant for the proton transfer quenching (kq) was calculated to be (1.5 ± 0.2) × 109 M-1 s-1 from the luminescence lifetime of the dye in the absence of HPO42- (τ0 ) 610 ns). The same results are obtained if the luminescence intensity is monitored instead of the emission lifetime (see Figure 1S in the Supporting Information). The coincidence of results obtained from the intensity and lifetime Stern-Volmer quenching plots reflects the dynamic nature of the proton transfer reaction from the excited Ru(II) complex. The dual excited-state deactivation pathway (when available) of the excited acidic form of 2snamely, spontaneous decay and buffer-induced deprotonationsdescribed by eq 7, properly reproduces the experimental decrease of the lifetime in the pH 5-7 range observed (Figure 4B). τHA )

1 (109 /610) + kq[HPO42-]

(7)

However, at higher pH, it predicts a plateau with a lifetime of ∼60 ns, so that we must conclude that the observed 80 ns lifetime

Table 2. Ground-State pKa Values Determined from the Absorption Spectra, and (Apparent) pKap* Values Determined from the Emission Spectra, as a Function of pH for the Ru(II) Indicator Dyes at Different Phosphate Buffer Concentrations [buffer] (mM)

Figure 6. Time-resolved luminescence spectra (TRLS) of [Ru(pbbs)2(pyim)]2- (2) in 10 mM phosphate buffer solution at pH 7.0 upon laser excitation at 355 nm. The spectra have been obtained at 123 ns (A-) and 300 ns (HA), respectively, after the laser pulse.

must not be assigned to the excited basic form of complex 2 but rather to its phosphate buffer-quenched excited acid form. The observed 170-ns lifetime at pH g8 should correspond indeed to that of the excited basic form of the indicator dye. Incidentally, this lifetime would decrease at pH

Can Luminescent Ru(II) Polypyridyl Dyes Measure pH Directly?

Recommend Documents