Langmuir 2007, 23, 13203-13208
13203
Dependence of Fluorescence Quenching of a Poly(p-phenyleneethynylene) Polyelectrolyte on the Electrostatic and Hydrophobic Properties of the Quencher Palwinder Kaur, Mingyan Wu, Laura Anzaldi, and David H. Waldeck* Chemistry Department, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260
Cuihua Xue and Haiying Liu* Chemistry Department, Michigan Technological UniVersity, Houghton, Michigan 49931 ReceiVed July 30, 2007. In Final Form: September 12, 2007 This study investigates the fluorescence quenching of a poly(p-phenyleneethynylene) (1) based polyelectrolyte by positively charged and neutral macromolecules. This work shows that the change in the fluorescence yield of 1 depends on a number of factors, including electrostatic, hydrophobic, and energy transfer interactions with the quencher and also changes in the solution conditions such as concentration and ionic strength. The fluorescence quenching is attributed to the formation of aggregates that form upon addition of different quenchers to a solution of 1 and/or the solution conditions. The extent of 1’s aggregation is shown to depend on the type of interaction between the polymer and the quencher, the concentration of the polymer, and the ionic strength of the solution.
Introduction In recent years, conjugated polymer materials have been exploited as enabling materials for new technologies and applications, e.g., light-emitting diodes,1 lasers,2 and solar cells.3 One of these potential applications aims to exploit their fluorescence yield’s sensitivity to minute quantities of a quencher to create biosensors.4-8 The high sensitivity of the fluorescence to a quencher results from the high mobility of the electronic excitation along the polymer backbone, so that a single trap site on the polymer backbone can efficiently capture the excitation. The binding of a quencher to the polymer creates a trap site which then quenches the fluorescence via energy transfer or electron transfer or due to conformational changes induced on binding to the quencher. This phenomenon has commonly been referred to as the molecular wire effect in the literature5,9,10 A number of conjugated polymer materials are fluorescent (PPV, PPP, polythiophenes, and others) and have been demonstrated to perform as sensing agents for chemical and biological molecules.11-20 Attaching ionic side groups to the polymer (1) Zhang, C.; Broun, D.; Heeger, A. J. J. Appl. Phys. 1993, 73, 5177-5180. (2) Hide, F.; Diaz-Garcia, M. A.; Schwartz, B. J.; Andersson, M. R.; Pei, Q.; Heeger, A. J. Science 1996, 273, 1833-1836. (3) Yu, G.; Gao, J.; Hummelen, J.; Wudl, F.; Heeger, A. J. Science 1995, 270, 1789. (4) Chen, L.; McBranch, D. W.; Wang, R.; Whitten, D. Chem. Phys. Lett. 2000, 330, 27-33. (5) DiCesare, N.; Pinto, M. R.; Schanze, K. S.; Lakowicz, J. R. Langmuir 2002, 18, 7785-7787. (6) Heeger, P. S.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 1221912221. (7) McQuade, D. T.; Hegedus, A. H.; Swager, T. J. Am. Chem. Soc. 2000, 122, 12389- 12390. (8) Tan, C.; Pinto, M. R.; Schanze, K. S. Chem. Commun. 2002, 446-447. (9) Swager, T. M. Acc. Chem. Res. 1998, 31, 201-207. (10) Zhou, Q.; Swager, T. M. J. Am. Chem. Soc. 1995, 117, 12593-12602. (11) Pinto, M. R.; Schanze, K. S. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7505-7510. (12) Rininsland, F.; Xia, W.; Wittenburg, S.; Shi, X.; Stankewicz, C.; Achyuthan, K.; McBranch, D.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 1529515300. (13) Kim, I. B.; Dunkhorst, A.; Gilbert, J.; Bunz, U. H. F. Macromolecules 2005, 38, 4560-4562. (14) Liu, B.; Bazan, G. C. Chem. Mater. 2004, 16, 4467-4476.
backbone has been used to make the materials water-soluble and assist in binding fluorescence quenchers and/or analyte species. Although most of the research effort on these materials has explored the ability of different polymer backbones and receptor strategies to sense macromolecules, some workers have explored the fluorescence quenching mechanism. Fan et al.21 showed that the fluorescence quenching of a PPV-based polyelectrolyte by the protein cytochrome-c arises from a combination of fluorescence quenching by electron transfer and the formation of bound complexes between the cationic cytochrome-c and the anionic PPV. Liu et al.18 showed that the quenching in a poly(p-phenylene) (PPP) based anionic polyelectrolyte results from conformational changes that occur upon binding of the polyelectrolyte with a protein or dendrimer macromolecule. They further showed that electron and energy transfer processes need not be present. From such studies, it is clear that complex formation between the polyelectrolyte and the quencher is important for fluorescence quenching of the polyelectrolytes. Recently, Kim et al.22 demonstrated fluorescence quenching of PPE based polyeletrolytes and oligomers upon exposure to a number of proteins such as histones, lysozymes, myoglobin, hemoglobin, and so forth. They showed that the net negative charge of the PPE plays a significant role but is not the only factor controlling the interaction of proteins with these polyelectrolytes and their fluorescence quenching. A number of factors such as electrostatic interactions, (15) Kumaraswamy, S.; Bergstedt, T. S.; Shi, X.; Rininsland, F.; Kushon, S. A.; Xia, W.; Achyuthan, K.; McBranch, D. W.; Whitten, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7511-7515. (16) Disney, M. D.; Zheng, J.; Swager, T. M.; Seeberger, P. H. J. Am. Chem. Soc. 2004, 16, 13343-13346. (17) Wosnick, J. H.; Mello, C. M.; Swager, T. M. J. Am. Chem. Soc. 2005, 127, 3400-3405. (18) Liu, M.; Kaur, P.; Waldeck, D. H.; Xue, C.; Liu, H. Langmuir 2005, 21, 1687-1690. (19) Thomas, S. W., III; Joly, G. D.; Swager, T. M. Chem. ReV. 2007, 107, 1339-1386. (20) Zhang, T.; Fan, H.; Zhou, J.; Liu, G.; Feng, G.; Jin, Q. Macromolecules 2006, 39, 7839-7843. (21) Fan, C.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2002, 124, 5642-5643. (22) Kim, I. B.; Dunkhorst, A.; Bunz, U., H. F. Langmuir 2005, 21, 79857989.
10.1021/la7023007 CCC: $37.00 © 2007 American Chemical Society Published on Web 11/20/2007
13204 Langmuir, Vol. 23, No. 26, 2007
Kaur et al. emission spectra were measured on a Spex Fluorolog 0.22 m double spectrometer. Stern-Volmer Constant. The Stern-Volmer constants were obtained using the Stern-Volmer equation26 described as follows F0 ) 1 + Ksv[Q] F
Figure 1. Chemical structure of poly(p-phenyleneethynylene)-based polyelectrolyte (1).
hydrophobic interactions, electron transfer, energy transfer, and so forth are likely to contribute to the fluorescence quenching behavior, although sometimes one factor may be dominant. Harrison et al.23 showed that the quenching of cationic PPE conjugated polyelectrolyte is also dependent on the polyelectrolyte’s concentration. They showed that the quenching efficiency increased by a factor of 175 on decreasing the concentration from 10 µM to 1 µM. Although they attributed this effect to the aggregation of polyelectrolyte at higher concentration, no detailed experimental results were provided. Later, Kim et al.13 emphasized the importance of buffer choice when using these polelectrolytes for sensing heavy metal ions. They showed that the sensing efficiency of these polyelectrolytes can be dramatically improved by the correct choice of buffer. This work reports how the fluorescence quenching of a poly(ppolyphenyleneethynylene)-based polyelectrolyte (1) changes in the presence of different macromolecules (proteins, dendrimers, and surfactants) and under different solution conditions (concentration and ionic strength). The change in the optical properties and relative fluorescence quantum yield of 1 in the presence of different macromolecules is explained by the different kind of interactions that are present between 1 and macromolecules. The fluorescence quenching of the polymer in the presence of protein and/or dendrimer is attributed to the formation of polymer aggregates via electrostatic and hydrophobic interactions between multiple polymer strands and the quencher molecules. The effect of solution conditions on the sensitivity of 1 has been studied by varying the concentration of polymer and the ionic strength of the solution. These studies show that the Ksv displays a concentration dependence that can be understood to arise from quencher-induced aggregation. Experimental Section Material. Poly[2,5-bis(3-sulfonatopropoxy)-1,4-(p-phenyleneethynylene-alt-1,4-polyphenylene ethynylene] (identified as 1; Figure 1) is a polyelectrolyte with two negative charges per repeat unit and was prepared in a manner similar to that reported in the literature.24 Cytochrome-c was bought from Sigma and was used without further purification. PAMAM-3G was bought from Dendritech, Inc., and protonated using trifluoroacetic acid (see Supporting Information for structure of PAMAM-3G). Six-armed poly(ethylene oxide) hydroxy-terminated, dipentaerythritol core (PEG-OH), Mw 12 kDa, was purchased from Polymer Source Inc. DEM-3.5 G (Mw 6.8 KDa) was a gift from Professor T. Chapman at University of Pittsburgh. Size exclusion chromatography25 was used to determine the molecular weight of 1 in DMSO (a good solvent), 38100 Da with a polydispersity of 3.04. The experimental polymer solution was highly diluted (10-6-10-8 M in terms of polymer repeat unit). All concentrations in this paper are reported in terms of polymer repeat unit. At these concentrations, the effect of the polymer on solution properties, such as viscosity, can be neglected. Steady-State Spectroscopy. Steady-state absorption spectra were measured on an Agilent 8453 spectrometer and the steady-state (23) Harrison, B. S.; Ramey, M. B.; Reynolds, J. R.; Schanze, K. S. J. Am. Chem. Soc. 2000, 122, 8561-8562. (24) Kim, S.; Jackiw, J.; Robinson, E.; Schanze, K. S.; Reynolds, J. R.; Baur, J.; Rubner, M. F.; Boils, D. Macromolecules 1998, 31, 964-974. (25) GPC studies were performed by American Polymer Std., 8680 Tyler Blvd., Mentor, OH 44060.
(1)
with F0 being the fluorescence intensity of the conjugated polymer by itself and F being the fluorescence intensity of the conjugated polymer solution with a given concentration of quencher [Q]. Ksv is the Stern-Volmer constant and can be extracted from the slope of the graph that plots F0/F vs quencher concentration with intercept 1. The Stern-Volmer constants reported in this paper were obtained by fitting the data at very low quencher concentration ranging from 0.0 to 0.4 µM. Time-Dependent Fluorescence Spectroscopy. The time-resolved fluorescence data were collected using the time-correlated single photon counting method.19 The instrument response function was measured using a sample of colloidal BaSO4. The samples were excited at 438 nm using a diode laser (PIL043, A.L.S. GmbH), and the emission was collected at different wavelengths. The fluorescence decay curves were fit by a convolution and compare method using IBH-DAS6 analysis software. Other details of the TCSPC apparatus can be found in ref 27. Fluorescence Correlation Spectroscopy (FCS). FCS is a noninvasive single molecule method which obtains dynamic and kinetic information by following the fluctuation trajectory of fluorescence about the equilibrium state.28-32 FCS was performed on a homemade FCS instrument based on a Zeiss IM35 inverted microscope. Details of the instrumentation will be provided elsewhere.33 The sample was excited at 438 nm through an objective lens (Olympus UPlanfluor 40×/1.30 oil), and the fluorescence was collected by the same lens. The concentrations of the polymer solutions were controlled to be 5.2 × 10-8 M and 2.5 × 10-6 M. To avoid photobleaching and optical trapping, the laser power was kept low, 24 µW, as measured at the front of the objective lens. Each measurement lasted 2 to 5 min, during which the time trajectory of fluorescence was monitored and only those having stable fluorescence intensity were kept. The corresponding autocorrelation function G(t) was fit by eq 2 to extract the correlation time τD. N h is the average
( )( -1
G(t) )
1 t 1+ τD N h
1+
ωxy2t
)
ωz2τD
-1/2
(2)
number of fluorophores in the focal volume; ωxy is the radius of the focal spot in the transverse direction, and ωz is the Rayleigh range of the excitation beam (see ref 34 for details relating to eq 2). The correlation time τD is related to the translational diffusion coefficient D of the fluorophore by τD )
ωxy2 4D
(3)
The apparatus was calibrated and tested using a 10 nM Rhodamine 6G aqueous solution, assuming the diffusion coefficient D ) 4.14 × 10-6 cm2 s-1.35 The Stokes-Einstein approximation (eq 4) was (26) Lacowitz, J. R. Principles of Fluorescence Spectroscopy; Plenum Press: New York, 1983. (27) Liu, M.; Waldeck, D. H.; Oliver, A. M.; Head, N. J.; Paddon-Row, M. N. J. Am. Chem. Soc. 2004, 126, 10778-10786. (28) Koppel, D. E. Phys. ReV. A 1974, 10, 1938-1945. (29) Magde, D.; Elson, E.; Webb, W. W. Phys. ReV. Lett. 1972, 29, 705-708. (30) Pristinski, D.; Kozlovskaya, V.; Sukhishvili, S. A. J. Chem. Phys. 2004, 122, 14907-14915. (31) Van Rompaey, E.; Sanders, N.; De Smedt, S. C.; Demeester, J.; Van Craenenbroeck, E.; Engelborghs, Y. Macromolecules 2000, 33, 8280-8288. (32) Leng, X.; Sarchev, K.; Buffle, J. J. Colloid Interface Sci. 2002, 251, 64. (33) Yue, H.; Wu, M.; Xue, C.; Liu, H.; Waldeck, D. H. J. Phys. Chem. B, submitted. (34) Krichevsky, O.; Bonnet, B. Rep. Prog. Phys. 2002, 65, 251-297.
Fluorescence Quenching of PPE Polyelectrolyte
Langmuir, Vol. 23, No. 26, 2007 13205
of the solutions, the Stern-Volmer constant is decreased by a factor of nearly 2.3. Experiments at ionic strength higher than 60 mM were not performed because of poor solubility of 1 in these higher ionic strength solutions. These results imply that, when 1 is in a relatively unaggregated form, it is more effectively quenched by the protein than when it is aggregated. To explore whether cytochrome-c quenches 1 by inducing aggregation, quenching experiments were performed at concentrations where 1 (∼2 × 10-7 M) does not self-aggregate. To further ensure that 1 exists in an unaggregated form, the 1 solutions were first heated to dissociate any aggregates and then cooled to room temperature; see reference 36. Figure 4 shows the Figure 2. Emission spectra for 1 in water at different concentrations (2.0 × 10-6 (-), 8.8 × 10-8 (0), 1.0 × 10-8 (O) M) and at 50 mM Na3PO4 at 2.0 × 10-6 M 1 concentration (----). used to extract the hydrodynamic radius RH from the measured diffusion coefficient where η is the shear viscosity, T is the D)
k BT 6πηRH
(4)
temperature, and kB is Boltzmann’s constant.
Results and Discussion It has been reported in the literature that the photophysical properties of poly(p-phenyleneethynylene)-based conjugated polymers change with the temperature, concentration, ionic strength, and solution environment.36 The recent study by Kaur et al.36 has shown that PPE-based polyelectrolytes exist as independent strands at low concentration in good solvent and they aggregate at high concentrations and in the presence of salts, causing the fluorescence to quench. Figure 2 shows how the emission spectrum of 1 changes with concentration and ionic strength. 1 exists as an independent polymer strand at very low concentration (∼10-8 M in polymer repeat units) and is aggregated at higher concentration (∼10-6 M in polymer repeat units). Similarly, 1 aggregates at higher ionic strength and the fluorescence is highly quenched (Figure 2). Quenching of 1 by Cytochrome-c. Fluorescence quenching experiments were performed at different concentrations of 1 with ferric cytochrome-c. Figure 3A shows the dependence of the Stern-Volmer constant Ksv on the concentration of 1. As the concentration of 1 was increased, the apparent Stern-Volmer constant decreased by a factor of nearly 10 and then became constant. Given that 1 aggregates at concentrations above 10-6 M,36 these data show that 1 is more sensitive to the quencher in its unaggregated form. To corroborate this result further, quenching experiments were done at different ionic strengths for 10-6 M solutions of 1 (Figure 3B). On increasing the ionic strength
Figure 4. Absorption and emission spectra of 2 × 10-7 M aqueous solutions of 1 without (- - - -) and with (-) 0.4 µM ferric cytochrome-c. The emission of 1 in the presence of 0.4 µM ferric cytochrome-c has been magnified 50 times for clarity.
absorption and emission spectra of the heated 1 before and after addition of 0.4 µM cytochrome-c. On addition of ferric cytochrome-c to the unaggregated 1 solution, a new red-shifted peak was observed in the absorption spectrum. This red peak has been attributed to aggregate induced planarization of the poly(pphenyleneethynylene)backbone in the previous literature.36,38 Also, the addition of ferric cytochrome-c quenched the fluorescence of 1 by 98%. Hence the protein appears to induce polymer aggregation and quench the fluorescence. Fluorescence lifetime experiments of these unaggregated polymer solutions before and after the addition of cytochrome-c supports this interpretation. Dilute solutions of 1, in which the steady-state spectra indicate single-stranded fluorophores, have an average fluorescence lifetime of 550 ((5%) ps that is wavelength-independent. Upon addition of cytochrome-c, which quenches the emission yield by six times (600%), the average lifetime decreases to about 300 ps at 460 nm and 400 ps at 510 nm. Although some dynamic quenching is present in these
Figure 3. Panel A shows the dependence of the Stern-Volmer constant on the concentration of 1 at an intrinsic ionic strength of the polymer, and panel B shows the dependence on ionic strength at a 1 concentration of 10-6 M.
13206 Langmuir, Vol. 23, No. 26, 2007
solutions, the data indicate that a static quenching mechanism is dominant, a finding that is consistent with the presence of protein-induced aggregates. The fluorescence quenching mechanism of 1 in the presence of ferric cytochrome-c was elucidated by performing quenching experiments with different forms of cytochrome-c, namely, ferrous, apo, and denatured protein. It was found that the quenching by ferric and ferrous forms of the protein is similar, suggesting that electron-transfer quenching is not dominant. Apo cytochrome-c also quenched the fluorescence but with 45% of the efficiency of the native protein. Further, the denatured cytochrome-c39 quenched the polymer fluorescence but with 75% of the efficiency of the native protein. Comparison of these quenching efficiencies suggests that the energy transfer quenching, as well as the induced aggregation upon addition of cytochromec, governs the fluorescence quenching of 1. To further investigate the aggregation and the appearance of a red-shifted peak upon addition of ferric cytochrome-c to solutions of 1, the 1-cytochrome-c solution was passed through a 0.25 µm filter and the emission and absorption were recorded again. The absorbance of the filtered solution was attenuated nearly ten times as compared to that of the 1-cytochcrome-c solution before filtration. In order to make sure that the attenuation was not coming from 1 adsorbing to the filter, 1 in water was passed through a 0.25 µm filter. The absorption and emission of 1 in water did not change. This experiment suggests that the 1-cytochrome-c solution contains aggregates that are retained by the filter, implying that aggregates of >0.25 µm are formed upon addition of micromolar amounts of ferric cytochrome-c. As a caveat, it is possible that capillary forces present in the filtration could induce aggregation in the 1-cytochrome-c solutions; hence, this experiment demonstrates the relative propensity for aggregate formation in the two solutions, and it may be that pre-existing aggregates are smaller than 250 nm but aggregate into large particles during the filtration. To further corroborate the hypothesis of aggregate formation on addition of ferric cytochrome-c, FCS studies were performed. Figure 5 shows the autocorrelation for 1 in water and a best fit of eq 2 to the data. In the absence of cytochrome-c, the 1 correlation could be fit well by eq 2 with a correlation time of 686 µs, and using eqs 3 and 4 (η ) 0.89 cP and ωxy ) 0.39 µm), one finds an RH of 4.4 nm. Upon addition of 0.4 µM cytochrome-c to the solution, the autocorrelation obtained could not be fit well by eq 2. The fluorescence trajectory showed large spikes which
Kaur et al.
dominated the average fluorescence from the excitation volume. This latter behavior is similar to that observed for a control sample of 0.2 µm fluorecein beads and is not seen for a pure scattering sample. Hence, this behavior is interpreted to arise from fluorescence of very large particles that diffuse into the excitation volume; however, not enough trajectories could be collected to obtain an accurate diffusion constant. Figure 5 shows one of the best correlation functions that were found for this solution, and it gives an RH of 56 nm. This value is taken to represent a lower bound for the size. From the filtration and FCS experiments, it is clear that 1 forms large aggregates when it interacts with the ferric cytochrome-c. Since 1 is negatively charged and ferric cytochrome-c is positively charged, we infer that multiple polymer strands interact with the ferric cytochrome-c molecules electrostatically which leads to polymer aggregation. Importance of Electrostatic Binding. The ionic strength dependence (see Figure 2B and ref 36) implies that electrostatic interactions play an important role in the fluorescence quenching of 1 via aggregation of polymer strands. If this is true, then a quencher with a higher charge might form even bigger aggregates with 1, since it can bring more polymer strands together via electrostatic interaction. To test this hypothesis, quenching experiments with PAMAM 3G and 1 were performed. PAMAM 3G is similar in size to ferric cytochrome-c (diameter ) 3.0 nm); however, it has a charge of +32 e at neutral pH. A 10 µM PAMAM 3G and 2 × 10-7 M solution of 1 has an absorption spectrum with a red-shifted peak (see Figure 6A) that is similar to the spectrum observed upon addition of ferric cytochrome-c to a solution of 1, indicating that aggregates are formed. The fluorescence was also quenched by 30%, which is much smaller than the quenching found with ferric cytochrome-c, however (Figure 6A). Upon passing the 1-PAMAM-3G solution through a 0.2 µm filter, the absorbance did not change and the fluorescence also did not change, indicating that any aggregates formed are smaller in size than the 200 nm filter. To investigate whether electron transfer from the amine groups of PAMAM-3G contribute to the quenching, experiments were done at pH ) 2 where the amines are fully protonated. Since 1 aggregates in the presence of salts and buffers (please see ref 36), the pH experiments were performed with 1 in an aggregated state. Once in an aggregated state, the emission and absorption spectra are not affected by pH changes, so that the observed fluorescence yield reflects the change in protonation of PAMAM-
Figure 5. Panel A shows the autocorrelation function for 1 in water ([) and with 0.4 µM cytochrome-c (9). Panel B shows the same water data on an exponential scale.
Fluorescence Quenching of PPE Polyelectrolyte
Langmuir, Vol. 23, No. 26, 2007 13207
Figure 6. Panel A shows absorption and emission spectra of 2 × 10-7 M 1 in water (-) and with 10 µM PAMAM-3G (- - - -). Panel B shows autocorrelation function for 1 in water ([) and with 10 µM PAMAM-3G (9).
3G. The quenching efficiency was found to be the same as that observed at pH ) 7. The quenching experiments were also done at pH ) 12 where PAMAM-3G is not charged and the surface amine groups might act as electron donors. No quenching was seen at the higher pH. These data indicate that the quenching occurs by electrostatic aggregation of 1 with PAMAM-3G and it does not involve electron transfer. FCS experiments were performed to better quantify the size of the aggregates formed on addition PAMAM-3G to solutions of 1. Figure 6B shows the autocorrelation function for 1 in the presence of PAMAM-3G. The correlation time increased by a factor of nearly 1.4× (τPAMAM-3G ) 850 µs, RH ) 6 nm) as compared to 1 in water (τH2O ) 640 µs, RH ) 4.4 nm). These aggregates appear to be smaller than those formed by 1 in the presence of cytochrome-c (vide supra). Two observations were made for 1 in the presence of 10 µM PAMAM-3G: First, small aggregates form; and second, the fluorescence quenching efficiency is weaker for PAMAM-3G than for cytochrome-c, presumably because it does not have an energy transfer center. Importance of Hydrophobic Interactions. An important difference between PAMAM-3G and ferric cytochrome-c is the presence of hydrophobic interactions for ferric cytochrome-c. The importance of hydrophobic interactions was explored by performing quenching experiments with PEG-OH, which has a hydrophobic core and is neutral (Figure 7A). Solutions of 1 (2 × 10-7 M) with 50 mM PEG-OH show a red shift in the absorption spectrum but do not appear to have any aggregation of polymer strands. Also, the addition of PEG-OH to a 1 solution enhances the fluorescence of 1 rather than quenching it (Figure 7B). This behavior is similar to what has been observed for 1 with the ionic surfactant octadecyl trimethylammonium bromide (ODTMA), already reported in the literature.36 Further, FCS studies show that the correlation time increases by a factor of nearly three (τH2O ) 1920 µs, RH ) 13.2 nm), indicating the formation of particles larger than those formed in the PAMAM3G/1 solution. Hence, aggregates form, but they are not the quenched red-emitting homoaggregates of 1. It may be that the addition of PEG-OH forms heteroaggregates between 1 and (35) Culbertson, C. T.; Jacobson, C. S.; Ramsey, J. D. Talanta 2002, 56, 365-373. (36) Kaur, P.; Yue, H.; Wu, M.; Liu, M.; Treece, J.; Waldeck, D. H. J. Phys. Chem. B 2007, 111, 8589-8596. (37) Lavigne J. J.; Broughton D. L.; Wilson J. N. Macromolecules 2003, 36, 7409-7412. (38) Bunz, U. H. F. Chem. ReV. 2000, 100, 1605-1644. (39) Ferric cytochrome-c was heated to 353 K and then cooled back to room temperature. The denaturation was verified using CD and UV/vis spectroscopy.
Figure 7. A. Chemical structure of PEG-OH. B. Emission spectra of 1 without (-) and with (----) 50 µM PEG-OH.
PEG-OH as observed for the surfactant ODTMA and also similar to that reported by Lavigne et al. for sugar-substituted PPEs in the presence of surfactants.37 Importance of Hydrophobic and Electrostatic Interactions. In order to probe how combined hydrophobic and electrostatic forces affect the interaction between 1 and a quencher, fluorescence quenching experiments were performed in the presence of DEM3.5G. Figure 8A shows the chemical structure of DEM-3.5G dendrimer, which has both a hydrophobic component and a charged component. A solution of 8.9 mM DEM-3.5G and 2 × 10-7 M 1 displays a very broad and red-shifted emission band (Figure 8B), but is only quenched by 20% from that of a 1 solution. A new redshifted peak was observed in the absorption spectrum as well. The spectral changes indicate the formation of aggregates. Filtration of the 2 × 10-7 M 1 solution with 8.9 mM DEM-3.5G solution with a 200 nm filter caused the fluorescence to decrease by 95% and the absorbance to attenuate, corroborating the conclusion that large aggregates are formed. As with the ferric cytochrome-c system, FCS experiments on the aggregates showed large particles but could not be quantified. Hence, the combination of hydrophobic and electrostatic features into the quencher molecules leads to the formation of very large aggregates. These data indicate that the aggregation itself causes some fluorescence quenching, but it is weak compared to the quenching observed with cytochrome-c.
13208 Langmuir, Vol. 23, No. 26, 2007
Kaur et al.
Figure 8. A. Structure of DEM3.5G. B. Emission spectra of 1 in water (-) and in the presence of 8.9 mM DEM3.5G (---).
Conclusion These studies show how the interaction between 1 and different macromolecular quenchers changes the optical properties of 1. It is clear that the interaction of ferric cytochrome-c, PAMAM3G, and DEM3.5G induces aggregation of 1 in aqueous solutions and quenches the fluorescence of 1. In the case of the protein ferric cytochrome-c, the fluorescence is quenched by 98% and very large aggregates (>0.25 µm) are formed. The high quenching by ferric cytochrome-c results because it induces aggregation and provides a chromophore that can quench the fluorescence by energy transfer. Contributions from each of these factors were studied by observing the optical properties of 1 in the presence of different quenching partners. Experiments with PAMAM-3G and DEM3.5G showed that both electrostatic and hydrophobic interactions are important to induce aggregation in 1; however, these interactions are not sufficient to quench the fluorescence of 1 as effectively as ferric cytochrome-c, with its heme chromophore that acts as an energy transfer acceptor. Further, it appears that PEG-OH does not induce aggregation of multiple 1 strands but forms heteroaggregates with the 1 strands. These studies provide insight into the importance of electrostatic and hydrophobic interactions for the aggregation of polyelectrolytes. Because such multiple interactions will be present for analysis in complex biological systems, such aggregation will need to be controlled or mitigated before these materials prove useful as sensors. From these studies, it is clear that the sensitivity of the fluorescence of 1 to a quencher is affected by its aggregation state. When the polymer is unaggregated at low concentration (∼10-8 M), the solutions show their highest Stern-Volmer
constant, suggesting that the polymer is most sensitive when it is in its unaggregated form. This conclusion is corroborated by the ionic strength dependence studies of the Stern-Volmer constant. As the ionic strength increases, the polymer aggregates and the sensitivity decreases, indicating that the polymer is not very sensitive in its aggregated state. The studies with cytochromec, described above, show that the protein induces aggregation and supports the hypothesis that quencher-induced aggregation of 1 can explain its concentration-dependent Ksv. At low 1 concentration (independent strands), the addition of protein induces formation of aggregates of 1 with the protein; both the aggregate formation and the energy transfer quenching give rise to a highly sensitive response. In contrast, at high concentration the polymer strands are already aggregated and somewhat selfquenched, so that the addition of cytochrome-c causes a weaker quenching response than for the low concentrations. Whether the aggregates formed in the two cases (cytochrome-c induced aggregates and cytochrome-c associated with aggregates) are the same is not clear at this time. The fluorescence quenching of 1 depends on its aggregation state, which is a sensitive function of generic quencher properties and solution conditions. Acknowledgment. D.H.W. acknowledges support from the U.S. National Science Foundation (CHE-0415457). Supporting Information Available: Additional experimental material as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org. LA7023007
