Article pubs.acs.org/JPCA
Reversed Freeze Quench Method near the Solvent Phase Transition Aliaksandr Marchanka† and Maurice van Gastel*,†,‡ †
Institut für Physikalische und Theoretische Chemie, Rheinische Friedrich-Wilhelms-Universität Bonn, Wegelerstrasse 12, D-53115 Bonn, Germany ‡ Max-Planck-Institut für Bioanorganische Chemie, Stiftstrasse 34-36, 45470 Muelheim an der Ruhr, Germany S Supporting Information *
ABSTRACT: Freeze quenching is a general method for trapping reaction intermediates on a (sub)millisecond time scale. The method relies on a mixing and subsequent rapid freezing of solutions of reactants. If the reaction is limited by diffusion, it may be advantageous to initially mix the reactants under conditions where the reaction does not proceed, e.g., by mixing them at low temperature as solids. The temperature may then be raised close to the melting point of the solvent. Depending on the viscosity of the solvent, the temperature can be raised either by heating or by applying laser pulses of nanosecond length with concomitant conversion of light into heat. A reduction of the dead time and a good control of the reaction speed in comparison to the standard freeze quench technique has been achieved with this method. The feasibility of the method in combination with EPR spectroscopy is verified by examining the important prototypical reductions of benzoquinone and 2,6-dichlorophenolindophenol by ascorbate as representatives for two-step redox reactions. By using light pulses of a laser, the reaction could be driven with rates lowered by 4 orders of magnitude as compared to room temperature reaction rates. This has allowed the observation of previously unobserved radical intermediates: the reduction of DCPIP by ascorbate is found to be strongly pH dependent. It proceeds via two one-electron steps at low pH, whereas at neutral pH, the reduction of DCPIP by ascorbate proceeds in a 1:2 stoichiometry followed by a disproportionation of the ascorbate radicals.
■
INTRODUCTION Many chemical reactions proceed via a series of metastable reaction intermediates or transition states. The elucidation of the reaction pathway of such reactions is often of crucial importance for understanding the function of catalyst molecules in the fields of catalysis and biochemistry. It is frequently a complicated task and a full understanding of the reaction mechanism generally requires the study of all accessible reaction intermediates by synthetic and spectroscopic methods, or the study of modified reaction analogues. The accessibility of reaction intermediates for spectroscopic investigations critically depends on the time constants of growth and decay associated with the intermediate. Often, an activation barrier is present for one of the initial intermediates, whereas subsequent steps may proceed as activationless processes. Under these conditions, the rate-limiting step is the formation of the initial intermediate. Detection of such species and subsequent intermediates in large quantity may be difficult if not unfeasible if activation barriers are absent in subsequent steps. Freeze quenching techniques in combination with spectroscopic studies are regularly employed to trap and investigate reaction intermediates. Freeze quenching is a general method, which proceeds in a sequence of three stages. In the first stage, called the mixing stage, separate solutions of the reactants are rapidly and thoroughly mixed by injecting them into a mixing chamber. During the second stage, called the aging stage, the © 2012 American Chemical Society
mixed solution exits the mixing chamber and the chemical reaction proceeds. In the third stage, cryofixation of the reaction mixture occurs by freezing the solution onto a cold plate or into a cold liquid, e.g., a cold isopentane solution. The time that the reaction is allowed to proceed, called the aging time, is determined by the time-of-flight of the mixed solution measured from the outlet of the mixing chamber up to the contact with cryogenic medium.1−3 The minimum time-offlight depends on the speed with which the mixed solution exits the mixing chamber and the distance to the cryogenic medium. Taking into account the freezing time upon contact with the cryogenic medium and the time the solution is present within the volume of the mixing chamber, the minimum aging time typically amounts to several milliseconds.2,4 By using high pressure and small mixing volumes of nanoliter size, the dead time has been reduced to 137 μs by Cherepanov and de Vries in an experiment referred to as freeze-hyperquenching.1 Freezehyperquenching is presently the fastest method by which the shortest aging times have been obtained. Although freeze quench methodology has appealing features, in particular the ease of use and the possibility of controlling the aging time, the minimum mixing time of the freeze Received: January 17, 2012 Revised: February 29, 2012 Published: March 13, 2012 3899
dx.doi.org/10.1021/jp300555x | J. Phys. Chem. A 2012, 116, 3899−3906
The Journal of Physical Chemistry A
Article
benzoquinone 22−24 and 2,6-dichlorophenolindophenol (DCPIP) 25,26 by ascorbate (vitamin C) as important representatives for two-step redox reactions, whereby paramagnetic one-electron oxidized and reduced intermediates are investigated by electron paramagnetic resonance (EPR) spectroscopy.
quenching technique is still rather large as compared to typical reaction times in chemistry (nanosecond time scale or faster) for it to be generally applicable to the investigation of catalytic mechanisms. Freeze quenching is presently applicable to processes that occur on a time scale larger than 137 μs, such as the reactivity of metal centers in biological systems,1 the catalytic cycles of CoA reductase5 and heterodisulfide reductase,6 the bioreduction of molecular oxygen,7,8 features of bovine liver catalase,9 the reaction of the reduced nitric oxide synthase with oxygen,10,11 or the characterization of tyrosil12,13 and thyil radicals.14−16 It would still be desirable to decrease the dead time even further. With decreased dead times, shorterlived intermediates such as the peroxy intermediate in wild type ribonucleotide reductase17−19 or radical intermediates in reductive epoxide opening reactions20,21 may become accessible for spectroscopic measurements. In an idealized picture for the class of reactions with an activation barrier for the initial intermediate (k2 ≫ k1 in eq 1), one may still be able to trap intermediates immediately following activation of the reaction. If the process associated with k2 involves a transition state where two reactants come together by diffusion, as opposed to, e.g., internal conversion or intramolecular rearrangement reactions, the intermediate may be kinetically stabilized by slowing down the diffusion of molecules in the solvent. This may be accomplished by increasing the viscosity of the solvent or by lowering the temperature. To obtain an estimate for the amount of intermediate that can in principle be trapped, the rate equations associated with an idealized two step reaction given in eq 1 can be trivially solved k1
k2
N1(t ) → N2(t ) → N3(t )
■
MATERIALS AND METHODS Sample Preparation. Separate solutions of 30 mM benzoquinone (Sigma-Aldrich), 120 mM sodium ascorbate (Roth) in pure ethanol, 70%:30% ethanol−water or 60%:40% ethanol−water, and 6 mM 2,6-dichlorophenolindophenol (Sigma-Aldrich) in 70%:30% ethanol−water have been prepared under anaerobic conditions. The pH of the sodium ascorbate solutions and DCPIP solution were 8.1 and 7.8, respectively. Additionally, separate solutions of DCPIP and ascorbate at pH 2.2 have been prepared in the same manner as at neutral pH. All solutions have been frozen in liquid nitrogen. Dispersion has been carried out at 77 K with an IKA UltraTurrax T 25 homogenizer with an S25N-10G dispersion tool in a glovebox under anaerobic atmosphere. Mixtures of (1) benzoquinone and ascorbate to end concentrations of 15 mM and 60 mM and (2) DCPIP and ascorbate to end concentrations of 3 and 60 mM at neutral pH or a low pH of 2.2 have been prepared. In addition, hydrated fullerene HyFn, prepared by ultrasonification as reported in ref 27 has been used as a photosensitizer. The grain size after dispersion was typically 10−50 μm. Thereafter, the frozen samples were transferred into an X-band EPR tube (inner diameter 2.8 mm) immersed in a liquid nitrogen bath. Subsequently, the EPR tube was quickly transferred from the liquid nitrogen dewar to the precooled cryostat of the EPR spectrometer. The temperature is set to 120 K to evaporate the liquid nitrogen. The helium in the cryostat ensures an oxygen-free atmosphere during the measurements. Additionally, a few experiments with a pink solution of 2 mM DCPIP in H2O without ascorbate at a pH value of 2.2 have been performed as a reference measurement. cw EPR Measurements. Continuous-wave (cw)-EPR spectra in the range of 230−240 K were recorded on a Bruker ESP 300E EPR spectrometer equipped with a 4102ST X-band resonator and an Oxford ESR900 helium gas-flow cryostat. cwEPR spectra at 145 and 180 K were recorded on a Bruker Elexsys E580 FT EPR spectrometer equipped with a MD4 Xband EPR/ENDOR resonator, an Oxford CF935 helium gasflow cryostat and Oxford ITC-5035 temperature controller. Laser excitation has been performed with a Continuum Surelite OPO Plus laser pumped by a Surelite II Nd:YAG laser at 10 Hz repetition rate. The laser energy was 11 mJ/pulse at 460 nm, 13 mJ/pulse at 532 nm, and 12 mJ/pulse at 560 nm. The temperature range at which cw-EPR measurements were carried out was 160−180 K for pure ethanol, 145−160 K for 70%:30% ethanol−water, and 230−240 K for 60%:40% ethanol−water. After the temperature was set below the melting point of the solvent, the cryostat was allowed to equilibrate for at least 15 min before the measurement. In experiments in which the temperature of the cryostat is increased without the use of laser pulses, the resonator frequency changed by several megahertz before stabilization of the new temperature. The microwave frequency was 9.66 GHz. The microwave power was 2 mW, and the modulation amplitude was 0.2 mT. In experiments with the aim to measure the time evolution of the reaction intermediate as a function of aging time, the magnetic field was set to the position of
(1)
Neglecting back-reactions, the general solution for the intermediate N2, with educt N2 and product N3 and with starting conditions N1(0) = N, N2(0) = 0, and N3(0) = 0 is N2(t ) =
k1 N (e−k 2t − e−k1t ) − k 2 + k1
(2)
From eq 2, it becomes clear that N2(t) remains essentially 0 if k2 ≫ k1. However, if the processes associated with k1 and k2 are diffusion limited, then k2 ≈ k1 and N2 reaches a maximum value of N/e (=0.37N) at time t = 1/k2. In other words, 37% of the total amount of educt may be trapped in the intermediate state, if the reaction is limited by diffusion. Inspired by this observation, we here examine the feasibility of inverting the three stages of freeze quenching. In reversed freeze quenching, the reactants are initially mixed under conditions where the reaction does not proceed, for example, by mixing and homogenizing them at low temperature as solid powders. The temperature is then raised close to the melting point of the solvent. Depending on the temperature dependence of the viscosity near the phase transition, heat may be added to induce the reaction, or laser pulses of nanosecond length may be applied with concomitant conversion of the laser light into heat. The conversion of light into heat may even be optimized by adding an otherwise unreactive photosensitizer molecule to the frozen solutions. By elimination of the mixing chamber and the time-of-flight, the aging time is expected to be significantly reduced. Moreover, the reaction can be straightforwardly controlled, by switching on and off the light pulses. The feasibility investigation is performed for the reduction of 3900
dx.doi.org/10.1021/jp300555x | J. Phys. Chem. A 2012, 116, 3899−3906
The Journal of Physical Chemistry A
■
Article
RESULTS AND DISCUSSION Reduction of Benzoquinone with Ascorbate. As a model for the reaction described by eq 1, the reduction of benzoquinone by ascorbate has been examined. Literature data reveal that the reduction reaction of ascorbate with benzoquinone is of first order in each reagent:37
maximum EPR signal and the signal was observed as a function of time. Additionally, a series of cw EPR spectra was measured at different stages of aging. In the latter, two-dimensional experiment, the recording time of one EPR spectrum amounted to six seconds. ESE Detected EPR/ENDOR Measurements. ESE detected EPR, ENDOR and three pulse electron spin echo envelope modulation (3P-ESEEM) spectra were recorded on a Bruker Elexsys E580 FT EPR spectrometer. The Hahn echo pulse sequence is shown in Scheme 1A.28,29 The pulse lengths
BQ + AscH− + H+ ⇄ BQ•− + AscH• + H+ ⇄ BQH 2 + Asc
Scheme 1
(3)
The one-electron reduction potential for benzoquinone (BQ/ BQ•−) amounts to +78 mV (vs NHE) at pH 7.0,38 whereas the two electron reduction potential amounts to +280 mV at pH 7.0.24,39,40 The redox potential of ascorbate amounts to +47 mV.24 The lifetimes of the radical intermediates at room temperature are below 0.1 s and require the use of stopped-flow techniques for measurement of the radical intermediates.24,37 In combination with EPR spectroscopy, only the radical intermediate species, BQ•− and AscH•, are observable as representatives of the reaction intermediates. The kinetics of radical generation and decay of the benzoquinone−ascorbate mixture in 60%:40% ethanol−water studied by reversed freeze quenching, essentially determined by the thawing kinetics of the frozen mixture, is shown in Figure 1a. Almost immediately following the temperature jump from 230 to 240 K, the amplitude of the EPR signal increases with a time constant of 190 s and reaches a maximum 250 s after the temperature jump. Afterward, the EPR signal decays with a time constant of 450 s. The cw-EPR spectrum of the radical is shown in Figure 1b. The spectrum has a characteristic isotropic g value of 2.0047 and a hyperfine pattern of five lines with intensity ratio 1:4:6:4:1, which derives from the four protons of the 1,4-benzosemiquinone radical anion, BQ•−.22,24 The shape of the EPR spectrum is typical for that of an organic radical in liquid solution, which is able to rotate on a (sub)nanosecond time scale. Ascorbate radicals have not been observed. For the benzoquinone−ascorbate mixture in ethanol, the recorded cw-EPR spectrum as a function of time at 180 K is shown in Figure 1c in a two-dimensional representation. The width of the EPR spectrum is larger than that of the mixture in 60%:40% ethanol−water at 240 K and the hyperfine structure is absent, indicating that the molecules in cold ethanol do not rotate on a time scale faster than nanoseconds. The growth and decay at 180 K requires more than 20 min. The spectrum slightly shifts with time, indicating that the temperature is not completely constant during the experiment. These observations indicate that the diffusion and rotation of molecules solved in ethanol at 180 K is largely frozen out. Subsequently, experiments with laser excitation have been performed. Figure 2 shows the EPR signal as a function of time at 145 K with pulsed laser irradiation at 560 nm of (a) a benzoquinone−ascorbate mixture and (b) only benzoquinone in 70%:30% ethanol−water. All experimental parameters are the same for both measurements. For the mixture (a), the EPR signal of the benzosemiquinone radical grows immediately after switching on the laser flashes with two time constants of 3.6 ± 0.1 and of 75 ± 4 min. Surprisingly, an increase of the benzosemiquinone signal in (b) has been observed with a similar short time constant of 3.8 ± 0.1 min. The long time constant for growth of 75 ± 4 min is absent in (b). Rather, the amplitude of the radical signal reaches its maximum intensity
were 16 and 32 ns; the delay between pulses was 200 ns. Pulse ENDOR measurements were performed by using a Davies ENDOR sequence (Scheme 1B).28,29 A soft microwave pulse was used for preparation (pulse length 128 ns) and a standard Hahn echo scheme was used for detection (pulse sequence π/2 - π with pulse lengths 64 and 128 ns, τ = 300 ns). The detection pulses were applied 2 μs after the end of the radio frequency π pulse of 18 μs length. At 34 GHz (Q-band), the pulse lengths were 36 and 72 ns; the delay between pulses was 400 ns for the ESE detected EPR pulse sequence. For the Davies ENDOR measurements at Q-band, lengths of the microwave pulses were 300 ns−150 ns−300 ns; the τ value was 700 ns. The three pulse ESEEM pulse sequence is shown in Scheme 1C. It consists of three π/2 microwave pulses of 36 ns length with a delay τ between the first and the second pulse of 300 ns. The delay t between the second and third pulse was stepwise scanned from 300 ns to 4 μs with an 8 ns interval. DFT Calculations. DFT calculations of DCPIP and ascorbate were performed with the ab initio, DFT, and semiempirical SCF-MO package ORCA.30−34 Calculations were performed using the BP86 functional, the TZVP basis set with TZV/C auxiliary basis set and a spin-unrestricted formalism.35 The convergence criteria for the SCF procedure were 10−7 Hartree for the change in total energy and 10−4 for the DIIS error. The DFT calculations have been performed on two ascorbate radicals and three DCPIP radicals; see Schemes 1 and 2 in the Supporting Information for details. The model geometries of DCPIP and ascorbate were constructed using Gaussview.36 The geometries were optimized with ORCA. Molecular orbital plots have been constructed using the orca_plot program from the ORCA package and Gaussview. 3901
dx.doi.org/10.1021/jp300555x | J. Phys. Chem. A 2012, 116, 3899−3906
The Journal of Physical Chemistry A
Article
Figure 2. Doubly integrated EPR signal of a benzoquinone−ascorbate mixture (a) and frozen solution of benzoquinone (b) in 70%:30% ethanol−water at T = 145 K, and excited at 560 nm. The beginning of laser flashing is indicated with arrows.
From the experiments with a temperature jump by heat, it has become clear that the proposed methodology allows an increase in the lifetime of the radical intermediates by 4 orders of magnitude (190 s) as compared to room temperature (
