3948
OPTICS LETTERS / Vol. 37, No. 19 / October 1, 2012
Optical limiting as result a of photoinduced electron transfer in hybrid systems with CdSe/ZnS quantum dots, C60, and Perylene Vladimir V. Danilov,1,2,* Anastasia S. Panfutova,1 Artem I. Khrebtov,1 Stefano Ambrosini,3 and Dmitry A. Videnichev1 1 2
S. I. Vavilov State Optical Institute, 199034 St. Petersburg, Russia
St. Petersburg State Transport University, 191031 St. Petersburg, Russia 3 University of Trieste, 34100 Trieste, Italy *Corresponding author:
[email protected]
Received June 27, 2012; revised August 1, 2012; accepted August 9, 2012; posted August 10, 2012 (Doc. ID 171516); published September 18, 2012 We investigated the interaction of an intense laser radiation with colloidal solutions containing CdSe/ZnS core shell quantum dots (QDs; mean size 3.4 nm), fullerene C60 , and Perylene. These materials would give rise to the photoinduced electron transfer and charge separation on the QDs and thus the optical limiting effect. Results confirm the intended aim, obtained by means of intermediate metastable products of reversible photochemical reactions, i.e., ion radicals of hybrid systems containing semiconductor nanoparticles. © 2012 Optical Society of America OCIS codes: 160.4236, 190.4720.
Optical limiting (OL) is a nonlinear effect, according to which above a threshold value, the intensity of a light beam at the output of a nonlinear device (the optical limiter) remains constant with increasing intensity input. This effect is exploited in devices to protect optical sensors (and human eyes). The most important mechanisms explaining OL for molecular systems are the “reverse saturation absorption” (RSA) and the “two-photon absorption” [1]. OL occurs thanks to the existence of metastable states that have an absorption cross section greater than the ground state, at the wavelengths of the incident laser radiation. The most studied OL systems are mixtures of fullerenes and similar compounds [1]. We published in our previous papers that ion radicals, in the form of metastable reaction products of photoinduced electron transfer, could extend the spectral range and improve the efficiency of the OL effect [2,3]. This was shown to depend on the difference between the absorption spectra of the ion radicals and those of their parent compounds. It was already shown that thin films mixtures of semiconductor quantum dots (QDs) and fullerene C60 give rise to charge separation during the resonance excitation in the range of QD excitonic transition [4,5]. In these cases, electron transfer is possible with the formation of ion radical pairs, where QDs act as electron donor and C60 plays the role of electron acceptor. In [4,5] the efficiency of charge separation provided with the external field. In this regard, photoactive systems based on nanostructures with efficient charge separation in absence of external fields are very interesting, especially for OL. Such nanoreactors could be realized in the form of the so-called “contact complexes,” on the surface of semiconductor nanoparticles in solutions under solubilization [6]. Perylene (Per) is a good electron donor and former studies showed the possibility of electron transfer from Per to the QD [7]. This would help to increase the number of nonrecombining radical anions C60 on the QDs. The maximum of the absorption band for Per• lies at λ 540 nm, where the extinction coefficient ε of Per is 5 × 10−4 cm−1 M−1 [8]. This makes Per• the ideal 0146-9592/12/193948-03$15.00/0
candidate to act as optical limiter for the wavelength of the second harmonic of a YAG:Nd3 laser. For our experiments we used the following solubilization agents: fullerene C60 (“Astrin”, Russia), Perylene (Merck), and CdSe/ZnS core-shell QDs (Institute of Physical and Chemical Problems of Belarus State University). The mean size of the QDs was 3.4 nm, their absorption peak lied at λ 540 nm. All experiments were performed on toluene solutions with stock concentrations of 4 × 10−4 M for C60 , 5 × 10−3 M for Per and 10−6 M for the QDs. Four solutions were tested: the first and third solutions contained pure materials, QDs (solution 1) and C60 (solution 3). The second and fourth solutions contained mixtures, respectively, 0.75 C60 0.25 QDs (solution 2) and 0.75 C60 0.25 QDs 1.1 mM Per (solution 4). Tested solutions were put in a 1 cm quartz cell (linear transmittance 45%) and placed in the focal plane of a system consisting of two lenses with 37 mm focal length. The main light source was a frequency-doubled Q-switched pulse-periodic YAG:Nd3 laser (λ 532 nm, τpulse 7 ns, beam diameter 8 mm, repetition rate up to 10 Hz). Laser power was measured by means of a standard photometer. Calibrated neutral glass filters allowed variation of the radiation power at the cell input. The power of transmitted pulses was monitored within 1.5 mrad angle by the detection unit, which included a collecting lens with a focal length of 650 mm, a 1 mm wide aperture, and a silicon detector. Absorption and fluorescence of stock solution were measured by means of a UV-probe 3600 spectrophotometer and MPF-44 spectro-fluorimeter. The fluorescence of stock QD solution showed biexponential attenuation with time constants τ1 3.5–4 ns and τ2 ≈ 18 ns. The first, short component is usually associated with the fundamental excitonic transition in QDs. The second, longer component of the fluorescence curve is still discussed in literature [9]. Figure 1 shows the absorption spectra of tested solutions. The wavelength of the exciting laser falls into the resonant absorption band range of the exciton transition © 2012 Optical Society of America
October 1, 2012 / Vol. 37, No. 19 / OPTICS LETTERS
Fig. 1. Absorption spectra of tested solutions. Curve 1: CdSe/ZnS QDs; curve 2: fullerene C60 ; and curve 3: Perylene. In all cases, the solvent is toluene.
(λabs 530 nm), while the absorption band for Per lies outside this range, around 495 nm. Figure 2 shows transmittance and OL effect of tested solutions upon laser excitation. Curve 1 refers to solution 1 (QDs). We can interpret this curve by the RSA model [1]. When QDs in the ground state absorb a photon, they undergo the reaction 1 S3 ∕2 h-1 Se. When the ground state becomes depleted, with increasing impinging energy density, we see an increased transmittance of the solution (bleaching) [10]. From the excited state 1 Se, transitions to higher excited states n Se are possible, and if the cross section of this reaction is greater than the cross section of the excitonic transition, i.e., σ IN > σ 01 , then OL is observed. Curve 3 refers to solution 3 (fullerene C60 ) and shows the classical OL effect played by C60 . Curve 2 refers to solution 2 (0.75 C60 0.25 QDs). We see the disappearance of the transmittance peak visible in curve 1. This is due to the deactivation of the 1 Se state, because of quicker electron transfer to C60 and subsequent recombination of electron-hole pair. Curve 4 refers to solution 4 (0.75 C60 0.25 QDs perylene 1.1 mM) and shows significantly increased OL. This can be explained as following. Molecules of Per passivate the QD surface; this raises the local concentration of donor-acceptor pairs and thus improves the OL efficiency. The formation of contact
Fig. 2. (Color online) Dependence of the solution transmittance on the energy of the incident radiation Ein J λexit 532 nm. 1: CdSe/ZnS QDs; 2: 0.75 C60 0.25 QDs; 3: C60 ; 4: 0.75 C60 0.25 QDs Per1.1 mM. The initial transmittance of all solutions is 45%.
3949
complexes between QDs and Per was demonstrated by preliminary results on the quenching of the photoluminescence after the introduction of Per in the QD solution. The analysis of Stern–Volmer curves for the binary systems QD C60 and QD Per (not shown) confirms that the quenching time constants of the QD photoluminescence do not depend on the excitation wavelength but their value in the range [103 –105 M−1 ] depends on used materials. Please notice that the values for the quenching time constants substantially exceed the typical values for the diffusion mechanism of photoluminescence quenching [11]. To confirm these data, parallel time resolved photoluminescence measurements on QD solutions were performed by means of the scanning laser microscope “PicoQuant.” We added stock Per solution to QD solution and verified a sharp quenching effect on the dark component of QD photoluminescence. Its time constant decreased from about 18 ns down to 4–5 ns. We interpret this fact as the formation of contact complexes among the QDs and Per: the passivation of the QD surface by Per molecules is achieved through mutually induced dipole forces (dispersion forces) between the reactants. Such a hybrid system QDs; C60 ; Per creates a nanopseudophase (Fig. 3), which enhances the charge separation efficiency on the QDs. We propose the following reaction to take place: • QDs; C60 hν → C•− 60 ; QDs • • Per → C•− C•− 60 ; QDs 60 QDs Per
(square brackets indicate the contact-complex in chemistry notation). We propose the formation of the radical cation Per• to be the reason of the observed increment in the limiting efficiency at λ 532 nm. Please note that the radical anion C•− 60 absorbs radiation in the value range [0.9–1.1 μm] (ε 18000 M−1 cm−1 [7]) and this significantly extends the spectral range of the optical limit. To test such reaction course we performed the experiment, whose result is shown in Fig. 4. We shined solution 4 with the laser beam (λ 1064 nm, energy E 1064 10−3 J, pulse duration τ 7 ns) and measured the transmittance of the solution (Fig. 4, curve 1). Then we shined the solution with a second light (λ 532 nm,
Fig. 3. Graphical representation of the photochemical processes occurring in the semiconductor nanophotoreactor under intense light exposure.
3950
OPTICS LETTERS / Vol. 37, No. 19 / October 1, 2012
of both the first and second harmonics of a YAG:Nd3 laser. This extends the spectral range of the optical limiting effect already achieved by a mixture of QDs; C60 ; Per solutions. We proposed a mechanism of photoreaction course and fully corroborated it. The authors thank Dr. G. M. Ermolaeva for the assistance granted during the measurements.
Fig. 4. Dependence of the transmittance of solution 4 (0.75 C60 0.25 QDs Per 1.1 mM), under different illumination conditions. Curve 1: only λ 1064 nm. Curve 2: combined action of λ 532 nm and λ 1064 nm light sources.
energy E 532 ∼ 10−5 J) and the transmittance dropped at half of the previous value (Fig. 4, curve 2). These results clearly indicate that the laser light at λ 1064 nm is not absorbed by the tested solution in absence of the secondary light beam at λ 532 nm. In other words, when the tested solution was illuminated by the secondary light radiation at λ 532 nm, the intensity of both the first and second harmonics of the laser dropped at half of the original value, at the output of the cell. Such a result fully confirms the proposed course for the photoreaction. In conclusion, we showed that the photoinduced electron transfer reaction from Perylene molecules to CdSe/ZnS core shell QDs effectively limits the intensity
References 1. J. D. Swalen and F. Kajzar, Nonlinear Opt. 105, 2981 (2001). 2. M. V. Gryaznova, V. V. Danilov, O. V. Khapova, A. I. Khrebtov, and T. A. Shakhverdov, Quantum Electron. 34, 407 (2004). 3. M. V. Gryaznova, V. V. Danilov, and T. A. Shakhverdov, Opt. Spectrosc. 105, 874 (2008). 4. A. P. Brown and P. V. Kamat, J. Am. Chem. Soc. 130, 8890 (2008). 5. R. C. Shallcross, G. D. D’Ambruoso, B. D. Korth, H. K. Hall, Jr., Z. Zheng, J. Pyun, and N. R. Armstrong, J. Am. Chem. Soc. 129, 11310 (2007). 6. V. V. Danilov, Opt. Zh. 74, 65 (2008). 7. D. V. Konarev, N. V. Drichko, V. N. Semkin, and A. Graja, Synth. Met. 103, 2384 (1999). 8. T. A. Shahverdov, Opt. Spectrosc. 29, 315 (1970). 9. A. O. Orlova, V. E. Andrianov, V. G. Maslov, P. S. Parfenov, and A. V. Baranov, Opt. Spectrosc. 108, 807 (2010). 10. S. Dneprovsky, E. A. Zhukov, M. V. Kozlova, T. Wumaier, D. S. Hieu, and M. V. Artemyev, Solid State Phys. 52, 1809 (2010). 11. J. A. Barltrop and J. D. Coyle, Excited States in Organic Chemistry (Wiley, 1978).
