ARTICLE pubs.acs.org/JPCA
Temporary Anion States of Pyrimidine and Halopyrimidines Alberto Modelli,*,†,‡ Paola Bolognesi,§ and Lorenzo Avaldi§ †
Dipartimento di Chimica “G. Ciamician”, Universita di Bologna, via Selmi 2, 40126 Bologna, Italy, and ‡Centro Interdipartimentale di Ricerca in Scienze Ambientali (CIRSA), Universita di Bologna, via S. Alberto 163, 48123 Ravenna, Italy § CNR-IMIP, Area della Ricerca di Roma 1, 00016 Monterotondo Scalo, Italy ABSTRACT:
The empty-level electronic structures of pyrimidine and its 2-chloro, 2-bromo, and 5-bromo derivatives have been studied with electron transmission spectroscopy (ETS) and dissociative electron attachment spectroscopy (DEAS) in the 05 eV energy range. The spectral features were assigned to the corresponding anion states with the support of theoretical calculations at the ab initio and density functional theory levels. The empty orbital energies obtained by simple Koopmans’ theorem calculations, scaled with empirical equations, quantitatively reproduced the energies of vertical electron attachment to π* and σ* empty orbitals measured in the ET spectra and predicted vertical electron affinities close to zero for the three halo derivatives. The total anion currents of the halo derivatives, measured at the walls of the collision chamber as a function of the impact electron energy, presented intense maxima below 0.5 eV. The mass-selected spectra showed that, in this energy, range the total anion current is essentially due to halide fragment anions. The DEA cross sections of the bromo derivatives were found to be about six times larger than that of the chloro derivative. The absolute cross sections at incident electron energies close to zero were evaluated to be 10161015 cm2.
1. INTRODUCTION Pyrimidines are an important class of organic molecules mainly due to the fact that cytosine, thymine, and uracil, the building blocks of DNA and RNA bases, are pyrimidine derivatives. Halogenated pyrimidines have long been employed in biomedical research as efficient Auger-electron-emitting molecules.1 Furthermore, the incorporation of halogenated pyrimidines into DNA to replace thymidine is known to sensitize the cell to ionizing radiation in the radiotherapy of tumors. Indeed, one of the mechanisms invoked to explain such a radio-sensitizing effect is the interaction of secondary low-energy electrons, following the Auger cascade produced by the decay of the inner-shell vacancies of the heavy halogen atom,24 with the DNA/RNA bases. Observation of the biological effects of soft X-rays in cells5 has raised the question of whether the occurrence of fragmentation processes is directly induced by the radiation or the bonds break as a result of electron swarms produced following the X-ray absorption. The macroscopic effects of photoabsorption processes in living cells can, in principle, be tracked down to the microscopic scale, where the initial processes on the elementary constituents are the same as those studied in molecular physics and photochemistry. Therefore, it is expected that gas-phase experiments of model molecules can provide insight into the physical and chemical properties of biological molecules. This has led, for instance, to the extensive investigation of the structure and dynamics of biomolecules and their solvated complexes.69 r 2011 American Chemical Society
In particular, several efforts have recently been devoted to experimental and theoretical studies1014 of the electronic structures of pyrimidine and some of its halo derivatives. The complex Auger decays of the carbon and nitrogen 1s shells were observed10 in an electron impact experiment at 1000 eV and interpreted using the Green’s function-based second-order algebraic diagrammatic construction method.15,16 The valence filledlevel structure of the same series of compounds was characterized11 by means of ultraviolet photoelectron spectroscopy, and the spectral features were assigned to the corresponding molecular orbitals (MOs) with the aid of density functional theory (DFT) calculations in conjunction with the hybrid B3LYP functional to reproduce the correct σ/π energy sequence. Ionizations from the core levels were measured12 by X-ray photoelectron spectroscopy (XPS). The inner-shell ionization energies are a local probe of the chemical environment and charge density of an atom and make it possible to establish relationships between the chemical shift and chemical properties with site dependence. Finally, detailed near-edge X-ray absorption fine structure (NEXAFS) and theoretical investigations of the inner-shell photoabsorption of pyrimidine and some halopyrimidines have also been reported,13,14 together with the analysis of the fragmentation following inner-shell excitations. Received: July 11, 2011 Revised: August 25, 2011 Published: August 29, 2011 10775
dx.doi.org/10.1021/jp206559d | J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A Scheme 1
However, specific complementary studies of these molecular systems to elucidate their valence empty-level structures, equally important from both the theoretical and reactivity points of view, have not been reported. An important improvement in the detection and characterization of unstable gas-phase anions came about with the electron transmission spectroscopy (ETS) apparatus devised by Sanche and Schulz,17 which favors the study of temporary anion formation in relatively large molecular systems of chemical interest18 and is still one of the most suitable means for measuring negative vertical electron affinities (EAvs). The ETS technique takes advantage of the sharp variations in the total electronmolecule scattering cross section caused by temporary capture of electrons with appropriate energy and angular momentum into empty MOs, the process referred to as shape resonance.18,19 Electron attachment is rapid with respect to nuclear motion, so that temporary anions are formed in the equilibrium geometry of the neutral molecule. The measured vertical attachment energies (VAEs) are the negative of the EAvs. These data are thus complementary to the ionization energies obtained by photoelectron spectroscopy. Only the combination of these two techniques can provide a complete picture of the structures of the frontier orbitals, which determine the reactivity properties of a molecule. Important additional information on temporary anion states can be obtained by dissociative electron attachment spectroscopy (DEAS),1921 which measures the yield of negative fragments as a function of the incident electron energy. When suitable energetic conditions prevail, the decay of unstable molecular anions formed by resonance can follow a dissociative channel that generates long-lived negative fragments and neutral radicals, in kinetic competition with simple re-emission of the extra electron. Thus, DEAS reveals possible dissociative decay channels of the molecular anions observed in ETS. An application of DEAS, in conjunction with ETS, that is of particular interest involves its use as a probe for intramolecular electron-transfer processes, when capture of an incident electron into an empty MO mainly localized on a functional group causes dissociation of a remote bond, with formation of a negative fragment. In unsaturated halohydrocarbons, the maximum yield of halide negative fragments generally occurs close to the energy of the lowest π* resonance observed in ETS.2230 Herein, we extend the study of the electronic structures of pyrimidine (1), 2-chloropyrimidine (2), 2-bromopyrimidine (3), and 5-bromopyrimidine (4), represented in Scheme 1, with the characterization of their empty-level structures by means of ETS and DEAS. We assigned the spectral features to the corresponding anion states with the support of appropriate ab initio and DFT calculations. Comparison of the VAEs measured in ETS with the energies of the peaks in the DEA spectra, as well as the relative DEA cross sections, are expected to provide more insight into the dissociative mechanism and the role played by the π* and σ* resonances in the chloro and bromo derivatives. At variance
ARTICLE
with most of the DEA studies reported in the literature, which have not examined the important quantitative aspect of determining the absolute or relative dissociative cross sections, in the present study, the absolute cross sections for the observed halide fragment anion yields were also evaluated.
2. EXPERIMENTAL SECTION Our electron transmission apparatus is in the format devised by Sanche and Schulz17 and has been previously described.31 To enhance the visibility of the sharp resonance structures, the impact energy of the electron beam is modulated with a small ac voltage, and the derivative of the electron current transmitted through the gas sample is measured directly by a synchronous lock-in amplifier. In the derivative signal, each resonance is characterized by a minimum and a maximum. The energy of the midpoint between these features is assigned to the VAE. The present spectra were obtained using the apparatus in the “highrejection” mode32 and are, therefore, related to the nearly total scattering cross section. The electron beam resolution was about 50 meV (fwhm). The energy scale was calibrated with reference to the (1s12s2) 2S anion state of He. The estimated accuracy is (0.05 or (0.1 eV, depending on the number of decimal digits reported in the tables. The collision chamber of the ETS apparatus was modified33 to allow for ion extraction at 90° with respect to the electron beam direction. These ions were then accelerated and focused toward the entrance of a quadrupole mass filter. Alternatively, the total anion current can be collected and measured with a picoammeter at the walls of the collision chamber (about 0.8 cm from the electron beam). Although the negative ion current at the walls of the collision chamber can, in principle, be affected by spurious trapped electrons, these measurements are more reliable than the current detected through the mass filter because of kinetic energy discrimination in the anion extraction efficiency in the latter. In a previous test34 with monochloroalkanes, our relative total anion currents reproduced to within 1% the ratios in the absolute cross sections reported by Pearl and Burrow.35 The DEAS data reported here were obtained with an electron beam current almost twice as large as that used for the ET experiments. The energy spread of the electron beam increased to about 100 meV, as evaluated from the width of the SF6 signal at zero energy used for calibration of the energy scales. The relative total anion currents were evaluated from the peak heights, normalized to the same electron beam current and sample pressure (measured in the main vacuum chamber by means of a cold cathode ionization gauge) for all compounds. Preliminary measurements showed that the total anion current reading was linearly proportional to the pressure, at least in the range of (14) 105 mbar range. The absolute cross sections were evaluated from comparison of the absolute cross sections found by Burrow and co-workers in chloroalkanes35 and chloroalkyl benzenes36 with our measurements on the same compounds. The average conversion factor (standard deviation = (25%) between the two sets of values was applied to the present bromo derivatives. The collision cell temperature was kept at about 70 °C. All compounds were commercially available from SigmaAldrich with purity higher than 95% and were used without further purification. The calculations were carried out with the Gaussian 09 suite of programs.37 The virtual orbital energies of the neutral molecules 10776
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A
ARTICLE
Table 1. HF/6-31G(d)//MP2/6-31G(d) VOEs, Scaled Values,a and Measured VAEs (eV) HF/6-31G(d)//MP2/6-31G(d) VOE
scaled VOE
ETS VAE
Pyridineb π*3 (b1)
9.552
4.76
4.48
π*2 (a2) π*1 (b1)
3.763 3.296
1.01 0.71
1.18 0.72
Figure 1. Representation of the empty π* MOs of pyrimidine, as obtained from HF/6-31G(d) calculations (nitrogen atoms in positions 1 and 3).
Pyrimidine core excited
5.5
π*3 (b1)
9.134
4.49
4.26
π*2 (b1)
3.205
0.65
0.82
π*1 (a2)
2.883
0.44
0.39
3.85 2.74
2-Chloropyrimidine core excited
5.2
π*3 (b1) σ*CCl
8.629 5.727
4.16 2.61
π*2 (b1)
2.891
0.44
0.48
π*1 (a2)
2.379
0.11
2-Bromopyrimidine core excited
5.0
π*3 (b1)
8.567
4.12
3.72 1.80
σ*CBr
4.722
1.51
π*2 (b1)
2.822
0.40
0.50
π*1 (a2)
2.360
0.10
5-Bromopyrimidine core excited
a
5.4
π*3 (b1)
8.512
4.09
3.84
σ*CBr
4.221
1.00
1.12
π*2 (b1)
2.789
0.38
0.48
π*1 (a2)
2.378
0.11
b
See text. VAEs from ref 40.
were obtained using HartreeFock (HF) calculations with the standard 6-31G(d) basis set (which does not include diffuse functions) and the MP2/6-31G(d)-optimized geometries. The vertical electron affinity (EAv) was calculated as the difference between the total energy (including only electronic contributions or the zero-point vibrational energy as well) of the neutral and the lowest anion state, both in the optimized geometry of the neutral state, using DFT calculations with the B3LYP38 hybrid functional and the standard 6-31+G(d) or 6-311+G(d) basis set (which include the minimum addition of diffuse functions). The adiabatic electron affinity (EAa) was obtained as the energy difference between the neutral and the lowest anion state, each in its optimized geometry. The same theoretical method was used to evaluate the thermodynamic energy threshold for production of halide fragment anions. The energies of the excited states of the vertical anion of pyrimidine were obtained with timedependent (TD) B3LYP/6-31G+(d) calculations.39
3. RESULTS AND DISCUSSION 3.1. Empty Level Structure: ET Spectra and Calculated VAEs. Like benzene, its aza derivatives pyridine and pyrimidine
Figure 2. Derivative of transmitted current as a function of electron energy in gas-phase pyrimidine and halopyrimidines 24. Vertical lines indicate the VAEs.
contain three empty π* MOs. Replacement of one or two benzene CH groups with nitrogen atoms [to give pyridine and pyrimidine (1), respectively] lowers the molecular symmetry to C2v, so that the degenerate e2u (π*) lowest unoccupied MO (LUMO) of benzene splits into b1 and a2 components. Because of the larger electronegativity of the nitrogen atom relative to a CH group, the VAE (0.72 eV40) of the b1 (π*) LUMO of pyridine is 0.4 eV smaller than that (1.12 eV31) of the benzene LUMO; that is, the EAv of pyridine is 0.4 eV larger than that of benzene. Further replacement of a CH group of pyridine with a nitrogen atom to give pyrimidine increases the EAv by 0.3 eV (see Table 1). A pioneering ET study by Nenner and Schulz41 considered the temporary anions of benzene and its diaza derivatives (including pyrimidine), but did not display the ET spectrum of 1. However, the onsets (0.25 and 0.77 eV, respectively) reported for the first and second resonances are in line with the corresponding VAEs (0.39 and 0.82 eV; see Table 1) found in the present work. A representation of the three empty π* MOs of 1 as obtained by HF/6-31G(d) calculations (essentially equal localization properties and energy sequence were obtained by the B3LYP method) is given in Figure 1, in order of increasing energy. The (π*1) LUMO of pyrimidine exhibits a2 symmetry. Figure 2 reports the ET spectra of pyrimidine and its halo derivatives 24. The measured VAEs are given in Table 1. In addition to the two sharp resonances below 1 eV, which display 10777
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A some evidence of vibrational structures, the ET spectrum of pyrimidine shows a third resonance at 4.26 eV associated with electron capture into the highest-lying b1 (π*3) MO. A signal at even higher energy (5.5 eV) is ascribed to a core-excited resonance, that is, electron capture accompanied by excitation of a valence electron. Consistently, TD-B3LYP/6-31G+(d) calculations predicted the lowest π f π* transition of the vertical anion to occur at 5.1 eV, which, when added to 0.4 eV (i.e., the π*1 VAE measured in the ET spectrum), gives 5.5 eV. The same TD-B3LYP calculations also predicted the π*1 f π*2 excitation energy of the vertical anion to be 0.39 eV, in excellent agreement with the difference measured between the first two VAEs of 1 (see Table 1). A theoretical approach adequate for describing the energetics of unstable anion states involves difficulties not encountered for neutral or cation states. The most correct method is, in principle, the calculation of the total scattering cross section with the use of continuum functions, although complications arise for an accurate description of electronmolecule interactions.42 The first VAE can be obtained as the energy difference between the lowest-lying anion and the neutral state (both with the optimized geometry of the neutral species), but the description of resonance anion states (unstable with respect to electron loss) with standard bound-state methods poses a serious problem. The use of a finite basis set formed with Gaussian functions has the effect of confining the system within a box, accounting in some way for the fact that, during a resonance process, the extra electron is confined to the molecule by a potential barrier.43 However, a proper description of the spatially diffuse electron distributions of anions requires a basis set with diffuse functions.44,45 On the other hand, as the basis set is expanded, the wave function ultimately describes a neutral molecule and an unbound electron in the continuum,28,43,46,47 as this is the state of minimum energy. To decide a priori which basis set (if any) can give a reliable description of both the energy and nature of resonance processes is a delicate task.48 Generally, the more unstable the anion state, the larger the need to augment the basis set with diffuse functions,48 thus increasing the risk that the singly occupied MO (SOMO) of the anion is described as a diffuse function of no physical significance with regard to anion formation. The Koopmans’ theorem (KT) approximation49 neglects correlation and relaxation effects. However, Chen and Gallup50 and Staley and Strnad47 demonstrated the occurrence of good linear correlations between the π*CdC VAEs measured in alkenes and benzenoid hydrocarbons and the corresponding virtual orbital energies (VOEs) of the neutral molecules obtained with simple KTHF calculations, using basis sets that do not include diffuse functions. Later, it was shown28 that the neutral-state π* VOEs obtained with B3LYP/6-31G(d) calculations also give a good linear correlation with the corresponding VAEs measured over a variety of unsaturated compounds, including heterosubstituted hydrocarbons. Table 1 gives the HF/6-31G(d) π* VOEs for the MP2/ 6-31G(d)-optimized geometries of pyridine and pyrimidines 14, in addition to the lowest σ* VOE for the halo derivatives. Also reported in Table 1 are the π* VOEs scaled with the empirical linear correlation (VAE = 0.64795 VOE 1.4298) found by Staley and Strnad47 using the same theoretical method and the σ* VOEs scaled with the linear correlations found by Burrow and coworkers51,52 with HF/6-31G(d) calculations [σ*CCl VAE = 0.901 VOE 2.55; σ*CBr VAE = (VOE 3.23)/0.99]. The scaled π* VOEs of pyridine and pyrimidine are in good quantitative agreement with the measured values (see Table 1). In particular, in both compounds, the first VAE is reproduced
ARTICLE
within experimental limits. This gives confidence in the reliability of the simple computational method employed for the interpretation of the ET spectra of the halopyrimidines. The scaled π*2 VOEs of the halopyrimidines 24 reproduce the energy of the first resonance displayed by their ET spectra within 0.1 eV (see Table 1). The scaled π*1 VOEs of all three halopyrimidines were calculated to be about 0.1 eV; that is, the ground vertical anion state was predicted to lie close to zero energy. Consistently, the formation of the first anion state was not observed in the ET spectra of 24 (see Figure 2). Stable anion states are, in fact, not accessible in ETS, and formation of anion states that lie only slightly above zero energy are hidden by the intense signal of the incident electron beam. As mentioned above, the VOE of the lowest σ* MO, with mainly (antibonding) Chalogen character, can be scaled with empirical linear correlations found for chloro51 and bromo52 hydrocarbons. The ET spectrum of 2-chloropyrimidine displays a resonance centered at 2.74 eV (see Figure 2 and Table 1), within the same energy range as the σ*CCl VAE (2.46 eV) of chlorobenzene.40 The scaling procedure leads to a σ*CCl VAE of 2.61 eV, in excellent agreement with the value measured in the ET spectrum. The ET spectra of the two bromopyrimidines 3 and 4 display a σ*CBr resonance at 1.80 and 1.12 eV, respectively, to be compared with a σ*CBr VAE e 1.8 eV in bromobenzene.30 The reason for this sizable difference between the σ*CBr VAEs in the two isomers 3 and 4 is not obvious. The CBr bond distance of 3 calculated at the MP2/6-31G(d) and B3LYP/6-31+G(d) levels is 1.9048 and 1.8976 Å, respectively, and the corresponding distance in 4 is 1.8916 and 1.8906 Å. Given that the empty σ* MO has CBr antibonding character, other conditions being the same, a lower energy would be expected in the isomer where the CBr distance is larger (in contrast with experiment). Moreover, at a very qualitative level, a larger EA should probably be expected for the isomer in which the CBr bond is closer to the (more electronegative) nitrogen atoms (i.e., in the 2-bromo isomer), again in contrast with the measured σ*CBrVAEs. In any event, it is at least comforting not only that the KT calculations (see Table 1) account for the σ*CBr energy difference observed in the ET spectra of 3 and 4, but also that their absolute VAEs are satisfactorily reproduced by the scaled σ*CBr VOEs (1.51 and 1.00 eV in 2-bromo- and 5-bromopyrimidine, respectively). It can also be noted that the relative energies of the σ* anion states of the halopyrimidines 24 are consistent with NEXAFS data previously reported.14 The C1sσ* excitation energy (where C is the carbon atom attached to the halogen atom) of the neutral state molecules was found to decrease in the same order (2 > 3 > 4) as the σ* VAEs measured in the ET spectra. Moreover, the energy difference between this electronic transition and the lowest C1sπ* transition was found14 to be 2.0, 1.3, and 0.6 eV in compounds 24, respectively, in line with a significantly lower energy of the σ*CBr MO in 4. These energy differences nicely fit the same trend of the differences between the lowest σ* and π* VAEs of 24 (2.6, 1.7, and 1.0 eV, respectively; see Table 1) and are about 0.5 eV smaller. Given that the neutral state obtained by excitation of a 1s electron to an empty MO can be thought of as the species produced by electron capture into an empty MO of the core-ionized cation, comparison between the two sets of data suggests that the stabilization caused by C1s ionization is larger for the σ*Chalogen MO than for the lowest-lying π* MO. To obtain an independent evaluation of the energy of the (experimentally not observed) first anion state of the 10778
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A
ARTICLE
Table 2. B3LYP Energies (eV) Relative to the Ground Neutral Statea 6-31+G(d)
6-311+G(d)
Pyridine vertical anion
0.932 (0.712)
0.912 (0.709)
adiabatic anion
0.770 (0.585)
0.742 (0.549)
Pyrimidine vertical anion
0.531 (0.340)
0.504 (0.313)
adiabatic anion
0.329 (0.150)
0.295 (0.144)
2-Chloropyrimidine vertical anion
0.084 (0.103)
0.048 (0.140)
adiabatic anion
0.193 (0.336)
0.241 (0.384)
[M Br]• + Cl
0.091 (0.000)
0.080 (0.012)
2-Bromopyrimidine vertical anion
0.057 (0.132)
0.005 (0.197)
adiabatic anion [M Br]• + Br
0.234 (0.379) 0.075 (0.008)
0.314 (0.460) 0.345 (0.452)
5-Bromopyrimidine
a
vertical anion
0.067 (0.123)
0.025 (0.171)
adiabatic anion
0.174 (0.329)
0.228 (0.384)
[M Br]• + Br
0.249 (0.160)
0.184 (0.260)
Values in parentheses include zero-point vibrational energy corrections.
halopyrimidines, the EAvs of pyridine, pyrimidine, and haloderivatives 24 were calculated as the total energy difference between the ground state of the neutral molecule and the anion, both with the geometry optimized for the former. In turn, this constitutes a benchmark for the reliability of the same theoretical method to evaluate the thermodynamic thresholds for production of the halide fragment anion observed in the DEA spectra discussed in the next section. Table 2 reports the results obtained with DFT calculations using the B3LYP functional and the standard 6-31+G(d) and 6-311 +G(d) basis sets, which include the minimum addition of diffuse functions (s- and p-type diffuse functions at the non-hydrogen atoms). As a first comment, it is important to note that, at variance with benzene, for which the SOMO is described as a diffuse σ* MO,28 in the present series of compounds (including pyridine, whose ground anion state is the most unstable), both the LUMO of the neutral molecules and the SOMO of the anions are correctly described as valence π* MOs. Table 2 shows that the energies calculated for the first vertical anion states of pyridine and pyrimidine nicely reproduce the corresponding measured π*1 VAEs. Inclusion of the zero-point vibrational energies of the neutral and anion states reduces by about 0.2 eV the anion energy and leads to calculated values equal to (pyridine) or slightly lower than (pyrimidine) the experimental VAEs. It seems reasonable that this trend continues on going from pyrimidine to its haloderivatives (where the stability of the first anion state further increases); that is, inclusion of zero-point vibrational energy corrections likely leads to a slight underestimation of the anion energy. The energies of the first vertical anions of the three halopyrimidines 24, relative to the corresponding neutral state, were found to be close to zero. For all three halides, the vertical anion was predicted to be stable by 0.10.2 eV when zero-point vibrational energies were included, and slightly unstable when only electronic contributions were taken into account. The results obtained with this theoretical approach are thus in good agreement with those obtained by the scaled KT VOEs.
Figure 3. Total anion currents as a function of the incident electron energy measured in gas-phase halopyrimidines 24.
Table 2 also reports the energies of the adiabatic anion states, where the anions have their optimized geometries. As expected for conjugated π systems, geometrical relaxation upon addition of an extra electron is relatively small, as well as the consequent anion stabilization. The calculated EAas are 0.20.3 eV larger than the corresponding EAvs. 3.2. DEA Spectra. When a temporary molecular anion is formed at an energy higher than that of its dissociation products, a kinetic competition between dissociation and simple detachment of the extra electron takes place. The dissociative cross section, σDEA, is given by the equation20 Td σ DEA ¼ σA exp Ta where σA is the attachment cross section and Ta and Td are the lifetime of the molecular anion with respect to electron detachment and the time required to allow the nuclei to reach the separation at which bond dissociation can occur, respectively. Figure 3 displays the total yield of negative ions measured at the collision chamber for compounds 24 as a function of the incident electron energy, in the 04 eV energy range. Mass analysis revealed that the total anion current was essentially due only to the halide negative fragments Cl or Br. For all three halopyrimidines, the total anion current displayed the most intense peak at zero energy, followed by a partially overlapped signal at 0.4 eV in the chloro derivative 2 or a shoulder at 0.2 0.3 eV in the bromo derivatives 3 and 4. The two low-energy signals present in each DEA spectrum are thus ascribed to dissociative electron attachment to the π*1 and π*2 MOs. In agreement, the absolute DEA cross sections of saturated linear bromo-53 and choroalkanes26 peak at 0.6 and 1.4 eV, respectively, and are much smaller than the present signals (see below). In bromides 3 and 4, the contributions from the two anion states are better resolved in the mass-selected Br currents, as shown in Figure 4. The peak energies observed in the total and mass-selected currents of the DEA spectra are reported in Table 3. In 2-chloropyrimidine, the relative intensity of the zero-energy signal is reduced on going from the total to the Cl current. This might be due to the presence of molecular anions that are sufficiently long-lived to reach the walls of the collision chamber (where the total current is measured), but not to be detected with the mass filter. However, a contribution to the different relative intensities in 10779
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A
ARTICLE
Table 3. Peak Energies (eV) in the Mass-Selected and Total Anion Currents Measured in the DEA Spectra of Halopyrimidines 24, along with Absolute Total Cross Sectionsa) mass-selected m/e
peak energy
total peak energy
cross section (1016 cm2)
2-Chloropyrimidine 35 Cl
0.44
0.4
0.66
35 Cl
0.0 sh.
0.0
1.29
2-Bromopyrimidine
Figure 4. m/e = 79 anion current as a function of the incident electron energy measured in gas-phase bromopyrimidines 3 and 4.
the mass-selected spectrum can also come from kinetic energy discrimination in the anion extraction efficiency. Formation of halide anions at zero energy requires a Chalide dissociation energy of the neutral molecule smaller than the EA of the halogen atom (3.614 and 3.364 eV for the chlorine54 and bromine55 atoms, respectively). Table 2 reports the thermodynamic threshold energies for production of the halide fragment anions in the three halopyrimidines as obtained from DFT calculations. In agreement with the observed anion currents, the B3LYP/6-311+G(d) results, with inclusion of zero-point vibrational energy corrections, predict that production of halide fragments at zero energy is thermodynamically possible. Measurements of the absolute intensity of (usually sharp) zero energy peaks is a delicate task because the observed peak height can be easily influenced by the energy spread of the incident electron beam. Moreover, because the cross sections of different species can differ by several orders of magnitude, contributions from traces of impurities cannot be excluded. A more detailed discussion of possible reasons that render questionable the strength of the near-zero energy intensity in electron beam experiments is available in the literature.56 Regardless of these difficulties for the zero-energy peaks, most of the DEA studies reported in the literature are not concerned with quantitative measurements, although determination of the absolute or at least relative dissociative cross sections is an important aspect. The last column of Table 3 reports the absolute cross sections (as evaluated from the peak heights, normalized to the same electron beam current and sample pressure) obtained from comparison of absolute cross sections reported in the literature35,36 with our measurements on the same organochlorides (see the Experimental Section). The average conversion factor between the two sets of values, previously applied to bromobenzenes,30 is now applied to the present halopyrimidines. It can be noted that, for symmetry reasons, σ*/π* mixing would not occur in the rigid (planar) equilibrium structure of the neutral molecules, so that production of halide fragment anions relies on vibronic coupling or geometrical distortion of the anion on the time scale of the π* resonances. The DEA cross section (1.29 1016 cm2; see Table 3) of the zero-energy signal of 2-chloropyrimidine was found to be equal to that of bromobenzene at 0.66 eV, which, in turn, was found30 to be 4.13 times as large as that of chlorobenzene at 0.75 eV, whereas at 0.4 eV, the cross section of 2 was evaluated to be 0.66 1016 cm2. For 2-bromopyrimidine, the intensities of the
79 Br
0.32
0.2 sh.
5.2
79 Br
0.01
0.0
7.98
79 Br
0.34
0.2 sh.
6.0
79 Br
0.01
0.0
8.01
5-Bromopyrimidine
a
Evaluated uncertainty is (25%, see text.
peaks at zero energy and 0.2 eV were measured to be about 6 and 8 times larger, respectively, than the corresponding signals of the chloro analogue 2. The absolute DEA cross sections of 5-bromopyrimidine at zero energy and 0.2 eV were found to be equal to and about 15% larger than, respectively, those of the corresponding signals of the 2-bromo isomer (see Table 3). Although the difference can fall within the estimated experimental error, the larger cross section at 0.2 eV measured in 4 would be consistent with the lower energy of its σ*CBr resonance. Other conditions being the same, this should, in fact, favor π*/σ* mixing.
4. CONCLUSIONS The present work elucidates for the first time the valence empty-level structures of pyrimidine and halopyrimidines. The energies of vertical formation of π* and σ* temporary anion states measured with electron transmission spectroscopy were satisfactorily reproduced by calculations at the HF level on the neutral molecule, once the calculated virtual orbital energies were scaled with appropriate empirical linear equations as reported in the literature. The (π*) ground anion state of pyrimidine is about 0.4 eV more unstable than the neutral ground state, whereas the vertical electron affinities of the chloro and bromo derivatives were found to be close to zero. As expected, a low-energy σ* anion state was observed in the chloro and bromo derivatives, but the σ* anion state of 5-bromopyrimidine was, surprisingly, sizably more stable than that of its 2-bromo isomer. Dissociative electron attachment spectroscopy revealed that the first two π* anion states of the halopyrimidines can follow dissociative decay channels, leading to formation of the halide fragment anion. The total anion currents, measured at the walls of the collision chamber as a function of the impact electron energy, present sharp and intense maxima at zero energy and around 0.3 eV. The total anion yields for 2-bromo- and 5-bromopyrimidine were found to be about 6 times larger than that measured for the 2-chloro derivative. An evaluation of the total anion current absolute cross sections was obtained from comparison of absolute cross sections reported in the literature with our measurements on the same compounds. Application of this procedure led to a large absolute cross section (on the order of magnitude of 10161015 cm2) for the negative current 10780
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A observed at zero energy in the bromopyrimidines. Thus, if the initial mechanism for direct DNA damage is bond cleavage by low secondary electrons, the measured values of the cross sections suggest that bromopyrimidines are much more effective as radiosensitizers. These findings are completely in line with the experimental observations in bromouracyl by Abdoul-Carime et al.57 The present theoretical and experimental results complete the knowledge of the electronic structure of this class of molecules, which are the building blocks of DNA/RNA bases and several pharmaceutical compounds, and contribute to the understanding of the physical and chemical origin of the photoinduced processes and damage in these large biological molecules. Although photoemission data1114 can help in explaining the role of these halopyrimidines in the production of electron swarms in larger biomolecules, ETS and DEA data can provide useful hints on dissociative processes leading to single- or multiple-strand rupture in DNA/RNA molecules.
’ AUTHOR INFORMATION Corresponding Author
*Tel.: +39 051 2099522. Fax: +39 051 2099456. E-mail: alberto.
[email protected].
’ ACKNOWLEDGMENT The authors thank the Italian Ministero dell’Istruzione, dell’Universita e della Ricerca, for financial support (PRIN 2009SLKFEX). ’ REFERENCES (1) Bloomer, W. D.; Adelstein, S. J. Nature (London) 1977, 265, 620. (2) O’Donoghue, J. A.; Wheldon, T. E. Phys. Med. Biol. 1996, 41, 1973. (3) Watanabe, R.; Nikjoo, H. Int. J. Radiat. Biol. 2002, 78, 953. (4) Hofer, K. G. Acta Oncol. 2000, 39, 651. (5) Fayard, B.; Touati, A.; Abel, F.; Herve du Penhoat, M. A.; Despiney-Bailly, I.; Gobert, F.; Ricoul, M.; L’Hoir, A.; Politis, M. F.; Hill, M. A.; Stevens, D. L.; Sabatier, L.; Sage, E; Goodhead, D. T.; Chetioui, A Radiat. Res. 2002, 157, 128. (6) Wang, F. J. Mol. Struct. (THEOCHEM) 2005, 728, 31. (7) Fujii, K.; Akamatsu, K.; Yokoya, A. J. Phys. Chem. B 2004, 108, 8031. (8) Sanche, L. Mass Spectrom. Rev. 2002, 21, 349. (9) Plekan, O.; Feyer, V.; Richter, R.; Coreno, R.; de Simone, M.; Prince, K. C.; Trofimov, A. B.; Gromov, E. V.; Zaytseva, I. L.; Schirmer., J. Chem. Phys. 2008, 347, 360. (10) Storchi, L.; Tarantelli, F.; Veronesi, S.; Bolognesi, P.; Fainelli, E.; Avaldi, L. J. Chem. Phys. 2008, 129, 154309. (11) O’Keeffe, P.; Bolognesi, P.; Casavola, A. R.; Catone, D.; Zema, N.; Turchini, S.; Avaldi, L. Mol. Phys. 2009, 107, 2025. (12) Bolognesi, P.; Mattioli, G.; O’Keeffe, P.; Feyer, V.; Plekan, O.; Ovcharenko, Y.; Prince, K. C.; Coreno, M.; Amore Bonapasta, A.; Avaldi, L. J. Phys. Chem A 2009, 113, 13593. (13) Bolognesi, P.; O’Keeffe, P.; Feyer, V.; Plekan, O.; Prince, K.; Coreno, M.; Mattioli, G.; Amore Bonapasta, A.; Zhang, W.; Carravetta, V.; Ovcharenko, Y.; Avaldi, L. J. Phys.: Conf. Ser. 2010, 212, 012002. (14) Bolognesi, P.; O’Keeffe, P.; Ovcharenko, Y.; Coreno, M.; Avaldi, L.; Feyer, V.; Plekan, O.; Prince, K. C.; Zhang, W.; Carravetta, V. J. Chem. Phys. 2010, 133, 034302. (15) Schirmer, J.; Barth, A. Z. Phys. A 1984, 317, 267. (16) Tarantelli, F. Chem. Phys. 2006, 329, 11. (17) Sanche, L.; Schulz, G. J. Phys. Rev. A 1972, 5, 1672. (18) Jordan, K. D.; Burrow, P. D. Acc. Chem. Res. 1978, 11, 341.
ARTICLE
(19) Schulz, G. J. Rev. Mod. Phys. 1973, 45, 378, 423. (20) Illenberger, E.; Momigny, J. Gaseous Molecular Ions. An Introduction to Elementary Processes Induced by Ionization; Springer-Verlag: New York, 1992. (21) Balog, R.; Langer, J.; Gohlke, S.; Stano, M.; Abdoul-Carime, H.; Illenberger, E. Int. J. Mass Spectrom. 2004, 233, 267. (22) Dressler, R.; Allan, M.; Haselbach, E. Chimia 1985, 39, 385. (23) Stricklett, K. L.; Chiu, S. C.; Burrow, P. D. Chem. Phys. Lett. 1986, 131, 279. (24) Modelli, A.; Foffani, A.; Scagnolari, F.; Jones, D. Chem. Phys. Lett. 1989, 163, 269. (25) Bulliard, C.; Allan, M.; Haselbach, E. J. Phys. Chem. 1994, 98, 11040. (26) Modelli, A.; Venuti, M. J. Phys. Chem. A 2001, 105, 5836. (27) Modelli, A.; Venuti, M.; Szepes, L. J. Am. Chem. Soc. 2002, 124, 8498. (28) Modelli, A. Phys. Chem. Chem. Phys. 2003, 5, 2923. (29) Modelli, A.; Jones, D. J. Phys. Chem. A 2004, 108, 417. (30) Modelli, A. J.Phys. Chem. A 2005, 109, 6193. (31) Modelli, A.; Jones, D.; Distefano, G. Chem. Phys. Lett. 1982, 86, 434. (32) Johnston, A. R.; Burrow, P. D. J. Electron Spectrosc. Relat. Phenom. 1982, 25, 119. (33) Modelli, A.; Foffani, A.; Scagnolari, F.; Jones, D. Chem. Phys. Lett. 1989, 163, 269. (34) Modelli, A.; Guerra, M.; Jones, D.; Distefano, G.; Tronc, M. J. Chem. Phys. 1998, 108, 9004. (35) Pearl, D. M.; Burrow, P. D. J. Chem. Phys. 1994, 101, 2940. (36) Aflatooni, K.; Gallup, G. A.; Burrow, P. D. J. Chem. Phys. 2010, 132, 094306. (37) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N. ; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C; Jaramillo, J.; Gomperts, R. E.; Stratmann, O.; Yazyev, A. J.; Austin, R.; Cammi, C.; Pomelli, J. W.; Ochterski, R.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision A.02; Gaussian, Inc.: Wallingford CT, 2009. (38) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (39) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218. (40) Modelli, A.; Burrow, P. D. J. Electron Spectrosc. Relat. Phenom. 1983, 32, 263. (41) Nenner, I.; Schulz, G. J. J. Chem. Phys. 1975, 62, 1747. (42) Lane, N. F. Rev. Mod. Phys. 1980, 52, 29. (43) Guerra, M. Chem. Phys. Lett. 1990, 167, 315. (44) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio Molecular Orbital Theory; Wiley: New York, 1986. (45) Dunning, T. H., Jr.; Peterson, K. A.; Woon, D. E. Basis Sets: Correlation Consistent Sets in the Encyclopedia of Computational Chemistry; Schleyer, P. v. R., Ed.; Wiley: Chichester, U.K., 1998. (46) Heinrich, N.; Koch, W.; Frenking, G. Chem. Phys. Lett. 1986, 124, 20. (47) Staley, S. S.; Strnad, J. T. J. Phys. Chem. 1994, 98, 161. (48) Modelli, A.; Hajgato, B.; Nixon, J. F.; Nyulaszi, L. J. Phys. Chem. A 2004, 108, 7440. (49) Koopmans, T. Physica (Amsterdam) 1934, 1, 104. (50) Chen, D. A.; Gallup, G. A. J. Chem. Phys. 1990, 93, 8893. (51) Aflatooni, K.; Gallup, G. A.; Burrow, P. D. J. Phys. Chem. A 2000, 104, 7359. (52) Pshenichnyuk, S. A.; Asfandiarov, N. L.; Burrow, P. D. Russ. Chem. Bull., Int. Ed. 2007, 56, 1268. 10781
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
The Journal of Physical Chemistry A
ARTICLE
(53) Modelli, A.; Jones, D. J. Phys. Chem. A 2004, 108, 417. (54) Martin, J. D. D.; Hepburn, J. W. J. Chem. Phys. 1998, 109, 8139. (55) Blondel, C; Cacciani, P.; Delsart, C.; Trainham, R. Phys. Rev. A 1989, 40, 3698. (56) Graupner, K.; Graham, L. M.; Field, T. A.; Mayhew, C. A.; Fabrikant, I. I.; Miller, T. M.; Braun, M.; Ruf, M.-W.; Hotop, H. Int. J. Mass Spectrom. 2008, 277, 113. (57) Abdoul-Carime, H.; Huels, M. A.; Bruning, F.; Illenberger, E.; Sanche, L. J. Chem. Phys. 2000, 113, 2517.
10782
dx.doi.org/10.1021/jp206559d |J. Phys. Chem. A 2011, 115, 10775–10782
