An Evanescent-Field-Driven Self-Assembled Molecular Photoswitch

Jun 12, 2004 - The reversible photoswitching performances of 1 SAM and 2 SAM via the evanescence field using light of appropriate wavelengths have bee...
0 downloads 0 Views 171KB Size
Download PDF
5850

Langmuir 2004, 20, 5850-5857

An Evanescent-Field-Driven Self-Assembled Molecular Photoswitch for Macrocycle Coordination and Release Zhenxin Wang, Anne-Mette Nygård, Michael J. Cook,* and David A. Russell* School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich, Norfolk NR4 7TJ, United Kingdom Received January 13, 2004. In Final Form: April 30, 2004 Two self-assembled monolayer (SAM) films containing the photoswitchable 4-pyridylazophenoxy chromophore have been deposited onto a gold-coated glass substrate. One film contains the chromophore as a single component, 1 SAM, and the other is doped with a nonphotoactive component as a 1:1 mixture, 2 SAM. The reversible photoswitching performances of 1 SAM and 2 SAM via the evanescence field using light of appropriate wavelengths have been investigated by UV spectroscopic and electrochemical monitoring. In principle, the trans-form SAMs present a coordinating surface, the “on” state, that can be switched “off” in the cis form. This has been illustrated by immersing both the as-deposited (trans form) SAMs and the photoswitched (predominantly cis form) SAMs into solutions of cobalt and zinc tetraphenylporphyrin (CoTPP and ZnTPP, respectively) and an octaoctyl-substituted cobalt phthalocyanine. In a further phase of this study, the remote control of binding events at the surface of the SAMs has been demonstrated through evanescent-field-driven photoswitching of trans-form SAMs coordinated at the surfaces with examples of these metallomacrocycles. This photoswitching was undertaken with the constructs immersed in neat toluene, and the macrocycles were released from the surface into the solvent. The release was measured by spectroscopic monitoring of the material remaining on the constructs. The study was extended to develop an in situ release/coordination cycle. Thus, irradiation of a construct of ZnTPP bound to the surface of trans-form 2 SAM using waveguided light at 365 nm releases the macrocycle into a toluene solution of ZnTPP. Further irradiation of the SAM, now in its cis form, with waveguided 439 nm light regenerates the trans form, which recoordinates ZnTPP from the solution. The results demonstrate the potential for using waveguided light to control molecular events within and at the surfaces of SAM constructs.

Introduction We are currently exploring the use of waveguided light rather than conventional direct irradiation to photoexcite chromophores contained within self-assembled monolayer (SAM) films on planar gold-coated substrates. The model constructs for these studies contain suitably substituted chromophores attached by sulfur-gold bonds to a gold film supported on a glass substrate. Photoexcitation is achieved via the evanescent wave at the substrate surface. We earlier used this protocol to stimulate fluorescence from a SAM of a phthalocyanine and illustrated an application within an optical gas-sensing device operated using waveguided light from a “remote light source”.1 Other laboratories have also reported the use of the evanescent wave for measurement of fluorescence2 as well as for measurement of UV-vis3 and IR4 spectra. An attractive target for this type of remote photoexcitation is the photoswitching of the “azobenzene” moiety. This chromophore undergoes photon-driven reversible cis-trans interconversions and has been much investigated for potential applications in areas such as manipulation of liquid crystal phases,5 photoswitching,6 optical data storage,7 nonlinear optics,8 molecular shuttles,9 and * To whom correspondence should be addressed. E-mail: m.cook@ uea.ac.uk (M.J.C.), [email protected] (D.A.R.). (1) Simpson, T. R. E.; Revell, D. J.; Cook, M. J.; Russell, D. A. Langmuir 1997, 13, 460-464. (2) Asanov, A. N.; Wilson, W. W.; Oldham, P. B. Anal. Chem. 1998, 70, 1156-1163. (3) (a) Mendes, S. B.; Li, L.; Burke, J. J.; Lee, J. E.; Dunphy, D. R.; Saavedra, S. S. Langmuir 1996, 12, 3374-3376. (b) Kato, K.; Takatsu, A.; Matsuda, N. Chem. Lett. 1999, 31-32. (c) Ohno, H.; Yoneyama, S.; Nakamura, F.; Fukuda, K.; Hara, M.; Shimomura, M. Langmuir 2002, 18, 1661-1665. (4) Bentaleb, A.; Abele, A.; Haikel, Y.; Schaaf, P.; Voegel, J. C. Langmuir 1998, 14, 6493-6500.

electrooptic devices.10 The incorporation of this photoactive unit and related “azo” chromophores into LangmuirBlodgett films11 and SAMs,8-10,12-16 the latter normally supported on planar gold or gold nanoparticles, has been well studied. A number of these studies have included investigation of the photoisomerism of the chromophores in these formulations, undertaken using direct irradiation at appropriate wavelengths of light. In the main, these (5) (a) Pieraccini, S.; Masiero, S.; Spada, G. P.; Gottarelli, G. Chem. Commun. 2003, 598-599. (b) Ruslim, C.; Ichimura, K. Adv. Mater. 2001, 13, 37-40. (c) Prasad, S. K.; Nair, G. G. Adv. Mater. 2001, 13, 40-43. (d) Sato, T.; Tsuji, K.; Kokuryu, E.; Wadayama, T.; Hatta, A. Mol. Cryst. Liq. Cryst. 2003, 391, 13-39. (6) Ikeda, T.; Tsutsumi, O. Science 1995, 268, 1873-1875. (7) (a) Liu, Z. F.; Hashimoto, K.; Fujishima, A. Nature 1990, 347, 658-660. (b) Muzikante, I.; Markava, E.; Gustina, D.; Gerca, L.; Rutkis, M.; Fonavs, E.; Stiller, B.; Brehmer, L. Ferroelectrics 2001, 258, 393404. (8) Lin, W.; Lin, W.; Wong, G. K.; Marks, T. J. J. Am. Chem. Soc. 1996, 118, 8034-8042. (9) (a) Murakami, H.; Kawabuchi, A.; Kotoo, K.; Kunitake, M.; Nakashima, N. J. Am. Chem. Soc. 1997, 119, 7605-7606. (b) Willner, I.; Pardo-Yissar, V.; Katz, E.; Ranjit, K. T. J. Electroanal. Chem. 2001, 497, 172-177. (10) (a) Campbell, D. J.; Herr, B. R.; Hulteen, J. C.; Van Duyne, R. P.; Mirkin, C. A. J. Am. Chem. Soc, 1996, 118, 10211-10219. (b) Caldwell, W. B.; Campbell, D. J.; Chen, K.; Herr, B. R.; Mirkin, C. A.; Malik, A.; Durbin, M. K.; Dutta, P.; Huang, K. G. J. Am. Chem. Soc. 1995, 117, 6071-6082. (11) (a) Seki, T.; Sakuragi, M.; Kawanishi, Y.; Suzuki, Y.; Tamaki, T.; Fukuda, R.-I.; Ichimura, K. Langmuir 1993, 9, 211-218. (b) Wu, A.; Talham, D. R. Langmuir 2000, 16, 7449-7456. (c) Tachibana, H.; Nakamura, T.; Matsumoto, M.; Komizu, H.; Manda, E.; Niino, H.; Yabe, A.; Kawabata, Y. J. Am. Chem. Soc. 1989, 111, 3080-3081. (d) Wang, R.; Jiang, L.; Iyoda, T.; Tryk, D. A.; Hashimoto, K.; Fugishima, A. Langmuir 1996, 12, 2052-2057. (e) Ve´lez, M.; Mukhopadhyay, S.; Muzikante, I.; Matisova, G.; Vieira, S. Langmuir 1997, 13, 870-872. (f) Matsumoto, M.; Miyazaki, D.; Tanaka, M.; Azumi, R.; Manda, E.; Kondo, Y.; Yoshino, N.; Tachibana, H. J. Am. Chem. Soc. 1998, 120, 1479-1484. (g) Shin, H. K.; Kim, J. M.; Kwon, Y. S.; Park, E.; Kim, C. Opt. Mater. 2003, 21, 389-394.

10.1021/la0498861 CCC: $27.50 © 2004 American Chemical Society Published on Web 06/12/2004

A Photoswitch for Macrocycle Coordination and Release

Langmuir, Vol. 20, No. 14, 2004 5851

experiments have been successful although difficulties have been reported for certain examples of SAM films supported on planar gold surfaces,14,15 vide infra. Notwithstanding this last point, we have been investigating the use of the evanescent field to excite and photoswitch functionally interesting analogues of azobenzene contained within such films. Thus, we have recently achieved reversible switching of a 2-pyridylazophenoxy moiety contained within a SAM obtained from 8,8′-[4,4′-di(2pyridylazo)phenoxy]diocta-1,1′-disulfide.17 It was shown that the SAM in the trans isomeric form provides a bidentate ligand that can chelate metal ions, such as Ni2+ and Co2+. In the cis form the film is inactive. In the present study18 we have examined two SAMs containing the 4-pyridylazophenoxy chromophore. The SAMs, obtained by deposition of compounds 1 and 2 onto gold, are referred to hereafter as 1 SAM and 2 SAM, Figure 1. We have investigated the reversible photoswitching performances of 1 SAM and 2 SAM via the evanescence field using light of appropriate wavelengths, as depicted in the idealized representation of the process for 2 SAM in Figure 2. In the trans form, or “on” position, the SAMs present a coordinating surface that can be switched “off” in the cis form. This is illustrated by coordination studies using a selection of metal ion containing macrocycles. In a further phase of this study, the remote control of binding events at the surface of 2 SAM has been demonstrated through an evanescent-field-driven release/coordination cycle. Experimental Section Reagents, Equipment, and Synthetic Methods. The chemical reagents and solvents (Aldrich) employed in the synthesis were used as supplied. The chromatographic separations were performed using silica 60 (70-230 mesh) from Fluka. TLC was performed using silica 60 F254 sheets from Merck. All solvents used in the SAM experiments were of spectroscopic grade. The melting points were obtained using an Olympus BH-2 polarizing microscope in conjunction with a Linkam TMS 92 thermal analyzer with a Linkam THM 600 cell. The 1H and 13C NMR spectra were recorded at 270 and 67.5 MHz, respectively, (12) (a) Yu, H.-Z.; Ye, S.; Zhang, H.-L.; Uosaki, K.; Liu, Z.-F. Langmuir 2000, 16, 6984-6954. (b) Yu, H.-Z.; Zhang, H.-L.; Liu, Z.-F.; Ye, S.; Uosaki, K. Langmuir 1998, 14, 619-624. (c) Sekkat, Z.; Wood, J.; Geerts, Y.; Knoll, W. Langmuir 1995, 11, 2856-2859. (d) Wang, Y.-Q.; Yu, H.-Z.; Cheng, J.-Z.; Zhao, J.-W.; Cai, S.-M.; Liu, Z.-F. Langmuir 1996, 12, 5466-5471. (e) Wang, R.; Iyoda, T.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. Langmuir 1997, 13, 4644-4651. (f) Hu, J.; Zhang, J.; Liu, F.; Kittredge, K.; Whitesell, J. K.; Fox, M. A. J. Am. Chem. Soc. 2001, 123, 1464-1470. (g) Wang, R.; Iyoda, T.; Jiang, L.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. J. Electroanal. Chem. 1997, 438, 213219. (h) Liu, Z. F.; Hashimoto, K.; Fujishima, A. J. Electroanal. Chem. 1992, 324, 259-267. (i) Kondo, T.; Kanai, T.; Uosaki, K. Langmuir 2001, 17, 6317-6324. (j) Liu, M.; Ushida, K.; Kira, A.; Nakahara, H. J. Phys. Chem. B 1997, 101, 1101-1104. (k) Saremi, F.; Tieke, B. Adv. Mater. 1998, 10, 389-391. (l) Oh, S. K.; Nakagawa, M.; Ichimura, K. J. Mater. Chem. 2002, 12, 2262-2269. (13) Zhang, J.; Whitesell, J. K.; Fox, M. A. Chem. Mater. 2001, 13, 2323-2331. (14) (a) Delamarche, E.; Michel, B.; Gerber, Ch.; Anselmetti, D.; Guentherodt, H.-J.; Wolf, H.; Ringsdorf, H. Langmuir 1994, 10, 28692871. (b) Jaschke, M.; Scho¨nherr, H.; Wolf, H.; Butt, H.-J.; Bamberg, E.; Besocke, M. K.; Ringsdorf, H. J. Phys. Chem. 1996, 100, 2290-2301. (15) Evans, S. D.; Johnson, S. R.; Ringsdorf, H.; Williams, L. M.; Wolf, H. Langmuir 1998, 14, 6436-6440. (16) (a) Tamada, K.; Akiyama, H.; Wei, T. X.; Kim, S. A. Langmuir 2003, 19, 2306-2312. (b) Manna, A.; Chen, P. L.; Akiyama, H.; Wei, T. X.; Tamada, K.; Knoll, W. Chem. Mater. 2003, 15, 20-28. (c) Akiyama, H.; Tamada, K.; Nagasawa, J.; Abe, K.; Tamaki, T. J. Phys. Chem. B 2003, 107, 130-135. (17) Wang, Z.; Cook, M. J.; Nygård, A.-M.; Russell, D. A. Langmuir 2003, 19, 3779-3784. (18) For a communication of aspects of this work see: Cook, M. J.; Nygård, A.-M.; Wang, Z.; Russell, D. A. Chem. Commun. 2002, 10561057.

Figure 1. Structures of and synthetic route to 1 and 2 and idealized representations of 1 SAM and 2 SAM.

Figure 2. Idealized representation of evanescent-wave-induced photoswitching, illustrated for 2 SAM. using a JEOL EX 270 spectrometer. The mass spectra were measured at the EPSRC service, University of Wales at Swansea, U.K. The low-resolution fast atom bombardment (FAB LSIMS) mass spectra were obtained using a VG AutoSpec instrument, with nitrobenzyl alcohol (NOBA) as the matrix. The accurate mass spectra (FAB LSIMS) were measured on a Finnegan MAT 900 instrument. The photoswitchable compounds 1 and 2 were prepared according to Figure 1. 8-Methylsulfonyloxyoctyl disulfide17 and 4-(4-pyridylazo)phenol19 were obtained following literature procedures. Of the compounds used in coordination experiments, 5,10,15,20-tetraphenyl-21H,23H-porphine cobalt and 5,10,15,20-tetraphenyl-21H,23H-porphine zinc were purchased from Aldrich, and 1,4,8,11,15,18,22,25-octaoctylphthalocyaninato cobalt(II) was a gift from Dr. I. Fernandes (University of East Anglia). 8-[4-(4-Pyridylazo)phenoxy]octyl Disulfide (1). 8-Methylsulfonyloxyoctyl disulfide (0.42 g, 0.92 mmol) and 4-(4pyridylazo)phenol (0.36 g, 1.83 mmol) were dissolved in anhydrous dimethylformamide (20 mL) under a nitrogen atmosphere, and excess K2CO3 (0.5 g, 4 mmol) was added. The mixture was stirred at room temperature under nitrogen for 7 days. The reaction mixture was poured onto water (50 mL) and extracted with ethyl acetate (3 × 30 mL). The solid that collected between (19) Buncel, E.; Keum, S.-R. Tetrahedron 1983, 39, 1091-1101.

5852

Langmuir, Vol. 20, No. 14, 2004

the two layers was recovered by filtration and dissolved in dichloromethane. The solution was washed with aqueous NaOH (10%) and dried (MgSO4) and the solvent removed to afford a crude product. The ethyl acetate extracts were dried (MgSO4) and concentrated. The residue was combined with the first batch of product and chromatographed (eluent toluene/methanol, 4:1). The first fraction eluted following unconsumed starting material was the monocoupled product 8-[4-(4-pyridyl)azophenoxy]octyl8′-methylsulfonyloxyoctyl 1,1′-disulfide (3), recovered as a pale yellow solid and used in the synthesis of 2, see below, without further purification. Yield: 0.10 g, 18%. 1H NMR (270 MHz, CDCl3): δ 8.75 (d, J ) 6.2 Hz, 2 H), 7.93 (m, 2 H), 7.64 (m, 2 H), 6.99 (m, 2 H), 4.20 (t, J ) 6.6 Hz, 2 H), 4.03 (t, J 6.6, 2 H), 3.00 (s, 3 H), 2.67 (m, 4 H), 1.8-1.6 (m, 8 H) and 1.5-1.2 (m, 16 H). The subsequent fraction from the chromatographic separation provided 1 as a yellow powder after recrystallization from dichloromethane/light petroleum ether. Mp: 105-106 °C. Yield: 0.33 g, 52%. 1H NMR (270 MHz, CDCl3): δ 8.75 (d, J ) 6.3 Hz, 4 H), 7.93 (m, 4 H), 7.64 (m, 4 H), 6.99 (m, 4 H), 4.03 (t, J ) 6.5 Hz, 4 H), 2.67 (t, J ) 7.3 Hz, 4 H), 1.81 (m, 4 H), 1.66 (m, 4 H) and 1.5-1.2 (m, 16 H). 13C NMR (75 MHz, CDCl3): δ 163.00 (Cq), 157.58 (Cq), 151.32 (CH), 146.84 (Cq), 125.66 (CH), 116.19 (CH), 114.91 (CH), 68.40 (CH2), 39.03 (CH2), 29.02 (CH2), 29.05 (CH2)2, 29.10 (CH2), 28.29 (CH2) and 25.82 (CH2). MS (FAB): [M + H] m/z calcd 685.3358, found 685.3348. 8-[4-(4-Pyridylazo)phenoxy]octyl-8′-phenyloxyoctyl Disulfide (2). 3 (0.09 g, 0.16 mmol) and excess phenol (0.60 g, 6 mmol) were dissolved in anhydrous DMF (10 mL) under a nitrogen atmosphere. An excess of K2CO3 was added, and the mixture was left stirring at room temperature. The reaction was monitored by TLC (eluent toluene/MeOH, 4:1) and 1H NMR spectrometry. After 7 days the reaction mixture was poured into water (50 mL) and extracted with ethyl acetate (3 × 30 mL). The combined organic extracts were dried (MgSO4), filtered, and concentrated. Repeated column chromatography (eluent toluene/ MeOH, 4:1 and 9:1) allowed the isolation of an orange solid. This crude product was recrystallized from DCM/petroleum ether (bp 40-60 °C) to afford 2 as an orange powder. Mp: 72-73 °C. Yield: 0.040 g, 45%. 1H NMR (270 MHz, CDCl3): δ 8.75 (m, 2 H), 7.93 (m, 2 H), 7.65 (m, 2 H), 7.24 (m, 2 H), 6.98 (m, 2 H), 6.89 (m, 2 H), 4.03 (t, J ) 6.3 Hz, 2 H),), 3.92 (t, J ) 6.4, 2 H), 2.67 (m, 4 H), 1.8-1.6 (m, 8 H) and 1.5-1.2 (m, 16 H). 13C NMR (75 MHz, CDCl3): δ 162.73 (Cq), 158.95 (Cq), 157.33 (Cq), 151.08 (CH), 146.57 (Cq), 129.26 (CH), 125.47 (CH), 120.35 (CH), 116.04 (CH), 114.73 (CH), 114.33 (CH), 68.31 (CH2), 67.66 (CH2), 38.99 (CH2), 29.12 (CH2), 29.08 (CH2), 29.01 (CH2)2, 28.27 (CH2), 25.89 (CH2) and 25.81 (CH2). MS (FAB): [M + H] m/z calcd 580.3031, found 580.3033. Formation of the 1 SAM and 2 SAM Films on Gold. Goldcoated glass substrates (cut to the dimensions of a standard microscope slide) were used for the formation of the SAMs. In each case, the glass substrate was first cleaned and the metals chromium (to aid adhesion) and gold were thermally evaporated (Edwards Auto 306 vacuum evaporator) onto the surface as previously described.1,20 Quartz glass substrates (Chemglass, U.S., or Crystran Ltd., U.K.) were used for the UV-vis measurements and ITO glass substrates (Crystran Ltd.) for the electrochemical measurements. All substrates were coated with 1 nm of Cr followed by 8 nm of Au. Surface roughness was measured by conventional methodology21a using the equation Ipa ) (2.69 × 105)n3/2AD1/2Cv1/2, where n is the number of electrons involved (n ) 1 in the Fe(CN)63-/Fe(CN)64- system), A is the electrode surface area (cm2), C is the bulk concentration of Fe(CN)63- (mol/cm3), D ) 7.6 × 10-6 is the diffusion coefficient of Fe(CN)63- (cm2/s),21b and v is the scan rate (V/s). Ipa is in amperes. The calculated Ipa for the electrodes used in the present work, with a two-dimensional surface area A ) 6 cm2 under the conditions of measurement where C ) 5.0 × 10-7 mol/cm3 and (20) Revell, D. J.; Chambrier, I.; Cook, M. J.; Russell, D. A. J. Mater. Chem. 2000, 10, 31-37. (21) (a) Finklea, H. O. Electrochemistry of Organized Monolayers of Thiols and Related Molecules on Electrodes. In Electroanalytical Chemistry; Bard, A. J., Rubinstein, I., Eds.; Marcel Dekker: New York, 1996; Vol. 19, pp 109-335. (b) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2001; p 813.

Wang et al. v ) 0.1 V/s, is 7.04 × 10-4 A. Experimental measurements gave IpaE ) 8.09 × 10-4 A, leading to a surface roughness factor of 1.15. The SAMs were formed by immersing freshly prepared goldcoated substrates for 15 h in ethyl ethanoate solutions of the disulfides 1 (3.07 × 10-5 M) and 2 (3.50 × 10-5 M). The coated substrates were then washed in ethyl ethanoate and dried in a stream of argon. Spectroscopic and Electrochemical Characterization. The UV-vis spectra of the photoswitch monolayer were obtained using a Hitachi U-3000 spectrophotometer, with the corresponding reference (a bare gold-coated substrate) placed in the reference beam. The cyclic voltammograms were recorded using a µAutolab II electrochemical analyzer (Eco Chemie, The Netherlands) with an in-house-built electrochemical cell. A gold-coated ITO glass substrate (ca. 4 cm2, bare or coated with a SAM) was used as the working electrode, a saturated Ag/AgCl electrode as the reference electrode, and a platinum disk as the counter electrode. During the electrochemical experiment, a phosphate buffer solution (pH 6.92, 0.05 M) was used. Photoswitching Experiments. The freshly prepared SAMs (in the trans form) were irradiated using UV light (λ1 ) 365 nm) via an evanescent wave for 9 h to obtain the cis form. The light was applied through optical fibers to the shorter edge of the glass slide. The cis to trans isomerization was achieved with irradiation using evanescent wave visible light (λ2 ) 439 nm) again for 9 h. Prior to spectral or electrochemical measurement the SAMs were washed in solvent, dried in a stream of argon, and stored in clean amber sample jars with airtight lids. Coordination Experiments. Investigation of the coordination of macrocycles to SAM surfaces in their trans and cis forms was undertaken typically by immersing freshly prepared films in a deoxygenated solution of cobalt 5,10,15,20-tetraphenylporphyrin (CoTPP), zinc 5,10,15,20-tetraphenylporphyrin (ZnTPP), or 1,4,8,11,15,18,22,25-octaoctylphthalocyaninato cobalt(II) [(C8H17)8CoPc] (ca. 1.6 × 10-4 M in toluene) for 8-10 h and then rinsing the film with toluene (5 × 2 mL) prior to UV-vis spectroscopic characterization. Evanescent-Field-Driven Release of Coordinated Macrocycles. Substrates bearing trans-form SAMs, the surfaces of which were coordinated to CoTPP and ZnTPP, were lowered vertically into toluene contained in a 60 mL bottle to a level such that the top edge of the substrate protruded above the solvent surface. Light at 365 nm was delivered to the edge of the substrate over a total of 10 h. Intermittently, the construct was withdrawn from the solvent and the extent of the release of CoTPP or ZnTPP from the surface was monitored by the decrease in the absorbance of the respective Soret band. Evanescent-Field-Driven Release/Coordination Cycle. A trans-form 2 SAM was surface coated with ZnTPP as above (see Coordination Experiments). The previous experiment (see Evanescent-Field-Driven Release of Coordinated Macrocycles) was repeated, but here the construct was lowered into a solution of ZnTPP in toluene (1.6 × 10-4 M). Light at 365 nm was delivered to the edge of the substrate as described above for 120 min. The construct was removed from the solution and washed with toluene (2 × 5 mL) and the release of ZnTPP from the surface monitored by UV-vis spectroscopy. The construct was reimmersed into the solution of ZnTPP and further illuminated with light at 439 nm for 180 min. After removal of the construct from the solution and washing as above, the Soret band absorbance from recoordinated ZnTPP was measured. The release/coordination sequence was then repeated.

Results and Discussion Design and Synthesis of Presursors for 1 SAM and 2 SAM. Zhang, Whitesell, and Fox13 reported the selfassembly of a thiol-terminated azobenzene derivative on gold nanoparticles and observed trans to cis photoisomerization. However, there are reported difficulties with certain constructs on planar gold surfaces. Earlier scanning probe microscopy studies of SAMs formed from azobenzene-derivatized thiols showed the chromophore

A Photoswitch for Macrocycle Coordination and Release

Langmuir, Vol. 20, No. 14, 2004 5853

Figure 3. Spectra obtained of 1 SAM (left-hand side) and 2 SAM (right-hand side) as the as-deposited films, trans form (solid line), after irradiation with waveguided 365 nm light for 9 h, “cis form” (broken line), and after further irradiation with waveguided 439 nm light for 9 h (dotted line).

units closely packed, presenting little free volume to allow photoisomerization to occur.14 In line with these observations, a systematic study by Evans et al.15 of SAMs prepared from both azobenzene-derivatized thiols and disulfides on planar gold showed that the photoisomerization did not occur. SAMs from the two sources showed no noticeable differences. However, they also established that “mixed monolayers” containing the azobenzene molecules and short ethyleneoxy-based “spacing molecules” did exhibit photoswitching. These monolayers were prepared from solutions containing two appropriately substituted thiol derivatives. More recently, and during the course of the present study, Tamada and co-workers presented a series of papers concerned with further “mixed” SAMs, obtaining them by deposition of mixed disulfides. These were successfully photoisomerized on both planar gold surfaces and gold nanoparticle surfaces.16 In the present work, we sought to compare the photoswitching behavior of the single-component film 1 SAM and the mixed-component film 2 SAM. We did this through synthetic work which utilized a disulfide precursor as outlined in Figure 1. Thus, the precursor bismesylate disulfide was reacted with 4-(4-pyridylazo)phenol under conditions which gave both the symmetrical disulfide product 1 and the product of partial reaction, i.e., the unsymmetrical disulfide 3. The reaction is not clean because there are also byproducts formed, believed to be products of alkylation at the pyridyl nitrogen atoms. However, repeated chromatography satisfactorily separated 1 and 3. The latter was then reacted with phenol under basic conditions to afford 2, a material designed to deposit as a 1:1 “mixed SAM”. Photoisomerization of 1 and 2 in Solution. The more stable and dominant isomer of azobenzenes is the trans form, and UV spectral data for 1 and 2 as solutions in ethyl ethanoate are characteristic of this isomer. Thus, both exhibit a strong π-π* absorption at 352 nm, typical for a trans-azobenzene analogue,13 a weaker band at 437 nm (n-π*), and a shoulder at 252 nm, which appears on a broader band at shorter wavelength. After UV irradiation of the solutions at 365 nm for 30 min, the photoisomerization of the trans form into the cis isomer was manifested by changes in both the 352 and 437 nm bands. The π-π* transition band for both 1 and 2 is blue-shifted, to 334 nm for 1 and 323 nm for 2, with lowering of intensity. Conversely, the 437 band remains unchanged in energy, but the intensity is increased. The 252 nm shoulder remains unchanged. The reverse (cis to trans) photoisomerization was also achieved but using visible light at

439 nm. For both 1 and 2 there was greater than 95% recovery of the peak intensity of the original 352 nm band within 30 min of irradiation. These latter isomerizations were also achieved by thermal relaxation within 18 h, on storing the solutions in the dark at room temperature. UV-Vis Spectroscopic Characterization and Photoswitching of 1 SAM and 2 SAM. The UV-vis spectra of the as-deposited 1 SAM and 2 SAM are shown in Figure 3. The π-π* transition band is clearly observed and is shifted from 352 nm in the solution-phase spectra to 349 nm in the spectrum of 1 SAM and 354 nm in that for 2 SAM. Photoisomerization of the chromophore using 365 nm light was attempted separately with 3 h of direct irradiation of the film or with 9 h of irradiation with waveguided light. The latter is illustrated in Figure 2. Both methods gave the same results. For 1 SAM, the 349 nm band shows a decrease in intensity with no evident change in λmax. This observation is similar to that observed after photoswitching of the SAM containing the isomeric 4-(2-pyridylazophenoxy) chromophore we described earlier.17 The behavior contrasts with that of 2 SAM. In this case the corresponding absorption band, at 354 nm, shows a decrease in intensity and the appearance of a band at 339 nm, Figure 3. There is no evidence of a band ca. 1020 nm to the blue for either SAM as was observed for the cis forms of compounds 1 and 2 in the solution phase. The present spectral changes observed for 1 SAM and 2 SAM appear similar to the spectral changes assigned to trans to cis photoisomerizations of azobenzenes supported as SAMs on gold nanoparticles referred to above.13,15 Evidence that the spectral changes are not a consequence of photodecomposition was forthcoming from a second set of irradiation experiments with 439 nm light, using both waveguided (9 h) and direct (3 h) illumination. The spectral data in Figure 3 show substantial recovery of the bands of the trans isomer. The intensity of the 349 nm band in 1 SAM is restored to 80-85% of that in the as-deposited film and that of the 354 nm band of 2 SAM to 92-95%. Comparable restorations of band intensities occurred on storing the SAMs in the dark for two weeks. These restorations of the trans form of the SAMs are lower, though only marginally so in the case of 1 SAM, than those of the precursor compounds 1 and 2 in the solution phase. This may be a consequence of some steric constraints within the films. Unfortunately, the coincidence, or near coincidence, of the π-π* transition from the asdeposited (trans) SAMs and from the photoswitched films precludes estimates of the ratios of the cis to trans forms contained within them. However, hereafter we shall refer

5854

Langmuir, Vol. 20, No. 14, 2004

Wang et al. Table 1. UV-Vis Spectroscopic Data for Metallomacrocycles Coordinated to 1 SAM and 2 SAM construct

λmaxa (nm)

Atransa

Acis/Atransb

1 SAM + CoTPP 2 SAM + CoTPP 1 SAM + ZnTPP 2 SAM + ZnTPP 2 SAM +(C8H17)8CoPc

438 438 432 432 705, br

0.014(5) 0.018(0) 0.016(0) 0.014(0) 0.002(2)

0.30 ( 0.01 0.24 ( 0.02 0.43 ( 0.01 0.16 ( 0.02 0.17 ( 0.05

aλ max and the absorbance Atrans were measured for the macrocycle on the trans form of the SAM. b Acis/Atrans is the ratio of the absorbances measured for the macrocycle coordinated onto preformed cis and trans forms of the SAMs.

centers on the electrodes (Γ) using the equation22

Γ ) Q/nFA or ip ) n2F2ΓAv/4RT

Figure 4. Cyclic voltammogram of the 2 SAM-modified gold electrode before (solid line) and after (dotted line) irradiation with waveguided 365 nm light for 9 h. The broken line shows the cyclic voltammogram after a further 9 h of waveguided irradiation with 439 nm light. The data were obtained from measurements of the SAM immersed in 0.05 M phosphatebuffered saline (pH 6.92) with a potential sweep rate of 20 mV/s between 0 and -0.5 V.

to them as “cis”- and “trans”-form films according to the predominant isomer. The greater extent of the recovery of the spectrum of the trans form of 2 SAM after the second photoswitching indicates that the reversible behavior of this SAM is superior to that of 1 SAM. The time required for photoisomerization using waveguided light, viz., 9 h, was considerably longer than for direct illumination of the SAM surface, indicating that further optimization of the delivery of the waveguided light may be possible. Electrochemical Characterization of SAM-Modified Gold Electrodes. Electrochemical studies were undertaken to probe further the trans to cis photoisomerization process and to compare surface coverages within the 1 SAM and 2 SAM constructs. The cyclic voltammograms of the SAM-modified gold electrodes were recorded before and after light irradiation and measured in 0.05 M phosphate-buffered saline (pH 6.92) with a potential sweep rate of 20 mV/s between 0 and -0.5 V. No faradic current was observed when a bare gold film substrate was used as the working electrode. The as-deposited (trans-form) SAMs both exhibited good electrochemical behavior. The oxidation and reduction peaks due to the electrochemical redox of 1 SAM were observed at -0.085 and -0.375 V, those for 2 SAM at 0.028 and -0.39 V. A representative cyclic voltammogram (for 2 SAM) is shown in Figure 4. For both SAMs the peak current varied linearly with the scan rates (shown as an inset in Figure 4). The shapes of the cyclic voltammograms for both SAMs changed after photoisomerization with UV light, with both the cationic peak current and anodic peak current decreasing. This result is comparable to that obtained for the 2-pyridylazophenoxy SAM investigated recently17 but differs from those obtained for azobenzene-containing SAMs.7,12e,g-i For the latter the cis-form SAMs exhibit better electrochemical behavior. Following irradiation with visible light, the original CV curves were essentially reproduced for both 1 SAM and 2 SAM. Electrochemical measurements of SAMs on gold surfaces provide estimates of surface coverage of redox active

where Q is the integrated charge passing on each reduction peak at 10 mV/s or the peak current, A is the electrode surface area, F is the Faraday constant, and n is the number of electrons involved in the electrode reaction. Measurements performed on 1 SAM and 2 SAM gave values of Γ ) 5.4 × 10-10 and 2.6 × 10-10 mol‚cm-2, respectively. In this approach, the values of A are the geometric surface area multiplied by the surface roughness factor of 1.15 (see the Experimental Section). It is encouraging that the relative number of redox active centers as determined by the method is approximately twice the number in 1 SAM than in 2 SAM. Furthermore, the value of Γ obtained for 1 SAM is in close agreement with the value obtained for the SAM formed from the isomeric compound we investigated earlier using identical protocols.17 Coordination of Metalated Macrocycles to the 1 SAM and 2 SAM Surfaces. The potential for axial ligation of appropriate metalated macrocycles to nitrogen ligands is well established and has in some cases been applied to SAM films containing, for example, pyridine23 or imidazole24 groups at the surface. In this work, we investigated first the coordination of ZnTPP and CoTPP. Thus, as-deposited (trans) and photoisomerized (cis) form films of both 1 SAM and 2 SAM were immersed for typically 8-10 h in solutions of the materials in toluene. The surface-modified constructs were then rinsed with toluene (5 × 2 mL) and investigated by UV-vis spectroscopy. The λmax data for the principal absorption band of the surface-bound ZnTPP and CoTPP, the Soret band, are collected in Table 1, and do not vary significantly for the different constructs. The third column of the table gives the Soret band absorption intensities for the macrocycles bound to trans-form 1 SAM and trans-form 2 SAM, Atrans. These are significantly higher than for material bound to the cis-form films: the ratios of the absorbance intensities from material bound to the cis-form and trans-form films, Acis/Atrans, are presented in the fourth column of the table. Representative spectra for the compounds bound to the trans and cis forms of the SAMs are shown in Figure 5AC. That the Acis/Atrans ratios are generally quite small is in line with expectations that the number of accessible binding sites is much reduced after trans to cis photoswitching. However, we note that the Acis/Atrans data may not directly measure the ratio of material bound to the (22) Wang, J. Analytical electrochemistry, 2nd ed.; Wiley: New York, 2000. (23) Zhang, Z.; Hu, R.; Liu, Z. Langmuir 2000, 16, 1158-1162. (24) Offord, D. A.; Sachs, S. B.; Ennis, M. S.; Eberspacher, T. A.; Griffin, J. H.; Chidsey, C. E. D.; Collman, J. P. J. Am. Chem. Soc. 1998, 120, 4478-4487.

A Photoswitch for Macrocycle Coordination and Release

Langmuir, Vol. 20, No. 14, 2004 5855

Figure 5. Spectra of constructs of macrocycles surface bound to SAMs in the trans form (solid line) and the photoswitched cis form (broken line): (A) CoTPP/1 SAM; (B) CoTPP/2 SAM; (C) ZnTPP/2 SAM; (D) (C8H17)8CoPc/2 SAM.

Figure 6. Spectral monitoring of the release of coordinated macrocycles from the surfaces of SAMs in the trans form (solid line) arising from irradiation of the constructs in toluene using waveguided light at 365 nm. The broken line shows the reduced Soret band absorption after 10 h of irradiation. The insets show reduction in the absorbance as a function of time. The data refer to (A) CoTPP on 1 SAM, (B) ZnTPP on 1 SAM, (C) CoTPP on 2 SAM, and (D) ZnTPP on 2 SAM.

surfaces because of possible differences in the orientation of the macrocycle in the trans- and cis-form films with respect to the spectrometer light source. Nevertheless, the smaller Acis/Atrans ratios measured for 2 SAM are

indicative that there are fewer available coordination sites available in the cis form than is the case for 1 SAM; i.e., 2 SAM undergoes more complete photoisomerization. The difference in the ratios for CoTPP and ZnTPP suggests

5856

Langmuir, Vol. 20, No. 14, 2004

Wang et al.

Figure 7. A schematic representation of an evanescent-wavedriven release/recoordination cycle applied to 2 SAM immersed in a solution of ZnTPP in toluene.

that the latter can be regarded as the more discriminating probe for binding sites. Finally, we investigated binding of a non-porphyrin macrocycle, (C8H17)8CoPc, to the trans and cis forms of 2 SAM. The surface-bound material was monitored through the intensity of the Q-band absorption in the 700 nm region. Unlike the Soret band absorptions from the porphyrin derivatives, the Q-band is substantially broadened, Figure 5D, relative to that of the solution-phase spectrum, a phenomenon we have observed before in our work on single-component phthalocyanine SAMs.20,25 The Acis/Atrans value is 0.17, similar to that of ZnTPP, but there is a higher error in the measurement because of the weak, broad-band absorption from the macrocycle on the cis form of the SAM. Evanescent-Field-Driven Release of Coordinated CoTPP and ZnTPP from Trans-Form SAMs. Having established that the surfaces of 1 SAM and 2 SAM in their trans form present surfaces for binding macrocycles to form heterocomponent constructs, experiments were undertaken to investigate the release of surface-coordinated material using 365 nm light, i.e., at the wavelength used to effect the trans to cis photoisomerization. Substrates supporting trans-form 1 SAM and 2 SAM surfaces coated with coordinated CoTPP and with ZnTPP were placed vertically in a 60 mL jar. This contained sufficient toluene such that the top end of the substrate was just above the surface of the solvent and available to receive and waveguide 365 nm light along the long axis of the substrate. The constructs were monitored by UV-vis spectroscopy over a period of 10 h. In all four cases there was a significant decrease in Soret band intensity over the first 3 h of irradiation, after which time there was little change. Figure 6 shows the absorption spectra at time zero, i.e., of material coordinated onto the trans form of the SAM, and after 10 h. The insets show the intensity changes with time. The ratios of the absorbance after 300 min, A300, to that at time zero, Atrans, are 0.46 for CoTPP on 1 SAM, 0.33 for CoTPP on 2 SAM, 0.40 for ZnTPP on 1 SAM, and 0.22 for ZnTPP on 2 SAM. To demonstrate that the reductions of absorption intensities were not merely a consequence of dissolution of material from the surface, a parallel set of experiments was undertaken in (25) Cook, M. J.; Hersans, R.; McMurdo, J.; Russell, D. A. J. Mater. Chem. 1996, 6, 149-154.

Figure 8. Soret absorption band from the construct of 2 SAM supporting coordinated ZnTPP before irradiation (line 0) and after alternating administrations of light at 365 nm over 120 min (lines 1 and 3, i.e., after the first and third photoswitching) and at 439 nm over 180 min (lines 2 and, 4 i.e., after the second and fourth photoswitching). The inset identifies more clearly the changes in absorbance arising from the photoswitching processes.

which 365 nm light was not applied to the samples. In these experiments, the ratios of the absorbance after 180 min to that at time zero were 0.92 for CoTPP-coordinated films and 0.90 for ZnTPP-coordinated films; i.e., only ca. 10% of material was lost by dissolution. In a further set of experiments solutions of both CoTPP and ZnTPP in toluene containing 2% pyridine (to mimic the SAM-bound macrocycles) were irradiated directly with 365 nm light over 3 h. The intensities of the Soret bands were reduced by less than 1% for CoTPP and less than 6% for ZnTPP. These results demonstrate that the much more significant losses of intensity of the Soret band in the film spectra are not consequences of photodecomposition. Thus, the A300/Atrans values reported above arise primarily from loss of material as a result of photoswitching. The results show that the release of coordinated material from 1 SAM is less complete than from 2 SAM and that ZnTPP is more readily released than CoTPP. An Evanescent-Field-Driven Release/Coordination Cycle. Figure 7 shows an idealized cycle that involves processes identified in separate experiments above. The sequence was investigated using a preformed construct of ZnTPP at the surface of trans-form 2 SAM. Photoswitching was undertaken using the protocol described in the previous section. However, to promote recoordination of material, the construct was immersed into a solution of ZnTPP in toluene rather than neat toluene. UV-vis spectroscopy was used to monitor the absorbance of material at the SAM surface after each irradiation, the results of which are summarized in Figure 8. The absorbance for the Soret band of a ZnTPP surface bound to trans-form 2 SAM before illumination, point zero, was 0.014 (see the inset in Figure 8). This fell to 0.006 after

A Photoswitch for Macrocycle Coordination and Release

120 min of irradiation with 365 nm light (switching number 1 in the inset in Figure 8). Irradiation with waveguided light at 439 nm for 180 min (switch 2) restores the trans form of the SAM, which recoordinates ZnTPP. This release/coordination cycle was repeated to give a low and a high absorbance at switching numbers 3 and 4, respectively. Conclusions Two types of SAM film incorporating the photoswitchable 4-pyridylazophenoxy chromophore have been assembled onto gold surfaces. One film contains the chromophore as a single component, 1 SAM, and the other is doped with a nonphotoactive component as a 1:1 mixture, 2 SAM. Spectroscopic evidence indicates that both asdeposited films are in their trans isomeric form. Electrochemical experiments and UV-vis spectroscopy established that the azo groups in both 1 SAM and 2 SAM undergo photoisomerization to the cis form to a substantial extent using 365 nm irradiation. This was achieved using either direct illumination of the surface or the evanescent field, the latter arising when the gold-coated glass substrate that supports the SAM was used as a waveguide. The trans-form SAMs are recovered from the cis-form films by a second irradiation using 439 nm light using either illumination protocol. The extent of the recovery was more complete for 2 SAM than for 1 SAM. In their trans form, the SAMs present pyridine lone pairs at the surface, so presenting an “on” coordinating state. The surface is

Langmuir, Vol. 20, No. 14, 2004 5857

effectively switched “off” in the cis form because the coordinating pyridine groups are now directed away from the surface. This has been demonstrated by a series of coordination studies using zinc and cobalt tetraphenylporphyrins and an octaoctyl-substituted cobalt phthalocyanine. In these experiments the absorbance arising from bands characteristic of the bound macrocycles was measured after the trans and cis films were immersed into solutions containing the macrocycles. Absorbances arising from material on the photoswitched films were typically between 16% and 42% of those from material on the trans films. In a further aspect of the study, release of coordinated macrocycles from the trans-form films was demonstrated using light at 365 nm, waveguided through the substrate while immersed in neat toluene. These sets of experiments showed that, in terms of absorbances of bound material, that photoswitching-induced release of material was more efficient for the 2 SAM. Finally, the control of these various molecular events at the surface of 2 SAM was demonstrated through two consecutive evanescent-field-driven release and recoordination cycles using ZnTPP. Acknowledgment. This work was financially supported by the EPSRC (Grant No. GR/M07434) and the CEC (Contract HPRN-CT2000-00004). The EPSRC’s mass spectrometry service, Swansea, is gratefully acknowledged for the MS data for compound characterization. LA0498861