Double Action: Toward Phosphorescence Ratiometric Sensing of Chromium Ion

The phosphorescence double ratiometric sensor detects Cr(III) ions by taking advantage of reversible coordination binding and the biomimetic oxidation reaction. Photoluminescent sensors provide useful information about identification, concentration, and spatiotemporal fluctuation of analytes. Of particular importance is the development of photoluminescent sensors for oxophilic transition metal ions such as Fe and Cu.1, 2 These d-block metal ions are critically involved in many of biological and environmental processes, but the exact chemical mechanisms of their actions are yet to be fully elucidated. Therefore, information concerning these metal ions is of prime importance, and photoluminescent sensors are in great demand for this purpose. Unfortunately, however, creation of useful photoluminescence turn-on and ratiometric sensors is difficult because low-lying d−d transition states and low redox potentials of the paramagnetic transition metals efficiently quench photoluminescence emission.3 In addition, lack of selective receptors for specific metal ions poses great challenges to the development of sensors for these transition metal ions. Chromium is widely used in a variety of industrial applications, such as electroplating and tanning. Chromium complexes have also been employed as colorants, catalysts, and wood preservatives. This widespread consumption of chromium has resulted in significant bio- and environmental accumulation. It is considered that excessive Cr ions are toxic, and particularly associated with genotoxicity.4 High-valent Cr ions such as Cr(V)5, 6 and Cr(VI)7, 8 are strongly implicated in the toxic process, which is believed to involve production of reactive oxygen species. Studies on Cr-oxo model compounds have provided mechanistic insight into the biomimetic Cr-induced oxidation reactions.9-11 Information on identification and spatiotemporal fluctuation of chromium ion would be of enormous value in relating the established oxidation chemistry of Cr with chromium toxicity. Selective photoluminescent sensors for chromium ion are, therefore, in high demand. In this context, there have been many efforts to develop photoluminescence turn-on12-19 or ratiometric20, 21 sensors based on fluorescent platforms such as conjugated imine,12 rhodamine,15, 16, 19-21 phthalimide,13, 17 dansyl,14 and BODIPY.18 Despite these advances, the sensors still suffer from inadequacies particularly with respect to reversibility, response times, and selectivity. Most of all, identification of chromium ion with high fidelity in the presence of other divalent transition metal ions such as Cu(II) remains a challenging task. Herein, we report and demonstrate a novel strategy for the double ratiometric detection of chromium(III) ion. The first phosphorescent probe (YJ1) for chromium ion has been developed based on a cyclometalated Ir(III) complex ([Ir(dfppy)2phen]+) bearing two 2-(2,4-difluorophenyl)pyridine (dfppy) and 1,10-phenanthroline (phen) ligands. The cyclometalated Ir(III) complex was chosen as a signal transmitter because of high efficiency room temperature phosphorescence and marked robustness against photobleaching.22, 23 The Cr(III) ion-responsive bis(2-(2-(methylthio)ethylthio)ethyl)amino (BTTA) receptor was introduced at the 4-position of the phen ligand. The interaction between Cr(III) ion and the BTTA receptor produces a double-stage phosphorescence ratiometric response in acetonitrile (Figure 1). The first stage is a prompt and reversible green-to-orange phosphorescence change (Figures 1b,c), which is due to perturbation of the excited states of [Ir(dfppy)2phen]+ that is weakly interacting with the BTTA receptor. The second stage is a relatively slow conversion of the orange phosphorescence to green phosphorescence through a specific Cr-mediated oxidative cleavage (Figures 1c,d). The overall observable are thus sequential double phosphorescence ratiometric responses (i.e., green → orange → green) that are specific to Cr(III) ion. Therefore, this sensing strategy allows for highly selective and accurate identification of Cr(III) ion in the presence of other metal ions. (a) Schematic representation of double phosphorescence ratiometric response of YJ1 for Cr(III) ion. Time-resolved photoluminescence spectra of CH3CN solutions of YJ1 (10 μM) in the absence (b) and presence (c) of Cr(III) ion (1 equiv) and (d) after Cr(III)-induced oxidative cleavage. Inset photos are phosphorescence emission of the YJ1 solutions illuminated under 365 nm excitation. This phosphorescent sensor for Cr(III) ion has been prepared in a nine step synthesis. Details on the synthetic procedures are described in the Supporting Information (SI, Scheme S1). Despite the fact that the sulfur-rich BTTA receptor is known to bind soft metal ions such as Cu(I) ion in water,24-26 in our case we observed a strong response to Cr(III) ion instead of Cu(I) in acetonitrile and no response in buffers (SI, Figure S1). In fact, there was also negligible response to soft metals such as Cu(I) and Ag(I) in CH2Cl2, THF, and CH3OH (SI, Figure S2). The reason for the Cr(III) ion specificity of YJ1 is not clear but may be due to the tighter Cr(III) binding of the tertiary amine of BTTA in acetonitrile. The Cr(III) receptor was introduced to the phen ligand through reductive amination between 4-formyl-1,10-phenanthroline and bis(2-(2-(methylthio)ethylthio)ethyl)amine. Substitution of the chlorides of the [Ir(dfppy)2(μ-Cl)]2 with the BTTA-functionalized phen ligand afforded YJ1. The probe has been characterized by using spectroscopic methods based on 1H, 13C, 19F, and {1H,1H} COSY NMR and an electrospray ionization mass (ESI MS) spectrometry (SI, Figures S19−S23), and is in agreement with the proposed structure. The positively charged complex YJ1 is highly soluble in a variety of common organic solvents such as acetonitrile up to a concentration of 10 mM. An air-equilibrated acetonitrile solution of YJ1 (10 μM) displays weak green phosphorescence emission with a peak wavelength at 515 nm (λex = 315 nm). Addition of Cr(ClO4)3 to the YJ1 solution evokes a rapid phosphorescence ratiometric change with a 8-fold turn-on and a 42 nm red-shift (Figure 2a).27 Corresponding photoluminescence quantum yield (PLQY; determined by a relative method using fluorescein standard) increases from 0.016 to 0.13 for Ar-saturated CH3CN solutions (10 μM) at room temperature. In contrast, there is no distinct change in the absorption bands of the singlet ligand-centered π−π* (1LC; < 330 nm) transition and the metal-to-ligand charge-transfer (1MLCT; 330−450 nm) transition after the addition of Cr(ClO4)3 because of low concentration (SI, Figures S3−S4).28 Cr titration has been carefully performed by measuring the prompt increase in the phosphorescence intensity by injecting a CH3CN solution of Cr(ClO4)3 (1 − 1.5 equiv). The titration isotherm plotting the phosphorescence intensity at 557 nm as a function of concentration of added Cr(III) ion indicates a 1:1 binding stoichiometry (Figure 2a, inset). Binding of Cr(III) ion is also supported by the appearance of a peak at m/z = 265.1 in the ESI MS spectrum which corresponds to [K2CrIII(YJ1)(PF6)]5+ (calcd m/z = 265.0; SI, Figure S5). The ratiometric Cr(III) response is fully reversible as demonstrated by restoration of the original green phosphorescence after subsequent addition of a strong metal chelator, N,N,N′,N′-tetrakis(2-picolyl)ethylenediamine (TPEN; Figure 2b). A reference Ir(III) complex (Ir5F; i.e., [Ir(dfppy)2phen]PF6) lacking the BTTA receptor shows no change in the phosphorescence spectrum upon the addition of Cr ion (10 equiv; SI, Figure S6), supporting that the ratiometric response of YJ1 is due to the Cr chelation of BTTA.29 Reversible phosphorescence ratiometric response to Cr(III) ions. (a) Change in phosphorescence spectrum of YJ1 (10 μM, CH3CN) with the continuous addition of Cr(ClO4)3 (0 − 1.5 equiv). Inset graph is a titration isotherm plotting prompt increase in phosphorescence intensity as a function of total concentration of Cr(ClO4)3 (1 − 1.5 equiv). (b) Phosphorescence spectra of YJ1 (10 μM, CH3CN) in the absence and presence of Cr(III) ion (5 equiv) and after subsequent addition of a strong metal chelator, TPEN (10 equiv). (c) Lippert-Mataga plot for the phosphorescence spectra of YJ1, the Cr(III)-bound form of YJ1, and Ir5F in DMSO, CHCl3, EtOAc, CH2Cl2, and CH3CN. Solvent polarity parameter (f) is defined as f = (ϵ−1)/(2ϵ+1)−(n2−1)/(2n2+1), where ϵ and n are dielectric constant and refractive index, respectively: DMSO (0.1352), CHCl3 (0.1481), EtOAc (0.1998), CH2Cl2 (0.2184), and CH3CN (0.3055). Inset photo is phosphorescence emission of the Cr(III)-bound YJ1 in various solvents. (d) Photoluminescence decay profiles of YJ1 (20 μM, CH3CN, deaerated) observed at 630 nm after 375 nm nanosecond laser excitation in the absence and presence of Cr(ClO4)3 (20 equiv) and after subsequent addition of TPEN (50 equiv). The phosphorescence ratiometric response of YJ1 is a consequence of an excited-state interaction between [Ir(dfppy)2phen]+ and BTTA. Actually, photoluminescence excitation (PLE) spectra of YJ1 in Cr(III)-free and -bound states are nearly identical, except tiny increase in the MLCT absorption band (410 − 500 nm; SI, Figure S7). In order to gain insight into a mechanism of the phosphorescence modulation upon Cr binding, we performed quantum chemical calculations based on density functional theory and time-dependent density functional theory (DFT/TD-DFT; uB3LYP/LANL2DZ:6-31+G(d,p)//uB3LYP/LANL2DZ:6-31+G(d,p)) for the triplet geometry of YJ1. The tertiary amino group of BTTA is quaternized by proton to mimic the electronic structure of a Cr-bound form of YJ1. Details on calculation results are listed in SI, Figure S8 and Table S1. The TD-DFT results obtained under the conductor-like polarizable continuum model (C-PCM, acetonitrile) indicate that the intraligand charge-transfer transition from BTTA to phen (ILphenCT) mainly constructs the lowest triplet state (T1 = 2.30 eV) in the metal-free YJ1, whereas the Ir(III)-to-phen ligand charge-transfer (MLphenCT) and the dfppy ligand-to-phen ligand charge-transfer (LdfppyLphenCT) transitions become dominant in the lowest triplet state (T1 = 1.60 eV) in the protonated form. Actually, similar phosphorescence ratiometric change of a 5-fold turn-on and a 42 nm red-shift is experimentally observed when HClO4 (10 equiv) is added to an YJ1 solution (10 μM, CH3CN; SI, Figure S9). Subsequent addition of Bu4NOH (20 equiv) restores the original phosphorescence (SI, Figure S9), further supporting that the Lewis basic tertiary amine of BTTA plays a key role in the phosphorescence ratiometric response. Specifically, the small red shift of 42 nm may be a consequence of +3 charge of the proximal [Cr(BTTA)] ionophore. In addition to this, Cr binding to the tertiary amine destabilizes the non-phosphorescent ILphenCT transition state, yielding the MLCT transition state being the lowest state at which an efficient phosphorescent transition occurs. Strong solvatochromism in the phosphorescence spectrum is therefore expected in the Cr-bound form due to the enhanced MLCT character.30-34 As anticipated, a Lippert-Mataga plot reveals prominent positive solvatochromism (−7670 cm−1) in the phosphorescence of the Cr-bound YJ1 (Figure 2c). In sharp contrast, the metal-free form of YJ1 (−403 cm−1) and Ir5F (−633 cm−1) do not display such strong solvatochromism. The metal-free form of YJ1 (20 μM, Ar-saturated CH3CN) exhibits a biphasic decay of the phosphorescence observed at 630 nm with time constants of 40 ns and 1.7 μs (Figure 2d). The fast decay component (i.e., 40 ns component) has been further identified by picosecond transient photoluminescence measurements (SI, Figure S10). Presence of the fast component may indicate occurrence of an emission-quenching process. For its origin we presume photoinduced electron transfer (PeT) from BTTA to Ir(IV) in the photoexcited state of YJ1. Actually, a large positive driving force for the PeT of −ΔGPeT = 1.35 eV is calculated through the Rehm-Weller equation by applying the reduction potential of [Ir(dfppy)2phen]+ (−1.39 V vs. SCE; SI, Figure S11), the oxidation potential of BTTA moiety (1.20 V vs. SCE; SI, Figure S11) and photoexcitation energy (3.94 eV).35, 36 Addition of Cr(III) ion (20 equiv) to the YJ1 solution clearly eliminates the fast decay component, yielding an apparent mono-exponential decay with a time constant of 1.5 μs (Figure 2d). Subsequent addition of TPEN (50 equiv) regenerates the fast component (29 ns), the result of which provides additional evidence for the reversibility for Cr binding. Taking the steady-state and time-resolved photophysical results into account, the reversible phosphorescence ratiometric response of YJ1 to Cr(III) ion can be rationalized by a consequence of enhanced MLCT contribution and suppression of PeT in the stabilized triplet state of the Cr-bound form. Air-equilibrated CH3CN solutions of the Cr-bound form of YJ1 display slow but complete conversion of the yellow phosphorescence (λems = 557 nm) into the green phosphorescence (λems = 515 nm; Figure 3a). Photoluminescence quantum yield (PLQY) of an Ar-saturated CH3CN solution of the green phosphorescence is 0.018 ± 0.004,37 which is comparable to the PLQY (0.016) of the Cr-free YJ1. The green phosphorescence observed at 630 nm follows a double-exponential decay with time constants of 40 ns and 0.92 μs which are also similar to those of the Cr-free form (SI, Figure S12). This second phosphorescence ratiometric response is spontaneous and the rate is proportional to temperature; increasing temperature from 0 °C to 50 °C results in a 14-fold enhancement in the observed rate (1/tobs) (SI, Figure S13). Applying the temperature-dependent rate constants into the Eyring-Polanyi equation yields enthalpy (ΔH‡) and entropy (ΔS‡) for the Cr-induced reaction to be 36.7 kJ mol−1 and −62.9 J K−1 mol−1, respectively. In order to figure out a mechanism underlying this conversion, we measured X-band CW EPR spectra of CH3CN solutions (1 mM) of fresh Cr(ClO4)3, the metal-free form of YJ1, a fresh mixture of Cr(ClO4)3 (1 equiv) and YJ1, and a mixture of Cr(ClO4)3 (1 equiv) and YJ1 which was left for 48 h at room temperature for full conversion. As shown in Figure 3b, a CH3CN solution containing free Cr(III) ion has two peaks at g values of 1.97 and 4.32, whereas a CH3CN solution of YJ1 is EPR-silent. The fresh mixture has EPR signals with g values of 1.98, 4.27 and 8.55, which indicates the Cr ion captured by YJ1 is high spin Cr(III). In contrast, no EPR signals are observed after the oxidative cleavage, suggesting that the oxidation state of the Cr ion changed to +2 or +4 and that a Cr-mediated reaction at the [Cr(BTTA)] ionophore occurred. Actually, a positive mode ESI MS spectrum of the 48 h-old mixture has three prominent peaks at m/z = 767.3, 781.3, and 797.3 (Figure 3d). These peaks correspond to [Ir(dfppy)2(phen-R)]+ with the R being methyl (calcd m/z = 767.1), formyl (calcd m/z = 781.1), and carboxylic acid (calcd m/z = 797.1) in place of BTTA (refer to inset structures in Figure 3d). We also observe a tiny peak at m/z = 256.7 that can be assigned to be a cleaved part of the [Cr(BTTA)] ionophore, [K2CrII(BTTA)(ClO4)]2+ (calcd m/z = 256.4). Observation of this Cr(II) species is in accordance with the EPR results (Figure 3b). In contrast, such oxidative cleavage does not occur in the presence of other transition metal ions such as Cu(II) as evidenced by the ESI MS spectra (SI, Figure S14). The oxidative cleavage at the benzylic position is also supported by an appearance of a formyl peak (δ = 10.63 ppm) in the 1H NMR spectrum of the 72 h-old mixture (1 mM, CD3CN; SI, Figure S15). The yield of the oxidation reaction is quantitative as revealed by 1H NMR. Collectively, the data strongly indicate occurrence of an oxidative cleavage of [Cr(BTTA)]. The Cr center is critical to the oxidation because protonated YJ1 does not undergo such an oxidation reaction (SI, Figure S16). In addition, since phosphorescence spectrum of Ir5F that lacks BTTA is almost identical to that of the oxidized products, the second stage phosphorescence ratiometric response can be reasonably ascribed to the Cr-promoted oxidative cleavage. It should be underscored that any additional oxidants and base are not required for the cleavage. Therefore, the reaction mechanism for the oxidative cleavage may involve activation of molecular oxygen by the Cr center. Actually, the oxidative cleavage did not occur when a CH3CN solution of Cr-bound YJ1 (10 mM) was left for 24 h under anaerobic conditions (SI, Figure S17). Such molecular oxygen activation and subsequent oxidation are frequently found in biological systems, representative examples of which are methanol production by methane monooxygenase38 and hydroxylation of taurine by taurine α-glutarate dioxygenase (TauD).39 Similar reactions are also found in biomimetic systems,40 such as generation of high-valent Fe(IV)(O) species and their catalytic oxidation of substrates.41 Based on the spectroscopic evidences described above, we propose a mechanism for our Cr-mediated oxidative cleavage, which is depicted in Figure 3c.42 Biomimetic phosphorescence ratiometric response to Cr(III) ions. (a) Phosphorescence spectra of YJ1 (10 μM, CH3CN) in the absence and presence of Cr(ClO4)3 (1 equiv) and after full oxidative cleavage (24 h). (b) X-band CW EPR spectra of Cr(ClO4)3 (1 mM), YJ1 (1 mM), a freshly prepared mixture of YJ1 (1 mM) and Cr(ClO4)3 (1 mM), and a mixture of YJ1 (1 mM) and Cr(ClO4)3 (1 mM) incubated for 48 h allowing for full oxidative cleavage. Refer to Experimental for measurement conditions. (c) Proposed mechanism of the oxidative cleavage of YJ1. Species in the square brackets are proposed intermediates. The appearance of the methyl-terminated product is yet to be explained by this mechanism and remains elusive. There is also a possibility for free-radical autooxidation. (d) Full ESI MS spectrum of the crude product of the Cr(III)-mediated oxidative cleavage of YJ1. Peaks corresponding to oxidized products and the cleaved Cr(III) ionophore are magnified and compared with theoretical isotopic distributions (black bars). Marked peaks at 573.4 (*), 613.9 (**), and 654.8 (***) corresponds to [Ir(dfppy)2]+ (calcd m/z = 573.1), [Ir(dfppy)2·(CH3CN)]+ (calcd m/z = 614.1), and [Ir(dfppy)2·(CH3CN)2]+ (calcd m/z = 655.1), respectively. The double-stage phosphorescence ratiometric response of YJ1 allows for detection of Cr(III) ion with excellent accuracy.43 We have acquired phosphorescence spectra of YJ1 (10 μM, CH3CN) in the presence of various metal ions (1 equiv), after subsequent addition of Cr(ClO4)3 (1 equiv) into the mixture, and after additional 12 h at room temperature (Figures 4a−c). Each of the phosphorescence spectra can be characterized by a phosphorescence intensity ratio of 557 nm vs. 520 nm (I557/I520) and integrated phosphorescence intensity. As shown in Figure 4a, Na(I), Ca(II), Mn(II), Co(II), Ni(II), Zn(II), Ag(I), Cd(II), Hg(II) ions exert almost no effect on the I557/I520 and integrated phosphorescence intensity values, whereas Cu(II), Pd(II), and Pb(II) ions influence the values. Subsequent addition of Cr(ClO4)3 shifts the points to the top right corner of the map plotting I557/I520 and integrated phosphorescence intensity, indicating that metal ions bound to the BTTA ligand were displaced by Cr(III) ion except Hg(II). The points that correspond to Fe(III) and Hg(II) ions experience small but observable changes.44 The Cr(III)-induced oxidative cleavage finally generates the second phosphorescence ratiometric response except Fe(III) ion. The double-stage phosphorescence ratiometric response can be quantified by defining a fill factor (FF, FF = a phosphorescence intensity ratio of I557/I520 × integrated phosphorescence intensity). As shown in Figure 4d, the fill factors are not affected by Na(I), Ca(II), Mn(II), Co(II), Ni(II), Zn(II), Ag(I), and Cd(II) ions. Fe(III) and Hg(II) ions seem to bind stronger than Cr(III) ion. Noticeable results are Cu(II), Pd(II) and Pb(II) ions, which evoke an increase in fill factors similar to Cr(III) ions. Subsequent oxidative cleavage specific to Cr(III) clearly allows one can discriminate presence of Cr(III) ion over these metal ions by observation of the FF changes being increased and then decreased. In particular, signal specificity toward Cr(III) ion over Cu(II) is worthwhile because Cu(II) ion usually binds most strongly in usual cases. Double-stage phosphorescence ratiometric response of YJ1 for Cr(III) ions. Plot of phosphorescence intensity ratios (I557 nm/I520 nm) vs integrated phosphorescence intensities of YJ1 (10 μM, CH3CN) in the presence of various metal ions (1 equiv) (a), after subsequent addition of Cr(III) ion (1 equiv) into the mixture (b), and after additional 12 h (c). (d) Corresponding fill factor (FF, FF = phosphorescence intensity ratio × integrated phosphorescence intensity) of YJ1. (e) Photo showing Cr(III) ion-selective double-stage phosphorescence ratiometric response of YJ1 (10 μM, CH3CN). To summarize, we have developed a novel strategy to detect Cr(III) ion selectively. The new phosphorescent sensor containing the sulfur-rich BTTA receptor exhibits the double phosphorescence ratiometric response to Cr(III) ion. Steady-state and time-resolved photophysical experiments have established a mechanism for the first reversible phosphorescence change which involves modulation of PeT and the ILCT transition state. In the Cr-bound state, the Cr(III) center evokes a biomimetic oxidation reaction by activating molecular oxygen to cleave the [Cr(BTTA)] ionophore. This biomimetic oxidative cleavage produces the second phosphorescence ratiometric response. Taking the double-stage phosphorescence ratiometric response, the probe successfully serves to discriminate Cr(III) ion among divalent transition metal ions such as Cu(II). The novel sensing strategy developed in this work can be extended to future photoluminescence sensors targeting oxophilic transition metal ions. Despite limited compatibility in aqueous environments, we envision that our strategy would provide the first and valuable guidance to the development of a new sensing methodology for transition metal ions. Supporting Information is available from the Wiley online library or from the author. This research was financially supported by National Research Foundation (NRF) of Korea funded by the Ministry of Education, Science and Technology (MEST) through the CRI, GRL (2010-00353), and WCU program (R31-2008-000-10010-0) and by Ewha Womans University (RP-Grant 2010). Y.Y. acknowledge Prof. Soo Young Park at Seoul National University and Prof. Joan S. Valentine at Ewha Womans University for use of a TCSPC system and helpful comments. Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.