Interaction between Encapsulated Excited Organic Molecules and

Jul 12, 2011 - File failed to load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js .... The nitroxide is kept free in aqueous solu...
2 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/Langmuir

Interaction between Encapsulated Excited Organic Molecules and Free Nitroxides: Communication Across a Molecular Wall Mintu Porel,† Steffen Jockusch,‡ M. Francesca Ottaviani,§ N. J. Turro,‡ and V. Ramamurthy*,† †

Department of Chemistry, University of Miami, Coral Gables, Florida 33124, United States Department of Chemistry, Columbia University, New York, New York 10027, United States § Department of Earth, Life and Environmental Sciences (DiSteVA), University of Urbino, Campus ex-Sogesta, Loc. Crocicchia, 61029 Urbino, Italy ‡

bS Supporting Information ABSTRACT: Communication between two molecules, one confined and excited (triplet or singlet) and one free and paramagnetic, has been explored through quenching of fluorescence and/or phosphorescence by nitroxides as paramagnetic radical species. Quenching of excited states by nitroxides has been investigated in solution, and the mechanism is speculated to involve charge transfer and/or exchange processes, both of which require close orbital interaction between excited molecule and quencher. We show in this report that such a quenching, which involves electronelectron spin communication, can occur even when there is a molecular wall between the two. The excited state molecule is confined within an organic capsule made up of two molecules of a deep cavity cavitand, octa acid, that exists in the anionic form in basic aqueous solution. The nitroxide is kept free in aqueous solution. 1H NMR and EPR experiments were carried out to ascertain the location of the two molecules. The distance between the excited molecule and the paramagnetic quencher was manipulated by the use of cationic, anionic, and neutral nitroxide and also by selectively including the cationic nitroxide within the cavity of cucurbituril. Results presented here highlight the role of the lifetime of the encounter complex in electronelectron spin communication when the direct orbital overlap between the two molecules is prevented by the intermediary wall.

’ INTRODUCTION Recently, we established through steady state and timeresolved EPR experiments the occurrence of electronelectron spin communication between two nitroxide molecules despite the intervening molecular wall between them.1,2 We also noted ultrafast singletsinglet energy transfer (FRET) between a confined excited donor and free acceptor.3 These two types of communications occur even when one of them is held tightly within a molecular capsule and the other resided outside in solution. Our recent studies on oxygen quenching (via exchange energy transfer process) of confined excited guest molecules revealed that interaction between acceptor oxygen and excited donor required partial opening of the capsule.4 Literature attributes the deactivation of excited molecules by nitroxides, similar to oxygen quenching, to either exchange and/or charge transfer processes, both requiring close contact between the excited guest and the nitroxide.511 This article deals with electronelectron spin communication between an excited molecule and a paramagnetic ground state molecule such as nitroxide separated by a well-defined molecular wall. The separation was accomplished by trapping the excited molecule within an organic capsule and positioning the nitroxide in the aqueous solution outside the capsule. The hostguest complex termed capsuleplex consists of two molecules of a cavitand commonly known as octa acid (OA; Scheme 1)12 and one or two molecules r 2011 American Chemical Society

of a guest (excited by irradiation). We have used three paramagnetic molecules, cationic (Tx), anionic (TQ), and neutral (T) nitroxides and eight guest molecules 18 (Scheme 1) to probe the insideoutside spin communication. Information on the location of nitroxide in relation to the capsule was obtained through its effect on the relaxation time of 1H NMR signals of the host and the guest in water. Electronelectron spin communication between the excited guest and the paramagnetic ground state molecules was monitored by the effect of the latter on the steady state emission and excited state decay kinetics of the confined guests. The size of the nitroxides, being much larger than oxygen, motivated us to examine if quenching of encapsulated excited molecules would require greater opening of the capsule or whether it would occur through a wall similar to radicalradical spin communication and singletsinglet energy transfer mentioned above. Triplettriplet energy transfer between a donor held within a carcerand and an acceptor in solution was previously established to be slower than that in free solution and to proceed without direct contact between the donor and acceptor.1315 This observation suggested that quenching by Received: June 6, 2011 Revised: July 9, 2011 Published: July 12, 2011 10548

dx.doi.org/10.1021/la202120u | Langmuir 2011, 27, 10548–10555

Langmuir

ARTICLE

Scheme 1. Structure of Hosts and Guests Examined in This Studya

a

The letters “A- J” in OA represent the corresponding protons.

nitroxide molecules also could occur across the capsular wall albeit at a much lower rate than in solution. We were interested in testing such a possibility and in fine-tuning the rate constant of quenching by controlling the lifetime of the encounter supramolecular complex. With the negatively charged capsule, use of three nitroxides with positive, negative, and no charge allowed us to vary the lifetime of the encounter complex. For examining the excited state spinradical spin transfer process, we carried out 1D and 2D 1H NMR, steady state and time-resolved absorption and emission, and EPR experiments. The results presented below establish that nitroxide molecules can indeed quench excited triplet molecules held within OA capsule, and the cationic nitroxide Tx can quench with much larger rate constants than either the neutral T or the anionic TQ. We show below that Coulombic interaction between the anionic capsule and Tx results in a longer lifetime for the encounter complex resulting in the quenching being essentially controlled by diffusion of the two species toward each other.

’ RESULTS AND DISCUSSION Capsuleplexes of guests 18 and host OA were prepared in water at pH ∼9.0 (borate buffer) and characterized by their 1D and 2D COSY and NOESY 1H NMR spectra. On the basis of 1H NMR titration experiments, we confirmed that 3 and 4 formed 2:2 (host to guest) complexes, while the rest formed 2:1

complexes. Readers are referred to an earlier publication of ours for details on characterization of these complexes.4 We provide below spectral details relating to location of the nitroxides with respect to capsuleplexes in water. In Figure 1, 1H NMR spectra of cavitand OA (with 1H signal assignments) and capsuleplex 1@OA2 and 1@OA2 in presence of Tx and Tx@CB8 are provided. The last one, Tx@CB8 (a representation of cavitandplex of Tx and CB8), refers to the inclusion of Tx within cucurbit[8]uril. Perusal of Figure 1 leads to the following conclusions: (a) Inclusion of 1 within the capsule as suggested by the upfield shift of the signal due to CH3 group of the guest. (b) Residence of the cationic nitroxide close to the capsule, probably in a preferred orientation as suggested by the broadening of the NMR signals of the host protons and CH3 of 1 to varying degrees by Tx (see Figure S1 in Supporting Information (SI) for NMR spectra with incremental addition of Tx). (c) For electron spin-proton spin communication, the nitroxide needed to be closer to the wall of the capsule as deduced from recovery of the line shape of the 1H NMR signals of the host and CH3 of 1 by cucurbit[8]uril (CB8) known to complex Tx. While broadening of NMR signals of OA by Tx was expected,16 differential broadening of signals due to various protons suggested Tx to adopt a specific orientation with respect to the capsule. To probe the location of Tx, we recorded the EPR spectra of Tx alone and in presence of 1@OA2 in water (Figure S2a and S2b in SI). By means of a computer-aided analysis of the spectra, we measured 10549

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir

Figure 1. 1H NMR spectra (500 MHz, D2O) of (a) OA, (b) 1@OA2, (c) 1@OA2+ Tx, and (d) 1@OA2+ Tx@CB8; [1] = 0.5 mM, [OA] = 1 mM, [Tx] = 1 mM, and [CB8] = 1 mM in 10 mM borate buffered D2O.

small variations in both the hyperfine coupling constant between the electron spin and the nitrogen nuclear spin (AN), which provided a measure of the environmental polarity of the probe.17,18 We also computed the rotational correlation time (τc, obtained by assuming a Brownian rotational diffusion model),19 which provided a measure of the interactions occurring between the probe and its environment. The very similar 16.8 and 16.7 G coupling constants for Tx in water and in the presence of OA, respectively, suggest Tx to be surrounded by water molecules even in the presence of OA, thus ruling out its inclusion within OA. Also, the relatively small change in rotational correlation time from 0.05 to 0.14 ns indicated Tx to be held weakly by the exterior walls of OA and not within it. We have shown previously that for a nitroxide encapsulated within a OA capsule the rotational correlation time increases from 0.05 to 1.35 ns.1 The observed 3-fold increase for Tx is much smaller than what is expected had Tx been included within OA capsule. To quantify the influence of nitroxide, relaxation times of each proton of the host and methyl proton of guest 1 in absence (T1,d) and presence (T1,obs) of nitroxides were measured by the inversionrecovery method. Values for 1/T1,p (relaxivity by nitroxide) for all protons calculated according to SolomonBloembergen equation are listed in Table 1 (for more details, see Tables S1 and S2 in SI).16,20,21According to the SolomonBloembergen equation, the relaxivity (1/T1,p) is proportional to the reciprocal of the sixth power of the distance between the proton and the paramagnetic center, in this case, nitroxide. The most important observation made is that Tx had a major influence on HA and HC protons of OA that are present at the wider rim of the OA cavitand. Its effect on all other host protons was smaller, suggesting that Tx stays closer to the middle region of the capsuleplex, probably being held by Coulombic interactions with benzoate anions (Figure 2). Consistent with this model, Tx had only a small effect on the methyl proton (of guest 1) positioned at the narrower end and far from the wider middle region of the capsule (Table 1). To test whether the observed

ARTICLE

relaxivity effect is due to closeness of Tx to the capsuleplex, Tx was moved away from side walls of the capsuleplex with the help of cavitand CB8 known to strongly bind Tx.22,23 As expected, upon addition of CB8, the 1/T1,p values for all host protons were much smaller (Table 1). Inclusion of Tx within CB8 was confirmed by its characteristic seven-line EPR spectrum (see SI Figure S2c). For details on the seven-line spectrum, readers are referred to earlier articles on this topic.24,25 Host systems such as calixarene and cyclodextrin were not as effective as cucurbituril in displacing Tx from the capsular exterior walls and therefore those results are presented in SI only (see Table S2 and Figures S3 and S4 for 1H NMR spectra and Figure S2d, S5, and S6 for EPR spectra in SI). In line with the expectation that the anionic nitroxide and capsuleplex would be repelled, TQ had smaller 1/T1,p values for all protons (last column in Table 1). More interestingly, its effect on HA and HC of OA was very small (compare with Tx in Table 1) suggesting that when TQ approaches the capsuleplex it avoids the wider rim substituted with benzoate anions prompting us to test the influence of neutral T. Perusal of 1/T1,p values listed in Table 1 for T reveals that its effect on all hydrogens are almost similar, and 1/T1,p values were much smaller than that for Tx and slightly larger than that for TQ. This suggests that T has no specific interaction with 1@OA2 and it approaches the capsule randomly and closer than TQ. The above observation that Tx is held in a specific geometry at the exterior of 1@OA2 was further supported by the effect of Txon the relaxation time of the CH3 protons of three substituted dibenzylketone guest molecules shown in Figure 3. We have established through extensive 1H NMR studies that these three dibenzylketone derivatives are held in the capsule in the geometry shown in Figure 3 (for 1H NMR spectra see Figure S7 in SI).26,27 It is important to recognize that the location of the CH3 group was distinctly different in each case, which could be inferred from the difference in 1H NMR chemical shift for the CH3 signal in CDCl3 in the absence of OA, and in water in the presence of 2 equiv of OA. The deeper the CH3 group moves from the middle of the capsule, the larger the Δδ for the CH3 group becomes (Figure 3). The Δδ value is the difference in 1H NMR chemical shift for CH3 signal in CDCl3 in the absence of OA, and in water in the presence of 2 equiv of OA. On the basis of the results presented for 1@OA2, we predict that the effect of Tx on the relaxivity of the CH3 groups of dibenzylketones (Figure 3) would be inversely related to Δδ values. As shown in Figure 3, this prediction is consistent with the observed results. We believe that the extent of relaxivity effects observed with the three systems is a reflection not only of the orientation of the nitroxide at the exterior of the capsule, but also of the lifetime of the encounter complex; the longer the lifetime, the larger the effect. On the basis of the extent and specificity of the influence of the three nitroxides on the relaxation of the protons, we reach the following conclusions: (a) The encounter complex between the capsuleplex and Tx has a preferred orientation and longer lifetime. (b) The encounter complex between capsuleplex and TQ is shorter-lived and the two anionic ends in the complex avoid each other. (c) Encounters between capsuleplex and T are random and the encounter complex has a lifetime intermediate between those of Tx and TQ. The above information allowed us to pursue our main goal of exploring spinspin communication between two molecules, one excited (triplet or singlet) and confined, and the other free and paramagnetic (organic radical). Guest molecules 15 with 10550

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir

ARTICLE

Figure 2. (a) Cartoon representation of [1@OA2]Tx complex and (b) partial chemical structure of OA. The letters AJ in OA represent relative positions of corresponding protons.

Table 1. 1H Relaxation Time (T1) for OA (Host) and 1 (Guest) and Relaxivity Parameters in the Absence/Presence of Nitroxide Radicals (Tx, T, and TQ; see Scheme 1) and CB8a T1,db

T1,obs c[Tx]

1/T1,pd[Tx]

1/T1,p [Tx] + CB8

1/T1,p [T]

1/T1,p [TQ]

HA

2.42

0.07

HB

0.56

0.23

14.28

0.2

2.86

0.07

2.56

0.13

1.33

HC

0.76

0.07

0.08

11.11

0.15

2.86

HD

0.11

0.35

0.15

2.85

0.03

2.38

0.37

HE HF

1.14 1.11

0.14 0.14

6.25 6.25

0.07 0.05

2.33 2.44

0.16 0.24

HI

0.38

0.19

2.63

0.2

1.22

0.15

HJ

0.35

0.12

5.56

0

1.49

0.26

HCH3

0.82

0.38

1.43

0.09

0.78

0.06

protons

The error limit on T1 values is (10% and on 1/T1 is between (10% and 12%. b T1,d = relaxation time in the absence of the paramagnetic center; Relaxation time for HG and HH could not be measured as these proton signals merged with residual H2O signal. c T1,obs = observed relaxation time in the presence of paramagnetic center. d T1,p = relaxation time caused by the paramagnetic species. 1/T1,p = 1/T1,obs  1/T1,d. 1/T1,p µ r6, r = distance between nucleus and paramagnetic center. [1] = 0.5 mM, [OA] = 1 mM, [Tx] = 0.5 mM, [CB8] = 0.5 mM, [T] = 0.5 mM, and [TQ] = 0.5 mM in 10 mM borate buffered D2O. a

very high rate constants of intersystem crossing from S1 to T1 show phosphorescence at room temperature in aqueous solution, while guests 68 with low rate constants for the same process only fluoresce. Quenching of triplet states in pyrene was monitored by its TT absorption due to its lack of phosphorescence. Generally, excited singlet and triplet states of aromatics and carbonyl compounds are quenched by nitroxides in solution at room temperature with rate constants >2  109 M1 s1.5,2830 Quenching details for molecule 4,40 -dimethylbenzil (1) as a representative of all guests are presented below. Experimental data related to deactivation by Tx for all guests (18) are provided as Figures S8 to S20 in SI. The room temperature emission spectrum of 1@OA2 in water (borate buffer) consisting of fluorescence, delayed fluorescence (480 to 520 nm), and phosphorescence (530 to 670 nm) is shown in Figure 4a. As illustrated in the figure, the phosphorescence intensity is significantly reduced in presence of nitroxide Tx compared to that in nitrogen saturated buffer solution. The emission intensities

depend on the concentration of Tx in water (as notable from Figure S8 in SI). Consistent with the steady state emission data, excited triplet state decay profiles in Figure S9 in SI also exhibit Tx concentration dependence. The rate constants for quenching were obtained from the slope of the plots of the pseudo-firstorder decay rate constants (1/τ) of 4,40 -dimethylbenzil triplet states vs the Tx concentration (for phosphorescence decay traces at various concentrations of Tx, see Figure S9 in SI). Quenching rate constants obtained by quenching of phosphorescence intensity and decay are identical within limits of experimental error ((1.5 ( 0.2)  109 M1 s1). Decay traces provided in Figure S9 (SI) suggest that the quenching is a dynamic process. Had it been a fully static process, no decrease in triplet lifetime would have been observed with increasing Tx concentration. The linear lines with nearly identical slopes in SternVolmer plots with points taken from both steady state emission and lifetime data (Figure 4c) are consistent with a dynamic process and lend additional support. To further confirm that the quenching is a dynamic process and the quencher Tx is not permanently held by 1@OA2, we 10551

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir

ARTICLE

Figure 3. Cartoon representations of (a) R-methyl dibenzyl ketone@OA2; (b) para-methyl dibenzyl ketone@OA2; and (c) R-octyl dibenzyl ketone@OA2. The red ball represents the CH3 group at the corresponding guest molecule. Relaxation times for CH3 protons are given. Δδ is the difference in 1H NMR chemical shift for CH3 signal in CDCl3 in the absence of OA, and in water in the presence of 2 equiv of OA.

Figure 4. (a) Phosphorescence quenching of 1@OA2 by Tx and recovery of phosphorescence upon addition of CB8; [1] = 1  105 M, [OA] = 2  105 M, [Tx] = 2  105 M, and [CB8] = 2  105 M; (b) comparison of 1@OA2 quenching by Tx, T, and TQ; [Tx] = [T] = [TQ] = 2  105 M; and (c) SternVolmer plot for phosphorescence quenching of 1@OA2 by Tx using steady-state phosphorescence intensities (I0/I; red ball) and phosphorescence lifetimes (τ0/τ; blue triangle) quenching constant kq = (1.5 ( 0.2)  109 M1 s1, λex = 320 nm.

recorded the phosphorescence of 1@OA2 and Tx in presence of a cavitand CB8 known to strongly bind to Tx and form a cavitandplex Tx@CB8. We believe the full recovery of the 1@OA2 phosphorescence quenched by Tx on addition of CB8 as illustrated in Figure 4a is due to the inability of Tx@CB8 to interact with excited states of 1 included within OA capsule, thereby suggesting that only free Tx located at the exterior of OA capsule can quench the triplet of 1. Examination of quenching data in Table 2 reveals two sets of molecules, one (triplet 14 and 6) that is quenched mostly by a

dynamic process by Tx when enclosed within OA capsule and the other (triplet 5 and excited singlet 68) which when enclosed within OA capsule exhibits quenching by Tx from a static process only. All quenching constants in Table 2 are estimated from triplet decays (monitored by phosphorescence or T-T absorption) and thus correspond to a dynamic process. Decay curves for quenching of 18 by Tx are presented in SI (Figures S9, S11, S13, S15, S16b, S17b, S19b, and S20b). From the data presented in SI (Figures S8 to S20), it should be clear that in the case of triplet 14 and 6 both emission intensities and 10552

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir

ARTICLE

Table 2. Excited State Lifetime (τ0) and Nitroxide (Tx and T, Scheme 1) Quenching Data for Capsuleplexes Made up of Octa Acid (OA) and Organic Guests 1-8 Listed in Scheme 1 excited guests

state type

6 1

T1 T1

2 3 4 5 6 7 8

T1 T1 T1 T1 S1 S1 S1

life time in OA (μs) 922 596 151 65 17 4.8 0.34 0.05 0.01

kq (Tx) in

quenching by

kq (T) in OA

kq (T) in acetonitrile

OA (M1 s1)

T in OA (%)b

(M1 s1)

(M1 s1)

c

77

(2.3 ( 0.1)  109 (1.5 ( 0.2)  109

17

(1.6 ( 0.5)  107 (2 ( 1)  108

2.3  109

63

(1.1 ( 0.2)  10

10

(4 ( 0.8)  10

3.5  109

43

(1 ( 0.6)  10

29

quenching by x

a

T in OA (%) c

9

7

c

8

(7 ( 2)  10

d

(2 ( 0.5)  10

7

(6 ( 2)  10

d

13

e

0

-------

d

13

e

0

-------

1  1010

11

e

0

-------

1.3  1010

15

e

0

-------

9 9

7 7

1  1010

x b

Determined by comparing the luminescence intensity (λmax) of the capsuleplex recorded in absence and presence of 1 μM of T . Determined by comparing the luminescence intensity (λmax) of the capsuleplex recorded in absence and presence of 5 μM of T. c Since 6 did not show phosphorescence at room temperature, measurements could not be made. d kq could not be determined because of self-quenching of thioketones in solution. e kq could not be determined because the lifetime of the guest molecules did not change with addition of Tx. a

triplet decays (τT) are dependent on the concentration of Tx, while in the case of triplet 5 and excited singlet 68, only the emission intensity and not the excited state lifetime is influenced by Tx. This information suggests that quenching of triplet 14 and 6 is a dynamic process and that of triplet 5 and excited singlet 68 is a static process. Closer examination of Table 2 suggests that the lifetime of the excited state is likely to play a role in the overall deactivation of excited states by Tx. Guest molecules whose lifetimes are shorter than 5 μs are quenched only by a static process when they are enclosed within OA capsule. It is also important to note that these guest molecules, whose lifetimes span between 17 and 922 μs, are quenched by Tx with almost identical rate constants. To fully understand the quenching rate constant, we need to know what exactly the measured kq represents. The deactivation of excited guest by Tx listed in Scheme 2 involves the formation of the encounter complex between Tx and excited guest-OA capsuleplex (controlled by kdiff; diffusion constant), its decomplexation without quenching (controlled by kdiff; diffusion out of the encounter complex without deactivation), and deactivation of the excited guest by Tx within the encounter complex (kdeact).3133 Under these conditions, the measured quenching rate constant kq is given by eq 1 (Scheme 1. While kdiff would be the diffusion rate constant in water, which is 7.4  109 M1 s1 at 25 °C, the value of kdiff would be expected to be much smaller than kdiff due to the strong Coulombic attraction between negatively charged capsule and positively charged Tx. If we assume that kdeact . kdiff, then kq would be given by eq 2. Under this condition, taking spins of the triplet excited guest and the paramagnetic quencher into account, the measured quenching constant kq would be expected to be (1/3)kdiff. The quenching of the triplet by a doublet nitroxide will result in doublet and quartet state collision complexes at the appropriate spin statistical ratios of 1/3 and 2/3. Of these two types of collision complexes, only the doublet complex formed at (1/3)kdiff will lead to quenching. It is gratifying that the measured quenching constants (kq) for triplets of 14 and 6 are in this range. Thus, in the case of positively charged quencher, Tx, 1/3 of encounter leads to quenching. Since the lifetime as well as the intensity of emission were dependent on the concentration of Tx, quenching must be dynamic in character. We wish to point out that the quenching by

Scheme 2

Tx is not limited by capsule opening as in the case of oxygen quenching studied previously by us. Had the capsule opening been the limiting process, observed quenching should be in the range of 105 to 107 M1 s1 and also should have varied with the guest molecule.4 The fact that the observed quenching rate constants for 14 and 6 are in near-diffusion-controlled rate constant suggests that the quenching is limited only by the rate constant for the quencher and the capsuled guest to meet each other and not by capsule opening. Unlike the systems discussed above, the quenching of phosphorescence of 5 and fluorescence of 68 by Tx without any effect on their excited state lifetimes suggests these quenching to be static processes. Percentages of quenching for the same 10553

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir concentration of Tx are provided in Table 2. The varying % quenching with the lifetime of the excited states in the first set of compounds (triplet 14 and 6) and its constancy (∼13%) in the second set (triplet 5 and excited singlet 68) suggested that the majority of the quenching observed with the first set of guests is dynamic while that with the second set is static. Thus, when the excited state decay constant (kdecay) is larger than kq[Tx] the quenching by Tx cannot compete with internal conversion. This indirectly suggests that the capsular wall might have slowed the rate constant of deactivation (kdeact) by nitroxide with respect to that in solution. The above results suggest that Tx is able to quench with an apparent high rate constant only because kdiff is slow due to the strong Coulombic attraction between the negatively charged capsuleplex and the positively charged quencher that prolongs the lifetime of the encounter complex. If these conclusions were correct, use of neutral quencher such as T should change the scenario. Phosphorescence spectra of 1@OA2 alone and in the presence of Tx, T, and TQ are provided in Figure 4b. While Tx quenches significantly, TQ quenches the least. As expected, the ability of T to quench is in between those by Tx and TQ. Rate constants for quenching of the excited states of 18 included within OA capsule by the neutral nitroxide (T) are provided in Table 2 (see Figures S21 to S33 in SI for data). Triplet 5 and excited singlets 68 were not quenched by T, even by the static process. The quenching by T of triplet 14 and 6 included within OA capsule was approximately an order of magnitude lower than that by Tx. We attribute this to the difference in kdiff values between Coulombically held1 and {[email protected]} held by van der Walls forces. In solution where the excited aromatics and ketones are free, the quenching by neutral nitroxides is nearly diffusion limited and the quenching numbers for guests enclosed within OA capsule are about two orders of magnitude lower. This lower number is consistent with the expected lower rate of quenching by a mechanism requiring overlap between the excited guest and the quencher (charge transfer and/or exchange process). The reduction observed here is almost similar to that reported for triplettriplet energy transfer between the excited biacetyl incarcerated within Cram’s carcerand and free pyrene in benzene. Results observed here are consistent with the conclusion that quenching of excited states by nitroxides can occur in spite of them being separated by a molecular wall and the lack of direct orbital overlap between these two molecules. The upper limit for the quenching of triplet 14 and 6 by anionic nitroxide (TQ) was estimated to be ∼105 M1 s1.

’ CONCLUSION Deactivation of electronically excited (singlet and triplet) molecules by nitroxides in solution is known to occur at close to diffusion limits, and the mechanism, although not clearly established, is believed to involve exchange and/or charge transfer interaction. Deactivation by both mechanisms requires close interaction between the excited molecule and nitroxide. We have hereby shown that excited molecules enclosed within a negatively charged (COO) organic container could be deactivated by a cationic nitroxide at nearly diffusion controlled rate. Clearly, the presence of a wall between the excited molecule and nitroxide prevented neither exchange nor charge interaction between them. The high rate constant of deactivation is attributed to the slow rate constant of decomplexation of the encounter complex due to Coulombic interaction between the

ARTICLE

anionic capsule and cationic nitroxide. Thus, spinspin communication between an excited molecule and a free radical can occur even without direct overlap of their orbitals. Close interaction between these species has been controlled with the help of a second container (cucurbituril). The presence of a double wall between the excited molecule and the nitroxide completely arrested the deactivation process. Of the various secondary containers used (cucurbituril, calixarene, and cyclodextrin), the best control was achieved with cucurbit[8]uril. Thus far, we have shown communication between: (1) a confined excited singlet or triplet donors and a free acceptor (singletsinglet and triplet triplet energy transfers respectively), (2) a confined radical and a ‘free’ radical (spinspin exchange), and (3) an excited molecule and a ‘free’ radical despite the presence of an organic molecular wall between them. Our results on electron transfer across the octa acid molecular wall will be presented soon.

’ ASSOCIATED CONTENT

bS

Experimental details, 1H NMR, emission spectra, and decays of excited states as monitored by their emission or transient absorption. This material is available free of charge via the Internet at http://pubs.acs.org. Supporting

Information.

’ AUTHOR INFORMATION Corresponding Author

*[email protected].

’ ACKNOWLEDGMENT V.R. and N.J.T. are grateful to the National Science Foundation, USA, for generous financial support (CHE-0848017 and CHE-0717518, respectively). ’ REFERENCES (1) Chen, J. Y.-C.; Jayaraj, N.; Jockusch, S.; Ottaviani, M. F.; Ramamurthy, V.; Turro, N. J. J. Am. Chem. Soc. 2008, 130, 7206–7207. (2) Jockusch, S.; Zeika, O.; Jayaraj, N.; Ramamurthy, V.; Turro, N. J. J. Phys. Chem. Lett. 2010, 1, 2628–2632. (3) Gupta, S.; Adhikari, A.; Mandal, A. K.; Bhattacharyya, K.; Ramamurthy, V. J. Phys. Chem. C 2011, 115, 9593–9600. (4) Jayaraj, N.; Jockusch, S.; Kaanumalle, L. S.; Turro, N. J.; Ramamurthy, V. Can. J. Chem. 2011, 89, 203–213. (5) Gijzeman, O. L. J.; Kaufman, F.; Porter, G. J. Chem. Soc., Faraday Trans. 2 1973, 69, 727–737. (6) Hoijtink, G. J. Mol. Phys. 1960, 3, 67–70. (7) Hoytink, G. J. Acc. Chem. Res. 1969, 2, 114–120. (8) Murrell, J. N. Mol. Phys. 1960, 3, 319–329. (9) Watkins, A. R. Chem. Phys. Lett. 1980, 70, 262–265. (10) Kobori, Y.; Takeda, K.; Tsuji, K.; Kawai, A.; Obi, K. J. Phys. Chem. A 1998, 102, 5160–5170. (11) Mitsui, M.; Kobori, Y.; Kawai, A.; Obi, K. J. Phys. Chem. A 2004, 108, 524–531. (12) Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126, 11408–11409. (13) Farran, A.; Deshayes, K. D. J. Phys. Chem. 1996, 100, 3305–3307. (14) Romanova, S. Z.; Deshayes, K.; Piotrowiak, P. J. Am. Chem. Soc. 2001, 123, 11029–11036. (15) Parola, J. A.; Pina, F.; Ferreira, E.; Maestri, M.; Balzani, V. J. Am. Chem. Soc. 1996, 118, 11610–11616. (16) Kosen, P. A. Methods Enzymol. 1989, 177, 86–121. 10554

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555

Langmuir

ARTICLE

(17) Knauer, B. R.; Napier, J. J. J. Am. Chem. Soc. 1976, 98, 4395–4400. (18) Ottaviani, M. F.; Martini, G.; Nuti, L. Magn. Reson. Chem. 1987, 25, 897–904. (19) Budil, D. E.; Lee, S.; Saxena, S.; Freed, J. H. J. Magn. Res. A 1996, 120, 155–189. (20) Li, Y.; Lei, X.; Lawler, R. G.; Murata, Y.; Komatsu, K.; Turro, N. J. J. Phys. Chem. Lett. 2010, 1, 2135–2138. (21) Solomon, I.; Bloembergen, N. J. Chem. Phys. 1956, 25, 261–266. (22) Lagona, J.; Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L. Angew. Chem., Int. Ed. 2005, 44, 4844–4780. (23) Lee, J. W.; Samal, S.; Selvapalam, N.; Kim, H.-J.; Kim, K. Acc. Chem. Res. 2003, 36, 621–630. (24) Jayaraj, N.; Porel, M.; Ottaviani, M. F.; Maddipatla, M. V. S. N.; Modelli, A.; Da Silva, J. P.; Bhogala, B. R.; Captain, B.; Jockusch, S.; Turro, N. J.; Ramamurthy, V. Langmuir 2009, 25, 13820–13832. (25) Da Silva, J. P.; Jayaraj, N.; Jockusch, S.; Turro, N. J.; Ramamurthy, V. Org. Lett. 2011, 13, 2410–2413. (26) Sundaresan, A. K.; Ramamurthy, V. Photochem. Photobiol. Sci. 2008, 7, 1555–1564. (27) Gibb, C. L. D.; Sundaresan, A. K.; Ramamurthy, V.; Gibb, B. C. J. Am. Chem. Soc. 2008, 130, 4069–4080. (28) Schwerzel, R. E.; Caldwell, R. A. J. Am. Chem. Soc. 1973, 95, 1382–1389. (29) Green, J. A., II; Singer, L. A. J. Am. Chem. Soc. 1974, 96, 2730–2733. (30) Green, J. A., II; Singer, L. A.; Parks, J. H. J. Chem. Phys. 1973, 58, 2690–2695. (31) Evans, T. R. Singlet quenching mechanisms. J. Am. Chem. Soc. 1971, 93, 2081–2082. (32) Turro, N. J.; Lehr, G. F.; Butcher, J. A., Jr.; Moss, R. A.; Guo, W. J. Am. Chem. Soc. 1982, 104, 1754–1756. (33) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular Photochemistry of Organic Molecules; University Science Books: Sausalito, CA, 2010.

10555

dx.doi.org/10.1021/la202120u |Langmuir 2011, 27, 10548–10555