Molecular Cursor Caliper: A Fluorescent Sensor for Dicarboxylate

5 hours ago - We report here the fluorescent sensing of both aromatic and linear saturated dicarboxylate anions (DC2–) (as their tetrabutylammonium ...
0 downloads 0 Views 12MB Size
Subscriber access provided by Nottingham Trent University

Article

Molecular Cursor Caliper: A Fluorescent Sensor for Dicarboxylate Dianions Wei Chen, Chenxing Guo, Qing He, Xiaodong Chi, Vincent M Lynch, Zhiyun Zhang, Jianhua Su, He Tian, and Jonathan L. Sessler J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b07170 • Publication Date (Web): 22 Aug 2019 Downloaded from pubs.acs.org on August 22, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10

Molecular Cursor Caliper: A Fluorescent Sensor for Dicarboxylate Dianions †,‡,

†,









Wei Chen, ⊥ Chenxing Guo, ⊥ Qing He, Xiaodong Chi, Vincent M. Lynch, Zhiyun Zhang, Jianhua ‡ ,‡ ,† Su, He Tian,* and Jonathan L. Sessler* †

Department of Chemistry, The University of Texas at Austin, 105 E. 24th Street, Stop A5300, Austin, TX 78712-1224, USA



Key Laboratory for Advanced Materials and Joint International Research Laboratory of Precision Chemistry and Molecular Engineering, Feringa Nobel Prize Scientist Joint Research Center, East China University of Science & Technology, Shanghai 200237, P. R. China

ABSTRACT: We report here the fluorescent sensing of both aromatic and linear saturated dicarboxylate anions (DC2- ) (as their tetrabutylammonium salts) with different lengths and shapes in acetonitrile using a single fluorescent probe, i.e., the bis-calix[4]pyrrole-appended 9,14-diphenyl-9,14-dihydrodibenzo[a,c]phenazine (DPAC-bisC4P) incorporating a vibration-induced emission (VIE) phenazine core. Fluorescence titration studies revealed that treating DPAC-bisC4P with dicarboxylate guests capable of forming pseudo-macrocyclic host–guest complexes via multiple hydrogen bonding interactions between the dicarboxylates and calix[4]pyrrole moieties, led to a blue-shift in the emission of the phenazine core. The binding-based fluorescence tuning features of DPACbisC4P allows the underlying binding events and inferred structural changes to be monitored in the form of different chromaticity outputs. The analyte-induced differences in the fluorescence response to DC2- cover a wide range within the chromaticity diagram and can be visualized readily. The present system thus functions as a rudimentary dicarboxylate anion sensor. It highlights the potential benefits associated with combining a tunable VIE core with non-covalent binding interactions and thus sets the stage for the development of new fluorescent chemosensors where a single chemical entity responds to different analytes with a high level of tunability.

(a)

Analyte A Analyte B Analyte C

...

Single VIE-active Fluorescent Probe

...

...

Output

Fluorescent sensing plays a vital role in the detection of chemical and biological analytes. It is a technique that is favored for its high sensitivity, real-time readout capability, design flexibility, and potential for tunability.1 Continued efforts are being made to exploit various photophysical phenomena in the creation of fluorescent sensors and to improve their sensitivity.2 In the context of this effort, so-called ratiometric fluorescent probes have attracted considerable attention since their dual (or multiple) emission features provide an internal reference, which improves the sensitivity and abets quantitative analysis.3 However, the analyte-dependent tunability of most known ratiometric fluorescent chemosensors is limited in terms of the resulting chromaticity output. Expanding the differences in fluorescence emission produced in response to different analytes (cf. Figure 1a) might allow closely related analytes to be more easily distinguished.4 Unfortunately, achieving such a goal remains challenging. Here we report a vibration-induced emission (VIE)-based fluorescent sensor system, namely the bis-calix[4]pyrrole-appended 9,14-diphenyl-9,14-dihydrodibenzo[a,c]phenazine (DPAC-bisC4P), that allows for dicarboxylate anion (DC2- ) sensing in acetonitrile. As detailed below, this system is capable of discriminating a variety of individual dicarboxylate guests (as their tetrabutylammonium salts); it does so as the result of each DPAC-bisC4P⊃DC2- host–guest complex producing a unique chromaticity readout. Furthermore,

the collection of these individual readouts covers a wide range (from blue to orange) within the Commission Internationale de L’Eclairage (CIE) chromaticity diagram.5

Input

1. INTRODUCTION

C te aly An B e yt al An eA yt al An

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Different Chromaticities

(b)

Figure 1. (a) Our approach to realizing analyte-dependent differential chromaticity outputs using a VIE-active fluorescent chemosensor. (b) Mechanism of proposed fluorescence tuning and dicarboxylate anion sensing.

Vibration-induced emission (VIE) is a recently reported fluorescence phenomena wherein the reversible bent-to-planar motion of two aromatic moieties is used to produce multiple emissions from a single chromophore system.6 One VIE chromophore of particular interest is 9,14-diphenyl-9,14-dihydrodibenzo[a,c]phenazine (DPAC), which was recently reported by one of our groups.6b It has been found that covalent and metal-

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

based coordination strategies are effective in fine-tuning the excited-state conformations of DPACs, allowing a variety of colorful fluorescent images to be produced.7 The emission color change seen for these VIE probes is ascribed to alterations in the molecular configuration, rather than to a difference in the underlying recognition mode. As a consequence, we felt that VIE-based approaches could be exploited to create anion receptor-based fluorescent sensors. To the extent this proved true, it could lead to new single-luminophore ratiometric sensors for what is an important class of analytes. Such a strategy could also provide a useful complement to the recent report of coordination-based VIE sensors for selected dianions.7b In fact, a greater range of readouts was obtained for the present anionresponsive system than for previous VIE probes. The basic strategy used to create our DPAC-based, anionrecognition VIE sensor system (DPAC-bisC4P) is shown in Figure 1. We chose calix[4]pyrroles as the anion receptor subunits. Calix[4]pyrroles per se possess high affinities toward carboxylate anions in organic media, a property that has been exploited to create simple fluorescent chemosensors and stimuliresponsive polymeric materials.8 In the present study, dicarboxylate anions (DC2-) were chosen as the targeted guests. These dianions are ubiquitous metabolites (e.g., succinate, glutarate, etc.) and are key enzyme regulators.9 Not surprisingly, therefore, considerable effort has been devoted to developing dicarboxylate sensors.10 However, new systems, particularly those that rely on new mechanisms, remain of interest. As detailed below, we have now prepared a first supramolecular, anion bindingbased system where VIE tuning is used to differentiate closely related DC2– analytes via production of a unique fluorescent signature; this allows discrimination to be achieved by simple visualization using the unaided eye. (a) N

POCl3

N

R

N

N

CHO

DMF

R

N

O COOH

EDCI, DMAP

NH HN

O R

NH HN

N

N

NH HN

DCM

C4P-COOH

DPAC-bisC4P, R =

–H2COOC

DPAC-monoC4P, R = –H

(b)

N H

N

NH HN NH HN

Crystal Structure:

O O

CH2OH

DPAC-bisOH, R = –CH2OH DPAC-monoOH, R = –H

DPAC-bisCHO, R = –CHO DPAC-monoCHO, R = –H

NH HN

N

MeOH

DPAC

+

NaBH4

N

N

H

H

N H

Putative DC2– Binding Site

N H N

O

H

H

N

N

H N

O

DPAC-bisC4P (c)

Aliphatic DC2–: O

C22– (n = 0),

C102– O

n

Aromatic DC2–:

O

O

O

O

O

C42– – C82– (n = 2 – 6),

O O

(n = 8),

C122– (n = 10)

O O

O O

O

p-Ph2–

O

m-Ph2–

O

o-Ph2–

Figure 2. (a) Synthesis of the receptor DPAC-bisC4P and the control compound DPAC-monoC4P. (b) Crystal structure of DPACbisC4P (most hydrogen atoms and solvent molecules are omitted for clarity). (c) Dicarboxylate anionic guests (DC2- ) considered in

Page 2 of 10

this study. Unless noted otherwise, all DC2- were studied in the form of their corresponding tetrabutylammonium (TBA+) salts.

2. RESULT AND DISCUSSION 2.1 Synthesis and Characterization. The syntheses of DPAC-bisC4P and the corresponding monofunctionalized control system, DPAC-monoC4P, are shown in Figure 2a. Both syntheses started with Vilsmeier-Haack formylation of DPAC. Functionalization was found to take place predominantly at the para positions of the 9,14-diphenyl groups to give DPAC-bis/monoCHO. The formyl groups were reduced by sodium borohydride to yield the key precursors, DPACbis/monoOH. DPAC-bisC4P and DPAC-monoC4P were then obtained by esterification of these alcohol intermediates with C4P-COOH using EDCI/DMAP. Separately, the tetrabutylammonium (TBA+) salts of the dianions considered in this study were prepared by adding two equivalents of TBAOH to the corresponding dicarboxylic acids in accord with literature procedures.11 All intermediates and final products were fully characterized by 1H NMR and 13C NMR spectroscopy, as well as by high resolution mass spectrometry (HRMS) (see the Supporting Information for more details). The molecular structures of DPAC-bis/monoCHO and DPAC-bisC4P were further confirmed by means of single crystal X-ray diffraction analyses (cf. Figure 2b and Figures S34–S36). 2.2 Response to Linear Saturated Dicarboxylates. The binding properties of DPAC-bisC4P for DC2- were initially investigated by subjecting the DPAC-bisC4P host to UV–vis spectroscopic titrations with various aliphatic dicarboxylate anionic guests (as their respective tetrabutylammonium salts, TBA2(CH2)n(CO2)2, cf. Figure 2c) in acetonitrile. Acetonitrile was chosen as the solvent since it allowed for solubilization of the full range of species considered in the present study. In the absence of a guest, the absorption maximum for DPAC-bisC4P is seen at ca. 350 nm (Figure S1). This key feature is ascribed to the DPAC chromophore rather than the C4P moieties, which display little in the way of appreciable absorption in the UV– vis spectral region. No discernible absorption features were observed above 400 nm. These results are consistent with previous studies of DPAC systems.7a Upon the addition of linear saturated dicarboxylate guests, little in the way of change in the absorption band was observed, leading us to suggest that the DC2guests have a minimal effect on the ground state optical features of DPAC. In contrast to the above, the emission spectra of DPACbisC4P showed remarkable changes upon the addition of DC2- . The associated changes were studied in detail in the case of the succinate dianion (C42- ) since it is a pivotal metabolite in many organisms.12 As can be seen from inspection of Figure 3a, UV irradiation (lex = 350 nm) of DPAC-bisC4P in acetonitrile gives rise to a major emission feature at 600 nm, along with a very minor band around 495 nm. Upon the addition of C42- , the intensity of the 600 nm emission band decreases while that of the band at 495 nm increases. These changes continued to the point where roughly 4 equiv of C42- are added, at which point they plateau. The increase in the higher energy band outweighs the decrease in the lower energy band such that there is a net enhancement in the emission intensity.

ACS Paragon Plus Environment

Page 3 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Figure 3. (a) Fluorescence emission spectral changes seen upon the addition of increasing quantities of C42- to a solution of DPAC-bisC4P (lex: 350 nm; [DPAC-bisC4P]: 10 µM; solvent: acetonitrile). (b) Chromaticity coordinates (CIE) of DPAC-bisC4P in acetonitrile for solutions containing various quantities of C42- (from 0 to 4.0 equiv). Inset: photos of acetonitrile solutions of DPAC-bisC4P containing 0 and 4.0 equiv of C42- under irradiation with a 365 nm UV light. (c) Normalized fluorescence emission spectra of DPAC-bisC4P in the presence (as well as absence) of various linear saturated dicarboxylates at the end of titrations. Solvent: acetonitrile; [DPAC-bisC4P]: 10 µM; lex: 350 nm. (d) Chromaticity coordinates (CIE) of the host–guest complexes generated from DPAC-bisC4P and DC2- (viz. C22-, C42-–C82-, C102-, and C122- ) along with DPAC-bisC4P in acetonitrile. Inset: Fluorescence images of DPAC-bisC4P⊃DC2- and DPAC-bisC4P recorded in acetonitrile upon irradiation with a 365 nm UV light.

By considering the I495/I600 fluorescence intensity ratio, the effect of adding C42- could be quantified in a ratiometric manner. A linear correlation with the C42- concentration over the 0– 5 µM range was seen. This allowed a limit of detection (LOD) of 34.4 nM to be determined (Figure S21). Each emission spectrum was also converted to its corresponding spot on the Commission Internationale de L’Eclairage (CIE) chromaticity diagram (cf. Figure 3b). Upon adding C42to DPAC-bisC4P in acetonitrile, the chromaticity moved in a straight trajectory from the orange-red region to the blue-green region. The emission colors recorded in the presence of 0, 0.6, 0.9, 1.2, and 4.0 equiv of C42- were orange-red (0.49, 0.42), pale yellow (0.37, 0.41), coral reef white (0.31, 0.40), pale green (0.28, 0.39), and aquamarine (0.21, 0.38), respectively.13 Probe DPAC-bisC4P was also found to produce a differential fluorescence response when exposed to other linear saturated dicarboxylates (cf. Figure S3c, d). In the case of oxalate (C22- ), the intensity of the higher energy peak increased during the titration. Its maximum was also blue-shifted 486 nm. Moreover, an obvious trend towards a red-shifted high energy emission was seen as the length of the chain increased: Glutarate (C52- ) 500 nm; adipate (C62- ) 525 nm; heptanedioate (C72- ) 560 nm; octanedioate (C82- ) 563 and 584 nm, and decanedioate (C102- ) 590 nm. In the case of dodecanedioate (C122- ), the red

shift was such that the major orange-red emission feature at 600 nm seen in the presence of the dianion was barely discernible from what was seen in its absence. These latter modest changes did not detract from an overall negative correlation between the I495/I600 ratio and the chain length of the alkyl dicarboxylates from C42– to C122– (cf. Figure S3). In addition, the change in emission intensity ratios or absolute intensity allowed the concentrations of DC2- to be quantified. The LOD values for C22–, C52-, C62–, C72–, and C82- were calculated to be 93.9, 18.3, 14.5, 60.1, and 79.4 nM, respectively, in acetonitrile (cf. Figures S20 and S22–S25). To test whether DPAC-bisC4P could be used to differentiate various linear alkyl dianions, collectively DC2- , the emission spectra at the end of each of the above titrations were integrated and normalized. As shown in Figure 3d, the nature of the final emission spectrum in question was dictated by the length of the alkyl chain. On passing from the shortest dianion, C22-, to the longest, C122-, the emission colors switched from downy white (0.22, 0.32), to aquamarine (0.21, 0.38), feijoa (0.23, 0.41), primrose (0.30, 0.44), goldenrod (0.37, 0.46), sweet corn (0.40, 0.46), sea buckthorn (0.43, 0.43), and orange (0.46, 0.42), successively. In each case, a unique and distinct guest-dependent fluorescent signature was produced. This allowed a distinction to be made between ostensibly similar analytes.

ACS Paragon Plus Environment

Journal of the American Chemical Society 2.3 Studies of Response Mechanism, Binding Stoichiometry, and Speciation. As a working hypothesis, we considered that the observed DC2- -induced tuning of the emissive features of DPAC-bisC4P could be explained in terms of mechanism analogous to that invoked in the case of the earlier non-supramolecular DPAC fluorophore systems.6 Specifically, we suggest that in the absence of the DC2- guests, the DPAC fluorophore embedded within the DPAC-bisC4P receptor is relatively free of constraint. The excited state would thus be free to undergo vibration-induced planarization. Such putative planarization would lead to effective elongation of the p-conjugation pathway, thus giving rise to the observed orange-red emission. In contrast, upon the addition of an appropriately sized DC2guest, a pseudo-macrocyclic complex, DPAC-bisC4P⊃DC2- , would be formed as the result of calix[4]pyrrole-mediated dianion recognition. Formation of this non-covalently linked host– guest complex would, in turn, impose various degrees of constraint on the DPAC core thus inhibiting its planarization in the excited state. Depending on the exact nature of the complex formed, different energy levels for the global minima in the S1 excited state would be expected, giving rise to the observed analyte-specific emission features (cf. Figure 1b). We thus propose that DPAC-bisC4P may be considered as a molecular cursor caliper that depending on the extent of DC2- -induced opening gives rise to a different fluorescent response.

recorded upon the addition of AcO- and C42- , respectively (cf. Figure 4). In contrast to the dramatic fluorescence changes produced when DPAC-bisC4P was treated with C42- , the fluorescence emission spectra in the case of control studies were not appreciably changed over the course of titration. These findings are ascribed to the fact that both a homoditopic host and dianionic guest is required to produce the pseudo-macrocyclic complex. Most systems involving saturated linear dicarboxylates except for C22- and C42- were well-behaved. For instance, both the associated Job plot analysis and binding isotherms were consistent with 1:1 binding (cf. Figure S14). The binding constants (K1:1), determined by means of fluorescence spectral titration curve fitting, are summarized in Table 1 (cf. Figures S7–S12 in the Supporting Information for details of the fitting).17 When DPAC-bisC4P was titrated with the smallest DC2guest (e.g., C22–) a Job plot analysis proved consistent with formation of a 1:2 complex (cf. Figure S13). The situation with the next largest guest studied, C42- proved more nuanced. Although clear isosbestic behavior was seen over the course of the titration and Job plots consistent with formation of a 1:1 host– guest complex between DPAC-bisC4P and C42- (cf. Figures 3a and 5 inset), the binding isotherm proved sigmoidal, as would be expected for a 1:2 binding event. Recent studies have revealed the limitation of Job plots in supramolecular systems.14 In fact, the use of a 1:1 model gave a poor fit to the spectroscopic data. In contrast, a good fit was obtained using a 1:2 binding model as shown in Figure 5. [C42–]/[DPAC-bisC4P] 0.0 1.0

0.8

1.6

2.4

3.2

4.0

4.8

5.6

K1:1 = (5.3 ± 3.6) × 105 M–1 K1:2 = (2.1 ± 0.9) × 106 M–1

0.8

(I-I0)/(IF-I0)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 10

R2 = 0.9990

0.6

0.4

0.2

0.0 0.0

8.0x10-6 1.6x10-5 2.4x10-5 3.2x10-5 4.0x10-5 4.8x10-5 5.6x10-5

[C42–]

Figure 4. (a) Fluorescence spectroscopic titration of DPACbisC4P with AcO-. lex = 350 nm; [DPAC-bisC4P] = 10 µM; Solvent: acetonitrile. (b) Fluorescence spectroscopic titration of DPAC-monoC4P with C42-. lex: 350 nm; [DPAC-monoC4P]: 10 µM; Solvent: acetonitrile.

To gain additional insight into the proposed complexationinduced change in the emission features of the DPAC core present in DPAC-bisC4P, we performed two control experiments wherein the fluorescence emission spectra of receptor DPACbisC4P and the control compound, DPAC-monoC4P, were

Figure 5. Least-squares nonlinear fitting of the normalized change in fluorescence at 495 nm obtained as a function of concentration on the basis of the fluorescence titration of DPAC-bisC4P with C42-. The lower x axis gives the absolute concentration of the dianion, whereas the upper x axis shows the molar ratio of the dianion guest relative to that of the DPAC-bisC4P host. [DPAC-bisC4P] = 10 µM; solvent: acetonitrile. Inset: Job plot analysis corresponding to the formation the host–guest complex between DPACbisC4P and C42-. The ∆I values were determined at 495 nm. The total concentration of host (i.e., DPAC-bisC4P) and guest (i.e., C42–) was held constant at 1.0 × 10−5 M; solvent: acetonitrile.

The stepwise binding constants between DPAC-bisC4P and C42-, K1:1 and K1:2, were determined to be (5.3 ± 3.6) × 105 M–1 and (2.1 ± 0.9) × 106 M–1, respectively, using the OriginPro

ACS Paragon Plus Environment

Page 5 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

software following a standard fitting protocol for formation of a 1:2 host–guest complex.15 We also used the supramolecular.org web applet to fit the original data. A similar result was obtained where the stepwise binding constants, K1:1’ and K1:2', were determined to be (4.9 ± 1.3) × 105 M–1 and (2.2 ± 0.2) × 106 M–1, respectively (cf. Figure S4).16 The small discrepancy seen using these two binding models, which gave largely concordant values, is ascribed to the difference in the underlying fitting algorithms and corresponding error determinations. The fitting for C22- was performed in a similar manner, as shown in Figures S5 and S6. Table 1. Binding constants and limits of detection (LODs) for receptor DPAC-bisC4P and anionic dicarboxylate guests in acetonitrile at 293 K. anionsa

binding constantb (M-1)

limit of detection (LOD) (nM)

C22-

K1:1 = (1.8 ± 0.5) × 106 K1:2 = (1.2 ± 0.2) × 105

93.9

C42-

K1:1 = (5.3 ± 3.6) × 105 K1:2 = (2.1 ± 0.9) × 106

34.4

C52-

K1:1 = (4.6 ± 0.8) × 106

18.4

C6

2-

6

14.2

C7

2-

7

K1:1 = (5.0 ± 1.7) × 10

60.2

C8

2-

K1:1 = (1.9 ± 0.4) × 107

79.5

C102-

K1:1 = (3.1 ± 0.8) × 106

N.D.c

C122-

K1:1 = (1.8 ± 0.4) × 105

N.D.

2-

6

K1:1 = (8.3 ± 2.4) × 10

21.0

2-

K1:1 = (5.7 ± 1.4) × 107

7.55

p-Ph

m-Ph

K1:1 = (8.2 ± 0.6) × 10

a The countercation for the DC2- guests was tetrabutylammonium (TBA+) in all cases. bIn the case of C22– and C42–, the calculated stepwise binding constants (K1:1 and K1:2) are listed. cNot determined.

On the basis of the determined binding constants, we believe that, upon the addition of C22- or C42-, a pseudo-macrocycle resembling that shown in cartoon form in Figure 1b is still formed. Although the formation of a 1:2 host–guest complex was expected to form at higher Cn2- (n = 2 or 4) to DPACbisC4P ratios, even under these conditions a strong fluorescent response was still seen, leading us to conclude that an appreciable quantity of the proposed pseudo-macrocycle (or possibly another connection mode that gives rise to a similar pseudomacrocyclic structure) is formed and that the associated structural changes allows operational sensing of C22- and C42-. The binding stoichiometry for DPAC-bisC4P and AcO- was found to be 1:2 as inferred from a fluorescence spectroscopicbased Job plot analysis. This provides further support for the conclusions drawn above that this species will not produce a substantial effect on the excited state emissive features of the

DPAC core present in DPAC-bisC4P as seen by experiment. Moreover, a clear distinction between the fits to a 1:2 and 1:1 binding stoichiometry seen for AcO- and all DC2- guests other than C22- and C42- provides further confidence that the 1:1 nature of the latter complexes is correctly assigned. 2.4 Response to three phthalate isomers. We also tested the fluorescent response of DPAC-bisC4P to three constitutional isomers, i.e., terephthalate (p-Ph2- ), isophthalate (mPh2- ), and phthalate (o-Ph2- ) (cf. Figures 6 and S17). A distinct fluorescent response was seen upon the addition of p-Ph2-. In this case, the intensity of the emission band at 600 nm dramatically decreased while a relatively broad emission band in the blue-green region with local maxima at both 428 nm and 480 nm was seen to grow in. The entire process can also be visualized by the unaided eye as a change in the emission colors from orange-red to light blue. It is notable that the trajectory of this color change passes through the white-light region as reflected in CIE coordinates of (0.33, 0.32) being recorded when the number of p-Ph2- equiv reached 0.35 (cf. Figure 6a, b and Figure S17a). Moreover, higher energy transitions in the blue-light region were seen for DPAC-bisC4P⊃p-Ph2- (i.e., 428 nm and 480 nm) than for the DPAC-bisC4P⊃C42- complex (i.e., 495 nm). This latter difference is ascribed to the higher rigidity of the phenyl ring in p-Ph2- , which is expected to impose a greater level of constraint on the DPAC core upon complex formation than provided by the relatively flexible aliphatic chain of C42- . Addition of m-Ph2- to an acetonitrile solution of DPACbisC4P gave rise to a different pattern. Here, we observed less spectral broadening and a relatively greater amplification of blue-light emission peaked at 485 nm coupled with a reduced decrease in the intensity of the orange-red emission band (cf. Figure 6c, d). As true for most other dianion studies, a Job plot analysis (cf. Figure S16a, b) provided support for the expectation that 1:1 pseudo-macrocyclic complexes were being formed between DPAC-bisC4P and both p-Ph2- and m-Ph2-. The two carboxylate groups are closer in m-Ph2- than in p-Ph2- . Therefore, the complexation of m-Ph2- by DPAC-bisC4P is likely to induce a greater level of constraint to the DPAC fluorophore than p-Ph2- and DPAC-bisC4P. In contrast, greater flexibility should persist in the case of the DPAC-bisC4P⊃p-Ph2- complex, which accounts for the broader band seen in the blue-light spectral region. In other words, these two isomeric substrates are expected to have a different structural effect on the DPACbisC4P molecular caliper, which in turn gives rise to a different fluorescent response, as indeed seen by experiment. In stark contrast to p-Ph2- and m-Ph2- , little in the way of emission spectral changes were seen when o-Ph2- was added to acetonitrile solutions of DPAC-bisC4P (cf. Figure 6e, f and Figure S17c). We attributed this result to o-Ph2- being unable to support formation of a pseudo-macrocyclic structure. Indeed, in the case of the latter dianion, a Job plot analysis proved consistent with the formation of 1:2 (host:dianion) complex (cf. Figure S16c).

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

p-Ph2–

Page 6 of 10

m-Ph2–

o-Ph2–

Figure 6. Fluorescence titration of DPAC-bisC4P upon the addition of (a) p-Ph2-, (c) m-Ph2-, and (e) o-Ph2- associated with the photographs showing the fluorescence features of DPAC-bisC4P with various amounts of (b) p-Ph2-, (d) m-Ph2-, and (f) o-Ph2- under irradiation with a 365 nm UV light. lex: 350 nm; [DPAC-bisC4P]: 10 µM; Solvent: acetonitrile.

The binding constants of receptor DPAC-bisC4P with pPh2- and m-Ph2- were determined by standard spectral titration methods and listed in Table 1 (see Figures S18 and S19 in the Supporting Information for details of the fitting). Since similar linear correlations between the change in fluorescence and the p-Ph2- and m-Ph2- concentrations were observed, the limits of detection (LODs) of DPAC-bisC4P for p-Ph2- and m-Ph2could be determined as 21.0 nM and 7.57 nM, respectively (cf. Figures S26 and S27). Again, the distinct chromaticity readouts could be used to differentiate these three isomeric phthalate dianions. 2.5 Computational modeling studies. Unfortunately, all attempts to obtain single crystals of host–guest complexes proved unsuccessful. Therefore, a series of computation studies were carried out so as to gain insights into the binding events. Due to the high degree of freedom expected for these host–guest complexes, all computations were performed using semi-empirical methods at the PM6 level.17 Using an approach analogous to that of Cafeo et al.,18 all putative structures were refined initially by molecular mechanics (MMFF), followed by a semi-empirical (PM6) treatment. The optimized structures are shown in Figure 7 and Figures S28– S33. Due to the flexibility of the chains, the linear saturated DC2- guests were modeled using their most stable conformations in the absence of the DPAC-bisC4P host (i.e., their zigzag conformations) as the initial input for the respective host–guest structural optimization. The distance between the two C4P moieties, i.e., dC4P–C4P, is determined by the distance between the centroids of two calix[4]pyrrole subunits, wherein each centroid is defined by the four pyrrolic nitrogen atoms. In addition, as defined in Figure 7a, three angles were used to reflect the constraint of the host–guest complexes: Θb = ∠ C1−N1−N2−C2, ΘS1 = ∠C3−N1−N2, and ΘS2 = ∠C4−N2−N1. The values of Θb, ΘS1, and ΘS2, as well as dC4P–C4P, for all the DPAC-bisC4P⊃DC2- complexes are listed in Table S1. Although only minimal changes in the values of Θb, ΘS1, and ΘS2 were observed, the distance between the two C4P moieties

(dC4P–C4P) was found to vary monotonically as a function of the length of the DC2- guest. The effect of this displacement is expected to become more profound in the S1 excited state (from whence the observed emissive features are expected to derive) as the result of planarization of the DPAC fluorophore. As a result, the longer the DC2- guest, the greater the degree of DPAC planarization and the larger the Stokes shift. Such an interpretation is in accord with the recent chelation-based study of Zhou et al.7b It also matches well our experimental findings. (a)

(b) C1 C4 N2 N1 C2 C3

8.2 Å (c)

13.7 Å (d)

10.4 Å

10.5 Å

Figure 7. Optimized structures of the host–guest complexes at semi-empirical (PM6) level of the theory (a) DPACbisC4P⊃C42-. (b) DPAC-bisC4P⊃C82-. (c) DPAC-bisC4P⊃mPh2-. (d) DPAC-bisC4P⊃p-Ph2-. The distance between the two calix[4]pyrrole moieties (dC4P–C4P) is provided for each optimized structure. Most non-polar hydrogen atoms are omitted for clarity.

ACS Paragon Plus Environment

Page 7 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society (a)

9.7 Å

Θb = 147º; ΘS1 = 122º; ΘS2 = 113º

(b)

12.9 Å

Θb = 171º; ΘS1 = 165º; ΘS2 = 177º

(c)

8.2 Å

Θb = 161º; ΘS1 = 143º; ΘS2 = 143º

Figure 8. Optimized ground-state structure (S0) of (a) DPAC-bisC4P and first excited-state structures (S1) of (b) DPAC-bisC4P and (c) DPAC-bisC4P⊃C42- calculated at the PM6 semi-empirical level.

To gain further insights into the structure–property correlation for the observed photophysical behavior, we optimized the first excited-state structures (S1) of DPAC-bisC4P and DPACbisC4P⊃C42- by means of time-dependent semi-empirical method at the PM6 level (TD-PM6).19 In addition, the groundstate structure (S0) of receptor DPAC-bisC4P was also optimized at the same level for a direct comparison. As shown in Figure 8, the excited-state structure of DPAC-bisC4P shows a high degree of planarization as the values of Θb, ΘS1, and ΘS2 all being close to 180°. As a result of such planarization, the distance between two calix[4]pyrrole moieties (dC4P–C4P) also increases. It does so in a substrate-dependent manner, as would be expected given the fluorescence spectral differences seen for the various aliphatic dianions tested in this study. On the other hand, the ground-state structure of receptor DPAC-bisC4P is consistent with the saddled-shape seen for the native DPAC core. 2.6 Differentiating dicarboxylates. As elaborated above, DPAC-bisC4P gave rise to different fluorescent responses when exposed to linear saturated dicarboxylates, as well as to three isomeric phthalate dianions. In order to test quantitatively the ability of our system to differentiate between dicarboxylate anions, the fluorescent intensity data at 428 nm, 490 nm, 560 nm, and 600 nm for all the DPAC-bisC4P⊃DC2- complexes were extracted from the final emission spectra, and a three-dimensional coordinate plot was constructed based on the resulting I428/I600, I490/I600, and I560/I600 values. As shown in Figure 9a, the eleven carboxylate dianions considered in this study could be clearly discriminated by considering the individual fluorescence response at three different wavelengths. Note that the unique red emission peak (600 nm) attributed to DPAC could be used as an internal standard, which largely obviates concerns regarding external interference in the context of sensing. Most importantly, all the DPAC-bisC4P⊃DC2- complexes exhibited different fluorescence colors, which allows these eleven dicarboxylates to be differentiated by the naked eye (cf. Figure 9b). The present results thus not only support the conclusion that efficient dicarboxylate sensing is being achieved, but also provides a paradigm where a single VIE fluorophore is capable of translating the nuances of multiple analytes into unique color outputs.

Figure 9. (a) Three-dimensional fluorescence response plots corresponding to the I428/I600, I490/I600, and I560/I600 values for individual samples of DPAC-bisC4P in acetonitrile containing a sufficient quantity of the indicated dicarboxylate anions (DC2- ) to saturate the VIE response (typically