J. Phys. Chem. A 2010, 114, 10523–10530
10523
Femtosecond Study of the Deuteron-Transfer Dynamics of Naphtol Salts in Water M. J. Cox and H. J. Bakker* FOM Institute AMOLF, Science Park 104, 1098 XG Amsterdam, The Netherlands ReceiVed: February 16, 2010; ReVised Manuscript ReceiVed: August 17, 2010
We study the rate and mechanism of deuteron transfer from the photoacids 1-naphtol-4-sulfonate (1-NPS) and 2-naphtol-3,7-disulphonate (2-NPS) to acetate base in aqueous (D2O) solution. The photoacids are activated by excitation with 100 fs laser pulses at 267 nm. The electronic absorption and stimulated emission spectra of the photoacid and the conjugate photobase and the vibrational absorption spectra of the hydrated deuteron and the acetate base are probed with broad-band delayed 100 fs pulses at visible and mid-infrared wavelengths, respectively. A significant fraction of the deuteron transfer events are observed to occur on a timescale of 75%) or 2-NPS (Aldrich, >95%) dissolved in D2O (Aldrich, 99.9%). The structural formulas of these compounds are presented in Figure 1. The concentration of the acetate base (CH3COONa, Aldrich, 99%) was varied (0, 0.5, 1, and 2 M). The sample is contained in a rotating temperature-controlled cell between two CaF2 windows separated by a 50 µm spacer. The 267 nm pump pulses are focused in the sample using a CaF2 lens with a focal length
10.1021/jp105506a 2010 American Chemical Society Published on Web 09/09/2010
10524
J. Phys. Chem. A, Vol. 114, No. 39, 2010
Cox and Bakker focusing lens. As a result, only isotropic absorption changes are detected, which implies that the measured signals are not affected by molecular reorientation. Here it should be noted that CaF2 is a weakly birefringent material in the extreme ultraviolet (corresponding to ∼11 nm/cm). This birefringence can make the polarization of the 267 nm excitation pulses slightly elliptical (with the short axis having a maximal amplitude of 3% of the long axis), but this ellipticity is too small to obtain an effect of the molecular reorientation on the data. A computer-controlled translation stage sets the relative delay between the pump and probe pulses, and a chopper blocks every other pump pulse for background subtraction. The instrument response function of the pump-probe setup is formed by the cross correlation of the excitation pulse and the visible/infrared probe pulse and has a width of ∼200 fs. Experimental Results
Figure 1. Molecular structures of the two investigated photoacids and stationary emission spectra of the two photoacids following excitation at 267 nm. The blue and red lines are measured for the photoacids dissolved in water and represent the emission spectra of the excited conjugated photobases (deprotonated form), and the gray lines are measured for the photoacids dissolved in methanol and represent the emission spectra of the excited photoacids (protonated form).
of 40 cm to a focus with a diameter of ∼200 µm. We attenuated the pump pulses to about 1 µJ per pulse to avoid multiphoton effects. We also checked that the measured signals originate from the naphtol-salt solutions and not from the CaF2 windows of the sample cell. The visible probing pulses are focused into the sample with a lens with a focal length of 10 cm to a focus with a diameter of ∼100 µm. The transmitted visible probe light is detected frequency-resolved with an optical spectrograph in combination with a silicon 2 pixel × 256 pixel detector array. The infrared probing pulses are split before the sample with a 50% beam splitter into probe and reference beams. Both probe and reference beams are focused into the sample by a gold-coated parabolic mirror (f ) 150 mm), but only the probe is in spatial overlap with the pump. The transmitted probe and reference beams are dispersed with an Oriel monochromator and are detected by two lines of an Infrared Associates 2 pixel × 32 pixel MCT (mercury-cadmium-telluride) detector array. The measurement of the reference allows for a frequency-resolved correction for shot-to-shot fluctuations in the probe-pulse energy. In all experiments, the pump polarization is rotated to the magic angle (54.7°) with respect to the polarization of the probe pulses using a half-wave plate and a polarizer placed after the
Figure 1 shows the stationary emission spectra of 1-NPS and 2-NPS dissolved in water and methanol following excitation by 267 nm pulses. In water, proton transfer occurs within a few hundred picoseconds, which is much shorter than the excited-state lifetime of ∼10 ns.29,34 As a result, the stationary emission spectrum is completely dominated by the deprotonated state of the photoacid (i.e., the conjugate photobase). In contrast, in methanol there is very little proton transfer within the excitedstate lifetime, with the result that the emission spectrum is dominated by the photoacid. The excited photoacids show an emission band near 350 nm (in methanol), and the conjugated photobases show an emission band centered at 440 nm (in water). Figure 2 shows transient spectra in the visible and midinfrared regions of the spectra for 1-NPS and 2-NPS. The visible transient spectra (top panels) show the decay of a negative differential absorption signal centered at 370 nm with increasing delay. This signal is due to the decay of the stimulated emission band of the excited photoacids. Simultaneously, the differential absorption shows a rise to increasing negative values at ∼440 nm, as a result of the rise of the stimulated emission band out of the excited state of the conjugate photobase. It should be noted that the signals may also contain contributions from excited-state absorption, but these are not observed because the signal is dominated by the stimulated emission of the photoacid (band at 370 nm) and the conjugate photobase (band at 440 nm). The mid-infrared spectra (lower panels) show a broadband absorption at early delays. This signal contains both the absorption of the loose proton/deuteron that is still bound to the excited photoacid and the absorption of the hydrated proton/deuteron.20,21 With increasing delay, the broad-band absorption decays because of the uptake of the deuteron by the acetate base. This latter process leads to the rise of a narrowband absorption at 1710 cm-1 of the carbonyl vibration of acetic acid.13-22 The transient visible spectra show an isosbestic point at ∼400 nm. The presence of such an isosbestic point means that the differential absorption signal of the photoacid (band at 370 nm) shows the same dynamics with opposite sign as the differential absorption of the photobase (band at 440 nm). Hence the spectral dynamics reflects the interconversion of the two species due to deuteron transfer. The presence of an isosbestic point also implies that only the amplitudes and not the shapes of the bands change in time. The kinetics of the interconversion can thus be obtained by monitoring the transient absorption at a particular wavelength as a function of the delay. In the upper panel of Figure 3, the dynamics of the population of excited 1-NPS and
Deuteron-Transfer Dynamics of Naphtol Salts in Water
J. Phys. Chem. A, Vol. 114, No. 39, 2010 10525
Figure 2. Transient visible (top panel) and mid-infrared (bottom panel) spectra after excitation with a 100 fs 267 nm pulse.
Figure 3. Differential absorption at a wavelength of 440/450 nm (1NPS/2-NPS) as a function of delay after optical excitation with a 100 fs 267 nm pulse. The signals show the decay of the population of the excited photoacid, probed via the dynamics of the stimulated emission of the excited conjugate photobase. The solid lines are calculated with the model described in the Appendix.
2-NPS in the absence of the acetate base are shown, as derived from the differential absorption signals of the conjugate photobase probed at 440/450 nm for 1-NPS/2-NPS. It is clearly seen that the photoacid population decays significantly faster for 1-NPS than for 2-NPS. In the lower panel of Figure 3, the decrease in differential absorption at 440 nm is presented for a solution of 1-NPS and four different concentrations of acetate base. The rate of deuteron release is observed to speed up with increasing base concentration. In addition, the decay of the photoacid population becomes strongly nonexponential at high base concentrations. Figure 4 shows the differential absorption in the mid-infrared at 1900 cm-1 (5.2 µm) at four different base concentrations. The signals observed consist of two components. The first component is relatively strong with a maximum amplitude of ∼20 mOD and is observed only near time delay zero. The second component is much weaker and has a maximum amplitude of ∼0.5 mOD. The strong component observed near delay time zero is not connected to deuteron transfer because its dynamics are independent of the base concentration. This component can also not be caused by multiphoton ionization of the sample because the amplitude is observed to be linear in the energy of the excitation pulse. This signal component results from the excitation of high-energy mixed electronic/vibrational states in the naphtol salts by the 267 nm pulse. These high-energy states show a broad-band absorption in the mid-infrared that is associated with transitions within the dense manifold of highenergy states. Because these transitions are electronic in nature, the absorption of these transitions is large (∼20 mOD) in comparison to the absorption of the vibrations of the hydrated deuteron and the carbonyl vibration (∼0.5 mOD). The fast signal component was not observed in previous studies of the proton/ deuteron transfer dynamics of the photoacid pyranine: 8-hydroxy-1,3,6-pyrenetrisulfonic acid trisodiumsalt (HPTS).13-22 Apparently, the early time excited-state absorption in the mid-
10526
J. Phys. Chem. A, Vol. 114, No. 39, 2010
Figure 4. Differential mid-infrared absorption at 1900 cm-1 as a function of delay after excitation with a 100 fs 267 nm pulse. The solid lines are calculated with the model described in the Appendix.
infrared of 1-NPS and 2-NPS is markedly different from the excited-state absorption of HPTS. This difference is probably connected to the fact that the naphtol photoacids are excited with higher-energy photons (267 nm) than HPTS (400 nm). The strong electronic mid-infrared signal rapidly vanishes because of relaxation within the naphtol salts from the excited high-energy states to lower-energy excited states. Hence, at later delays (>10 ps), the measured transient mid-infrared absorption shown in Figure 4 consists of only the slow component. This slow component represents the absorption of both the loosely bound deuterons and the released, hydrated deuterons. This broad-band mid-infrared absorption is associated with the vibrations of the hydration structures of the deuteron in liquid water.36 In the absence of base, the absorption signal shows only a very small decrease in the delay-time range of 300 ps, which can be explained as follows. The hydration structure of a deuteron that is loosely bound to the photoacid is very similar to the hydration structure of a fully hydrated deuteron. As a result, these two species show a similar broad-band vibrational absorption spectrum. These absorption spectra are not exactly the same, as is discussed in ref 21, but at 1900 cm-1 their cross sections are very similar. Therefore, at this frequency the signal is not affected by the transfer from loosely bound deuterons to fully hydrated deuterons, meaning that the decay of this signal in the presence of acetate base directly reflects the uptake of the deuteron by the base. Figure 5 presents the transient mid-infrared absorption at a frequency of 1710 cm-1 as a function of delay, after subtraction of the transient broad-band mid-infrared absorption. This broadband mid-infrared absorption represents both the fast-decaying electronic absorption of high-energy states of 1-NPS and 2-NPS
Cox and Bakker
Figure 5. Differential mid-infrared absorption at 1710 cm-1 as a function of delay after excitation with a 100 fs 267 nm pulse. The absorption corresponds to the carbonyl vibration of acetic acid. The lower panel shows the same results over a shorter time interval. The solid lines are calculated with the model described in the Appendix.
and the infrared absorption of loosely bound and hydrated deuterons. After subtraction, the remaining transient absorption at 1710 cm-1 is a narrow band that represents the carbonyl vibration of acetic acid. This band rises as a result of the uptake of the deuteron by the acetate base. The carbonyl signals presented in Figure 5 show a very rapid rise in which the amplitude strongly increases with the base concentration. Unfortunately, it is not possible to determine the time constant of this rise because the data at early delays in Figure 5 contain substantial scatter. This scatter results from the fact that the data shown in Figure 5 are obtained by subtracting the signal measured at 1710 cm-1 in the absence of base (broad-band mid-infrared electronic and deuteron absorption only) from the signals measured with acetate base (broadband mid-infrared electronic and deuteron absorption + carbonyl vibrational absorption). Nevertheless, it is clear from the data that the initial rise of the carbonyl signal occurs on a timescale of 3 ps), the rate at which the carbonyl signal increases is similar for 2-NPS and 1-NPS. In Figure 5, it is seen that the amplitude of the fast initial rise of the carbonyl signal is much smaller for 2-NPS than for 1-NPS.
Deuteron-Transfer Dynamics of Naphtol Salts in Water Interpretation and Discussion On short timescales, a fast initial rise of the carbonyl signal is observed for which the amplitude strongly increases with base concentration. A similar fast rise has been observed in previous studies of the proton/deuteron transfer of HPTS to different carboxylate bases.13-20 This fast signal has been assigned to proton/deuteron transfer in acid-base complexes in which the (proton/deuteron)-donating hydroxyl group of the photoacid is directly hydrogen bonded to the accepting carboxylate base. By analogy to those results, we also assign the fast rise of the carbonyl signal to deuteron transfer within contact 1-NPS/2NPS-acetate acid-base complexes. The amplitude of the fast initial rise of the carbonyl signal is much smaller for 2-NPS than for 1-NPS, which indicates that for 2-NPS the fraction of directly hydrogen-bonded acid-base pairs is smaller than for 1-NPS. Interestingly, for delays >3 ps the rate at which the carbonyl signal increases is similar for 2-NPS and 1-NPS, indicating that at these later delay times 1-NPS and 2-NPS show similar reaction rates for the deuteron transfer to acetate. We describe the deuteron transfer with a model (described in the Appendix) that is an extension of a recently described model for proton transfer.21 In this model, the deuteron/proton transfer takes place via two parallel channels. In one of these channels, the deuteron is directly conducted from the photoacid to the base via a short-living hydrogen-bonded wire of D2O molecules that temporarily connects the acid and the base (direct conduction channel). In the other reaction channel, the deuteron is first transferred to the D2O solvent and at a later time is taken up by a base (scavenging channel). The model also includes the effects of mutual diffusion of the photoacid and the base. The relative importance of the reaction channels will strongly depend on the base concentration. At high base concentrations, there is a significant probability that an acetate base is located close to the photoacid and direct (conducted) transfer to the acetate base becomes possible. This latter reaction can be much faster than the deuteron transfer to the water solvent that forms the first step of the scavenging channel. If the base concentration is low, then the water solvent will form the primary acceptor of the deuteron, meaning that the scavenging channel will be dominant. To describe the conduction channel, we view the solution as a distribution of acid-base configurations with different numbers of intervening D2O molecules. One of the acid-base configurations is the direct contact acid-base pair (no intervening D2O molecule). These direct contact pairs are expected to show the fastest deuteron transfer. Because the data indicate that 1-NPS and 2-NPS have different fractions of contact acid-base pairs, we include in the model that the fraction of direct contact acid-base pairs can deviate from a statistical fraction. The distribution of the other water-separated acid-base configurations is assumed to be statistical. The water wires that conduct the deuteron from the photoacid to the acetate base are short-lived configurations (transition states) of hydrogen-bonded water molecules. These short-lived configurations arise from the spontaneous fluctuations of the hydrogen bonds between the water molecules. Similar shortlived hydrogen-bonded wires of water molecules are present in bulk liquid water. In the present case, the spontaneous fluctuations of the water hydrogen bonds result in the short-lived connectivity between the photoacid and the base via a few water molecules. This connection allows the rapid transfer of the deuteron. This type of deuteron transfer via conduction is in line with previous results in refs 17-22.
J. Phys. Chem. A, Vol. 114, No. 39, 2010 10527 TABLE 1 kw (ps-1)
k0 (ps-1)
∆
f
1-naphtol-4-sulfonate 1/(110 ( 10) 1/(2.7 ( 0.5) 0.2 ( 0.05 5 ( 1 (1-NPS) 2-naphtol-3,7-disulphonate 1/(250 ( 30) 1/(2.3 ( 0.5) 0.2 ( 0.05 1.5 ( 0.5 (2-NPS)
The conduction channel is described with three parameters: a rate constant k0 representing the characteristic rate of conduction of the deuteron in case the acid and base are separated by one water molecule, a parameter ∆ that defines the decrease in the reaction rate with each additional intervening D2O molecule, and a factor f that equals the ratio of the actual concentration of direct contact acid-base pairs and the statistical concentration of direct contact pairs (which would be Pa). The direct contact acid-base pairs are assumed to react with a time constant of 200 fs. The rate of this reaction could not be determined from the data because the data at early delays show a strong additional contribution of an electronic absorption signal. The rates kp of different water-separated acid-base configurations are given by kp ) k0∆p - 1 (p g 1), with p being the number of intervening water molecules between the photoacid and the base. The dependence of the reaction rate kp on p expresses the fact that the transfer rate in the conduction channel is determined by the probability that the dynamics of the water molecules establish a short-lived water wire capable of conducting the deuteron. This probability will decrease by a factor of ∆ for every additional intervening water molecule. The scavenging channel is described with two parameters: a rate constant kw describing the proton/deuteron transfer rate to the water solvent and a parameter ku that when multiplied by the molar fraction of base-accepting sites Pa defines the reaction rate of hydrated deuterons with acetate ions in the solution. The two reaction channels show very different time dependencies. The scavenging channel will lead to a delayed rise of the carbonyl signal because first a significant concentration of hydrated deuterons has to be generated before the uptake of deuterons by the acetate base can become efficient. This characteristic explains the slow rise of the carbonyl signal at a relatively low base concentration of 0.5 M, as observed in Figure 5. The conduction channel will lead to a rate of deuteron transfer to the base that is initially very high and that decreases with increasing time delay. This decrease in reaction rate is due to the fact that at later times the reaction will become increasingly dominated by deuteron-transfer events within acid-base configurations in which the acid and the base are separated by a large number of intervening water molecules. Because the conduction channel becomes more important with increasing base concentration, this decrease in the reaction rate with time delay is best observed at a high base concentration of 2 M (Figure 5). A fit of the model to the data shown in Figures 3-5 yields the results given in Table 1. In this fit, the dynamics of the photoacid and the photobase (visible differential absorptions), the loosely bound deuteron, the hydrated deuteron, and the accepting acetate base (mid-infrared differential absorptions) are all fitted simultaneously. Deuteron transfer to the D2O solvent is seen to be significantly faster for 1-NPS than for 2-NPS: kw ) 1/(110 ( 10) versus 1/(250 ( 30) ps-1. The two naphtol salts strongly differ in the fraction of direct contact acid-base pairs. For 1-NPS, the population of direct contact pairs is ∼5 times larger than the population expected from a statistical distribution. An enhancement of the fraction of contact acid-base pairs has been observed before for another photoacid (HPTS).13-22 For HPTS, the fractions of direct contact
10528
J. Phys. Chem. A, Vol. 114, No. 39, 2010
pairs were observed to be 0.11, 0.24, and 0.39 at acetate concentrations of 0.5, 1, and 2 M, respectively.13 If the distribution of acetate would have been statistical, then these fractions would have been 0.02, 0.04, and 0.07. Hence, for HPTS the observed fraction of contact acid-base pairs is 5 to 6 times larger than statistically expected, which is quite similar to the present observations for 1-NPS. The large fraction of contact acid-base pairs observed for HPTS and 1-NPS can be explained by the fact that the hydrogen bond between the O-D group of the photoacid and the acetate base is much stronger than the hydrogen bond between this O-D group and a water molecule. For 2-NPS, the fraction of direct contact acid-base pairs is less enhanced (by a factor of ∼1.5 only) than for 1-NPS and HPTS (by a factor ∼5). This difference can be explained by the close proximity of an SO3- (sulfonate) group in the case of 2-NPS. Because of the nearby (ortho) location of the bulky and negatively charged sulfonate group (Figure 1), the formation of a hydrogen bond of the O-D group to acetate will be sterically hindered. The hydroxyl group of 2-NPS can also form an intramolecular hydrogen bond with the sulfonate group, which in principle could also contribute to a reduction of the fraction of contact acid-base pairs. However, the formation of such an intramolecular hydrogen bond would also reduce the deuteron-transfer rate at later delay times because the intramolecular hydrogen bond would hinder the formation of a water wire connecting the acid and the base. This latter notion does not agree with the observations. At later times, the rates of 1-NPS and 2-NPS are similar, as expressed by their values of the rate constant k0. For both 1-NPS and 2-NPS, we find k0 ) 1/(2.5 ( 0.5) ps-1 and ∆ ) 0.2 ( 0.05 showing that 1-NPS and 2-NPS possess the same deuteron-transfer rates when the O-D group of the photoacid is connected to the acetate base via one or more water molecules. Hence, the ortho location of the sulfonate group in 2-NPS affects only the fraction of contact acid-base pairs, which points to a steric hindrance effect. For 1-NPS and HPTS, this effect is much smaller or even absent because the closest sulfonate is in the meta position (HPTS) or in the para position (1-NPS, see Figure 1). It is also interesting that the rate of deuteron transfer by direct conduction to a nearby (acetate) base is not determined by the transfer rate of the photoacid to water. 1-NPS and 2-NPS have quite different values of kw but nevertheless have similar values of k0. This aspect can also be seen directly in the data. In Figure 5, it is seen that after the initial rise (associated with the direct contact acid-base pairs) the deuteron-transfer rate is quite similar for 2-NPS and 1-NPS. The value of k0 of 1/(2.5 ( 0.5) ps-1 for 1-NPS and 2- NPS is also similar to the value of k0 of 1/(2 ( 0.3) ps-1 found for deuteron transfer between HPTS and acetate.19,20 The rate of deuteron conduction to a nearby base thus appears to be quite independent of the nature of the photoacid. In contrast, the nature of the base does play a role: the deuteron transfer between HPTS and the weaker base chloroacetate21 has a k0 value of 1/(10 ( 2) ps-1, which is 4 to 5 times smaller than the value of k0 for the reaction between photoacids 1-NPS, 2-NPS, and HPTS and the acetate base. Hence, for these three relatively strong photoacids, the rate of deuteron transfer in water-separated acid-base configurations is determined by the strength of the base only. This notion is of course valid only for water-separated acid-base configurations for which the reaction via direct conduction is faster than the reaction via the scavenging channel. For configurations in which the acid and the base are separated by many water molecules, the reaction will be dominated by the scavenging channel, and the rate constant kw of deuteron transfer to (pure)
Cox and Bakker water is important because this transfer constitutes the first step of this channel. Conclusions We studied the mechanism of intermolecular aqueous deuteron transfer between photoacids 1-naphtol-4-sulfonate (1-NPS) and 2-naphtol-3,7-disulphonate (2-NPS) and the acetate base. We probed the transient electronic absorption and stimulated emission of the excited photoacid and photobase with femtosecond visible pulses and the vibrational absorptions of the loosely bound deuteron, the hydrated deuteron, and the carbonyl group of acetic acid with femtosecond mid-infrared pulses. We find that the proton/deuteron transfer involves two competing reaction channels. In the first channel, the proton is transferred from the photoacid to the water solvent and is later taken up from the solvent by an acetate ion (scavenging channel). In the second channel, the proton is transferred from the excited photoacid to acetate within a transition-state complex in which the acid and the base are connected by a short-lived water wire consisting of a few hydrogen-bonded water molecules (direct conduction channel). We analyzed the data with a model that incorporates both reaction channels. We find that deuteron transfer to the pure D2O solvent (no base added) is significantly faster for 1-NPS than for 2-NPS: kw ) 1/(110 ( 10) versus 1/(250 ( 30) ps-1. We find that with increasing base concentration a larger fraction of the deuteron transfer occurs in direct contact photoacid-acetate base pairs. These transfers are very fast with a time constant of
