Ultrafast Photochemistry of a Manganese-Tricarbonyl CO-Releasing

Publication Date (Web): February 4, 2013. Copyright © 2013 American Chemical Society. *E-mail: [email protected]. Cite this:J...
0 downloads 0 Views 2MB Size
Letter pubs.acs.org/JPCL

Ultrafast Photochemistry of a Manganese-Tricarbonyl CO-Releasing Molecule (CORM) in Aqueous Solution Philipp Rudolf,† Florian Kanal,† Johannes Knorr,† Christoph Nagel,‡ Johanna Niesel,‡ Tobias Brixner,† Ulrich Schatzschneider,‡ and Patrick Nuernberger*,† †

Institut für Physikalische und Theoretische Chemie and ‡Institut für Anorganische Chemie, Universität Würzburg, Am Hubland, 97074 Würzburg, Germany ABSTRACT: Ultraviolet irradiation of a manganese-tricarbonyl CO-releasing molecule (CORM) in water eventually leads to the liberation of some of the carbon monoxide ligands. By ultraviolet pump/mid-infrared probe femtosecond transient absorption spectroscopy in combination with quantum chemical calculations, we could disclose for the exemplary compound [Mn(CO)3(tpm)]+ (tpm = tris(2-pyrazolyl)methane) that only one of the three carbonyl ligands is photochemically dissociated on an ultrafast time scale and that some molecules may undergo geminate recombination.

SECTION: Biophysical Chemistry and Biomolecules

I

studied in the gas phase via transient ionization with different laser wavelengths13−15 or by transient electron diffraction16 and in solution with infrared17−20 or X-ray21,22 transient absorption. These studies corroborate the great interest in the initial steps of CO photolysis from metal carbonyls and underline the multifaceted reaction pathways that may occur. In the gas phase, depending on the excitation wavelength used, several to all CO ligands can be liberated, and even quantum control of the photochemistry of Fe(CO)5 was demonstrated;23,24 in the liquid phase, usually one but in some solvents also two of the CO ligands are released on an ultrafast time scale after absorption of one UV photon.18−20 Among other substances studied in solution are, for example, metal carbonyls with only CO ligands25−28 (even with two29−31 or three32 bonded metal atoms) but also complexes with, for example, cyclopentadienyl33−35 and more complex organic ligands,36 which may even perform photochelation upon CO dissociation.37,38 Although the variety of ultrafast studies on metal carbonyls in solution is vast, aqueous environments have been omitted so far for metal carbonyls (in contrast to CO release from heme proteins pioneered by Anfinrud and co-workers39,40 in water). Very recently, the Kubarych group41−43 has investigated the CO vibrations of metal carbonyl complexes in water with femtosecond MIR pulses. Yet, the ultrafast dynamics of CO photolysis from a metal carbonyl in the biologically most relevant solvent water have not been reported in the literature,

n recent years, in addition to nitric oxide (NO) and hydrogen sulfide (H2S), carbon monoxide (CO) has been identified as the third gasotransmitter in higher organisms, including humans, that is endogenously generated from heme by the activity of heme oxygenase enzymes and is widely studied for its signaling function and cytoprotective activity. For the exploitation of the latter for therapeutic applications in medicine, CO-releasing molecules (CORMs) are investigated as a safe-to-handle in situ source of carbon monoxide.1−6 To serve as a CO prodrug, CORMs have to be stable in air and water, be soluble in physiological buffer, and specifically accumulate in the targeted tissue. Various trigger mechanisms are studied to initiate the CO release in a controlled fashion, including ligand exchange reactions in solution,7,8 enzyme activation (ET-CORMs),9 and photoactivation in the so-called PhotoCORMs.10−12 Especially the latter would allow for a dark-stable CORM prodrug to distribute in the body only to be activated in a spatially and temporally precisely controlled way by application of a focused light pulse. While the general CO release and biological activity profiles of a significant number of (Photo)CORMs have been explored in detail, virtually nothing is known to date about the primary photophysical and photochemical steps involved in the carbon monoxide release from the metal coordination sphere. In particular, although many CORMs can release almost all of the carbonyl ligands bound to the metal, in most cases, it remains unknown whether this actually occurs in a stepwise or concerted fashion. The ultrafast dynamics of CO release in metal carbonyls has been investigated for many compounds and under various conditions. Probably the substance studied in most detail is iron pentacarbonyl Fe(CO)5, whose photodissociation has been © 2013 American Chemical Society

Received: December 12, 2012 Accepted: January 28, 2013 Published: February 4, 2013 596

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

central wavelength of either 264 or 347 nm (see pump pulse spectra in Figure 1). TDDFT calculations (see Experimental Methods for details) have been performed for the [Mn(CO)3(tpm)]+ cation in the gas phase; the results are displayed in Figure 2. The three highest occupied molecular orbitals (HOMOs), which are almost degenerate, show a high electron density at the manganese atom and exhibit mostly Mn 3d character, with small contributions from metal−carbon bonding. We further performed TDDFT calculations of the first 30 singlet excited states and identified the major transitions that can be induced with the pump pulse spectra of Figure 1. The three HOMOs are mostly centered on the manganese atom and are constituted by the three doubly occupied Mn d orbitals, as expected for a low-spin 3d6 system. The electron density of the accessible LUMOs is either located on the tpm ligand (LUMO and LUMO+1) or delocalized over all three CO ligands (LUMO+2 to LUMO+5). Whereas the former might indicate that excitation will not lead to CO photolysis, the latter might imply the contrary, namely, that several CO ligands could be released as the excess energy introduced by a 264 nm photon could be sufficient for such a process.35 For the two LUMOs (LUMO+2 and LUMO+3) reached upon 347 nm excitation, a high electron density is found at the metal and one or two of the CO ligands. The associated antibonding character indicates that CO will be photochemically dissociated by 347 nm excitation. The TDDFT results for the CORM can be compared to the band assignments of well-studied manganese tricarbonyls for which the photochemical behavior is known. For CpMn(CO)3 compounds (Cp = cyclopentadienyl derivatives of the form C5H4Y, the simplest being C5H5) in alkane solution, a UV absorption band in the region at around 325−350 nm is observed, which was assigned to a Mn → Cp charge transfer (CT) transition, also comprised of some Mn → CO CT character. A further UV band at around 250 nm is assigned to an intraligand π → π* transition in the arene ring.48 A quantum yield of 1, with no dependence on the UV wavelength, has been found for the release of one CO ligand from CpMn(CO)3 in various organic solvents;49 in the nonpolar solvent isooctane, it is 0.65.50 As a remarkable aspect, it is reported for several solvents that irradiation of the dicarbonyl leads to further photochemical CO release.51 We therefore have performed

despite the high significance with regard to the CO release from PhotoCORMs in envisaged biological applications. In this Letter, we present results from ultrafast ultraviolet pump/mid-infrared (MIR) probe transient absorption measurements on the PhotoCORM [Mn(CO)3(tpm)]Cl (tpm = tris(2pyrazolyl)methane) (Figure 1). This complex undergoes

Figure 1. The studied CORM and its absorption spectrum (black) in D2O solution, together with the spectrum of the pump pulse, which is centered at either 264 (blue) or 347 nm (red), respectively.

photoinduced CO release upon UV irradiation and has been shown to be tissue-selective and cytotoxic against cancer cells.44 UV irradiation eventually causes the average release of two44 or even all three CO ligands45 per molecule. Our study is comprised of ultrafast experiments with different UV pump wavelengths and is complemented by linear absorption spectroscopy both in the UV and in the MIR in combination with DFT and TDDFT calculations. Our experiments have been performed with the CORM dissolved in heavy water because the absorption of MIR probe light in the 1800−2100 cm−1 spectral region is much lower for D2O than that for H2O.46 The UV absorption in the solvent is negligible for wavelengths above 200 nm.47 The steady-state ultraviolet absorption spectrum of the compound is shown in Figure 1. Two distinct absorption bands are found at 265 and 348 nm, together with a strong absorption extending into the far-UV. We will both identify the underlying transitions in TDDFT calculations and explore the influence on the photochemical CO release for excitation into the two bands. Hence, we performed experiments with pump pulses at a

Figure 2. TDDFT calculations for the dominating electronic transitions induced with the employed pump pulses. Whereas the three involved HOMOs exhibit an electron density mostly located at the manganese atom, the character of the accessible LUMOs varies strongly. 597

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

Figure 3. (Top) Infrared absorption spectrum of the CORM dissolved in D2O in the spectral region of the CO vibrations and spectral position of CO vibrations as obtained from DFT calculations. (Below) DFT results for various possible photoproducts.

Figure 4. Transient absorption of the CORM, recorded with 264 nm pump and MIR probe pulses. Transients for selected wavenumbers after 347 nm excitation are analyzed in Figure 5 to explore the involved dynamic processes.

1958 cm−1 correspond to the symmetric and the antisymmetric CO stretching vibration, respectively. The later is degenerate due to the C3v symmetry of the molecule. When the CORM is excited by a UV photon, several product molecules could be formed, ranging from an unaltered CORM

experiments with two UV pump pulses to test if this is also possible for our CORM in aqueous solution. The infrared absorption, recorded via Fourier transform infrared (FTIR) spectroscopy, of the CO vibrations of the CORM in D2O is displayed in Figure 3. The bands at 2051 and 598

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

Figure 5. Transient absorption for exemplary wavenumbers. (a−d) Recorded for delays of up to 1.5 ns and with 347 nm pump pulses; note the change from a linear (−15 to 50 ps) to a logarithmic abscissa (>50 ps). Negative (a and b) and positive (c and d) signals persist for the full delay. (e,f) A shorter time scale to visualize cooling contributions that decay within the first 20 ps; (f) was recorded with 264 nm excitation.

ground-state bleach signals at 1958 and 2051 cm−1 (blue in Figure 4), which agree with the band positions of the FTIR measurements (Figure 3, top panel). These bleach signals persist for delay times up to several nanoseconds (Figure 5a and b, where the bands are probed at 1951 and 2052 cm−1, respectively). Thus, they evidence ultrafast CO liberation. Two positive transient absorption signals appear in the spectral regions at around 1980 and 1857 cm−1 and are observable over at least several nanoseconds (Figure 5c and d, where the bands are probed at 1982 and 1853 cm−1, respectively). By comparison to the DFT calculations in Figure 3 and to FTIR measurements of long-lived intermediates,45,57 it can be deduced that a stable dicarbonyl species is formed (most likely as a singlet), that is, only one CO ligand has been photochemically dissociated on an ultrafast time scale and has been replaced by a D2O molecule. The energy needed to cleave the Mn−CO bond in Mncontaining CORMs58 is less than half of the energy provided by absorption of a UV photon. We do not see indications for the formation of a monocarbonyl, that is, that two CO ligands are released simultaneously. Thus, the excess energy will be distributed among the vibrational modes of the dicarbonyl product, and it would be sufficient for several quanta of the CO stretch mode. In the literature, it was shown that UV excitation of hexacarbonyls leads to a significant number of molecules with vibrational hot bands in the CO stretch;27 however, such hot CO bands are absent in the photolysis of dicarbonyls.33 From our data, we can infer that the excess energy has to go to low-frequency modes of the dicarbonyl photoproduct. Before vibrational relaxation is over, these modes couple to the CO stretch27 of the newly formed dicarbonyl. This gives rise to a broadened absorption that narrows upon cooling, as is evident both in the spectral region at roughly 1980−2000 cm−1 and also for the isolated product band at 1857 cm−1 (yellow regions in Figure 4). To illustrate this, the latter is analyzed by probing red-shifted from the absorption band center at 1840 cm−1 (Figure 5e). In a biexponential fit, two time constants of (0.6 ± 0.1) and (10 ± 1) ps are found. The former would fit solvation time scales in water,59 but its assignment is aggravated due to processes for very short pump−probe delays described earlier; however, the latter evidences cooling of the hot photoproduct molecules. For hexacarbonyls in alkane, the low-frequency modes lose their energy somewhat slower within a few tens of picoseconds.27,28 The faster vibrational relaxation of our CORM dicarbonyl photoproduct can be attributed to effective

to mono- and dicarbonyl species with additional solvent molecules as ligands. We have calculated (Gaussian 0952 and density functional method B3LYP 6-311+G(d,p) with the optimal scaling factor from ref 53) the wavenumbers of the CO vibrations in a variety of putative photoproducts. Higher multiplicity states were higher in energy and therefore were not considered. The results are summarized in Figure 3 and confirm that photoproducts with one or two released CO ligands should be observable by a band on the low-energy side of the CORM’s CO stretch spectrum displayed in the top graph of Figure 3. Another possibility that might occur after UV excitation is the cleavage of a Mn−N bond. In case it does not directly reform, a carbene isomer with a Mn−C bond could be formed. Compared to the N-coordinated pyrazole, a carbene isomer photoproduct is calculated (left table in the middle row of Figure 3) to show both the symmetrical and antisymmetrical CO stretching vibrations at much lower wavenumbers (1978 versus 2036 cm−1 and 1915/1920 versus 1943 cm−1), but probably more importantly, the significantly reduced difference between the positions of the two sets of CO bands (only 61 cm−1 in the carbene isomer versus 90 cm−1 for the pyrazole binding mode) argues against the formation of such a species. The dynamics following photolysis are monitored by femtosecond transient absorption. Figure 4 shows the MIR transient absorption signal recorded with 264 nm pump pulses. The signal at negative pump−probe delay times is due to the perturbed free-induction decay54,55 and will not be discussed any further. For positive delay times, a strong absorption is observed at all probe wavenumbers that vanishes within the first few picoseconds. This contribution is not due to the CORM but instead obscures the very early dynamics of the compound. It matches the absorption of solvated electrons in D2O solution that can be generated at this pump wavelength56 and was observed in transient measurements on the pure solvent as well. Furthermore, a pump-induced transmission change in the CaF2 windows might add to this strong absorption at early delay times.33 Within these first few picoseconds, solvation of a photoproduct will occur, and hence, free coordination sites due to released CO ligands are filled by D2O molecules on this time scale. The dynamics of the CO release after 347 nm excitation is very similar to the behavior with 264 nm pulses. The important features of these dynamics will be discussed by means of absorption changes at exemplary wavenumbers. For positive delay times, the MIR transient absorption exhibits two distinct 599

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

them would interact with both if possible), we could not see any emergence of a new product absorption. Arguments for this lack of sequential UV-induced CO release might be comprised of the following: First, the dicarbonyl can be repumped, but no bond is broken at all. Second, the dicarbonyl can be repumped and leads to the release of a ligand, but it is the new D2O ligand that is photochemically dissociated (as, e.g., discussed for CpMn(CO)3 in ref 49). In that case, the rebinding dynamics of the aqua ligand is not observable within the time resolution, identical to the replacement of CO by D2O after the first pump pulse. Then, the probed signal would behave as if no repumping had occurred. Third, the dicarbonyl intermediate does not absorb UV light and hence cannot be repumped. This is partially supported by strong red shifts observed for the electronic absorption spectrum of CpMn(CO)3 after photolysis of one CO;50 although, it is unlikely that further UV absorption is not possible. This should be directly verified in future UV pump−UV probe transient absorption measurements. In conclusion, UV excitation of an exemplary CORM in aqueous solution leads to ultrafast photolysis of one CO ligand, which may geminately recombine for some of the molecules. The liberated CO will go to the solvent environment and a water molecule will take the free coordination site. If the CORM is excited in the higher-energy UV band, some molecules will not dissociate, but the excitation might be located mostly at the trispyrazolyl ligand, as is suggested by TDDFT calculations and also in agreement with the experimental observations. Excess energy, both in the dicarbonyl photoproduct and in the CORM if still comprising three CO ligands after excitation, is transferred to lowfrequency vibrational modes. No CO hot bands are observed, that is, the remaining CO ligands are not vibrationally excited through the UV excitation of the CORM. Direct re-excitation with a second UV pulse did not lead to observable monocarbonyl species; hence, a further interaction with UV light seems insufficient to release more CO ligands. Therefore, further loss of CO most likely necessitates manganese oxidation, as suggested by Berends and Kurz.45 The elucidation of the primary processes presented here contributes to a comprehensive understanding of the biological activity of CORMs.

water-assisted cooling. In this regard, water is a unique solvent as it has manifold possibilities for reorientational dynamics and a highly dynamical solvation shell, resulting in a higher acceptance for vibrational energy. This leads to the faster relaxation of hot CO ligands in aqueous41,42 compared to alkaneous solutions.27,28 As discussed, UV excitation of a CORM does not lead to the ultrafast release of two CO ligands. Moreover, it does not even lead to the release of one CO ligand for some of the molecules. Although CO may dissociate, geminate recombination can restore the initial CORM by rebinding the photolyzed CO. Measurements for Cr(CO)6 in alkanes have shown that this becomes less likely the smaller the solvent molecules.28 Such behavior is recognized by looking at dynamics associated with wavenumbers of the intact CORMs; as a first aspect, a bleach recovery can be detected, as is evident for early times in the signal of the antisymmetric CO stretch (Figure 5a). Note that geminate recombination occurs within 1 ps accompanied by a broadening of the absorption due to the coupling of the CO band to the low-frequency modes, and therefore, the recovery reflects again the cooling dynamics rather than recombination dynamics.28 Whereas the bleach recovery may also be partially due to the cooling of the overlapping dicarbonyl product absorption at around 1980 cm−1, the second aspect is more obvious for the symmetric CO stretch centered at 2051 cm−1; none of the possible products after photolysis were calculated to absorb at wavenumbers above 2000 cm−1. Therefore, a line broadening due to recombined CORMs is the origin of the positive signal at 2060 cm−1 (Figure 5f). It is observed around the negative ground-state bleach (caused by CORMs that lost a CO ligand) and persists as long as the CORM is vibrationally hot. Beyond that, the possibility of an excitation without any bond cleavage might exist for CORMs, as suggested by our TDDFT calculations (Figure 2). There is no direct way to distinguish this scenario from geminate recombination on the grounds of a single transient absorption map. Nevertheless, two experimental findings indicate that this might occur by comparing data with different pump wavelengths. The signature of excited yet intact CORM on the blue-energy side of the symmetric CO stretch ground-state bleach (Figure 5f) seems more pronounced in the data with pump pulses at 264 nm than for those at 347 nm. If the recombination dynamics can be considered to be independent of the pump wavelength, 264 nm excitation might indeed involve LUMOs located mostly at the tpm ligand and hence lead less often to CO release. The second finding is that the yield of photolysis is not three times higher at 264 nm than that at 347 nm, although the absorption is (Figure 1), which further supports that not every UV excitation will lead to CO release. Recording of a photolysis excitation spectrum (i.e., measuring the amount of released CO as a function of pump wavelength) could further substantiate this but has not been performed so far. In order to find out whether the dicarbonyl photoproduct can release a further CO upon UV excitation, we have employed a further UV pulse, which may repump the products 200 ps after the initial pump event. By splitting either the 264 or the 347 nm pump pulse with an interferometer setup, we obtained pump pulse pairs temporally separated by 200 ps (for a similar three-pulse experiment, see ref 60) and probed 200 ps after the repump pulse. Although the energies were chosen so that more than a quarter of the molecules in the probed volume were excited by pump or repump pulse (and hence also a few of



EXPERIMENTAL METHODS The CORM was synthesized according to the procedure as described.44 The UV absorption spectrum of Figure 1 was recorded with a JASCO V-670 spectrometer in a cuvette with two 4 mm CaF2 windows and a sample thickness of 100 μm. The same cuvette was also used for the time-resolved studies with sample concentrations of 3−5 mM. The FTIR spectrum of Figure 3 was recorded in a commercial JASCO FT/IR-4100 spectrometer. With few modifications, the employed femtosecond transient absorption setup is similar to the one outlined elsewhere.61 Briefly, a 1 kHz Ti:sapphire amplifier is the main light source. Pump pulses at either a 264 or 347 nm central wavelength were generated as the second harmonic of the output of a noncollinear optical parametric amplifier and had pulse durations of 50 fs and pulse energies of 1−2 μJ. For the pump−repump experiments, the pump pulse was split into two fractions in an interferometer, and both pulses were directed into the sample with a delay of 200 ps. The tunable MIR probe pulses were obtained from a home-built two-stage optical 600

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

(9) Romanski, S.; Kraus, B.; Schatzschneider, U.; Neudörfl, J.-M.; Amslinger, S.; Schmalz, H.-G. Acyloxybutadiene Iron Tricarbonyl Complexes as Enzyme-Triggered CO-Releasing Molecules (ETCORMs). Angew. Chem., Int. Ed. 2011, 50, 2392−2396. (10) Jackson, C. S.; Schmitt, S.; Dou, Q. P.; Kodanko, J. J. Synthesis, Characterization, and Reactivity of the Stable Iron Carbonyl Complex [Fe(CO)(N4Py)](ClO4)2: Photoactivated Carbon Monoxide Release, Growth Inhibitory Activity, and Peptide Ligation. Inorg. Chem. 2011, 50, 5336−5338. (11) Ward, J. S.; Lynam, J. M.; Moir, J. W. B.; Sanin, D. E.; Mountford, A. P.; Fairlamb, I. J. S A Therapeutically Viable PhotoActivated Manganese-Based CO-Releasing Molecule (Photo-CORM). Dalton Trans. 2012, 41, 10514−10517. (12) Pierri, A. E.; Pallaoro, A.; Wu, G.; Ford, P. C. A Luminescent and Biocompatible PhotoCORM. J. Am. Chem. Soc. 2012, 134, 18197−18200. (13) Bañares, L.; Baumert, T.; Bergt, M.; Kiefer, B.; Gerber, G. Femtosecond Photodissociation Dynamics of Fe(CO)5 in the Gas Phase. Chem. Phys. Lett. 1997, 267, 141−148. (14) Bañares, L.; Baumert, T.; Bergt, M.; Kiefer, B.; Gerber, G. The Ultrafast Photodissociation of Fe(CO)5 in the Gas Phase. J. Chem. Phys. 1998, 108, 5799−5811. (15) Trushin, S. A.; Fuss, W.; Kompa, K. L.; Schmid, W. E. Femtosecond Dynamics of Fe(CO)5 Photodissociation at 267 nm Studied by Transient Ionization. J. Phys. Chem. A 2000, 104, 1997− 2006. (16) Ihee, H.; Cao, J.; Zewail, A. H. Ultrafast Electron Diffraction of Transient [Fe(CO)4]: Determination of Molecular Structure and Reaction Pathway. Angew. Chem., Int. Ed. 2001, 40, 1532−1536. (17) Snee, P. T.; Payne, C. K.; Mebane, S. D.; Kotz, K. T.; Harris, C. B. Dynamics of Photosubstitution Reactions of Fe(CO)5: An Ultrafast Infrared Study of High Spin Reactivity. J. Am. Chem. Soc. 2001, 123, 6909−6915. (18) Portius, P.; Yang, J.; Sun, X.-Z.; Grills, D. C.; Matousek, P.; Parker, A. W.; Towrie, M.; George, M. W. Unraveling the Photochemistry of Fe(CO)5 in Solution: Observation of Fe(CO)3 and the Conversion between 3Fe(CO)4 and 1Fe(CO)4(Solvent). J. Am. Chem. Soc. 2004, 126, 10713−10720. (19) Besora, M.; Carreón-Macedo, J.-L.; Cowan, A. J.; George, M. W.; Harvey, J. N.; Portius, P.; Ronayne, K. L.; Sun, X.-Z.; Towrie, M. A Combined Theoretical and Experimental Study on the Role of Spin States in the Chemistry of Fe(CO)5 Photoproducts. J. Am. Chem. Soc. 2009, 131, 3583−3592. (20) Nguyen, S. C.; Lomont, J. P.; Zoerb, M. C.; Hill, A. D.; Schlegel, J. P.; Harris, C. B. Chemistry of the Triplet 14-Electron Complex Fe(CO)3 in Solution Studied by Ultrafast Time-Resolved IR Spectroscopy. Organometallics 2012, 31, 3980−3984. (21) Ahr, B.; Chollet, M.; Adams, B.; Lunny, E. M.; Laperle, C. M.; Rose-Petruck, C. Picosecond X-ray Absorption Measurements of the Ligand Substitution Dynamics of Fe(CO)5 in Ethanol. Phys. Chem. Chem. Phys. 2011, 13, 5590−5599. (22) Wernet, P. et al. Mapping Chemical Bonding of Reaction Intermediates with Femtosecond X-Ray Laser Spectroscopy. Proceedings of the XVIII International Conference on Ultrafast Phenomena; Lausanne, Switzerland, 2012; in press. (23) Assion, A.; Baumert, T.; Bergt, M.; Brixner, T.; Kiefer, B.; Seyfried, V.; Strehle, M.; Gerber, G. Control of Chemical Reactions by Feedback-Optimized Phase-Shaped Femtosecond Laser Pulses. Science 1998, 282, 919−922. (24) Bergt, M.; Brixner, T.; Kiefer, B.; Strehle, M.; Gerber, G. Controlling the Femtochemistry of Fe(CO)5. J. Phys. Chem. A 1999, 103, 10381−10387. (25) Simon, J. D.; Xie, X. Photodissociation of Cr(CO)6 in Solution: Direct Observation of the Formation of Cr(CO)5(MeOH). J. Phys. Chem. 1986, 90, 6751−6753. (26) King, J. C.; Zhang, J. Z.; Schwartz, B. J.; Harris, C. B. Vibrational Relaxaton of M(CO)6 (M=Cr, Mo, W): Effect of Metal Mass on Vibrational Cooling Dynamics and Non-Boltzmann Internal Energy Distributions. J. Chem. Phys. 1993, 99, 7595−7601.

parametric amplifier with subsequent difference-frequency mixing. All beams were spatially overlapped in the sample and had parallel or magic angle polarization directions. A time resolution of 300 fs was measured with a germanium wafer placed inside of a cuvette. The MIR spectrum of the probe was read out for each laser pulse with a MIR spectrometer and a 32 pixel MgCdTe array. Because every second pulse was blocked by a mechanical chopper, the change in absorption was calculated from two consecutive pulses and then averaged. DFT calculations for determination of the wavenumbers of the vibrational modes were performed with the software package Gaussian 09,52 using the density functional Becke three-parameter hybrid method in combination with the Lee− Yang−Parr correlation functional (B3LYP) and the 6-311+G(d,p) basis set. The aqueous surrounding was included by a polarizable continuum model (PCM) in the self-consistent reaction field (SCRF). TDDFT calculations for exploring the electronic levels involved in the excitations were performed with ORCA 2.6 on a Linux cluster using TZV(P) or better basis sets on Mn, N, and O and a VDZ basis set on all other atoms with the Grid4 and TightSCF options.62,63 The first 30 singlet excited states were calculated in the gas phase.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We are grateful to Dominik Gehrig and Heiko Hildenbrand for their help. P.N. acknowledges the Deutsche Forschungsgemeinschaft for financial support within the Emmy-Noether program.



REFERENCES

(1) Alberto, R.; Motterlini, R. Chemistry and Biological Activities of CO-Releasing Molecules (CORMs) and Transition Metal Complexes. Dalton Trans. 2007, 1651−1660. (2) Motterlini, R.; Otterbein, L. E. The Therapeutic Potential of Carbon Monoxide. Nat. Rev. Drug Disc. 2010, 9, 728−743. (3) Mann, B. Carbon Monoxide: An Essential Signalling Molecule. In Topics in Organometallic Chemistry; Metzler-Nolte, N., Jaouen, G., Eds.; Springer: Berlin, Germany, 2010; Vol. 32, pp 247−285. (4) Schatzschneider, U. PhotoCORMs: Light-Triggered Release of Carbon Monoxide from the Coordination Sphere of Transition Metal Complexes for Biological Applications. Inorg. Chim. Acta 2011, 374, 19−23. (5) Romão, C. C.; Blättler, W. A.; Seixas, J. D.; Bernardes, G. J. L. Developing Drug Molecules for Therapy with Carbon Monoxide. Chem. Soc. Rev. 2012, 41, 3571−3583. (6) Rimmer, R. D.; Pierri, A. E.; Ford, P. C. Photochemically Activated Carbon Monoxide Release for Biological Targets. Toward Developing Air-Stable PhotoCORMs Labilized by Visible Light. Coord. Chem. Rev. 2012, 256, 1509−1519. (7) Zhang, W.-Q.; Atkin, A. J.; Fairlamb, I. J. S.; Whitwood, A. C.; Lynam, J. M. Synthesis and Reactivity of Molybdenum Complexes Containing Functionalized Alkynyl Ligands: A Photochemically Activated CO-Releasing Molecule (PhotoCO-RM). Organometallics 2011, 30, 4643−4654. (8) Atkin, A. J.; Fairlamb, I. J. S.; Ward, J. S.; Lynam, J. M. CO Release from Norbornadiene Iron(0) Tricarbonyl Complexes: Importance of Ligand Dissociation. Organometallics 2012, 31, 5894− 5902. 601

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602

The Journal of Physical Chemistry Letters

Letter

(27) Dougherty, T. P.; Heilweil, E. J. Ultrafast Transient Infrared Absorption Studies of M(CO)6 (M = Cr, Mo or W) Photoproducts in n-Hexane Solution. Chem. Phys. Lett. 1994, 227, 19−25. (28) Lian, T.; Bromberg, S. E.; Asplund, M. C.; Yang, H.; Harris, C. B. Femtosecond Infrared Studies of the Dissociation and Dynamics of Transition Metal Carbonyls in Solution. J. Phys. Chem. 1996, 100, 11994−12001. (29) Zhang, J. Z.; Harris, C. B. Photodissociation Dynamics of Mn2(CO)10 in Solution on Ultrafast Time Scales. J. Chem. Phys. 1991, 95, 4024−4032. (30) Owrutsky, J. C.; Baronavski, A. P. Ultrafast Infrared Study of the Ultraviolet Photodissociation of Mn2(CO)10. J. Chem. Phys. 1996, 105, 9864−9873. (31) Steinhurst, D. A.; Baronavski, A. P.; Owrutsky, J. C. Transient Infrared Spectroscopy of Mn2(CO)10 with 400 nm Excitation. Chem. Phys. Lett. 2002, 361, 513−519. (32) Vergeer, F. W.; Hartl, F.; Matousek, P.; Stufkens, D. J.; Towrie, M. First Direct Observation of a CO-Bridged Primary Photoproduct of [Ru3(CO)12] by Picosecond Time-Resolved IR Spectroscopy. Chem. Commun. 2002, 1220−1221. (33) Dougherty, T. P.; Heilweil, E. J. Transient Infrared Spectroscopy of (η5−C5H5)Co(CO)2 Photoproduct Reactions in Hydrocarbon Solutions. J. Chem. Phys. 1994, 100, 4006−4009. (34) Glascoe, E. A.; Kling, M. F.; Shanoski, J. E.; DiStasio, R. A., Jr; Payne, C. K.; Mork, B. V.; Tilley, T. D.; Harris, C. B. Photoinduced βHydrogen Elimination and Radical Formation with CpW(CO)3(CH2CH3): Ultrafast IR and DFT Studies. Organometallics 2007, 26, 1424−1432. (35) Dougherty, T. P.; Heilweil, E. J. Photochemistry of Rh(CO)2(acetylacetonate) and Related Metal Dicarbonyls Studied by Ultrafast Infrared Spectroscopy. J. Phys. Chem. 1994, 98, 9396−9399. (36) Ridley, A. R.; Stewart, A. I.; Adamczyk, K.; Ghosh, H. N.; Kerkeni, B.; Guo, Z. X.; Nibbering, E. T. J.; Pickett, C. J.; Hunt, N. T. Multiple-Timescale Photoreactivity of a Model Compound Related to the Active Site of [FeFe]-Hydrogenase. Inorg. Chem. 2008, 47, 7453− 7455. (37) Jiao, T.; Pang, Z.; Burkey, T. J.; Johnston, R. F.; Heimer, T. A.; Kleiman, V. D.; Heilweil, E. J. Ultrafast Ring Closure Energetics and Dynamics of Cyclopentadienyl Manganese Tricarbonyl Derivatives. J. Am. Chem. Soc. 1999, 121, 4618−4624. (38) To, T. T.; Heilweil, E. J.; Duke, C. B., III; Ruddick, K. R.; Webster, C. E.; Burkey, T. J. Development of Ultrafast Photochromic Organometallics and Photoinduced Linkage Isomerization of Arene Chromium Carbonyl Derivatives. J. Phys. Chem. A 2009, 113, 2666− 2676. (39) Lim, M.; Jackson, T. A.; Anfinrud, P. A. Binding of CO to Myoglobin from a Heme Pocket Docking Site to Form Nearly Linear Fe−C−O. Science 1995, 269, 962−966. (40) Lim, M.; Jackson, T. A.; Anfinrud, P. A. Ultrafast Rotation and Trapping of Carbon Monoxide Dissociated from Myoglobin. Nat. Struct. Biol. 1997, 4, 209−214. (41) King, J. T.; Ross, M. R.; Kubarych, K. J. Water-Assisted Vibrational Relaxation of a Metal Carbonyl Complex Studied with Ultrafast 2D-IR. J. Phys. Chem. B 2012, 116, 3754−3759. (42) King, J. T.; Arthur, E. J.; Brooks, C. L., III; Kubarych, K. J. SiteSpecific Hydration Dynamics of Globular Proteins and the Role of Constrained Water in Solvent Exchange with Amphiphilic Cosolvents. J. Phys. Chem. B 2012, 116, 5604−5611. (43) King, J., T.; Kubarych, K. J. Site-Specific Coupling of Hydration Water and Protein Flexibility Studied in Solution with Ultrafast 2D-IR Spectroscopy. J. Am. Chem. Soc. 2012, 134, 18705−18712. (44) Niesel, J.; Pinto, A.; Peindy N’Dongo, H. W.; Merz, K.; Ott, I.; Gust, R.; Schatzschneider, U. Photoinduced CO Release, Cellular Uptake and Cytotoxicity of a Tris(pyrazolyl)methane (tpm) Manganese Tricarbonyl Complex. Chem. Commun. 2008, 1798−1800. (45) Berends, H.-M.; Kurz, P. Investigation of Light-Triggered Carbon Monoxide Release from Two Manganese PhotoCORMs by IR, UV−Vis and EPR Spectroscopy. Inorg. Chim. Acta 2012, 380, 141−147.

(46) Falk, M.; Ford, T. A. Infrared Spectrum and Structure of Liquid Water. Can. J. Phys. 1966, 44, 1699−1707. (47) Barrett, J.; Mansell, A. L. Ultra-Violet Absorption Spectra of the Molecules H2O, HDO and D2O. Nature 1960, 187, 138. (48) Geoffroy, G. L.; Wrighton, M. S. Organometallic Photochemistry; Academic Press: New York, 1979. (49) Strohmeier, W.; von Hobe, D. Quantenausbeute als Funktion der Wellenlänge bei der Photochemischen Bildung von Metallcarbonylderivaten. Z. Phys. Chem. 1962, 34, 393−400. (50) Giordano, P. J.; Wrighton, M. S. Photosubstitution Behavior of Dicarbonyl(η5-cyclopentadienyl)pyridinomanganese and -rhenium and Related Complexes. Inorg. Chem. 1977, 16, 160−166. (51) Strohmeier, W.; von Hobe, D.; Schönauer, G.; Laporte, H. Einfluss des Zentralatoms und der Donatoren auf die Photochemische Bildung von Metallcarbonyl-Derivaten. Z. Naturforsch., B: Chem. Sci. 1962, 17, 502−504. (52) Frisch, M. J. et al. Gaussian 09, revision A.1; Gaussian Inc.: Wallingford, CT, 2009. (53) Andersson, M. P.; Uvdal, P. New Scale Factors for Harmonic Vibrational Frequencies Using the B3LYP Density Functional Method with the Triple-ζ Basis Set 6-311+G(d,p). J. Phys. Chem. A 2005, 109, 2937−2941. (54) Hamm, P. Coherent Effects in Femtosecond Infrared Spectroscopy. Chem. Phys. 1995, 200, 415−429. (55) Nuernberger, P.; Lee, K. F.; Bonvalet, A.; Polack, T.; Vos, M. H.; Alexandrou, A.; Joffre, M. Suppression of Perturbed Free-Induction Decay and Noise in Experimental Ultrafast Pump−Probe Data. Opt. Lett. 2009, 34, 3226−3228. (56) Laenen, R.; Roth, T.; Laubereau, A. Novel Precursors of Solvated Electrons in Water: Evidence for a Charge Transfer Process. Phys. Rev. Lett. 2000, 85, 50−53. (57) Huber, W.; Linder, R.; Niesel, J.; Schatzschneider, U.; Spingler, B.; Kunz, P. C. A Comparative Study of Tricarbonylmanganese Photoactivatable CO Releasing Molecules (PhotoCORMs) by Using the Myoglobin Assay and Time-Resolved IR Spectroscopy. Eur. J. Inorg. Chem. 2012, 3140−3146. (58) Vummaleti, S. V. C.; Branduardi, D.; Masetti, M.; Vivo, M. D.; Motterlini, R.; Cavalli, A. Theoretical Insights into the Mechanism of Carbon Monoxide (CO) Release from CO-Releasing Molecules. Chem.Eur. J. 2012, 18, 9267−9275. (59) Pal, S. K.; Peon, J.; Bagchi, B.; Zewail, A. H. Biological Water: Femtosecond Dynamics of Macromolecular Hydration. J. Phys. Chem. B 2002, 106, 12376−12395. (60) Buback, J.; Kullmann, M.; Langhojer, F.; Nuernberger, P.; Schmidt, R.; Würthner, F.; Brixner, T. Ultrafast Bidirectional Photoswitching of a Spiropyran. J. Am. Chem. Soc. 2010, 132, 16510−16519. (61) Rudolf, P.; Buback, J.; Aulbach, J.; Nuernberger, P.; Brixner, T. Ultrafast Multisequential Photochemistry of 5-Diazo Meldrum’s Acid. J. Am. Chem. Soc. 2010, 132, 15213−15222. (62) Schäfer, A.; Horn, H.; Ahlrichs, R. Fully Optimized Contracted Gaussian Basis Sets for Atoms Li to Kr. J. Chem. Phys. 1992, 97, 2571− 2577. (63) Neese, F. The ORCA Program System. WIREs Comput. Mol. Sci. 2012, 2, 73−78.

602

dx.doi.org/10.1021/jz302061q | J. Phys. Chem. Lett. 2013, 4, 596−602