Open Access
Add to BibSonomyAbstract
Trans-to-cis photoisomerization efficiency of azobenzene dye is artificially modified from 0.09 to 0.38 when dye molecules are placed close to gold nanoparticle films with different structures. Transient fluorescence and surface enhanced Raman scattering measurement verify that the enhancement and reduction of photoisomerization efficiency come from the competition between enhanced local optical field from surface plasmon resonance and the accelerated nonradiative decay of excited dye molecules. The photoisomerization efficiency can be further modified by controlling the distance between azobenzene dye and gold films. Our finding can be applied to improve the performance of photoisomerization effect in photochemistry and photonics.
©2011 Optical Society of America
1. Introduction
Reversible trans-cis photoisomerization exists widely in organic materials. Upon absorbing a photon, molecules isomerize from the thermally stable trans isomers to metastable cis isomers and vice versa. Based on the significant differences of two isomers in absorption spectra and optical nonlinearity, such photo-sensitive molecular structure change gives rise to applications in molecular switch, optical storage, protein folding triggering and holography [1–4]. In principle, performance improvement in these applications is expectable if the efficiency of photoisomerization can be artificially optimized. Block of trans-to-cis photoisomerization was observed by depositing azobenzene dye onto surface of gold film [5]. The block can be removed when azobenzene dye is lifted away from gold film surface. By exciting the molecule with a phase and amplitude re-shaped femtosecond laser, a 15% enhancement or 3% reduction of photoisomerization efficiency of cyanine dye were also achieved [6]. However, to our knowledge, it is still hard to continuously modify photoisomerization efficiency in a large range with simpler technical requirement.
In this paper, we report our approach to significantly manipulate photoisomerization quantum efficiency (PQE) by using metallic nanostructure substrates. The surface plasmon resonance (SPR) of gold nanoparticles enhances the exciting pump field, meanwhile the energy transfer from azobenzene dye to gold particles accelerates the decay of excited molecules back to thermal-stable trans isomers. Combination of SPR and energy transfer thus changes the PQE. We also show that by controlling the distance between gold nanoparticle film and azobenzene dye, the PQE can be precisely modified.
2. Experiment method and results
Two kinds of gold nanoparticle films are synthesized. The gold island film with particle size about 30 nm was produced by directly sputtering gold onto glass substrate; on the other hand, self-assembled gold films with particle size of 10 nm, 30 nm and 50 nm were prepared according to literature methods [7,8].
Figure 1(a) and 1(b) are the Atomic Force Microscopy (AFM) images of self-assembled gold film with particle size of 30nm and gold island film. Figure 1(c) shows the size dispersion of 30nm gold nanoparticles. Significant difference can be observed. For self-assembled gold film, the particles are closely packed, and for gold island film, there are clear dark areas, indicating that particles are not tightly touched; some are even separated from each other. The occupancy of Au is about 0.75, estimated from the image. As a result, we will see below that the corresponding SPR of the structures differs from each other.
Fig. 1 (a) AFM image of self-assembled film of 30nm-gold particles. (b) AFM image of gold island film. (c) Size dispersion of 30nm gold nanopartilces (d) The absorption coefficients of gold island film(blue) and self-assembled films of 10nm(black), 30nm(red) and 50nm(green) gold particles. The red dash line indicates the wavelength of 532nm.
The absorptions of all gold films are measured with a clean bare glass substrate as the reference. The absorption coefficient spectra were then obtained by dividing the measured absorbance with the thickness of film and the occupancies (0.75 for island film, 1.00 for self-assembled films). The results are plotted in Fig. 1(d). A clear absorption peak around 530 nm shows up for gold island film, indicating the uncoupled SPR. On the contrary, for self-assembled gold films, the SPR around 530 nm is suppressed due to the severe dipole-dipole interaction between touched particles, leading to an additional absorption peak in the near infrared. We will show later that the SPR strength around 530 nm is a key factor to achieve photoisomerization efficiency manipulation.
The azobenzene dye we used here is Disperse Red 1 (DR1). 10 wt% DR1 doped Poly (methyl methacrylate) (PMMA) thin films with thickness about 20 nm were prepared by dip coating. The absorption spectra of DR1 films on Au films do not have noticeable difference compared to the film on glass substrate (hereafter called pure DR1).
Rau’s Method, which is a pump-probe technique, was used to measure the photoisomerization quantum efficiency (PQE) ϕtc [9]. According to Rau, the reciprocal of absorbance change ratio A 0/(Apss-A 0) at pump intensity Ip is:
Here A 0 is the initial absorbance without pump and Apss is the absorbance after the system reaches photo-stationary state at pump Ip. εt and εc are the extinction coefficients of trans and cis isomers. εt and A 0 can be obtained through absorbance measurement without pump. ϕtc and ϕct are PQE of trans-to-cis and cis-to-trans isomerization respectively. τc is the lifetime of cis isomer. The equation tells that A 0/(Apss-A 0) changes linearly with 1/F, i.e., inverse of pump intensity. ϕtc can be deduced from the slope of the linear curve if other parameters like τc and εc are known (all of PQE mentioned below refers to ϕtc). We simplify the process by assuming that (εc-εt) is not sample-dependent, because experimentally we found that εt is almost the same in different environment and εt is always much larger than εc at λ = 532nm, the pump and probe wavelength. Then ϕtc can be deduced by comparing the slope with that of pure DR1.Fisher’s method was applied to deduce ϕtc of pure DR1 [10]. The principle of Fisher’s method is to do Rau’s measurement at two independent combinations of pump-probe wavelengths. A thin film of DR1 in PMMA on glass substrate was used. We carried out Rau’s measurement at 473-473 nm and combined the data together with that from 532 to 532 nm measurement to deduce ϕtc of pure DR1.
Pump-probe detection setup was used for the Rau’s and Fisher’s measurements, as well as transient absorption change to deduce τc. A circularly polarized 532nm CW beam from a diode-pumped solid laser was split to an intense pump beam and a weak probe beam. The probe beam was chopped before sample and the transmitted light intensity was detected by a photo diode and recorded by a lock-in amplifier. When transient absorption change is measured, the signal from the photo diode was recorded by a digital oscilloscope.
Figure 2(a) shows, for example, the transient absorption change of DR1 on self-assembled film of 10 nm gold particles. From it, a thermal relaxation lifetime of cis isomer of 1.56 second was obtained. Measured lifetimes of all samples are listed in Table 1 . No obvious change of lifetime happens between pure DR1 and DR1 on various gold nanostructures. This is not a surprising result due to the fact that DR1 molecules is not in physical contact with gold nanoparticles, so that additional phonon oscillation from gold nanoparticles to cis isomer is negligible, leading to unaffected relaxation characteristic of cis isomer.
Fig. 2 (a) The transient absorbance of DR1 on self-assembled film of 10nm-gold particles. Pump light was blocked at t = 0 and cis isomer begins to relax to trans isomer thermally. Line is the exponential decay fitting. (b) The relationship between the reciprocal of absorption change ratio with 1/F for pure DR1(green), DR1 on gold island film(red) and DR1 on self-assembled film of 10nm(purple), 30nm(blue) and 50nm(magenta) gold particles. Lines are linear fittings. Lines are the linear fitting.
Table 1. Final cis Occupancies (FCO), cis Isomer Lifetimes and trans-to-cis Photoisomerization Quantum Efficiencies of All Samples
Figure 2(b) is the Rau’s measurement results for five samples. They have obviously different slopes and intercepts. Meanwhile, a ϕtc = 0.20 for pure DR1 was obtained by Fisher’s method. With τc measured, all ϕtc can be deduced, they are summarized in Table 1 together with final cis isomer occupancies. The final cis occupancy (FCO) can be obtained from the intercepts.We can conclude from the Table 1 that for dye on gold island film, PQE goes up to 0.38 and the final occupancy of cis isomer is enhanced to 70% at infinite pump power. On the contrary, for DR1 on self-assembled gold particle films, PQE is significantly reduced. It is 0.15, 0.11 and 0.09 when gold particle size is 10, 30 and 50 nm respectively. Meanwhile, the final occupancy of cis isomer is also suppressed to be 24%, 19% and 9% respectively. The larger the gold nanoparticle size, the stronger the suppression.
PQE modification comes from the competition between local field enhancement and nonradiative decay acceleration. Two additional experiments, surface enhanced Raman Scattering (SERS) and transient fluorescence decay, were done to give more concrete evidences of the mechanism.
SERS experiment was used to check the local field enhancement from SPR. We used 514.5nm CW laser as the light source for SERS because its wavelength is close to the SPR peak wavelength. As Raman signal of a pure 20 nm DR1/PMMA film cannot be directly detected, we used a 1-μm-thick DR1/PMMA film as the reference and scaled down the Raman signal to a 20 nm thick film. Figure 3(a) shows the Raman spectra of four samples. The vibrational mode of N = N double bond is around 1400cm−1. Clearly DR1 on gold island film has the largest Raman signal enhancement, which is more than 100 × . For self-assembled films, the enhancement factor is much less and even no obvious enhancement can be detected when gold particle size is 50nm. This supports that DR1 on gold island film “feels” much larger optical field.
Fig. 3 (a) SERS spectra of DR1 on gold island film (red), DR1 on self-assembled film of 10nm (green), 30nm (blue) and 50nm (purple) gold particles. 1400 cm-1 is the N = N Raman mode. (b) Transient fluorescence decay of RhB on glass substrate (pink), gold island film (orange), self-assembled films of 10nm (green), 30nm (blue) and 50nm (red) gold particles.
Simple explanation can be applied by comparing the particle arrangement of two kinds of films. With particles closely packed in self-assembled films, the surface plasmonic field of individual gold nanoparticle is coupled, leading to a coupled SP Resonance in the near-infrared. This will dramatically decrease the local field enhancement at 514.5 nm. On the other hand, in the case of gold island film, the particles are relatively isolated, SP Resonance is around 532nm so that local field enhancement at 514.5 nm is obvious. Coupled SPR also changes local field which surrounds the metallic particles. Optical field simulation of touched metal particles showed that the field is centralized near the touching point which is called “hot spot”, where only a few dye molecules will be influenced [11,12]. On the contrary, for separated metal particles, electrical field spread over much larger area, thus more azobenzene dye molecules will be affected.
To demonstrate the influence of Au films on lifetime of excited molecules, fluorescence quenching of dye molecules on two kinds of Au films was measured by using Rhodamine B (RhB) as the model molecule, because DR1 has almost no observable fluorescence. Thin films of RhB in Polyvinyl Pyrrolindone (PVP) were prepared. Film thickness is again about 20 nm. A 371 nm pulsed semiconductor laser was used as the light source to excite fluorescence, and its pulse width is 50 ps. The fluorescence at 580 nm was collected by a photomultiplier and recorded by time-to-amplitude-conversion (TAC) technique. Figure 3(b) shows the transient fluorescence decay curves of five samples. Fluorescence from free RhB has a clear exponential decay with a lifetime of 2 ns, meanwhile RhB fluorescence on all metal films are severely quenched. However, we can still see that quenching is the strongest for 50nm-gold particle film, and the least quenching occurs on gold island film. This agrees with Feldmann’s finding that, the nonradiative decay rate of fluorescent dye on gold nanoparticles increases as gold particle size grows [13].
Combining information on local field enhancement and fluorescence quenching, it is clear that gold island film has the largest field enhancement, but least quenching rate. The strongly enhanced local field in gold island film excites more trans isomers to upper potential energy surface, which induces the relaxation path change of excited molecule, favoring relaxation along torsional motion coordinate to increase cis isomer [14]. Meanwhile, energy transfer induced nonradiative decay carries excited molecule immediately to trans ground state before converting to cis isomer, which leading to reduction of cis isomer number.
In order to achieve continuous PQE modification when gold nanostructure is fixed, we inserted a spacing layer between DR1/PMMA film and gold film, and investigated the spatial influence of local field/energy transfer. PVP film was used as the spacing layer and its thickness is between 10 and 80 nm.
Figure 4(a) shows the PQE change of DR1 on gold island film and self-assembled film of 30nm-gold particles as a function of spacing thickness. For DR1 on self-assembled film, PQE rises as the spacing thickness increases. When the spacing is above 30 nm, PQE almost recovers to the value of free molecule. For DR1 on gold island film, PQE decreases with a much slower rate and approaches to the free molecule value when spacing thickness is 80 nm. So it is clear that as spacing thickness increases, the influence of gold films to DR1 molecules decreases, which is analogous to the result given by Ref. [15]. The different recovering rates of PQE come from the different effective ranges of local field, which were evidenced by SERS enhancement with various thicknesses of the spacing layers. The results are shown in Fig. 4(b). In the case of 30 nm synthesized gold film, no noticeable SERS can be detected when the spacing thickness is over 10 nm. On the contrary, SERS signal is still clear in DR1/gold island film as the spacing layer thickness is as large as 30 nm due to its widespread local field as mentioned before. Note that the decay of PQE follows more or less the decay of Raman enhancement, confirming that the influence of SPR is more important in this case. Thus, the photoisomerization efficiency can be precisely modified using spacing layers with different thicknesses.
Fig. 4 (a) Changes of PQE as a function of spacing between dye and metallic nanostructures. (b) Changes of SERS enhancement when DR1 is isolated from gold nanostructures with a spacing layer. Blue points: DR1 on self-assembled film of 30nm-gold particles; Magenta points: DR1 on gold island film. The red dash line in (a) indicates PQE of free DR1molecule.
3. Conclusions
In conclusion, we find that the photoisomerization quantum efficiency of DR1 molecule can be modified in a large range by placing dye molecules on gold nanoparticle films with different structures. With spacing layer of variant thicknesses, photoisomerization efficiency can be further modified, verifying that precise control for photoisomerization efficiency is also possible.
Acknowledgments
Dr. Shen Xu and Dr. Jiong Shan contributed equally to this work. This work is supported in part by National Natural Science Foundation of China (grant # 60638010, 10874033, 60907011, 61078052, 11074051) and National Basic Research Program of China (973 Program) under the grant (No. 2011CB921802).
References and links
1. S. Loudwig and H. Bayley, “Photoisomerization of an individual azobenzene molecule in water: an on-off switch triggered by light at a fixed wavelength,” J. Am. Chem. Soc. 128(38), 12404–12405 (2006). [CrossRef] [PubMed]
2. T. Ikeda and O. Tsutsumi, “Optical switching and image storage by means of azobenzene liquid-crystal films,” Science 268(5219), 1873–1875 (1995). [CrossRef] [PubMed]
3. D. C. Burns, F. Z. Zhang, and G. A. Woolley, “Synthesis of 3,3′-bis(sulfonato)-4,4′-bis(chloroacetamido)azobenzene and cysteine cross-linking for photo-control of protein conformation and activity,” Nat. Protoc. 2(2), 251–258 (2007). [CrossRef] [PubMed]
4. X. Pan, C. S. Wang, C. Y. Wang, and X. Q. Zhang, “Image storage based on circular-polarization holography in an azobenzene side-chain liquid-crystalline polymer,” Appl. Opt. 47(1), 93–98 (2008), http://www.opticsinfobase.org/abstract.cfm?uri=ao-47-1-93. [CrossRef]
5. M. J. Comstock, N. Levy, A. Kirakosian, J. Cho, F. Lauterwasser, J. H. Harvey, D. A. Strubbe, J. M. Fréchet, D. Trauner, S. G. Louie, and M. F. Crommie, “Reversible photomechanical switching of individual engineered molecules at a metallic surface,” Phys. Rev. Lett. 99(3), 038301 (2007). [CrossRef] [PubMed]
6. G. Vogt, G. Krampert, P. Niklaus, P. Nuernberger, and G. Gerber, “Optimal control of photoisomerization,” Phys. Rev. Lett. 94(6), 068305 (2005). [CrossRef] [PubMed]
7. G. Frens, “Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions,” Nature 241, 20–22 (1973).
8. Y. K. Park and S. Park, “Directing close-packing of midnanosized gold nanoparticles at a water/hexane interface,” Chem. Mater. 20(6), 2388–2393 (2008). [CrossRef]
9. H. Rau, G. Greiner, G. Gauglitz, and H. Meier, “Optimal control of photoisomerization,” J. Phys. Chem. 94(17), 6523–6524 (1990). [CrossRef]
10. E. Fischer, “The calculation of photostationary states in systems A ⇔ B when only A is known,” J. Phys. Chem. 71(11), 3704–3706 (1967). [CrossRef]
11. M. Futamata, Y. Maruyama, and M. Ishikawa, “Local electric field and scattering cross section of Ag nanoparticles under surface plasmon resonance by finite difference time domain method,” J. Phys. Chem. B 107(31), 7607–7617 (2003). [CrossRef]
12. H. Xu and M. Käll, “Surface-plasmon-enhanced optical forces in silver nanoaggregates,” Phys. Rev. Lett. 89(24), 246802 (2002). [CrossRef] [PubMed]
13. E. Dulkeith, A. C. Morteani, T. Niedereichholz, T. A. Klar, J. Feldmann, S. A. Levi, F. C. van Veggel, D. N. Reinhoudt, M. Möller, and D. I. Gittins, “Fluorescence quenching of dye molecules near gold nanoparticles: radiative and nonradiative effects,” Phys. Rev. Lett. 89(20), 203002 (2002). [CrossRef] [PubMed]
14. B. Dietzek, B. Brüggemann, T. Pascher, and A. Yartsev, “Mechanisms of molecular response in the optimal control of photoisomerization,” Phys. Rev. Lett. 97(25), 258301 (2006). [CrossRef]
15. P. Viste, J. Plain, R. Jaffiol, A. Vial, P. M. Adam, and P. Royer, “Enhancement and quenching regimes in metal-semiconductor hybrid optical nanosources,” ACS Nano 4(2), 759–764 (2010). [CrossRef] [PubMed]
