Chemical Reactions Triggered Using Electrons Photodetached from

Aug 15, 2014 - The photodetached electrons, in turn, initiate covalent chemistry, inducing C–C bond formation following electron-capture by CO van d...
0 downloads 12 Views 581KB Size
Letter pubs.acs.org/JPCL

Chemical Reactions Triggered Using Electrons Photodetached from “Clean” Distributions of Anions Deposited in Cryogenic Matrices via Counterion Codeposition Ryan M. Ludwig and David T. Moore* Lehigh University, Department of Chemistry, 6 East Packer Avenue, Bethlehem, Pennsylvania 18015, United States S Supporting Information *

ABSTRACT: Application of matrix isolation spectroscopy to ionic species is typically complicated by the presence of neutral contaminants during matrix deposition. Herein we demonstrate that simultaneous deposition of balanced currents of counterions with massselected ions of interest generates “clean” distributions of matrix-isolated metal carbonyl anions, where the only bands appearing in the CO-stretching region of the vibrational spectrum arise from ions. (Neutrals are initially absent.) Photodetachment by mild irradiation with visible light leads to complete conversion of the anions into their corresponding neutral species. The photodetached electrons, in turn, initiate covalent chemistry, inducing C−C bond formation following electron-capture by CO van der Waals dimers to produce trans-OCCO−. The initial clean distribution of ions enables clear connections to be drawn between the spectral changes occurring at each experimental step, thus demonstrating the potential of the counterion codeposition technique to facilitate detailed studies of chemistry involving ions and electron transfer in cryogenic matrices. SECTION: Spectroscopy, Photochemistry, and Excited States

M

diagnostic uses of photodetachment; there are some examples where wavelength filters were used to distinguish anionic species by their photodetachment thresholds, which tend to be significantly blue-shifted in matrices.12,13 The coupling of a vibrational spectroscopy matrix isolation apparatus to the output of a mass spectrometer to enable structural characterization of arbitrary ions of interest has long been recognized as an important goal in instrumental analytical and physical chemistry.14 Recent results from our group demonstrated the feasibility of using simultaneous deposition of equal currents of anions and cations for just such a purpose.15 Anionic copper carbonyl complexes in sufficient number densities for FTIR spectroscopic characterization were produced by simultaneously directing low-energy beams of monatomic copper anions (Cu−) and rare gas cations (Ar+ or Kr+) into a CO-doped argon matrix.15 That study demonstrated proof-of-concept for the technique but still was not an ideally “clean” source of matrix isolated ions because neutral copper carbonyls were also produced during deposition. Figure 1A shows an improved version of the experiment where the matrix deposition was conducted under completely darkened conditions at 10 K. (Full experimental details are given in Supporting Information.) The expected bands for the anionic mono-, di-, and tricarbonyl complexes are observed at 1733, 1780, and 1829 cm−1, respectively, as in previous studies,15,16 but now the neutral copper carbonyl peaks are

atrix isolation has long been used to stabilize transient species such as weakly bound complexes, radicals, and ions for spectroscopic characterization.1−4 The chemically inert, low-temperature matrices (typically solidified rare gases), are effective at cooling and trapping unstable or labile species by physically separating them from potential reaction partners while rapidly quenching their internal degrees of freedom.5,6 Although many matrix isolation studies of ions have been reported,7 the requirement that charge-balance be maintained throughout matrix deposition has limited the scope of ions that could be studied to those produced by intrinsically neutral sources such as laser ablation3 and microwave discharge,8 or to robust ions such as carbon chains that can survive deposition at very high kinetic energies.9 All of these methods for generation and deposition of matrixisolated ions tend to produce contaminated samples in that a large fraction of the species trapped in the matrix are neutrals produced during ion generation, or formed by secondary processes such as fragmentation or autoneutralization during deposition. Photodetachment, the process whereby an electron is released from a negatively charged atom or molecule following the absorption of a sufficiently energetic photon (in analogous fashion to photoelectron emission from neutrals), is a standard method used to distinguish charged from neutral species in matrices.10−13 Irradiation of the matrix with broadband UV−visible radiation induces selective depletion of the bands associated with ions; the interpretation is that anions are directly neutralized by photodetachment and the electrons released then combine with and neutralize cations. However, because of the contamination issues previously mentioned, few matrix studies have gone beyond such simple © 2014 American Chemical Society

Received: July 23, 2014 Accepted: August 15, 2014 Published: August 15, 2014 2947

dx.doi.org/10.1021/jz501547b | J. Phys. Chem. Lett. 2014, 5, 2947−2950

The Journal of Physical Chemistry Letters

Letter

be detected against the background of neutral bands, which were orders of magnitude more intense than the anions in the as-deposited matrix.16 Here the photodetachment seems be a quite gentle event, preserving the approximate relative intensities of the mono-, di-, and tri-carbonyls within the anionic and neutral manifolds. The dicarbonyl seems a bit more intense in the neutral spectrum, relatively, but it is not clear whether this represents a difference in population or in oscillator strength. Annealing to higher temperatures (Figure 1C−F) causes only minor quantitative changes in the spectrum; the 1975 cm−1 band gains intensity relative to the other bands with increasing temperature, and there is some loss of signal in the mono- and dicarbonyl bands relative to the tricarbonyl, suggesting that a fraction of the complexes increase their coordination number due to the addition of CO ligands from the matrix. This lack of qualitative changes upon annealing indicates that the matrix environment around the complexes was not significantly perturbed during photodetachment, again speaking to the relative gentleness of the process. It is interesting to consider the fate of the photodetached electrons in the matrix. Another pair of peaks at 1512.6 and 1515.5 cm−1 also appeared in the spectrum following the irradiation step; these have previously been assigned the transOCCO− species in argon matrices.18 This strongly suggests that some of the photodetached electrons ended up getting captured by neutral (CO)2 van der Waals complexes, which were present in high abundance in the 2% CO matrix (cf. 2140 cm−1 band19 in Figure S2 of the Supporting Information). This in turn means that the photoelectrons initiated the formation of a covalent C−C bond in the weakly bound neutral CO-dimer, and thus we have achieved the phototriggered cryochemistry scheme illustrated in Figure 2. (Note that although only the

Figure 1. FTIR spectra following codeposition of Cu− and Ar+ ionbeams (3 nA each) with 2% CO in Ar at 10 K, with subsequent processing steps; all spectra recorded at 10 K. Spectral regions corresponding to anionic and neutral copper carbonyl complexes are shaded in blue and red, respectively; Cu-center coordination numbers are indicated above corresponding peaks. (A) Sample after 2 h of deposition, showing only anionic copper carbonyl bands. (B) Sample after photodetachment by visible irradiation; new bands correspond to neutral copper carbonyl complexes as well as trans-OCCO− near 1516 cm−1. (C−F) Postdetachment spectra taken after annealing steps at 15, 20, 25, and 30 K, respectively.

conspicuously absent from the spectrum. (The peak at 1774 cm−1 was also previously assigned to an anionic dicarbonyl species.15) Evidently the removal of all sources of ambient light during the deposition phase protected the anions from premature photodetachment, so that neutral species were not produced. Note that this also implies that neutralization by electron transfer between adjacent pairs of oppositely charged ions does not occur to any significant extent. The implication is that the ions become trapped in matrix sites during the deposition phase before they can diffuse together under the influence of Coulomb attraction; this is qualitatively consistent with previous 10 K studies for this system, which showed trapping of uncoordinated Cu− ions at low CO concentrations.15 The benefit of having a clean source of anions can be appreciated from Figure 1B, which shows the spectrum following deliberate photodetachment of the anions using ∼10 min irradiation with light from a tungsten filament bulb (UV−visible spectrum of source in Figure S1 of Supporting Information). All of the bands corresponding to anionic species (including the 1774 cm−1 band) have been completely depleted and replaced in the spectrum by the corresponding neutral copper complexes, with peaks for the mono- and dicarbonyls at 2010 and 1890 cm−1, respectively, and a broad feature centered at 1980 cm−1 with sharp peaks at 1975 and 1985 cm−1 representing the tricarbonyls.15,17 Contrast this with the previous study, where the loss of the anions upon irradiation with visible light was noted but the neutral products could not

Figure 2. Cryochemistry scheme based on using electrons photodetached from the anionic monocarbonyl complex to induce formation of the trans-OCCO− by electron attachment to the van der Waals CO-dimer, which results in C−C bond formation. Positions of corresponding IR bands in Figure 1 and Figure S2 of the Supporting Information are indicated.

monocarbonyl is explicitly shown in Figure 2, it is assumed that photodetached electrons from all anionic copper carbonyl complexes can induce the same cryochemical process.) Finally, it should be noted that photodetachment of the trans-OCCO− did not occur for the source used in this study (a longer, 1 h irradiation with the same light source was done to confirm this), presumably because the photodetachment threshold is at shorter wavelengths. (Photodetachment using broadband UV irradiation has been observed.18) There is obviously a significant driving force for photodetached electrons to recombine with cations trapped in the 2948

dx.doi.org/10.1021/jz501547b | J. Phys. Chem. Lett. 2014, 5, 2947−2950

The Journal of Physical Chemistry Letters

Letter

of Ar+ counterions using a quadrupole ion-bender and guided through RF-ion optics to the deposition region; the beam currents were adjusted so the net current was zero. The ionbender ensured that there is no line-of-sight path from the source to the deposition region, so any neutral species from the source were prevented from contaminating the matrix sample. The ions were codeposited with argon matrix gas doped with 2% CO onto a KBr window maintained at 10 K using a closedcycle helium cryostat. Matrix deposition was carried out for 2 h under darkened conditions, after which transmission-mode IR spectroscopy experiments were performed. (The sample was protected from the IR beam and alignment laser from the spectrometer when not recording data.) For photodetachment, the sample was illuminated for ∼10 min with visible and nearIR light from a tungsten filament bulb. (See Figure S1 of the Supporting Information for spectrum.) Annealing steps were done by warming the sample to the indicated temperature for 30 min then cooling back to 10 K to record spectra.

matrix, but because those are rare-gas centers lacking vibrational bands,15 it is not possible to directly observe this in the spectrum. However, the apparent gentleness of the photodetachment from the spectra in Figure 1 indicates that this recombination cannot be happening in close proximity to the anionic complexes that served as the sources of the photoelectrons. Otherwise the huge amount of energy (>10 eV) released locally into the matrix on recombination would be expected to perturb the spectral bands of the (now neutral) copper carbonyl source complex, as was observed in our previous set of experiments when free Cu− ions underwent diffusion and spontaneous neutralization, and the local energy released shifted the stoichiometry of the observed neutral complexes to larger clusters.15 Instead, it seems that the cation neutralization is happening in remote sites following diffusion of the electron across multiple argon atoms in the matrix, which is reasonable given the relatively low ion abundance of ∼3 ions per 107 argon atoms. (See the Supporting Information for details.) This is also consistent with the expected behavior of electrons flowing freely in the conduction band of the argon solid,20 which lies ∼0.4 eV above the vacuum level.21 For the current case, the known gas-phase photodetachment thresholds of the anionic copper carbonyls are 0.95 and 1.02 eV for the diand tricarbonyls, respectively.22 Thus, for the anionic copper carbonyls in the current study, long-range diffusion following photodetachment should be possible using near-IR and visible light of wavelength