Complex Formation between Rhodamine B and Poly(sodium 4

21576

J. Phys. Chem. B 2006, 110, 21576-21581

Complex Formation between Rhodamine B and Poly(sodium 4-styrenesulfonate) Studied by 1H-NMR Ignacio Moreno-Villoslada,*,†,| Rodrigo Gonza´ lez,† Susan Hess,† Bernabe´ L. Rivas,‡ Toshimichi Shibue,§ and Hiroyuki Nishide| Instituto de Quı´mica, Facultad de Ciencias, UniVersidad Austral de Chile, Casilla 567, ValdiVia, Chile, Facultad de Ciencias Quı´micas, UniVersidad de Concepcio´ n, Casilla 160-C, Concepcio´ n, Chile, and Materials Characterization Central Laboratory and Department of Applied Chemistry, School of Science and Engineering, Waseda UniVersity, Tokyo 169-8555, Japan ReceiVed: June 27, 2006; In Final Form: July 26, 2006

A 1H NMR study is presented for the binding of rhodamine B (RB) to the polyanion containing aromatic groups poly(sodium 4-styrenesulfonate) (PSS), which is also evidenced by diafiltration. 1H NMR spectra showed an accentuated upfield shift of proton H6′ of the benzoic ring of RB at pH 7, indicating the stacking of RB onto PSS. The corresponding structure is proposed which is in accordance to Hunter and Sanders rules. At pH 2, an upfield shift of the xanthene protons of RB would indicate a highly condensed state for this molecule.

1. Introduction Molecular association involving aromatic-aromatic interactions is attracting much attention nowadays.1-13 As a result of the planar geometries of aromatic molecules, the molecular surface/volume ratio is high compared to that of spherical particles. Then, the aggregation of aromatic groups in water may produce the release of higher amounts of surface-solvating water molecules and, consequently, an increase in the favorable entropic and enthalpic contributions to the free energy by means of classical and nonclassical hydrophobic effects.1 In addition to these solvent contributions, site-specific interactions such as short-range electrostatic interactions, hydrogen bond formation, π-π interactions, or cation-π interactions may also contribute to the free energy and define the geometry of the complexes. These aromatic-aromatic interactions are found in water in many systems such as nucleic acids, proteins, porphyrins, semiconductors, molecular clips, and so forth1-13 and are one of the principal noncovalent forces governing molecular recognition and biomolecular structure. In particular, they are important in the stabilization of DNA and its association with intercalators. A change in the hydrogen bonding capacity of DNA bases as a consequence of π-π interactions has been described in recent literature.4,5 The vast majority of X-ray crystal structures of protein complexes with small molecules reveal bounding interactions involving aromatic amino acid side chains of the receptor and/or aromatic and heteroaromatic rings of the ligand. A raise of the pKa of substrates by means of π-π interactions with enzymes has been proposed as a mechanism of enzymatic catalysis.6 We have recently found a smart system composed by poly(4-sodium styrenesulfonate) (PSS) and rhodamine B (RB).14 The behavior of this system is pH dependent. Lowering the pH from 5 to 3, we have observed an abrupt change in the binding * FAX: 56-63-221597. E-mail: [email protected]. † Universidad Austral de Chile. ‡ Universidad de Concepcio ´ n. § Materials Characterization Central Laboratory, Waseda University. | Department of Applied Chemistry, Waseda University.

Figure 1. Molecular structures of RB and PSS.

constant and in the light absorption and emission properties of RB. Moreover, the changes described witness an increase of the RB pKa. This behavior is not found for other polyanions that do not contain aromatic rings,15 so the aromatic ring of PSS plays an important role in the overall interaction. NMR is a useful tool to probe the stacking of aromatic rings. The effect of stacking the rings is to place one ring in the shielding cone of the second, resulting in upfield shifts of 1H resonances for the stacked rings. In this paper, we analyze the binding of RB to PSS by 1H NMR and diafiltration and will propose a geometry for the π-stacking of RB onto PSS. 2. Experimental Section 2.1. Reagents. Commercially available PSS (Aldrich, synthesized from the para-substituted monomer) and RB (Sigma) were used to prepare the solutions in deionized distilled water or D2O (Acroˆs, 99.8% d-content). The structures of RB and PSS are shown in Figure 1. The pH was adjusted with minimum amounts of NaOH and HNO3 or DCl (Acroˆs, +99% d-content). 2.2. Equipment. The unit used for diafiltration studies consisted of a filtration cell (Amicon 8010, 10 mL capacity) with a magnetic stirrer, a regenerated cellulose membrane with a MWCO of 10 000 Da (Ultracel PLGC, 25 mm diameter), a reservoir, a selector, and a pressure source. Distilled water was deionized in a Simplicity millipore deionizer. The pH was controlled on Quimis Q 400 M2 or Horiba F-15 pH meters. UV-vis experiments including analyses of the diafiltered

10.1021/jp0640169 CCC: $33.50 © 2006 American Chemical Society Published on Web 10/05/2006

Complex Formation Between RB and PSS

J. Phys. Chem. B, Vol. 110, No. 43, 2006 21577

Figure 2. 1H NMR aromatic region spectra of solutions in D2O at pH 7.5 (1) and 2.2 (2) of different RB concentrations: (a) 1‚10-2 M; (b) 5‚10-3 M; (c) 1‚10-3 M; (d) 5‚10-4 M.

solutions were performed in a Heλios γ spectrophotometer at 293 K and 1 cm of path length. 1H NMR measurements were made in JNM-Lambda500 (JEOL, 500 MHz) and AVANCE600 (Bruker, 600 MHz) spectrometers. 2.3. Procedures. Details for diafiltration procedures may be found elsewhere,16-18 and experimental conditions are provided in the figure captions. 2D diffusion-ordered spectroscopy (DOSY) experiments were made under a stimulated echo sequence employing bipolar gradients and a longitudinal eddy current delay. Diffusion delays in the range 20-100 ms and gradient durations in the range 2-6 ms were searched for in 10-3 M RB and 10-2 M PSS solutions, to obtain appropriate curves (25 points) for inverse Laplace transformation. The same procedure was done by taking well-resolved signals of RB as reference in mixtures of 10-3 M RB and variable amounts of PSS. 3. Results and Discussion 3.1. RB Alone. RB is a zwitterionic molecule which undergoes self-aggregation at concentrations higher than 10-4 M. Its pKa at concentrations lower than 10-5 M is 3.2. Figure 2 shows the aromatic region of solutions at four different RB concentrations and pHs higher and lower than the pKa. According to Burghrdt et al.,19 assignments of the signals are related to the RB structure shown in Figure 1. Protonation of RB produces (a) a higher withdrawing effect on the benzoic ring and (b) a decrease in the electron density of the carboxylic group situated directly over the xanthene ring. This may explain the downfield shift of protons H3′, H4′, and H5′ in the benzoic group in the protonated molecule. On the other hand, the xanthene protons are upfield-shifted at increasing RB concentrations, due to π-π stacking. A structure for the dimer at [RB] ) 2‚10-3 M was deduced by other authors19 showing a faceto-face stacking of the xanthene groups. At higher concentrations, higher-order associations may be found. 3.2. RB-PSS at pH 7. The interaction of RB with PSS at this pH is evidenced by diafiltration. Diafiltration features are described elsewhere.16-18 The main magnitudes managed in diafiltration analyses are the filtration factor (F), defined as the

Figure 3. Diafiltration profiles of an RB 1‚10-4 M aqueous solution at pH 7 in the absence of PSS (+) and in the presence of different PSS concentrations: ([) 2‚10-4 M; (]) 4‚10-4 M; (9) 6‚10-4 M; (0) 8‚10-4 M; (2) 1‚10-3 M; (4) 2‚10-3 M; (see Table 1 for linear adjustments).

ratio between the volume in the filtrate and the constant volume in the diafiltration cell, the concentration in the filtrate of the low-molecular-weight species under study (LMWS) (cLMWSfiltrate), the concentration of free LMWS in the cell solution (cLMWSfree), the concentration of LMWS reversibly bound to the water-soluble polymer (WSP) (cLMWSrev-bound), the apparent dissociation constant (KLMWSdiss-WSP), defined as the ratio cLMWSfree/cLMWSrev-bound, the diafiltration parameters km, j, u, and V, and the polymer concentration in moles per liter of monomeric units (cP). km and j parameters (the absolute value of the slope of the curve ln cLMWSfiltrate versus F in the absence and in the presence of the WSP, respectively) are related to the strength of the interaction, while V and u are related to the amounts of LMWS reversibly or irreversibly bound to the polymer, respectively. Results of diafiltration of a 10-4 M solution of RB at pH 7 shown in Figure 3 and Table 1 indicate no significant interaction of RB with the diafiltration cell components (km close to 1). On the other hand, the amounts of RB irreversibly bound to PSS are negligible (u values close to 0), and the fraction of RB reversibly bound to PSS increases with

21578 J. Phys. Chem. B, Vol. 110, No. 43, 2006

Moreno-Villoslada et al.

TABLE 1: Results for Diafiltration of 10-4 M RB Solutions at pH 7 and Different PSS Concentrationsa expt

cP (M)

V

u

j

RB-03 PSS -RB-01 PSS -RB-02 PSS -RB-03 PSS -RB-04 PSS -RB-05 PSS -RB-06

2‚10-4 4‚10-4 6‚10-4 8‚10-4 1‚10-3 2‚10-3

0.91 0.85 0.85 0.92 0.94 0.95 0.94

0.09 0.15 0.15 0.08 0.06 0.05 0.06

0.63 0.49 0.41 0.37 0.29 0.18

km 0.84

KRBdiss-PSS b

linear adjustments for the experimental data

R2

2.4 ( 0.5 1.2 ( 0.2 0.8 ( 0.1 0.7 ( 0.8 0.5 ( 0.05 0.2 ( 0.03

y ) -0.84x - 9.5 y ) -0.63x - 9.6 y ) -0.49x - 9.9 y ) -0.41x - 10.0 y ) -0.37x - 10.1 y ) -0.29x - 10.4 y ) -0.18x - 10.9

1.00 0.99 0.99 0.99 0.99 1.00 1.00

a For linear adjustments: y ) ln 〈cRBfiltrate〉; x ) F; R2 ) linear regression factor. b KRBdiss-PSS is calculated following (j/1 - j) e KLMWSdiss-WSP e (kmj/km - j).

Figure 4. 1H NMR aromatic region spectra of solutions in D2O at pH 7 of (a) RB 1‚10-3 M; (b-d) RB 1‚10-3 M in the presence of different PSS concentrations: (b) 0.5‚10-3 M, (c) 2‚10-3 M, (d) 6‚10-3 M; (e) PSS 1‚10-2 M.

polymer concentration (decreasing values of j), which is reflected in the KRBdiss-PSS values. The association of RB with PSS is seen by 1H NMR by a general broadening and upfield shift of the RB bands (see Figures 4 and 5). This behavior is not found in analogous systems including polyelectrolytes with no aromatic groups, such as poly(sodium vinyl sulfonate) (PVS).15 In Figure 4, some representative spectra are shown for constant amounts of RB in the presence of variable amounts of PSS, while in Figure 5, the increase of the chemical shifts of the different RB aromatic protons for a set of experiments is shown as a function of PSS concentration. The upfield shift, and consequently the binding of RB to PSS, increases with increasing the WSP concentration but at a decreasing rate, due to a decreasing free RB in solution. This indicates an equilibrium binding in accordance to diafiltration observations. The only exception to the general upfield shift is the H3′ signal, and we interpret this fact as a higher

Figure 5. Upfield shift of RB aromatic signals ([RB] ) 1‚10-3 M in D2O at pH 7) as a function of PSS concentration: (]) H3′; (O) H4′; (4) H5′; (0) H6′; ([) H4H5; (9) H1H8; (2) H2H7.

Complex Formation Between RB and PSS

J. Phys. Chem. B, Vol. 110, No. 43, 2006 21579

Figure 6. COSY spectrum of the aromatic region of a RB 1‚10-3 M/PSS 2‚10-3 M solution in D2O at pH 7.

tendency of RB to protonate in the presence of PSS as evidenced by the change on its pKa.14 H4′ and H5′ have the same chemical shift for RB alone, but they are differently shifted in the presence of the polymer. To assign each signal, a COSY experiment has been done for the sample corresponding to spectrum c in Figure 4. The COSY spectra shown in Figure 6, allowed us to assign the signal at 7.53 ppm to H4′, since a clear correlation is seen with H3′. The correlation between H5′ and H6′ also appears clearly, which confirms the overlapping of H6′ with H1H8 at this PSS/RB ratio, following the upfield shift tendency evidenced in Figure 5. On the other hand, the signals for H4′ and H5′ overlap with those of the polymer protons appearing around 7.4 ppm, and for some PSS/RB ratios, they were not clearly distinguishable. To obtain the respective chemical shifts, COSY and DOSY experiments are useful. DOSY experiments at pH 7 showed diffusion coefficients for a 10-2 M PSS solution of 0.9‚10-10 m2 s-1 and of 9‚10-10 m2 s-1 for a 10-3 M RB solution. For mixtures of 10-3 M RB and variable amounts of PSS at pH 7, the diffusion delays and gradient durations were selected by observing the well-resolved peaks H3′ and H6′ of RB. This allowed us to obtain the 1D spectra at small gradient amplitudes, where all the signals are recuperated, and medium gradient amplitudes, where the signals for RB are lost, but not those of PSS, since the diffusion coefficient of RB is higher than that of PSS. Then, subtracting the latter 1D spectra from the former, the chemical shifts for H4′ and H5′ become observable. It can be seen in Figures 4 and 5 that the upfield shift in the RB signals is more accentuated for H6′ and H5′ protons (0.6 ppm for H6′), which witnesses the stacking of RB onto PSS by means of π-π interactions. Due to these π-π interactions, the pKa of RB changes,14 as pointed out by other authors considering other substrates.6 To propose a structure for the stacking of RB onto PSS at pH 7, we search to satisfy the following criteria: (a) explain the high shift value of H6′, (b) conserve the symmetry, (c) avoid unacceptable steric hindrances, (d) provide an adequate charge disposition, and (e) be in accordance with Hunter-Sanders

Figure 7. Proposed structure for the stacking of RB onto PSS at pH 7.

rules.13 The proposed geometry for the interaction is shown in Figure 7, where an edge-to-face stacking occurs between the RB electron-deficient benzoic ring (edge) and the PSS electron-rich benzenosulfonate ring (face), according to Hunter and Sanders rules. The contacting edge should contain H6′ and H5′, so that the negatively charged carboxylic group is placed at the opposite side of the negatively charged sulfonate group. Moreover, if the xanthene group of RB adopts an opposite disposition with respect to the polymer backbone avoiding steric hindrances, its positive charge would be located between the two negative charges, stabilizing the symmetrical structure. 3.3. RB-PSS at pH 2. The study of this system at acidic pH is more complex. Diafiltration experiments showed a quantitative binding yet at a PSS/RB ratio of 2/1 and low concentrations.14 Increasing the absolute concentration of both components results in precipitate formation, but by increasing the PSS/RB ratio from PSS/RB ) 4, precipitation is avoided. Precipitation is due to a charge balance between the negatively charged PSS and the positively charged protonated RB. On the other hand, fluorescence is quenched for PSS/RB ratios under 20, but it is visibly conserved at PSS/RB ) 100, so fluorescence quenching is not due to the binding of RB to PSS itself. This may mean that, at moderate and low PSS/RB ratios, RB is found in a highly condensed state, which may be avoided under high PSS/RB ratios. Although experiments are

21580 J. Phys. Chem. B, Vol. 110, No. 43, 2006

Moreno-Villoslada et al.

Figure 8. 1H NMR aromatic region spectra of solutions in D2O of RB 3.4‚10-3 M and PSS 2‚10-2 M at different pHs: (a) 6.9; (b) 5.2; (c) 4.6; (d) 4.1; (e) 2.2.

being done by other techniques in order to clarify this behavior, we show here the results obtained by 1H NMR. 1H NMR also supports the above hypothesis, since by increasing the pH, the signal/noise ratio decreases dramatically, so that the aromatic signals of RB are lost at 297 K at a PSS/ RB ratio of 6, even with an increase in the RB concentration to 3.4‚10-3 M. This can be seen in Figure 8. Increasing the pH produces an upfield shift of the xanthene protons. On the contrary, the signals of the RB benzoic ring are downfield shifted. The formation of higher-order structures in which RB is compacted within the polymer domain is congruent with these facts. The downfield shifting of H6′ and H5′ may indicate a reaccommodation of the polymer. Note that all the signals are upfield-shifted with respect to those found for RB selfassociation at acidic pH and high concentrations (see Figure 2). Lost signals at pH 2.2 and 297 K are recovered at 343 K with small changes in the chemical shifts. Conclusions RB binds to PSS at pH 7 by means of π-π interactions according to Hunter and Sanders rules. 1H NMR shows the interaction between the benzoic ring of RB and the aromatic ring of PSS in an edge (RB) to face (PSS) geometry. Diafiltration experiments also evidence the interaction between the polyanion and the zwitterionic molecule. At pH 2, an upfield shift of the xanthene protons of RB would indicate a highly condensed state for this molecule, which is in accordance with previous observations of fluorescence quenching.

Acknowledgment. The authors thank Fondecyt (grants no. 1030669 and no. 1060191) and the 21COE program “Practical Nano-Chemistry” at Waseda University from MEXT, Japan, for financial support. References and Notes (1) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed. 2003, 42, 1210. (2) Sokolov, A. N.; Frisˇcˇicˇ, T.; MacGillivray, L. R. J. Am. Chem. Soc. 2006, 128, 2806. (3) Nohra, B.; Graula, S.; Lescop, C.; Re´au, R. J. Am. Chem. Soc. 2006, 128, 3520. (4) Mignon, P.; Loverix, S.; Steyaert, J.; Geerlings, P. Nucleic Acids Res. 2005, 33, 1779. (5) Verse´es, W.; Loverix, S.; Vandemeulebroucke, A.; Geerlings, P.; Steyaert, J. J. Mol. Biol. 2004, 338, 1. (6) Mignon, P.; Loverix, S.; De Proft, F.; Geerlings, P. J. Phys. Chem. A 2004, 108, 6038. (7) Sinnokrot, M. O.; Sherrill, C. D. J. Phys. Chem. A 2003, 107, 8377. (8) Rashkin, M. J.; Waters, M. L. J. Am. Chem. Soc. 2002, 124, 1860. (9) Brunsveld, L.; Folmer, B. J. B.; Meijer, E. W.; Sijbesma, R. P. Chem. ReV. 2001, 101, 4071. (10) Reek, J. N. H. J. Am. Chem. Soc. 1997, 119, 9956. (11) Lokey, R. S. Nature (London) 1995, 375, 303. (12) Linse, P. J. Am. Chem. Soc. 1992, 114, 4366. (13) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5525. (14) Moreno-Villoslada, I.; Jofre´, M.; Miranda, V.; Gonza´lez, R.; Sotelo, T.; Hess, S.; Rivas, B. L. J. Phys. Chem. B 2006, 110, 11809.

Complex Formation Between RB and PSS (15) Moreno-Villoslada, I.; Jofre´, M.; Miranda, V.; Chandı´a, P.; Gonza´lez, R.; Hess, S.; Rivas, B. L.; Elvira, C.; San Roma´n, J.; Shibuhe, T.; Nishide, H. Polymer 2006, 47, 6496. (16) Moreno-Villoslada, I.; Miranda, V.; Chandı´a, P.; Villatoro, J. M.; Bulnes, J. L.; Corte´s, M.; Hess, S.; Rivas, B. L. J. Membr. Sci. 2006, 272, 137.

J. Phys. Chem. B, Vol. 110, No. 43, 2006 21581 (17) Moreno-Villoslada, I.; Miranda, V.; Oyarzu´n, F.; Hess, S.; Luna, M.; Rivas, B. L. J. Chil. Chem. Soc. 2004, 49, 121. (18) Moreno-Villoslada, I.; Miranda, V.; Gutie´rrez, R.; Hess, S.; Mun˜oz, C.; Rivas, B. L. J. Membr. Sci. 2004, 244, 205. (19) Ilich, P.; Mishra, P. K.; Macura, S.; Burghrdt, T. P. Spectrochim. Acta, Part A 1996, 52, 1323.