Thermal Cis-to-Trans Isomerization of Azobenzene ... - ACS Publications

Article pubs.acs.org/JPCC

Thermal Cis-to-Trans Isomerization of Azobenzene-Containing Molecules Enhanced by Gold Nanoparticles: An Experimental and Theoretical Study Evgenii Titov,†,‡ Liudmila Lysyakova,† Nino Lomadze,† Andrei V. Kabashin,§ Peter Saalfrank,*,‡ and Svetlana Santer*,† †

Experimental Physics, Institute of Physics and Astronomy, University of Potsdam, Karl-Liebknecht-Strasse 24-25, 14476 Potsdam, Germany ‡ Theoretical Chemistry, Institute of Chemistry, University of Potsdam, Karl-Liebknecht-Strasse 24-25, 14476 Potsdam, Germany § Aix Marseille University, CNRS, LP3 UMR 7341, Campus de Luminy, Case 917, 13288, Marseille Cedex 9, France S Supporting Information *

ABSTRACT: We report on the experimental and theoretical investigation of a considerable increase in the rate for thermal cis → trans isomerization of azobenzene-containing molecules in the presence of gold nanoparticles. Experimentally, by means of UV−vis spectroscopy, we studied a series of azobenzene-containing surfactants and 4-nitroazobenzene. We found that in the presence of gold nanoparticles the thermal lifetime of the cis isomer of the azobenzenecontaining molecules was decreased by up to 3 orders of magnitude in comparison to the lifetime in solution without nanoparticles. The electron transfer between azobenzene-containing molecules and a surface of gold nanoparticles is a possible reason to promote the thermal cis → trans switching. To investigate the effect of electron attachment to, and withdrawal from, the azobenzene-containing molecules on the isomerization rate, we performed density functional theory calculations of activation energy barriers of the reaction together with Eyring’s transition state theory calculations of the rates for azobenzene derivatives with donor and acceptor groups in para position of one of the phenyl rings, as well as for one of the azobenzene-containing surfactants. We found that activation barriers are greatly lowered for azobenzene-containing molecules, both upon electron attachment and withdrawal, which leads, in turn, to a dramatic increase in the thermal isomerization rate.

1. INTRODUCTION The catalytic activity of gold nanoparticles (Au NPs) has received considerable attention in many organic and inorganic reactions such as selective oxidation/reduction of organic compounds, carbon−carbon coupling, or synthesis of organic materials.1−5 In heterogeneous catalysis, the reacting molecules adsorb on the catalytically active solid surface followed by breaking and formation of chemical bonds on the surface and eventually desorption of newly formed products back into the liquid or gas phase. One of the explanations for the catalytic activity of gold nanoparticles is related to a charge transfer between the adsorbed molecule and the gold particle.6−8 One of prominent cases of Au NPs catalytic activity involves azobenzene molecules, which are known to offer lightcontrolled isomerization yielding a variety of applications in light-triggered switching of polymers, surface-modified materials, protein probes, molecular machines, holographic recording devices, and metal ion chelators, etc.9,10 The azobenzene molecule consists of two phenyl rings linked by azo group (NN) and can be reversibly switched between trans (more stable) and cis (less stable) isomers. The trans → cis isomerization in solution is induced by light illumination with appropriate wavelength in the UV range. The cis → trans isomerization, the back isomerization, can be achieved by © XXXX American Chemical Society

illumination with light of longer wavelength (in the blue range). Back isomerization also occurs readily at room temperature in the dark because of relatively low activation energy barrier ∼95 kJ mol−1 (∼1 eV/molecule).9 The thermal back isomerization typically takes several hours or days. However, the presence of gold nanoparticles was found to accelerate the back isomerization of azobenzene down to several minutes.11−14 This phenomenon was reported for azobenzene-derivatized sulfides adsorbed and spatially confined in gold nanoparticle aggregates,11 for azobenzene−oligoglycerol conjugates on cyclodextrin-coated gold nanoparticles,12 and finally for aqueous azobenzene solutions under the addition of Au NPs.13 An efficient interaction of azobenzene with gold in the latter case was possible because of the pseudo-naked surface of Au NPs (only Cl− left on the surface),13,15 conditioned by laser irradiation of chemically synthesized NPs in a small droplet. Hallett-Tapley et al. suggested that the rapid thermal cis → trans isomerization is due to an electron transfer from azobenzene to the gold nanoparticles.13 The charge transfer results in the formation of an azobenzene radical cation, which Received: March 13, 2015 Revised: July 3, 2015

A

DOI: 10.1021/acs.jpcc.5b02473 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C isomerizes far more readily than neutral azobenzene, probably through rotation about the NN bond.13 No striking evidence for the formation of an azobenzene radical cation could however be given. In this context it is interesting to note, though, that earlier spectrophotometrical16 and electrochemical17 investigations revealed that indeed a cis azobenzene radical anion isomerizes very fast to the trans isomer. We recently explored interactions of azobenzene-modified cationic surfactants with Au-based nanomaterials.18 The incorporation of an azobenzene moiety into the hydrophobic tail of a surfactant, having a charged headgroup, enables one to light-trigger the hydrophobicity of the surfactant in order to control the strength of interactions with other charged compounds with a variety of promising applications in photoreversible DNA compaction,19−22 remote control of the size of microgel particles triggered by light,23−25 and lightcontrolled structuring of polymer brushes.26−28 Profiting from electrostatic interactions between the positively charged surfactant head and negatively charged bare Au nanoparticles prepared by methods of laser ablation in deionized water,29−33 we have created novel plasmonic nanoarchitectures based on azobenzene-modified cationic surfactants−Au nanoparticle complexes. In this case, a light-tunable azobenzene unit in the surfactant can be used to initiate light-controlled nanoclustering of Au NPs, yielding a switch in the plasmonic absorption band and a related change in the solution color. Despite the different materials used (bare Au NPs and azobenzene-modified surfactants) and the different nature of involved chemical interactions, we also could observe a drastic acceleration of back cis → trans isomerization of azobenzene in the formed azobenzene-modified cationic surfactant−Au nanoparticle complexes.18 This paper is conceived as an attempt to clarify the nature of the observed acceleration of back cis → trans isomerization of azobenzene in the presence of Au NPs. In our tests, we vary the nature of the headgroup starting from toxic cationic trimethylammonium to less toxic polyamines,34 as well as the length of the hydrophobic tail. To gain theoretical insight into this phenomenon, we perform density functional theory (DFT) calculations of the activation energy barriers for isolated azobenzene-containing molecules in differently charged states and apply Eyring’s transition state theory to calculate isomerization rates. We also estimate the influence of a gold surface on the formation energies of ions of azobenzene derivatives. Since the direction of charge transfer between molecular switches and Au nanoparticles in solution is not known with certainty, we consider both (radical) anions and cations. The paper is organized as follows. In section 2 the studied molecules and experimental and computational details are described. In section 3 we present our results and discuss them. Section 4 concludes the work.

Table 1. Chemical Structure of the Studied Molecules, the Used Substituents, Abbreviations, and Atom Labeling

abbreviation

R′

R

4-R-AB AzoCn AzoEAn

H (CH2)3CH3 (CH2)3CH3

H, CH3, OCH3, NH2, CF3, NO2 O[CH2]nN+(CH3)3, n = 6, 8, 10, 12 O(CH2)6[NH(CH2)2]nNH2, n = 1, 2

experimentally). The cationic azobenzene-containing surfactants have not been studied before in the context of acceleration of thermal cis → trans isomerization by gold nanoparticles in contrast to small azobenzenes (similar to 4-RAB molecules considered here) which were investigated experimentally in refs 13 and 14 and were found to undergo rapid thermal back isomerization upon addition of Au NPs. However, the calculations have not been done before neither for small neutral molecules nor for long cationic surfactants. In the case of the 4-R-AB molecules substituents R which are either electron donors (CH3, OCH3, NH2) or acceptors (CF3, NO2) were considered in calculations. Hereafter the parent azobenzene molecule 4-H-AB is denoted as AB. For the 4-R-AB molecules we calculated the neutral, anionic, and cationic states, whereas for the AzoC6 surfactant molecule which is initially cationic, the cationic, neutral (one electron added), and dicationic (one electron removed) states were investigated. The choice of cationic azobenzene-containing trimethylammonium and polyamine surfactants was motivated by their important role in remote control of the size and properties of soft objects as mentioned above. The selection of molecules for theoretical investigation serves the same purpose (for AzoC6) but also to elucidate substituent effects for small azobenzenes similar to ref 13. 2.2. Experimental Details. The cationic azobenzenecontaining surfactants were synthesized according to the published procedure.19,34 At first, azo coupling of p-buthylanyline with phenol and subsequent reaction with appropriate dibromoalkane were carried out. The next step was quaternization with trimethylamine (AzoCn) or the reactions with ethylenediamine (AzoEA1) and diethylenetriamine (AzoEA2). For the trimethylammonium surfactants the length of the spacer between a charged head and azobenzene ring ranges between 6 CH2 and 12 CH2 groups (Table 1). The azobenzene-containing polyamines AzoEAn differ in the number of amine groups in the head to be 2 (AzoEA1) and 3 (AzoEA2). The surfactants were dissolved in water (Milli-Q) and kept in the dark for several days to ensure complete relaxation to the trans isomer state. 4-Nitroazobenzene (4-NO2AB) was purchased from Sigma-Aldrich and dissolved in methanol (Sigma-Aldrich). No effect on its thermal relaxation has been revealed when the solution in methanol was slightly diluted with water, to be the reference for the solution in methanol with added gold nanoparticles aqueous colloid. Bare Au NPs were prepared by femtosecond laser ablation from a gold target in deionized water.30,31 To homogenize the size dispersion of Au NPs, in some cases we additionally applied the technique of femtosecond laser fragmentation from already formed bare Au colloids.32,33 The final Au NPs were 10 ± 3 nm in diameter (as measured using TEM). Gold

2. MATERIALS AND METHODS 2.1. Studied Molecules. The chemical structure of the molecules investigated in this work along with used substituents and abbreviations as well as atom labeling are presented in Table 1. The experiments were carried out with neutral 4-nitroazobenzene molecule (4-NO2-AB) and cationic AzoCn and AzoEAn surfactant molecules, whereas the calculations were done for all the 4-R-AB molecules and the AzoC6 surfactant molecule (as a representative of a class of surfactants studied B


Article

The Journal of Physical Chemistry C

3. RESULTS AND DISCUSSION 3.1. Experimental Results. A series of azobenzenecontaining trimethylammonium surfactants, differing in the length of the hydrophobic tail, as well as azobenzene-containing polyamine surfactants, differing in the number of amine groups, were examined with respect to the kinetics of the thermal cis → trans isomerization in the presence of Au NPs. The absorption spectra of all used surfactants are similar, since the ring subtituent groups near the azobenzene unit are the same for all molecules. Figure 1a shows characteristic absorption spectra of

nanoparticles are partially oxidized with atomic percentages estimated as 88.7% for Au0, 6.6% for Au+, and 4.7% for Au3+.35,36 Those gold atoms that are bound to oxygen on the particle surface tend to accumulate positive or negative charge depending on the pH of the surrounding media. At basic conditions (pH ≈ 8 in a solution together with AzoCn surfactants) the nanoparticles are negatively charged. The colloids just produced were stored in darkness at 4 °C. Before mixing with the azobenzene-containing compounds, the gold nanoparticles were sonicated in order to destroy possible aggregates. UV−vis absorption spectroscopy (Cary 5000, Varian Inc.) was used to experimentally estimate the changes in the rate of the thermal cis → trans isomerization. In the experiments, a droplet of gold nanoparticles aqueous colloid was added into solutions of azobenzene-containing molecules previously irradiated with UV light (365 nm, VL-4.L, Vilber Lourmat) for 10 min. 2.3. Computational Details. Quantum chemical calculations were done using density functional theory (DFT) (e.g., see ref 37). The B3LYP hybrid functional38,39 and the 6-31G* basis set40,41 were employed. For the calculations, the Gaussian 09 program (revisions A.02 and D.01)42 was used. The neutral singlet 4-R-AB species were calculated with restricted B3LYP (RB3LYP), whereas for charged doublet 4-R-AB species unrestricted B3LYP (UB3LYP) was used. In the case of the AzoC6 molecule, RB3LYP was applied to treat the cationic singlet state and UB3LYP for neutral and dicationic doublet states, respectively. In a first step, geometry optimizations of the cis and trans isomers were performed. Only real frequencies were obtained for both isomers by normal mode analysis showing that optimized geometries correspond to minima on the potential energy surface. In a next step, transition states (TS) for the cis ↔ trans reaction were located by the synchronous transitguided quasi-Newton method (QST2 and QST3)43,44 and/or optimization to a transition state using the Berny algorithm as implemented in Gaussian 09. Normal mode analysis showed a single imaginary frequency for the TS. The activation parameters of the cis → trans isomerization (activation energy ΔE at 0 K and activation Gibbs free energy ΔG at 298.15 K) were calculated as a difference between values of corresponding parameters for TS and cis isomer:

Figure 1. Absorption spectra of the azobenzene-containing surfactant AzoC8 in the initial dark state (trans, black curve) and UV-irradiated state (cis, red curve) (a) and the kinetics of its cis → trans isomerization in the presence of gold nanoparticles (b). The inset in (b) shows the evolution of optical density at 376 nm with time (open squares), and the corresponding exponential fit (red curve).

Δ) = )(TS) − )(cis) ) = E, G

(1)

Note that the energy E is the point on the potential energy surface, i.e., the sum of the electronic energy and the energy of nuclear−nuclear repulsion. The isomerization rate k of the cis → trans reaction was calculated from the activation Gibbs free energy ΔG according to Eyring’s transition state theory:45 k=

⎛ ΔG ⎞ kBT exp⎜ − ⎟ h ⎝ kBT ⎠

azobenzene molecules illustrated by the cationic surfactant AzoC8 containing 8 CH2 groups in the spacer between the cationic head and the azobenzene unit. The photoisomerization behavior of the surfactant is described in detail elsewhere.19 In short, all surfactants in the dark state are characterized by an absorption band (π → π* transition of the azobenzene unit in the trans conformation) with a maximum at 353 nm (black curve in Figure 1a). Irradiation with UV light induces trans → cis photoisomerization accompanied by changes in absorption spectra (red curve in Figure 1a). The spectrum of the cis isomer is characterized by two absorption bands with maxima at 313 nm (π → π* transition) and at 437 nm (n → π* transition). The band with

(2)

Here kB and h are the Boltzmann and Planck constants, respectively, and T is the temperature. All the thermochemical calculations were done under ideal gas conditions at the temperature of 298.15 K and the pressure of 1 atm. C


Article


Table 2. Thermal cis isomer Lifetimes, cis → trans Isomerization Rates for Free Compounds in Solution without Gold (τ0, k0) and in the Presence of Gold Nanoparticles (τ, k) and the Acceleration Factor τ0/τ = k/k0

a

molecule

τ0 (min)

τ (min)

k0 (10−4 s−1)

k (10−4 s−1)

τ0/τ = k/k0

4-NO2-AB AzoCn (n = 6, 8, 10, 12)a AzoEAn (n = 1, 2)a

144 3720 3720

3.5 3.0 9.5

1.2 0.045 0.045

48 56 18

41 1240 392

The lifetimes for the series of AzoCn and AzoEAn surfactants were averaged, since they are approximately the same for each n.

maximum at ∼240 nm presented in the spectra of both isomers corresponds to the absorption of the π-conjugated benzene rings. The back cis → trans isomerization of azobenzene occurs via spontaneous thermal relaxation or can be driven by irradiation with visible light. In this work we studied spontaneous thermal cis → trans isomerization of azobenzene-containing compounds. The characteristic time of the thermal relaxation is referred to as the lifetime of the cis isomer, denoted as τ0 in the absence of Au NPs in what follows. The lifetime of the cis isomer of the surfactant molecules in an aqueous solution is ∼62 h (see Supporting Information, Figure 1). However, the lifetime of the cis isomer of the azobenzene-containing surfactants shortens significantly in the presence of Au NPs. Figure 1b shows the evolution of absorption spectra of the UVirradiated surfactant in the presence of Au NPs. The solution of the azobenzene-containing compound was first irradiated with UV light for 10 min in order to convert the trans isomer to the cis isomer. The irradiation time was chosen to get a so-called steady state at which a maximally achievable concentration of the cis isomer is obtained. For AzoC6, ∼90% of the surfactant molecules are in the cis conformation at the steady state.20 The absorption spectroscopy measurements were started immediately after Au NPs had been added into the UV-irradiated solutions of azobenzene-containing compounds. Figure 1b shows, for AzoC8, the spectrum change with time, revealing the transition from the initial cis to the final trans conformation. At the same time the gold plasmon peak located at 527 nm (523 nm for free Au NPs, gray curve) does not change significantly. The plots of the optical density at a fixed wavelength Dλ versus time t were used to estimate the thermal lifetime of the cis isomer. Specifically, the value of the optical density at λ = 376 nm was chosen for the azobenzenecontaining surfactants. In the case of 4-NO2-AB, the optical density at λ = 360 nm was used. To estimate the lifetime, an exponential fit of the optical density as a function of time was performed using the following equation: ⎛ t ⎞ ⎛ t⎞ Dλ(t ) = D0 + D1 exp⎜ − ⎟ + D2 exp⎜ − ⎟ + ... ⎝ τ⎠ ⎝ τ0 ⎠

In the presence of Au NPs azobenzene-containing surfactants undergo cis → trans relaxation far more readily than free cis isomers. The lifetime τ does not depend markedly on the number n of CH2 groups in the spacer. Azobenzene-containing polyamines show the same trend; the gold-enhanced cis → trans relaxation of AzoEA2 is approximately as fast as in Au NPs-containing AzoEA1 solution. In contrast to the cationic surfactants, 4-NO2-AB in solution is not expected to form ionic bonds with Au NPs. However, its thermal cis → trans relaxation is also enhanced by gold nanoparticles. At this point, we can hypothesize that the azobenzene unit can interact with the gold particle via the NN bond, which can accumulate or give away an electron. Interestingly, the lifetime of the cis isomer of the azobenzene-containing polyamines is longer than that of trimethylammonium surfactants, probably because of stronger ionic interaction of the former with the gold surface via its charged head, which inhibits the competitive interaction via the azobenzene NN bond, which in turn is responsible for the possible electron transfer. 3.2. Computational Results and Discussion. A possible hypothesis to explain the experimental findings is that an electron transfer between azobenzene-containing molecules and the surface of gold nanoparticles with the following formation of the state, different in charge from the initial, is the reason to promote thermal switching. To test this possibility, we investigated various charged states of azobenzenecontaining molecules, treated as gas phase species, considering neither the metal surface nor the solution, and compared the activation parameters and isomerization rates of the thermal cis → trans reaction with calculated ones for a state of reference. Specifically, we studied all 4-R-AB compounds of Table 1 with the neutral molecules being the reference, as well as AzoC6 for which the cation is the reference state. Single electrons were removed from or attached to these references, as outlined above. The energetic picture of location of the isomers and the TS is shown in Figure 2. The energy values of the isomers and the TS, shown in Figure 2, were calculated as described in section 2.3 and then recalculated relative to the value for the trans isomer in the state of reference, i.e., E = 0 for the trans isomer of neutral 4-R-AB species and the cationic state of AzoC6. The values of activation parameters (ΔE, ΔG) and absolute k and relative krel (which is the ratio k(ion)/k(neutral)) isomerization rates calculated for 4-R-AB molecules are presented in Table 3. The activation parameters and isomerization rates for the AzoC6 molecule are shown in Table 4. In the case of AzoC6 we defined krel as k(neutral or dication)/k(cation). Note that for AzoC6 the number of stable conformations, which are the minima on the potential energy surface, is most likely large because of the presence of carbon chain substituents, capable of adopting various conformations. The geometries considered in our work are shown in Supporting Information, Table 4. From Figure 2, Table 3, and Table 4 it is seen that both the electron attachment to and withdrawal from the studied

(3)

In the case of Au NP solutions a reliable fit was achieved using a sum of two exponents in eq 3 for the representation of optical density time dependencies, yielding two lifetimes. One lifetime is comparable with the lifetime of the cis isomer without gold, τ0, and therefore can refer to the thermal cis → trans relaxation of the free azobenzene-containing molecules. Being orders of magnitude shorter than τ0, the second lifetime, τ, shows the Au NP-enhanced isomerization of azobenzenecontaining molecules near a gold surface. The lifetimes τ0 and τ of the cis isomers, the corresponding isomerization rates k0 = 1/ τ0 and k = 1/τ, and the acceleration factor τ0/τ = k/k0 are shown in Table 2. D


Article


azobenzene-containing molecules lead to significant reduction in activation barriers of thermal cis → trans isomerization in comparison with species of reference (neutral 4-R-AB and cationic AzoC6), which is in agreement with previous findings for other azobenzenes.46,47 In particular for AB, ΔE = 0.26 eV in the anionic form and ΔE = 0.15 eV in the cationic form, while the barrier for the thermal isomerization in the neutral form is ΔE = 1.10 eV (here and below the values of activation energies are given per molecule). Thus, the activation energy is ∼4 times smaller in the anionic form and ∼7 times smaller in the cationic form, relative to the value for neutral AB. This reduction in barrier leads to a large increase in the isomerization rate of AB in ionic forms: ∼1013 times higher in the anionic form and ∼1015 times higher in the cationic form in comparison with the isomerization rate of neutral AB. The values of krel in Table 3 and Table 4 are given not very precisely but merely to show the order of magnitude of the enhancement effect. While the computed isomerization rates for the reference states are of reasonable order of magnitude when compared to experiment (in particular for AzoC6), we should stress that the calculated isomerization rates for species with an added/ removed electron and hence enhancement factors krel are much too large. This fact is not too surprising, since our calculations were done for isolated molecules and under ideal gas conditions. The experimental picture is much more complicated. Apart from the fact that the isomerization takes place in solution, one must also expect strong effects of the surface of Au NPs. While possible (integer) charge transfers to and from the surface are accounted for in our model calculations, steric hindrance by the surface and associated, altered energies and structures are not included. Also, the possible interaction of charged with other charged species or neutral molecules in solution is not taken care of. Nevertheless, we qualitatively reproduce the experimental finding of accelerated rates near Au NPs if we accept the hypothesis of the charge transfer from or to the surface. We also assume that trends due to different substituents should be meaningful (note that calculated krel for AzoC6 is bigger than the corresponding value for 4-NO2-AB, which is in (qualitative) agreement with the experiment (see Table 2)). From Table 3 one can then also conclude the following: (1) For neutral species the substitution of hydrogen atoms in para position leads to a reduction in activation energies. The acceptors (CF3, NO2) reduce the barrier more than donors (CH3, OCH3, NH2). Such a trend was revealed previously in ref 48. (Note that the activation energies of ref 48 and here are almost identical, but slight deviations in activation free energies and, hence, rates were obtained for the neutral species compared to the present work. In this context we note that in particular activation entropies are sensitive to restrictive computational settings (Gaussian keywords Opt=Tight, Integral(Grid=UltraFineGrid) were used here).) The most efficient in the reduction of barrier (lowering by ∼0.3 eV the activation energy) is the acceptor NO2. The isomerization rates do not change significantly when substituents are donors. A slight increase, less than 10 times, is observed in the case of electron donating substituents. Acceptors in para position result in a more significant increase of the isomerization rate: ∼10 times bigger for a CF3 group and ∼103 times bigger for a NO2 group. (Note, however, that certain σ-acceptors like fluorine when in ortho positions such as in o-fluoroazobenzenes can lead to exceptionally long-lived cis species.49)

Figure 2. Energies of the cis and trans isomers as well as the TSs of the studied molecules, relative to the energy of the corresponding trans isomer in the state of reference (E = 0 for the trans isomer of the neutral state of 4-R-AB and cationic state of AzoC6). The chemical groups on the bottom axis are the used substituents R. The inset shows the location of both isomers and the TS in the figure. Symbols are the data. Lines are merely shown to clarify the figure.

Table 3. Activation Parameters and Isomerization Rates for Neutral, Anionic, and Cationic States of the 4-R-AB Molecules molecule

a

ΔE (eV)a

AB 4-CH3-AB 4-OCH3-AB 4-NH2-AB 4-CF3-AB 4-NO2-AB

1.10 1.08 1.06 1.04 0.98 0.83


0.26 0.26 0.23 0.21 0.14 0.05


0.15 0.16 0.13 0.09 0.14 0.12

k (s−1)

ΔG (eV)a Neutral 0.99 0.98 0.98 0.96 0.92 0.79 Anion 0.22 0.23 0.19 0.18 0.11 0.06 Cation 0.10 0.11 0.11 0.08 0.10 0.09

krel

1.1 1.7 1.7 3.7 1.7 2.8

× × × × × ×

10−4 10−4 10−4 10−4 10−3 10−1

1.2 8.0 3.8 5.6 8.6 6.0

× × × × × ×

109 108 109 109 1010 1011

1013 1012 1013 1013 1014 1012

1.3 8.6 8.6 2.8 1.3 1.9

× × × × × ×

1011 1010 1010 1011 1011 1011

1015 1015 1015 1015 1014 1012

Energies are given in eV/molecule.

Table 4. Activation Parameters and Isomerization Rates for Cationic, Neutral, and Dicationic States of the AzoC6 Molecule ΔE (eV)a

ΔG (eV)a

1.16

1.04

0.22

0.19

k (s−1)

krel

Cation 1.6 × 10−5 Neutral 3.8 × 109

1014

1.2 × 1010

1015

Dication 0.21 a

0.16


E


Article


spectra.11 However, note that in ref 11 this was not discussed in connection with a possible electron transfer. In the case of the cation formation we expect that the electron will escape from the HOMO of the neutral cis isomer, which has essentially NN nonbonding character. Further inversion and/or rotation around the NN bond can be possible ways to stabilize the cationic state. The TS structures obtained for cations differ from each other in the sense of TS type (inversion/rotation) that depends on the used substituents (see Supporting Information, Table 1). The NN bond length for the cis isomer cations is only slightly affected (shortened by ∼0.02−0.05 Å) in comparison to the neutral cis isomers (Supporting Information, Table 1), which is one reason why the HOMO can be characterized as nonbonding. The isomerization of the cation radical through rotational pathways was hypothesized to be a possible mechanism of the thermal cis → trans isomerization of azobenzene-containing molecules in the presence of Au NPs.13 Since both situations with electrons attached to or removed from the switching molecules can lead to enhanced backward rates and since none of the two possibilities can be excluded a priori, it is instructive to estimate the “energy costs” for formation of anions and cations from the neutral state of 4-RAB, close to an Au NP. For this purpose, we consider first the relaxed electron affinities EA = E0 − E− and ionization potentials IP = E+ − E0, where E0 is the energy of neutral molecule, E− is the energy of anion, and E+ is the energy of cation (relaxed values mean that geometries of anions and cations were optimized). The obtained values are presented in Table 5 (and pictorially represented in Figure 2). (Note that we

(2) For anionic species a strong decrease in barrier for all investigated molecules is observed. All the activation energies are less than 0.3 eV compared to about 1 eV for neutral molecules. This large drop in the value of the barrier in comparison with neutral species results in a dramatic increase (∼1012−1014 times) of the isomerization rate of the thermal cis → trans isomerization, relative to neutral azobenzenecontaining molecules. As in the case of neutral molecules considered above, donors do not influence much the activation barriers and hence the isomerization rate in comparison with anionic AB. On the contrary, acceptors lower the barrier much more efficiently (as was observed for neutral species as well): the activation energy is only 0.14 eV in the case of a CF3 group and merely 0.05 eV in the case of a NO2 group. (3) For cationic species a strong decrease in barrier for all the molecules is found, too. All the activation energies are less than 0.2 eV. The isomerization rates for cationic forms of ABcontaining molecules are ∼1012−1015 times higher than for neutral molecules. We should note that donor and acceptor groups do not have much of an influence on the barrier relative to the value for parent cationic AB. Moreover, acceptors are not more efficient than donors in contrast to the situation observed for neutral and anionic species. The acceleration of isomerization upon electron attachment to and withdrawal from azobenzene-containing molecules can be discussed in qualitative terms using frontier orbital arguments, by considering the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of neutral cis AB (see Figure 3).

Table 5. Relaxed Electron Affinities and Ionization Potentials of cis Isomers of 4-R-AB

a

Figure 3. HOMO ((a, b) different views) and LUMO ((c, d) different views) of neutral cis AB. The molecules and molecular orbitals were visualized with Jmol, an open-source Java viewer for chemical structures in 3D (http://www.jmol.org/).

molecule (cis isomer)

EA (eV)a

IP (eV)a


0.72 0.68 0.61 0.48 1.16 1.66

6.96 6.82 6.62 6.34 7.24 7.47


used the basis set without diffuse functions, namely 6-31G*. The employing of diffuse functions leads, in particular, to larger electron affinities.) From Figure 2 and Table 5 it is seen that for the isolated species the ionization potentials are typically around 7 eV. The electron affinities for all studied 4-R-AB molecules are found to be positive, indicating that the formation of an anion from a neutral molecule is energetically favored compared to formation of a cation. However, this conclusion is true for the species in the gas phase only. On a metal surface the energetics can be most likely different, even if geometries were the same as in the gas phase (which surf they are not). The formation energies of an anion, ΔEsurf − = E− surf surf surf surf − E0 , and cation, ΔE+ = E+ − E0 , on a surface can be roughly estimated, neglecting geometric distortions and chemical bonding, from the following expressions (for an initially neutral surface):

In an (unrelaxed) one-electron picture we expect that upon formation of an anion the electron will occupy the LUMO of the neutral cis isomer, which has NN antibonding character. Then we can suppose that the rotation around NN bond should happen to minimize the unfavorable antibonding interaction. Indeed, the found TS structures of azobenzenecontaining molecules in the anionic state are rotation-type TSs with ωCNN′C′ ≈ 90° (see Supporting Information, Tables 1 and 2). Also, our calculations show that in the anionic state the N N bond length in the cis isomer is typically elongated by about 0.1 Å compared to the neutral molecule. Remarkably, the weakening of the NN bond of the azo group in the cis isomer in the vicinity of a gold nanoparticle was suggested in the literature based on the analysis of surface-enhanced Raman F


Article


of the thermal lifetime of the cis isomer upon the addition of 10 nm gold nanoparticles to aqueous solutions of azobenzenecontaining molecules. The thermal lifetime in the presence of Au NPs was found to be about 1000 times smaller for the trimethylammonium surfactants and 400 times smaller for polyamines. Also 4-nitroazobenzene was studied experimentally in methanol solution (partially containing water after addition of gold colloid), for which we observed a ∼40 times decrease in the thermal lifetime. Following the hypothesis of charge transfer between the azobenzene-containing molecules and gold nanoparticles, we performed DFT calculations of various charged states of azobenzene-containing molecules and found that both the electron attachment to and withdrawal from the azobenzenecontaining molecules lead to a pronounced decrease in the activation energy barrier of the thermal back isomerization. Applying Eyring’s transition state theory, we calculated absolute rates for thermal cis → trans isomerization and found that they are ∼1012−1015 times larger for azobenzene-containing molecules upon electron attachment and withdrawal. Such large numbers are not in quantitative agreement with the obtained experimental values; however, that was also not expected because of differences of the experimental conditions and the idealized gas-phase type theoretical models as adopted here. Finally, we made a simple estimate of the 4-R-AB ion formation energies in the presence of a gold surface and revealed that the energy for cation formation is smaller than for anion formation; for some of the molecules the cation formation energy was even negative. This result suggests an easier cation formation and, thus, corroborates the possible electron transfer from the azobenzene-containing molecule to the gold nanoparticle surface, proposed by Hallet-Tapley et al.13 Certainly, this is still a hypothesis, which should be verified.

ΔE−surf = −EA + Φ + Δim ΔE+surf = IP − Φ + Δim

(4)

Here, Φ is the work function of the surface and Δim the image charge term (Δim = −1/(4z) in atomic units, with z being the distance between ion and image plane). A positive ΔEsurf value corresponds to an energetically unfavorable situation with a smaller tendency to enforce a charge transfer and form an ion. To estimate the formation energies ΔEsurf and ΔEsurf on − + Au(111) surface, the values of Φ = 5.31 eV 50 (note, most probably, the work function of nanoparticles is different from the value used here which is for planar surface) and Δim = −1.50 eV (at z = 2.4 Å, as z we used the vertical distance between the surface and azo group as computed for azobenzene on an extended Au(111) surface in refs 51 and 52) were taken. The results are shown in Table 6. Table 6. Estimated Formation Energies of Anion ΔEsurf − and Cation ΔEsurf + of cis Isomers of 4-R-AB on a Gold Surface

a

molecule (cis isomer)

a ΔEsurf − (eV)

a ΔEsurf + (eV)


3.09 3.13 3.20 3.33 2.65 2.15

0.15 0.01 −0.19 −0.47 0.43 0.66


It is seen that the formation energy of a cation on the gold surface is smaller than the formation energy of an anion; i.e., the tendency to form a cation is greater. For 4-OCH3-AB and 4-NH2-AB the values of ΔEsurf + are even negative, indicating a possible spontaneous formation of a cation on the surface. We should remark that it was observed experimentally with scanning tunneling microscopy that AB does not photoisomerize on a planar gold surface.53 It is also interesting to note that as revealed by the already mentioned dispersioncorrected DFT calculations, the ground-state barrier of thermal cis → trans isomerization of AB along the rotational pathway is significantly reduced on Au(111) in comparison with the gas phase, to about 0.4 eV.52 Photoswitching of azobenzenes on an extended gold surface, however, can be achieved by spatial decoupling of the azo unit from the surface through, for example, attachment of tert-butyl groups to the phenyl rings of azobenzene53 or complexation of AB-containing conjugates to a functionalized surface.12 However, it is very important to note that the system studied here, with gold nanoparticles in solution and azobenzenecontaining species interacting with them, is quite different from the case of photosensitive molecules on a surface in vacuum. Thus, our estimate of formation energies should be considered only as a qualitative guide. It is nevertheless a worthwhile guide, since the complete simulation of the physical system by firstprinciples methods is currently out of reach.

■

ASSOCIATED CONTENT

S Supporting Information *

Kinetics of isomerization in an aqueous solution and the geometries of isomers and transition states of the studied azobenzene-containing molecules. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.5b02473.

■

AUTHOR INFORMATION

Corresponding Authors

*P.S.: e-mail, [email protected]. *S.S.: e-mail, [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was done in the framework of the IMPRS on Multiscale Bio-Systems. P.S. acknowledges support through Sonderforschungsbereich 658, “Molecular switches at surfaces”. E.T. thanks Prof. Dr. A. D. Boese and M. Utecht for valuable advice. We also acknowledge a financial contribution of the French National Research Agency (ANR) (Grant LASERNANOBIO, ANR-10-BLAN-919) and the A*MIDEX project (Grant ANR-11-IDEX-0001-02) funded by the “Investissements d’Avenir” French Government program, managed by the ANR. Finally, we acknowledge support from “Plan Cancer” of

4. CONCLUSIONS In this work we studied the effect of gold nanoparticles on the thermal cis → trans isomerization of various azobenzenecontaining molecules in aqueous solution. Experimentally, cationic azobenzene-containing trimethylammonium and polyamine surfactants were investigated. We have found a decrease G


Article


(21) Zakrevskyy, Y.; Cywinski, P.; Cywinska, M.; Paasche, J.; Lomadze, N.; Reich, O.; Löhmannsröben, H.-G.; Santer, S. Interaction of Photosensitive Surfactant with DNA and Poly Acrylic Acid. J. Chem. Phys. 2014, 140, 044907. (22) Zakrevskyy, Y.; Titov, E.; Lomadze, N.; Santer, S. Phase Diagrams of DNA-Photosensitive Surfactant Complexes: Effect of Ionic Strength and Surfactant Structure. J. Chem. Phys. 2014, 141, 164904. (23) Richter, M.; Zakrevskyy, Y.; Eisele, M.; Lomadze, N.; Santer, S.; v. Klitzing, R. Effect of pH, Co-Monomer Content, and Surfactant Structure on the Swelling Behavior of Microgel-AzobenzeneContaining Surfactant Complex. Polymer 2014, 55, 6513−6518. (24) Zakrevskyy, Y.; Richter, M.; Zakrevska, S.; Lomadze, N.; von Klitzing, R.; Santer, S. Light-Controlled Reversible Manipulation of Microgel Particle Size Using Azobenzene-Containing Surfactant. Adv. Funct. Mater. 2012, 22, 5000−5009. (25) Rumyantsev, A. M.; Santer, S.; Kramarenko, E. Y. Theory of Collapse and Overcharging of a Polyelectrolyte Microgel Induced by an Oppositely Charged Surfactant. Macromolecules 2014, 47, 5388− 5399. (26) Schuh, C.; Lomadze, N.; Rühe, J.; Kopyshev, A.; Santer, S. Photomechanical Degrafting of Azo-Functionalized Poly(methacrylic acid) (PMAA) Brushes. J. Phys. Chem. B 2011, 115, 10431−10438. (27) Lomadze, N.; Kopyshev, A.; Rühe, J.; Santer, S. Light-Induced Chain Scission in Photosensitive Polymer Brushes. Macromolecules 2011, 44, 7372−7377. (28) Kopyshev, A.; Galvin, C. J.; Genzer, J.; Lomadze, N.; Santer, S. Opto-Mechanical Scission of Polymer Chains in Photosensitive Diblock-Copolymer Brushes. Langmuir 2013, 29, 13967−13974. (29) Kabashin, A.; Delaporte, P.; Pereira, A.; Grojo, D.; Torres, R.; Sarnet, T.; Sentis, M. Nanofabrication with Pulsed Lasers. Nanoscale Res. Lett. 2010, 5, 454−463. (30) Kabashin, A. V.; Meunier, M. Synthesis of Colloidal Nanoparticles During Femtosecond Laser Ablation of Gold in Water. J. Appl. Phys. 2003, 94, 7941−7943. (31) Kabashin, A. V.; Meunier, M. Femtosecond Laser Ablation in Aqueous Solutions: A Novel Method to Synthesize Non-Toxic Metal Colloids with Controllable Size. J. Phys.: Conf. Ser. 2007, 59, 354−359. (32) Besner, S.; Kabashin, A.; Meunier, M. Two-Step Femtosecond Laser Ablation-Based Method for the Synthesis of Stable and UltraPure Gold Nanoparticles in Water. Appl. Phys. A: Mater. Sci. Process. 2007, 88, 269−272. (33) Maximova, K.; Aristov, A.; Sentis, M.; Kabashin, A. V. SizeControllable Synthesis of Bare Gold Nanoparticles by Femtosecond Laser Fragmentation in Water. Nanotechnology 2015, 26, 065601. (34) Venancio-Marques, A.; Bergen, A.; Rossi-Gendron, C.; Rudiuk, S.; Baigl, D. Photosensitive Polyamines for High-Performance Photocontrol of DNA Higher-Order Structure. ACS Nano 2014, 8, 3654−3663. (35) Sylvestre, J.-P.; Poulin, S.; Kabashin, A. V.; Sacher, E.; Meunier, M.; Luong, J. H. Surface Chemistry of Gold Nanoparticles Produced by Laser Ablation in Aqueous Media. J. Phys. Chem. B 2004, 108, 16864−16869. (36) Muto, H.; Yamada, K.; Miyajima, K.; Mafuné, F. Estimation of Surface Oxide on Surfactant-Free Gold Nanoparticles Laser-Ablated in Water. J. Phys. Chem. C 2007, 111, 17221−17226. (37) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and Molecules; Oxford University Press: Oxford, U.K., 1989. (38) Becke, A. D. Density-Functional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652. (39) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994, 98, 11623−11627. (40) Hehre, W. J.; Ditchfield, R.; Pople, J. A. Self-Consistent Molecular Orbital Methods. XII. Further Extensions of Gaussian-Type Basis Sets for Use in Molecular Orbital Studies of Organic Molecules. J. Chem. Phys. 1972, 56, 2257−2261.

the French Institute of Health and Medical Research (INSERM).

■

REFERENCES

(1) Hvolbæk, B.; Janssens, T. V. W.; Clausen, B. S.; Falsig, H.; Christensen, C. H.; Nørskov, J. K. Catalytic Activity of Au Nanoparticles. Nano Today 2007, 2, 14−18. (2) Haruta, M.; Kobayashi, T.; Sano, H.; Yamada, N. Novel Gold Catalysts for the Oxidation of Carbon Monoxide at a Temperature far Below 0 °C. Chem. Lett. 1987, 16, 405−408. (3) Bond, G. C.; Louis, C.; Thompson, D. T. Catalysis by Gold; Imperial College Press: London, 2006. (4) Zhou, X.; Xu, W.; Liu, G.; Panda, D.; Chen, P. Size-Dependent Catalytic Activity and Dynamics of Gold Nanoparticles at the SingleMolecule Level. J. Am. Chem. Soc. 2010, 132, 138−146. (5) Lin, C.; Tao, K.; Hua, D.; Ma, Z.; Zhou, S. Size Effect of Gold Nanoparticles in Catalytic Reduction of p-Nitrophenol with NaBH4. Molecules 2013, 18, 12609−12620. (6) Sanchez, A.; Abbet, S.; Heiz, U.; Schneider, W.-D.; Häkkinen, H.; Barnett, R. N.; Landman, U. When Gold Is Not Noble: Nanoscale Gold Catalysts. J. Phys. Chem. A 1999, 103, 9573−9578. (7) Ricci, D.; Bongiorno, A.; Pacchioni, G.; Landman, U. Bonding Trends and Dimensionality Crossover of Gold Nanoclusters on MetalSupported MgO Thin Films. Phys. Rev. Lett. 2006, 97, 036106. (8) van Bokhoven, J. A.; Louis, C.; Miller, J. T.; Tromp, M.; Safonova, O. V.; Glatzel, P. Activation of Oxygen on Gold/Alumina Catalysts: In Situ High-Energy-Resolution Fluorescence and TimeResolved X-ray Spectroscopy. Angew. Chem., Int. Ed. 2006, 45, 4651− 4654. (9) Bandara, H. D.; Burdette, S. C. Photoisomerization in Different Classes of Azobenzene. Chem. Soc. Rev. 2012, 41, 1809−1825. (10) Emoto, A.; Uchida, E.; Fukuda, T. Optical and Physical Applications of Photocontrollable Materials: Azobenzene-Containing and Liquid Crystalline Polymers. Polymers 2012, 4, 150−186. (11) Yoon, J. H.; Yoon, S. Photoisomerization of Azobenzene Derivatives Confined in Gold Nanoparticle Aggregates. Phys. Chem. Chem. Phys. 2011, 13, 12900−12905. (12) Nachtigall, O.; Kördel, C.; Urner, L. H.; Haag, R. Photoresponsive Switches at Surfaces Based on Supramolecular Functionalization with Azobenzene-Oligoglycerol Conjugates. Angew. Chem., Int. Ed. 2014, 53, 9669−9673. (13) Hallett-Tapley, G. L.; D’Alfonso, C.; Pacioni, N. L.; McTiernan, C. D.; González-Béjar, M.; Lanzalunga, O.; Alarcon, E. I.; Scaiano, J. C. Gold Nanoparticle Catalysis of the Cis-Trans Isomerization of Azobenzene. Chem. Commun. 2013, 49, 10073−10075. (14) Simoncelli, S.; Aramendia, P. F. Mechanistic Insight Into the ZE Isomerization Catalysis of Azobenzenes Mediated by Bare and CoreShell Gold Nanoparticles. Catal. Sci. Technol. 2015, 5, 2110−2116. (15) Bueno-Alejo, C. J.; D’Alfonso, C.; Pacioni, N. L.; GonzálezBéjar, M.; Grenier, M.; Lanzalunga, O.; Alarcon, E. I.; Scaiano, J. C. Ultraclean Derivatized Monodisperse Gold Nanoparticles Through Laser Drop Ablation Customization of Polymorph Gold Nanostructures. Langmuir 2012, 28, 8183−8189. (16) Neta, P.; Levanon, H. Spectrophotometric Study of the Radicals Produced by the Reduction of syn- and anti-Azobenzene. J. Phys. Chem. 1977, 81, 2288−2292. (17) Laviron, E.; Mugnier, Y. A Study of the Isomerization of cisAzobenzene Anion Radical in Dimethylformamide. J. Electroanal. Chem. Interfacial Electrochem. 1978, 93, 69−73. (18) Lysyakova, L.; Lomadze, N.; Neher, D.; Maximova, K.; Kabashin, A. V.; Santer, S. Light-Tunable Plasmonic Nanoarchitectures Using Gold Nanoparticle-Azobenzene-Containing Cationic Surfactant Complexes. J. Phys. Chem. C 2015, 119, 3762−3770. (19) Zakrevskyy, Y.; Kopyshev, A.; Lomadze, N.; Morozova, E.; Lysyakova, L.; Kasyanenko, N.; Santer, S. DNA Compaction by Azobenzene-Containing Surfactant. Phys. Rev. E 2011, 84, 021909. (20) Zakrevskyy, Y.; Roxlau, J.; Brezesinski, G.; Lomadze, N.; Santer, S. Photosensitive Surfactants: Micellization and Interaction with DNA. J. Chem. Phys. 2014, 140, 044906. H


Article

The Journal of Physical Chemistry C (41) Hariharan, P.; Pople, J. The Influence of Polarization Functions on Molecular Orbital Hydrogenation Energies. Theor. Chim. Acta 1973, 28, 213−222. (42) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revisions A.02 and D.01; Gaussian Inc.: Wallingford, CT, 2009. (43) Peng, C.; Schlegel, H. B. Combining Synchronous Transit and Quasi-Newton Methods to Find Transition States. Isr. J. Chem. 1993, 33, 449−454. (44) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. Using Redundant Internal Coordinates to Optimize Equilibrium Geometries and Transition States. J. Comput. Chem. 1996, 17, 49−56. (45) Eyring, H. The Activated Complex and the Absolute Rate of Chemical Reactions. Chem. Rev. 1935, 17, 65−77. (46) Füchsel, G.; Klamroth, T.; Dokić, J.; Saalfrank, P. On the Electronic Structure of Neutral and Ionic Azobenzenes and Their Possible Role as Surface Mounted Molecular Switches. J. Phys. Chem. B 2006, 110, 16337−16345. (47) Chapman, C.; Paci, I. Behavior of a Chemisorbed Azobenzene Derivative in an STM Environment: A DFT Study of Charged States and Electric Fields. Chem. Phys. Lett. 2011, 517, 204−210. (48) Dokić, J.; Gothe, M.; Wirth, J.; Peters, M. V.; Schwarz, J.; Hecht, S.; Saalfrank, P. Quantum Chemical Investigation of Thermal Cis-toTrans Isomerization of Azobenzene Derivatives: Substituent Effects, Solvent Effects, and Comparison to Experimental Data. J. Phys. Chem. A 2009, 113, 6763−6773. (49) Knie, C.; Utecht, M.; Zhao, F.; Kulla, H.; Kovalenko, S.; Brouwer, A. M.; Saalfrank, P.; Hecht, S.; Bléger, D. orthoFluoroazobenzenes: Visible Light Switches with Very Long-Lived Z Isomers. Chem. - Eur. J. 2014, 20, 16492−16501. (50) Lide, D. Electron Work Function of the Elements. In CRC Handbook of Chemistry and Physics; CRC Press: Boca Raton, FL, 2005. (51) Maurer, R. J. First-Principles Description of the Isomerization Dynamics of Surface-Adsorbed Molecular Switches. Ph.D. Thesis, Technische Universität, München, Germany, 2014. (52) Maurer, R. J.; Reuter, K. Bistability Loss as a Key Feature in Azobenzene (Non-)Switching on Metal Surfaces. Angew. Chem., Int. Ed. 2012, 51, 12009−12011. (53) Comstock, M. J.; Levy, N.; Kirakosian, A.; Cho, J.; Lauterwasser, F.; Harvey, J. H.; Strubbe, D. A.; Fréchet, J. M.; Trauner, D.; Louie, S. G.; et al. Reversible Photomechanical Switching of Individual Engineered Molecules at a Metallic Surface. Phys. Rev. Lett. 2007, 99, 038301.

I


Thermal Cis-to-Trans Isomerization of Azobenzene ... - ACS Publications

Recommend Documents