Article pubs.acs.org/JPCA
Temperature Effect on the Two-Photon Absorption Spectrum of Alltrans-β-carotene M. G. Vivas and C. R. Mendonca* Instituto de Física de São Carlos, Universidade de São Paulo, Caixa Postal 369, 13560-970 São Carlos, São Paulo, Brazil ABSTRACT: In this report, we investigate the influence of temperature on the twophoton absorption (2PA) spectrum of all-trans-β-carotene using the femtosecond white-light-continuum Z-scan technique. We observed that the 2PA cross-section decreases quadratically with the temperature. Such effect was modeled using a threeenergy-level diagram within the sum-over-essential states approach, assuming temperature dependencies to the transition dipole moment and refractive index of the solvent. The results show that the transition dipole moments from ground to excited state and between the excited states, which governed the two-photon matrix element, have distinct behaviors with the temperature. The first one presents a quadratic dependence, while the second exhibits a linear dependence. Such effects were attributed mainly to the trans→cis thermal interconversion process, which decreases the effective conjugation length, contributing to diminishing the transition dipole moments and, consequently, the 2PA cross-section.
I. INTRODUCTION All-trans-β-carotene (ATβC) is a biosynthetic π-conjugated molecule with noteworthy electronic and optical features, such as nonlinear refractive index,1 third-harmonic generation,2 electric conductive,3 excited state absorption,4 two-photon absorption (2PA),5,6 and so on. Among these properties, we can highlight its strong 2PA in the visible region as well as its ultrafast excited state dynamics.7 Such characteristics make this material a potential candidate for application in threedimensional optical storage and optical power limiting. Two-photon absorption is a third-order nonlinear optical effect where two photons are simultaneously absorbed by an atomic or molecular system, promoting a real electronic transition from a lower to an upper energy level, in a single quantum event.8 Due to the quadratic dependence with the irradiance as well as the distinct electric-dipole selection rules, the 2PA process has been extensively employed in various kinds of applications, from biology to physics.9−11 Although in the past few years the 2PA process has been widely investigated in organic compounds because of its great technological appeal, most of these studies were concentrated in intramolecular effects, such as the electronic delocalization,12 degree of molecular planarity,13,14 intramolecular charge transfer process,15 bond length alternation,16 cooperative enhancement between branches,17 and so on. However, there are only a few studies concerning the relationship between the 2PA processes and external effects, such as its dependence with the temperature.18,19 This kind of study can bring important information about the relationship between the molecular structure and the electronic property of organic materials, as well as possible applications of these compounds in photonic devices. In this context, we report here the effect of temperature on the 2PA cross-section of ATβC using a white-lightcontinuum (WLC) Z-scan technique. We observed a decrease in the 2PA cross-section as the temperature is increased, which © 2012 American Chemical Society
was modeled using a three-energy-level diagram within the sum-over-essential states approach. Such an effect was attributed to a trans→cis thermal interconversion process, which decreases the effective conjugation length contributing to a decrease in the 2PA cross-section magnitude.
II. EXPERIMENTAL SECTION We prepared ATβC/toluene solutions with concentrations of 8.5 × 10−5 and 1.7 × 10−3 mol L−1 for linear and nonlinear optical measurements, respectively. ATβC was purchased from Sigma-Aldrich with purity higher than 93% and its molecular structure is illustrated in Figure 1.
Figure 1. Molecular structure of all-trans-β-carotene.
For the optical measurements, the samples were placed in 2 mm thick quartz cuvettes. The linear absorption spectra were recorded using a Cary 17 UV−vis-NIR spectrophotometer. The 2PA spectra were obtained using the WLC Z-scan (ranging from 590 to 740 nm) technique. The femtosecond pump source was a Ti:sapphire chirped pulse amplified system (CPA2001, CLARK-MXR) delivering 150 fs pulses at 775 nm, with energies up to 0.8 mJ at a 1 kHz repetition rate. This beam was used to pump an optical parametric amplifier (OPA) that Received: April 19, 2012 Revised: June 1, 2012 Published: June 7, 2012 7033
dx.doi.org/10.1021/jp303789s | J. Phys. Chem. A 2012, 116, 7033−7038
The Journal of Physical Chemistry A
Article
Figure 2. (a) Linear absorption spectra of ATβC as a function of temperature. The region about 350 nm is highlight to emphasize the trans → cis thermal interconversion process. (b) Linear absorption spectra for the temperatures of 20 and 90 °C and after the cooling (90 → 20 °C).
peak”), and it is related to the presence of small amounts of cisisomer because the ATβC purchased from Sigma-Aldrich is not totally pure (purity >93%). Moreover, as reported by W. V. Doering et al.,22 carotenoid compounds coexist at room temperature as a mixture of isomers (trans and cis). It is observed in Figure 2a that as the temperature is increased (20 → 90 °C) there is a decrease in the magnitude of the “main peak” band, concomitantly with the increase of the “cis-peak” band. It is known from the literature that the increase of temperature in these compounds favors the trans → cis thermal interconversion process due to its small activation energy of approximately 30 kcal/mol.22,23 Activation energy on the order of tens of kcal/mol can be easily obtained by only gently warming the sample.22 The cis-isomers production with the increase of temperature, due to the thermal interconversion process, is highlighted by a box in Figure 2a at around 350 nm. Recent studies on the thermal and photochemical properties of carotenoids22,23 have shown that the heating of ATβC produces high amounts of 13cis-carotene, followed by 9-cis-β-carotene and an unidentified cis-isomer. As described in literature, the isomerization rate of ATβC depends on the solvent; for example, the isomerization rate of ATβC is higher in nonpolar solvents, such as petroleum ether and toluene, than in polar solvents as nitromethane.24,25 A blue-shift (approximately 8 nm) of the 11Ag− → 11Bu+ transition is observed with the increase of temperature in Figure 2. According to the literature,26,27 such effect is due to trans → cis thermal interconversion process that decreases the effective conjugation length of the β-carotene. Among the different isomers possible, probably the 13-cis-isomer is being produced in higher quantities because it has the lower activation energy as well as larger blue-shift as compared with the other isomers.22,27 To check if oxidative degradation of the material is occurring during the experiments, we show in Figure 2b the linear absorption spectrum at room temperature (20 °C), at the highest temperature (90 °C), and the spectrum after the cooling (90 → 20 °C). As can be seen, after cooling the “cispeak” is not altered, while for the “main peak” there is a certain recovery, although it does not completely return to its original value. To verify if the difference between absorption spectrum at room temperature and after cooling is indeed associated to trans → cis thermal interconversion and not due to degradation of the material, we calculated the difference between the normalized area under the absorption spectrum at room temperature and after cooling (black solid and dashed lines in Figure 2b). Such results, obtained for the “main peak” and “cispeak” bands, are the same considering the experimental error,
provides 120 fs pulses, tunable from 460 to 2600 nm. The WLC is produced by focusing a 1110 nm laser beam, from the OPA, with a f = 10 cm lens into a 4 cm thick quartz cell containing distilled water. More details about the experimental setup can be found in ref 6. In the open aperture Z-scan technique, 2PA cross-section is determined by translating the sample through the focal plane of a focused Gaussian beam, while transmittance changes in the far field intensity are monitored. For a 2PA process, the light field creates an intensity-dependent absorption, α = α0 + βI, in which I is the laser beam intensity, α0 is the linear absorption coefficient, and β is the 2PA coefficient. Far from one-photon resonances, the power transmitted through the sample due to a 2PA process, for each wavelength, is integrated over time (assuming a pulse with a Gaussian temporal profile) to give the normalized energy transmittance T (z ) =
1 π q0(z , 0)
∞
2
∫−∞ ln[1 + q0(z , 0)e−τ ]dτ
(1)
with q0 = βI0L(1 + (z 2/z 02))−1
(2)
where L is the sample thickness, z0 is the Rayleigh length, z is the sample position, and I0 is the laser intensity at the focus. The nonlinear coefficient β is obtained by fitting the Z-scan data with eq 1. The 2PA cross-section, σ2PA, is determined from σ2PA = (ℏωβ)/(N), where ℏω is the excitation photon energy, and N is the number of molecules per cm3. Usually the 2PA cross-section is expressed in units of Göppert-Mayer (GM), where 1 GM = 1 × 10−50 cm4 s photon−1. The sample temperature was measured close to the cell region where it was exposed to the excitation light. Moreover, the temperature of the sample was maintained by means of a programmed temperature controller (Eurotherm). The absorption measurements were carried out 5 min after reaching the desired temperature to guarantee thermal equilibrium. In addition, the cuvette was sealed to prevent evaporation of the solvent.
III. RESULTS AND DISCUSSION Figure 2 shows the linear (one-photon) absorption spectrum (1PA) of ATβC as a function of sample temperature. Such spectrum presents two characteristic optical absorption bands. The first one (called “main peak”) is located at 470 nm and is due to the π→π* transition (11Ag− → 11Bu+) from the πconjugated backbone (polyenic chain), presenting a characteristic vibronic structure (155 meV).20,21 The second absorption band (1Ag− → 1Ag+) appears at around 350 nm (called “cis7034
dx.doi.org/10.1021/jp303789s | J. Phys. Chem. A 2012, 116, 7033−7038
The Journal of Physical Chemistry A
Article
Figure 3. (a) 2PA cross-section spectrum of ATβC as a function of temperature obtained using the WLC Z-scan technique. (b) Z-scan signatures corresponding to the peak of 2PA allowed band (630 nm).
Figure 4. (a) 1PA and 2PA cross-section at the peak of the one- (470 nm) and two-photon (630 nm) allowed bands as a function of temperature. (b) 1PA cross-section at the peak of the “cis-peak” band (350 nm) as a function of temperature. The inset shows the thermal coefficient obtained for each process.
indicating that the rate of creation of cis-isomers is approximately equal to the rate of reduction of trans-isomers. In addition, Mercadante et al.28 observed that the thermally induced isomerization and not oxidative degradation is the main reaction that occurs during heating at atmospheric pressure and at temperatures lower than 100 °C. Therefore, the “main peak” band does not completely recover because the produced cis-isomers do not return to their lower energy configuration (trans). To verify as the temperature affects the 2PA cross-section spectrum of ATβC, the same kind of experiment was performed using the WLC Z-scan techniques. Figure 3 (a) shows the 2PA spectrum of ATβC (from 580 to 740 nm) as a function of sample temperature (from 20 to 90 °C). Such spectrum, reported recently,6 presents a strong 2PA band with maximum around 630 nm, as well as the intermediate state resonance enhancement effect.29 We also performed additional measurements using the conventional Z-scan technique at the wavelength corresponding to the 2PA band peak (630 nm) for distinct temperatures, as shown in Figure 3b. The solid lines show the fittings obtained employing eq 1. Although the 2PA excitation region shown in Figure 3a does not correspond to the “main peak” band from linear absorption spectrum, we observed a decrease of 2PA cross-section as a function of temperature that is analogous to the one observed for the 1PA spectrum. Two-photon transitions in centrosymmetric molecules, such as ATβC, obey electric-dipole selection rules distinct from those governing linear excitation processes.5 In this case, one-photon allowed transitions are forbidden by two-photon absorption. Therefore, the 2PA allowed band that appears in the nonlinear spectrum at around 630 nm does not
correspond to any excited state present in the linear absorption spectrum. Such techniques are complementary and, therefore, provide additional information concerning the molecular structure. A possibility in which the cis-isomer could also affect the 2PA spectrum would be if photoisomerization takes place during the nonlinear measurement. However, it is generally accepted that photoisomerization of isomers (mainly all-trans) takes place via the lowest triplet state,30,31 with a characteristic isomerization time smaller than nanoseconds.31 As in our measurements we are using 120 fs pulses with a repetition rate of 1 kHz, there is no time for photoisomerization to occur during the pulse interaction. Figure 4a shows the decrease of the 1PA (“main peak” band, diamonds) and 2PA (squares) cross-section as a function of temperature. In Figure 4b it is presented the increase of the 1PA cross-section (“cis-peak”) with the temperature. All values reported in Figure 4 are due only to thermally induced conformation changes of the chromophore, because the temperature effect on the solvent was subtracted by introducing temperature dependencies on the refractive index of solvent (toluene), as will be described later. As can be seen in Figure 4, the 1PA (“main peak” band) and 2PA cross-section decrease quadratically with the temperature, while the “cis-peak” presents a quadratic increase. As shown in the inset of Figure 4, the 1PA and 2PA thermal coefficients (∂σ/∂T) exhibit a linear dependence with the temperature. In general, however, two distinct behaviors can be observed in Figure 4a for the 1PA and 2PA cross-section. A smaller slope is observed for the temperature range of 20−45 °C, while for temperatures above 50 °C a higher slope was attained 7035
dx.doi.org/10.1021/jp303789s | J. Phys. Chem. A 2012, 116, 7033−7038
The Journal of Physical Chemistry A
Article
(corresponding to the complementary behavior for the cis-peak in Figure 4b). Such behavior can be easily observed by adjusting independent linear fits to the regions T < 45 °C and T > 50 °C. Proceeding in this way (data not shown), we obtained °C thermal coefficients of approximately ∂σT50 /∂T = 4.4 × 10−18 cm2/°C for the 1PA 1PA °C °C /∂T = 34 GM/°C and ∂σT>50 /∂T = cross-section and ∂σT
